RECORD: Stauffer, R. C. ed. 1975. Charles Darwin's Natural Selection; being the second part of his big species book written from 1856 to 1858. Cambridge: Cambridge University Press.

REVISION HISTORY: Transcribed (double key) by AEL Data, corrections by John van Wyhe 2.2007. RN2

NOTE: From the original dust jacket:

Charles Darwin's On the Origin of Species is unquestionably one of the chief landmarks in biology. The Origin (as it is widely known) was literally only an abstract of the manuscript Darwin had originally intended to complete and publish as the formal presentation of his views on evolution. Compared with the Origin, his original long manuscript work on Natural Selection, which is presented here and made available for the first time in printed form, has more abundant examples and illustrations of Darwin's argument, plus an extensive citation of sources. It had reached a length of over one quarter of a million words and was well over half completed when Darwin's writing was dramatically interrupted by the celebrated letter from Alfred Russel Wallace. After the brief preliminary announcement with Wallace, Darwin wrote out in eight months the new abstract of his views which appeared as On the Origin of Species. Darwin had originally intended this version to be An Abstract of Natural Selection, but his publisher insisted on a different title. The first two chapters of the manuscript version of Natural Selection became the two volumes of Variation of Animals and Plants under Domestication (1868). The following eight and a half chapters are published here under the title Natural Selection, which Darwin gave to this work in his letter to Asa Gray in 1857, published in the preliminary announcement of 1858.

The main interest of Natural Selection centres on the fact that On the Origin of Species was unique amongst Darwin's published books and formal scientific papers in appearing without a single footnote, so that the immediate sources of Darwin's facts and ideas were unknown. All this despite Darwin's firm belief in the desirability of publishing the references for these sources. For the first time the sources of information and, in the case of books, the edition used, are made known.

Professor Stauffer has edited Darwin's manuscript with meticulous care to make the lengthy text as readable as possible without departing from Darwin's own words, to clarify Darwin's source of citations and to explain the interrelationship of this work with other papers and published material. The editorial process that has been necessary to achieve these objectives is carefully explained in the introduction.

Help is given to the reader in relating this fuller version to the Origin by means of an index that is both an index of this volume and of the 1859 edition of the Origin.

The publication of this major work enables scholars and students in the history of science and all biologists interested in evolutionary processes and theories to examine the references and authorities for the statements made and conclusions reached in the Origin.

Reproduced with the permission of the Syndics of Cambridge University Library and William Huxley Darwin. This book is copyright of Cambridge University Press, and is reproduced with permission.

[page i]


[page ii]

[page iii]



University of Wisconsin, Madison

With the Guides to the texts, Collation with the first edition of the Origin
and Index prepared by
Mount Holyoke College, Massachusetts
Emeritus Fellow of St Catharine's College, Cambridge


Cambridge New York New Rochelle Melbourne Sydney

[page iv]

Published by the Press Syndicate of the University of Cambridge
The Pitt Building, Trumpington Street, Cambridge CB2 1RP
32 East 57th Street, New York, NY 10022, USA
10 Stamford Road, Oakleigh, Melbourne 3166, Australia
© Cambridge University Press 1975

First published 1975
First paperback edition 1987

British Library cataloguing in publication data
Darwin, Charles
[Natural selection. Ch. 3—11] Charles Darwin's Natural
selection: being the second part of his big species book written
from 1856 to 1858. 1.

Evolution 2. Natural selection
I. [Natural selection. Ch. 3-11] II. Title III. Stauffer, R.C.
IV. Smith Sydney, 1911- V. Rachootin, Stan
575.01'62 QH375

ISBN 0 521 34807 2

Transferred to digital printing 1999

[page v]


Commonly used symbols and abbreviation code page ix
Acknowledgements xi
General Introduction 1
Editorial considerations 15
Darwin's table of contents 25
III Possibility of all organic beings crossing 33
IV Variation under nature 92
V The struggle for existence 172
VI On natural selection 213
VII Laws of variation 275
VIII Difficulties on the theory 339
IX Hybridism 387
X Mental powers and instincts of animals 463
XI Geographical distribution 528
Appendices 567
Bibliography 587
Guides to the texts of the long and the short versions 630
Collation between the Origin and Natural Selection 635
Index 643

[page vi]

[page vii]

To Nora Barlow
On behalf of all who study her grandfather, Charles Darwin

[page viii]

[page] ix


L & L The Life and Letters of Charles Darwin, edited by Francis Darwin, 3 vols. (London, 1887)

NY New York edition of Life and Letters, 2 vols. (1888)

ML More Letters of Charles Darwin, edited by Francis Darwin and A. C. Seward, 2 vols. (London, 1908)


Reference numbers are those written on the individual folios and other pieces of paper, either given by Darwin or added later to unnumbered pieces

FC Refers to folios of fair copies of parts of the manuscript, where Darwin occasionally made changes from his earlier holographs

( ) Darwin's parentheses

< > Darwin's cancellations

[ ] Editor's additions to Darwin's text or notes

[?] Uncertain reading

/ Normal end of manuscript folio, note slip, etc.

// End of existing piece of manuscript where Darwin sheared off a portion to be transferred for use elsewhere

…. Unrestorable gap in text

s Seite, Darwin's usage for page citations in German sources

[page x]

[page xi]


Acknowledgements are inherently endless and incomplete, and I ask indulgence from the many friendly, helpful institutions and people whom I do not name here.

The dedication to Nora, Lady Barlow represents the great appreciation of many scholars to the whole Darwin family, especially Sir Robin Darwin and to the late Sir Alan Barlow for their preservation of the Darwin papers as an invaluable intact collection generously made available through their gifts to the Cambridge University Library, as well as deep personal gratitude for her friendly help.

Had this work appeared in two volumes, the second would have been dedicated to Dr Sydney Smith of St Catharine's College, Lecturer in Zoology in the University of Cambridge, for his persistent diplomatic and generous assistance to Darwinists and for his personal friendship. Peter Gautrey of the University Library, Cambridge, has been a warm helpful friend in many ways besides his invaluable contribution of careful copy reading of my final edited typescript against the original manuscript.

I am happy to record my indebtedness to my wife, Velma Mekeel Stauffer, specifically for inspired aid in deciphering some of the most illegible words in Darwin's handwriting and for many other hours of partnership in working together on the manuscript.

My colleagues and students of the Department of the History of Science of the Madison University of Wisconsin have supported my work with friendly encouragement and loyal patience. Helpful fellow scholars and friends include Loren Eiseley, John Brooks, William Steam, Bert J. Loewenberg, Thompson Webb, Jr, Thomas R. Buckman, the late Sir Gavin de Beer, Sten Lindroth, Mr O'Grady of the Linnean Society of London, and Mr Robinson of Down House.

Friendly and indispensable fellow workers on the manuscript include as research assistants Elizabeth Nash, who made the original typed transcript of chapter 11, and Alice Guimond who transcribed all the other chapters except five (which I did myself). M. J. S. Hodge and Albert A. Baker contributed valuable locations of Darwin's exact source and source editions. Graham Pawelec was inspired and inspiring in his editing of the bibliography. The Index and Concordance were complied under the supervision of Sydney Smith (for details see p. 630).

[page xii]

Edna Dahl, in typing the complete final manuscript, was never ruffled by the countless vagaries of my typing and handwriting, nor by the necessary editorial conventions unpredicted in standard secretarial training.

My indebtedness to institutions ranges widely indeed. My understanding of Darwin's field work basic for his research career, is founded on my own field experience supported for two summers by the Minnesota State Geological Survey under my father, Clinton R. Stauffer, and for three summers by the Woods Hole Oceanographic Institution under Alfred C. Redfield.

Background research was aided by library hospitality savante, gemutlich and kosmopolit in the Bibliothdque Nationale and the Museum National d'Histoire Naturelle in Paris and the University Libraries in Vienna and Uppsala.

West of the Channel my indebtedness for generous library assistance includes Widener, the Memorial Library of the Madison University of Wisconsin, the British Museum Reading Room and Manuscript Room, the British Museum (Natural History), the London Library, the Linnean Society of London, University College, London, and the Wellcome Institute of the History of Medicine. Then in Cambridge, England, Christ's College, the Balfour Library, the Botany School Library, and of course, happily indispensable, The University Library where I enjoyed borrowing privileges and much other aid thanks to H. R. Creswick, E. B. Ceadel, A. Tillotson, J. Claydon, and most of all P. Gautrey in the Anderson Room.

Financial support of the necessary research and editorial work came from the Wisconsin Alumni Research Foundation through the University Research Committee, travel funds from the American Philosophical Society, the essential fifteen months research grant NSF G‐13032 from the National Science Foundation, and an appreciated advance from the University Press, Cambridge. Without additional private support from departed Stauffer and Webb family this work would not have been completed.

Indispensable encouragement and cosmopolitan hospitality came from P. G. Burbidge, A. Winter and the Syndics of the Cambridge University Press and from the University Combination Room and the Senior Common Room of St Catharine's College, Cambridge.

To all, my warm and grateful thanks; and complete absolution for any errors still persisting despite their assistance. Any sins of omission or commission are my own.

July 10, 1973

[page break]

Chapter VII, folios 105 and 105A exemplify a discontinuity in the text resulting from cutting up the manuscript sheet. Presumably the missing middle third was used up in putting together the manuscript for Variation under Domestication.

[page break]

Sheared off portions of present folio 21 of chapter IX. Originally this was folio 6 of the earlier draft of this chapter. The missing portion of the quotation from Herbert's Amaryllidaceae can be restored from the text quoted in Variation under Domestication as well as from the original. See p. 399 n 4.

[page break]

Cancelled upper portion of folio 9, chapter V, showing Darwin's original beginning discussion of the struggle for existence. For transcription of this rejected passage see p. 569. For revised text, see chapter V, folio 9.

[page break]

Folio 94, chapter IX. In this sample folio of the manuscript, a sentence in the middle was cancelled before completion, and was reworded immediately thereafter.

[page] 1


On The Origin of Species was literally only an abstract of the manuscript Darwin had originally intended to complete and publish as the formal presentation of his views on evolution. Compared with the Origin, his original long manuscript work on Natural Selection, which is presented here, has more abundant examples in illustration of Darwin's argument plus an extensive citation of sources. It had reached a length of over one quarter of a million words and was well over half completed when Darwin's writing was dramatically interrupted by the celebrated letter from the other end of the world outlining Alfred Russel Wallace's astonishingly parallel but independently conceived theory of natural selection. Darwin felt obliged to change his plans for initial publication; and, after the brief preliminary announcement was presented jointly with Wallace's paper at the Linnean Society of London, he rapidly wrote out in eight months the new abstract of his views which appeared as the Origin of Species in 1859. But, he still planned to publish a more extensive account of his views on evolution, and he did not abandon his long manuscript, nor write on the unused backs of the sheets for drafting other new publications as he so often did with other manuscripts.1 As we shall see, the first two chapters of the manuscript became the two volumes of his Variation of Animals and Plants under Domestication (1868). The following eight and a half chapters are published here under the title, Natural Selection, which Darwin gave to this work in the 1857 letter to Asa Gray published in the preliminary announcement of 1858.2

Judging from my own experience in tracing and checking the references for the present work I believe any attempt on Darwin's part to recheck a normal number of references in order to include sources in footnotes for the Origin could have added a number of months to the time he needed to prepare his Origin manuscript for the copyist and then for the press. Because of the pressure to publish as quickly as possible a fuller statement of Darwin's evolutionary views once the preliminary announcement had been made at the July 1, 1858 meeting of the Linnean Society, the Origin of Species was, I believe, unique among Darwin's published

1 L & L, I, 121; NY, I, 99; e.g. Darwin MSS. 17 (i).

2 L & L, II, 128; NY, I, 480.



books and formal scientific papers in appearing without a single footnote. Therefore the Origin tantalizes us with questions as to what were the immediate sources of Darwin's facts and ideas, and Darwin was ahead of his critics in mentioning the desirability of publishing the references for these sources. Already on the second page of the first edition of the Origin he stated: 'This Abstract, which I now publish, must necessarily be imperfect. I cannot here give references and authorities for my several statements … No one can feel more sensible than I do of the necessity of hereafter publishing in detail all the facts, with references, on which my conclusions have been grounded; and I hope in a future work to do this.' The first part of this moral obligation to publish his sources Darwin satisfied in 1868 when he published his two volumes on Variation under Domestication. For his own selection of his other sources we must examine the present work on Natural Selection. One already published illustration of the use that can be made of this work is the specific confirmation of earlier speculation about the derivation of Darwin's ecological concept of the economy of nature from the Linnean dissertations.1

The manuscript proves to support Darwin's comment about it: 'I fear my M.S. for the bigger book… would be illegible,'2 … partly because of his handwriting and partly because his drastic revisions often obscure the continuity of the text. Transcription of a considerable portion of the manuscript seemed the best way to start studying its content, and I have completed the transcription and editing of the manuscript in order to make the work generally accessible and with the particular hope that it will promote informed analysis of other aspects of Darwin's work.

The first desideratum in introducing the manuscript would seem to be to supply the reader with the background information most useful for understanding the work. Certain historical material seems important here, and I have tried to present a reasonably full account of the immediate history of the manuscript as well as of the editorial procedure followed. Some important details relate specifically to individual chapters, and these will be presented separately before the parts of the text immediately concerned.

My editorial aim has been to take time not only to clarify the lengthy text to make it as readable as possible but also to clarify Darwin's source citations. His abbreviated references, in a form

1 R. C. Stauffer, 'Haeckel, Darwin, and Ecology' , Quart. Rev. Biol, 32 (1957), 138-144; 'On the Origin of Species: an unpublished Version', Science, 130 (1959), 1449-52, 'Ecology in the Long Manuscript Version of Darwin's Origin of Species and Linnaeus' Oeconomy of Nature', Amer. Phil. Soc, Proc, 104 (1960), 235-41.

2 L & L, II, 281; NY, II, 75.


natural to an unfinished draft, are often somewhat cryptic. Where necessary I have added to Darwin's notes just enough clues to key them clearly to the cumulative bibliography where fuller titles, editions, and dates have been included; and blank citations and important details have been filled in as far as possible from the clues Darwin left in his papers and his notebooks listing titles and dates for the books he had read.

The citation of almost 750 books and articles makes the bibliography of the long manuscript an extensive guide to the sources Darwin selected out of his very comprehensive reading as most valuable for his own purposes. It is valuable as well for the modern student of the evolution of biological thought, whether as scientist or historian; for in giving us his own selective reading list for the preceding century of natural science Darwin has pointed out a pathway offering a representative view of a scientific literature formidably vast for exhaustive examination by any single scholar.


The Natural Selection manuscript not only has a prominent place in a considerable sequence of Darwin papers touching on evolution, but it is related to many more of the notebooks, papers, letters and annotated books, journals, and pamphlets which, together with his memories of his extensive field experience particularly in South America, and his continuing observations and experiments, constituted Darwin's working materials for his writing. The Darwin family, Down House authorities, the British Museum (Natural History), and the University officials at Cambridge have done everything feasible to make these available to scholars. Nature1 published something of the contents of these papers, but scholars must proceed to the invaluable Handlist of Darwin Papers at the University Library Cambridge (Cambridge, 1960) for an indispensable survey, which gives a preliminary account and listing of the more than 150 major parts or groups of items in that part of the collection then already at Cambridge. The manuscript material here published from this part of the collection will be identified by the reference numbers published in this Handlist. Since this Handlist was published, important new portions of Darwin's papers have been located and made available in the University

1 'Darwin Manuscripts and Letters, Gifts to Cambridge and Down House', Nature, 150 (1942), 585.



Library Cambridge by Sir Robin Darwin.1 These latter items will be designated by the reference numbers for the sections of the collection in which they occur which were assigned in the hand-written 'Catalogue of the MSS, papers, letters, and printed books of Charles Darwin now at Gorringes, Downe, Kent, July, 1932', made by Mrs Catherine Ritchie Martineau.

In addition to these notebooks, note sheets and slips, scientific journals, and manuscript drafts included in these papers, we must note the extensive marginalia and note slips in Darwin's scientific library of books, journals, reprints and pamphlets. Those books with significant annotations are now in the University Library Cambridge together with Darwin's reprint collection, on loan from the Botany School by courtesy of the Professor of Botany.2

Darwin's papers and library constituted an interrelated set of working materials for Darwin, and when studied together they help reveal the development of his thought as Sydney Smith has elegantly shown3 Many of these papers, valuable as background for understanding the present work on natural selection have now been published. Four of the pocket note books filled with evolutionary evidence and ideas from July 1837 to July 1839 have been published by Sir Gavin de Beer.4 The two early drafts of Darwin's evolutionary argument written in 1842 and 1844 were published together in 1909 by Francis Darwin.5

For general background, the Life and Letters of Charles Darwin is of prime importance, particularly the chapter, entitled 'The Unfinished Book', which is devoted to the Natural Selection manuscript. Then besides the More Letters also published by Francis Darwin, and the complete Autobiography published by

1 Sir Gavin de Beer, M. J. Rowlands, and B. M. Skramovsky, 'Darwin's Notebooks on Transmutation of Species, Part vi', fir. Mus. not. Hist. Bull. (Hist. ser.) 3 (1967), 131-2.

2 For titles see the mimeographed 'Darwin Library: List of books received in the University Library Cambridge, March-May, 1961'; see also: Cambridge University. Botany School., Catalogue of the Library of Charles Darwin . . . comp. Henry William Rutherford (Cambridge, 1908). Peter Vorzimmer notes in Isis 54 (1963), 374, n. 11 that there are also about a quarter of a million words of Darwin's marginalia in his 2500 reprints.

3 'The Origin of "The Origin"as discerned from Charles Darwin's Notebooks and his Annotations in the Books he read between 1837 and 1842', Advmt. Sci., London, 16 (1960), 391-401.

4 B. M. Bull. 2 (1960) 25-183; 2 (1961) 185-200; 3 (1967) 129-76; see also 'A Transcription of Darwin's First Notebook on "Transmutation of Species"', ed. Paul II. Barrett, Mus. Comp. Zool. Harvard, Bull., 122 (1960) 245-96.

5 First published as The Foundations of the Origin of Species: Two Essays Written in 1842 and 1844 by Charles Darwin. Ed. Francis Darwin, (Cambridge, 1909); republished in Evolution By Natural Selection. With a foreword by Sir Gavin de Beer (Cambridge, 1958).


Nora Barlow, we should note the valuable chronological details in the Pocket Diary,1 kept by Darwin from 1838 to 1881, of which de Beer has published an old copy made by an amanuensis.2 More recently the long sought original diary has been found, so I have been able to rely upon that.


The more immediate background of what Darwin came to call 'my big book'3 starts before the middle of the 1850s. In 1853 Darwin's first major scientific honour came to him in the Royal Society's award of the Royal Medal in recognition of his books on the Geology of the Voyage of the Beagle and his comprehensive taxonomy of the barnacles.4 The latter work, which confirmed his position as a professionally qualified biologist,5 was then near enough to completion so that he could mention to Hooker his expectation to be at work on his 'species book' in a year or two.6 The next year he was ready to pack up his barnacle specimens, arrange for distributing copies of his publication; and, with the decks thus clear, he recorded in his Pocket Diary, for September 9, 1854: 'Began sorting notes for Species theory.' In March of 1855 he wrote to his second cousin and close college friend, William Darwin Fox, that 'I am hard at work at my notes collecting and comparing them, in order in some two or three years to write a book with all the facts and arguments, which I can collect, for and versus the immutability of species.'


These notes on his thoughts and on his extensive reading in the search for relevant facts formed an important part of Darwin's working materials along with his current observations and experi-

1 So designated in: Cambridge. University. Order of the Proceedings at the Darwin Celebration held at Cambridge June 22-June 24, 1909, with a Sketch of Darwin's Life. (Cambridge, 1909), p. [13] n. Francis Darwin referred to it as 'Diary' or 'Pocket-Book', e.g. L & L, I, p. iv; ML, I, p. xvii: 'Outline of Charles Darwin's Life based on his Diary, dated August, 1838', and Foundations, xiv, xvii.

2 'Darwin's Journal', B. M. Bull., 2 (1959), 1-21. Darwin's original diary is with the C.U.L. CD. MSS. item D 5.

3 L & L, II, 85; NY, I, 443.

4 Roy. Soc. London, Proc., 6 (1854), 355-6.

5 Cf. opinions of Hooker and Huxley, L & L, I, 346-8; NY, I,314-16.

6 L & L,II, 41; NY, I, 402.

7 L & L, II, 46; NY, I, 406; Christ's College (Cambridge) Library, Darwin-Fox letter no. 87.


ments and his memories of his fundamentally important field experience in South America.

Charles Darwin and his son Francis have both described his procedure in regard to his working papers;1 and examination of the extant manuscripts allows us to understand in some significant detail how his system actually worked. Initially and at least until 1839, Darwin jotted his notes and thoughts on his readings in small bound notebooks, such as those on evolution published by Sir Gavin de Beer. Later, he changed to the opinion of Alphonse de Candolle: 'The essential is to be able to compare, classify, and rearrange the materials up until the definitive writing, without being obliged to tear apart a notebook or to copy and recopy what one has written.'2 In his Autobiography, Darwin explained: 'I keep from thirty to forty large portfolios, in cabinets with labelled shelves, into which I can at once put a detached reference or memorandum' (p. 137). Thus for example his assembled notes and correspondence containing useful facts on the struggle for existence are still together in volume 46(i) of the Darwin Papers at Cambridge. Notes for other chapters of his evolution book are similarly grouped together. Finally he resolved even to select, separate, and sort out the many pages of his early evolution notebooks which had material he might use in his species book. Inside the front cover of Notebook B, the first of them, he wrote: 'All useful pages cut out. Dec. 7/1856.' The other three notebooks also have numerous pages cut out and have Darwin entries such as that inside the front cover of the second: 'All good References selected Dec 13 1856.'3 Of these selected reference notes, pp. 253-4 from the second notebook were attached to the verso of folio 14 of chapter VII of the Natural Selection manuscript along with other note slips including one stating: 'In Portfolio "Instinct" some excellent facts from Bachman on change of ranges in N. American Birds. . .' (See Appendix for chapter VII.) Similarly, page five selected and cut out of the third notebook, is also explicitly related to the Natural Selection manuscript by Darwin's pencilled classification: 'Ch IX Mongrels & Hybrids.'4 As we shall see, Darwin's scientific papers can provide clear identification of some of the references in his manuscript which he abbreviated too drastically to be self-explanatory.

Besides sorting his notes and selected letters into classified

1 L & L, I, 100, 151-2; NY, I, 80, 127-9; Autobiography, 137-8.

2 La phytographie. . . (Paris, 1880), p. 37; cf. L & I"III, 333; NY, II, 505-6.

3 de Beer, B. M. Bull., 2 (1960), 41, 82, 128, 160.

4 de Beer ed., B. M. Bull., 3 (1967), 157.


portfolios, Darwin also compiled useful surveys and an index of his reading in the form of notebooks listing, in roughly dated sequence, short titles of the books of both scientific and non-scientific which he had read.1 His papers include a long series of abstracts of books, pamphlets, and articles from scientific journals.2


Would the facts noted from thousands of pages of reading really support the theory of evolution by natural selection Darwin had sketched out in 1842 and developed in the 1844 essay? This question seems to have been in his mind when he wrote Hooker on March 26, 1854: 'How awfully flat I shall feel, if, when I get my notes together on species, &c., &c, the whole thing explodes like an empty puff-ball', and again in 1855: 'I should have less scruple in troubling you if I had any confidence what my work would turn out. Sometimes I think it will be good; at other times I really feel as much ashamed of myself as the author of the Vestiges ought to be of himself.'3

Robert Chambers' still anonymous Vestiges of Creation had indeed stirred up wide discussion of evolution among the reading public, but it had not persuaded experienced scientists that they might need to re-examine their adverse verdict against the muta-ability of species.4 From the point of view of desiring favourable consideration of the strictly scientific value of an evolutionary theory, Darwin could well write 'Lamarck… has done the subject harm, as has Mr. Vestiges.'5 Darwin could maintain that 'without speculation there is no good and original observation',6 but he was not interested in abstract theorizing for its own sake. In his mature period, works such as his grandfather's Zoonomia could only leave him 'much disappointed, the proportion of speculation being so large to the facts given'.7

In 1855 Darwin's tactical problem was clear. Previous discussion of evolution as presented in the works of Lamarck and in the still anonymous Vestiges of Creation had only led to its being rejected or ignored by the vast majority of scientists. To win much unprejudiced consideration of his views, Darwin had to succeed where others had failed.

1 Darwin MSS., items 119, 120, 128.

2 Darwin MSS., Vols. 71-5, item 116.

3 L & L, II, 44; NY, I, 404; ML no. 43.

4 Milton Millhauser, Just before Darwin: Robert Chambers and Vestiges (Middletown, Conn., 1959), Chs. 5, 6, esp. pp. 125, 148-9.

5 L, & L, II, 39; NY, I, 399.

6 L & L, II, 108; NY, I, 399.

7 Autobiog., p. 49.


Friends such as Charles Lyell and Lyell's brother-in-law Charles J. F. Bunbury, the squire of Mildenhall, certainly deserve credit for encouraging Darwin at this period. In a letter dated April 16, 1856, Bunbury wrote to Darwin: 'I am exceedingly interested by all you tell me about your researches & speculations on species & variation & distribution, & am delighted that you are going on working at the subject. I trust that you will not on any account give up the idea of publishing your views upon it; tho' neither you nor any one else may be able to unravel the whole mystery, or to command the universal assent of naturalists, still the research of one who has studied the whole question so long, & with such extensive knowledge & in so philosophical a spirit, cannot fail to be of very great advantage to science. The whole subject,—I mean every thing connected with the geography of plants & animals, including all the questions of distribution & variation, is to me particularly interesting & delightful; but how much we have yet to learn upon it! The difficulties which appear to attend upon each & every one of the theories,—of specific centres, of multiple creation, & of transmutation,—are so many, that what is most clear to me is the necessity of caution & candour, of avoiding dogmatism, & of giving a fair consideration to every fact & argument on any side. I say this, because the theory to which you lean is the most remote from that to which I incline, & yet I am quite ready to admit that your notion may be the right one.'1

Early in May Darwin was corresponding with Lyell and with Hooker about the former's urgent recommendation that Darwin publish a preliminary sketch of his views on evolution,2 and in his Pocket Diary Darwin recorded for May 14, 1856: 'Began by Lyell's advice writing Species Sketch.' But the initial doubts Darwin expressed to Hooker about publishing a preliminary announcement of his views without giving supporting evidence grew stronger. Meanwhile his letters began to dwell on problems concerning the geographical distribution of plants and animals. On July 8, 1856, he wrote to Lyell: 'I have just been quoting you in my essay on ice carrying seeds in the S. Hemisphere…

1 Darwin MSS. c. 40. c. The end of this letter with the signature is missing, but the reference to 'my Cape book' and Mildenhall as the place of writing both suggest Bunbury as the writer, and this is confirmed by Darwin's reply, given in the introduction to chapter XI, p. 528f. Bunbury had heard Darwin talk about his evolutionary views as early as Nov. 23, 1845, see The Life of Sir Charles J. F. Bunbury, Bart. Edited by his sister-in-law Mrs. Henry Lyell. (London, 1906), I, pp. 213-14. About Bunbury, see also the obituaries by J. D. Hooker, Roy. Soc. London, Proc. 46 (1889), xiii-xiv, and by J. W. Judd, Geol. Soc. London, Proc. Session 1886-7, pp. 39-40.

2 L & L, II, 67-71; NY, I, 426-30.


[See Natural Selection, chapter XI, folio originally numbered 39 (now 47).]

Hooker, with whom I have formerly discussed the notion of the world or great belts of it having been cooler. . . I think is much inclined to adopt the idea.—With modification of specific forms it explains some wondrous odd facts in distribution.

But I shall never stop if I get on this subject, on which I have been at work, sometimes in triumph, sometimes in despair, for the last month.'1

By mid-July he had so far enlarged his proposed scale of writing as to mention (apparently as already written) forty pages just on the influence of the glacial period on distribution.2 And this was the scale and scope of the first draft of chapter XI of the present manuscript.

As he soon explained to Lyell, 'I have found it quite impossible to publish any preliminary essay or sketch; but I am doing my work as completely as my present materials allow without waiting to perfect them. And this much acceleration I owe to you,'3

Thus Darwin was under way on actually writing his species book. The first chapter on stock breeding and on variation under domestication he left in an imperfect state,4 but he was sufficiently satisfied with the second chapter to record its completion on October 13, 1856 in his Pocket Diary. In November he wrote to Lyell, 'I am working very steadily at my big book',5 and as he finished each succeeding chapter or major section he continued to record his progress in his Pocket Diary by noting the dates, which appear at the start of the chapters in this edition.

Darwin not only wrote first drafts of his chapters, but he also revised them, rewrote, reorganized, expanded, and supplemented. On special points he consulted many authorities such as Hooker and Huxley by letter, and even had fair copies made of sections such as those on variations in large and small genera and on geographical distribution to send them to Hooker for his general opinion. Such details of the history of the manuscript will be covered in the appropriate chapter introductions.

By the spring of 1858 Darwin had completed his tenth chapter and had recently finished for chapter IV a major supplement on divergence, when, on June 18, his writing was interrupted by the arrival of Wallace's letter with its sketch of evolutionary processes in terms so surprisingly close to Darwin's own. Darwin's agreement,

1 C.U.L. CD. MSS.; 146, Lyell letter no. 54.

2 ML, no. 49 to Hooker, July 13, 1856.

3 L & L, II, 85: cf. 71, 84; NY, I, 443, cf. 430, 442.

4 ML no. 84.

5 L & L, II, 85; NY, I, 443


following the strong urging by Lyell and Hooker, to present along with Wallace's letter brief selections of his own writings which had been read in previous years by Hooker and Asa Gray is well known.

After the harrowing interval with both scarlet fever and diphtheria spreading from the village of Downe into his own house, with nurses sick as well as children, culminating in the death of his youngest child three days before his paper was presented at the Linnean Society meeting, Darwin started to write a formal article on his views for the Linnean Society. This article by March 1859 had grown into a complete book, well characterized by his proposed title: 'An Abstract of an Essay on the Origin of Species and Varieties through Natural Selection.' In regard to this title, he wrote Lyell: 'I am sorry about Murray [publisher of the Origin] objecting to the term Abstract, as I look at it as the only possible apology for not giving references and facts in full, but I will defer to him and you.'

When Wallace's letter interrupted Darwin's writing program on June 18, 1858, the long manuscript had covered about two thirds of the topics later presented in the Origin of Species. If we estimate the length of the surviving eight and a half chapters of Natural Selection at 225,000 words, and project to the fourteen chapters as in the Origin, this would indicate a length of about 375,000 words if the work had been completed. This would have made a book perhaps slightly longer than Murchison's Silurian System but certainly shorter than Lyell's Principles of Geology, and the scale does not seem inordinate considering the standards of the days of double-decker and triple-decker novels. Of the fourteen chapters of the Origin, nine had been preceded by extensive treatment in Natural Selection. The table below not only shows the correlation between the two works but also suggests some of the reorganization of the argument in the later work. In comparing the two works we can agree with Darwin's remark to Hooker that writing the Origin as an Abstract of his long manuscript 'has clarified my brains very much, by making me weigh the relative importance of the. several elements'.2 Yet in view of the great amount of writing on Natural Selection actually completed and the more than 1,800 pages which Darwin published just in the decade after 1858, the assertion that without the pressure arising from Wallace's 1858 letter Darwin would never have finished his Species Book seems unpersuasive.

In 1859 Darwin presented the first edition of the Origin as a preliminary announcement, simply an abstract of his work, stating

1 L & L, II, 153; NY, I, 508.

2 L & L, II, 138; NY, I, 494.


TABLE. Comparing Natural Selection and Origin of Species

Chapter MSS. Chapter 1859
I Variation under Domestication I Variation under Domestication
II Variation under Domestication (cont.)
III On possibility of all organisms crossing: on susceptibility to change Partly in IV
IV Variation in Nature II Variation under Nature
V Struggle for Existence III Struggle for existence
VI Natural Selection IV Natural Selection
VII Laws of Variation V Laws of Variation
VIII Difficulties in Transitions VI Difficulties on Theory
IX Hybridism & Mongrelism VIII Hybridism
X Instinct VII Instinct
Section on Geographical Distribution XI Geographical Distribution

that 'No one can feel more sensible that I do of the necessity of hereafter publishing in detail all the facts, with references, on which my conclusions have been grounded; and I hope in a future work to do this.' and 'My work is now nearly finished; but.. .it will take me two or three more years to complete it…'1

As we have seen, even the Natural Selection manuscript had been for Darwin a condensed form of the presentation he preferred for his material, and he recorded in his Pocket Diary that in January 1860, he 'Began looking over MS for work on Variation.' As he wrote to Asa Gray, this was to be 'the first part forming a separate volume, with index etc. of the three volumes which will make my bigger work'.2 By June he recorded the completion of the second chapter of the work eventually published in 1868 as The Variation of Animals and Plants under Domestication, and he continued to record his writing progress in his Pocket Diary until 1867 when in March he received the first proof. Thus instead of completing the Natural Selection manuscript he expanded the scale of his treatment, so that the two volumes on Variation represent the first two chapters of Natural Selection. He also published material from other parts of Natural Selection in Variation. There are now folios missing from the surviving Natural Selection manuscript and other folios with part of the text cut away. These gaps can often be related to topics which were treated in both works3 and it seems evident that he simply incorporated passages

1 Origin, pp. 2, [1].

2 L & L, II, 270; cf. II, 318; NY, II, 64, cf. II, 111.

3 E.g. Natural Selection, ch. 9, fols. 36 v and 40.


from the older manuscript into the new one by transferring what he had already written to save himself recopying.1 A further such transfer and incorporation of materials on variation from the first two chapters of Natural Selection would easily account for the fact that of those initial chapters only one folio (here published in the appendix) has been preserved with the remainder of the manuscript.2 It also could account for the fact that some few of the pages selected and cut out of the transmutation notebooks seem to be lost permanently. Unfortunately aside from the preliminary draft on Pangenesis practically none of the manuscript of Variation under Domestication seems to have survived. Apparently the rest of the two initial chapters of Natural Selection were thus used up and discarded.3

In 1867, when he finished writing his Variation under Domestication, he still considered this as the first part of his big Species Book, which was to be completed with two more works, and he still expected to publish the material covered by the Natural Selection manuscript, which he had so carefully saved and to which he then returned to write addenda.4 In the introduction after describing the scope of the two volumes to be published in 1868, he announced that the 'problem of the conversion of varieties into species.. .will form the main subject of my second work'.5 Here, 'after treating of the Variation of organisms in a state of nature, of the Struggle for Existence and the principle of Natural Selection, I shall discuss the difficulties which are opposed to the theory. These difficulties may be classed under the following heads: the apparent impossibility in some cases of a very simple organ graduating by small steps into a highly perfect organ; the marvellous facts of Instinct; the whole question of Hybridity; and, lastly, the absence, at the present time and in our geological formations, of innumerable links connecting all allied species.'6

This prospectus of the 'second work' fits the present manuscript, except that the latter does not include a discussion of missing fossil links. Instead it includes a section on the effects of the ice age as the only completed part of Darwin's fuller discussion of geographical distribution.

This section is the only portion of the manuscript which

1 See Natural Selection MS. ch. 9, fol. 136 v where Darwin wrote in regard to a missing note, presumably on a separate slip of paper: 'Note used in Domestic Animals Chapter 15, Crossing.'

2 Darwin MSS. vol. 51, see Robert C. Olby, 'Charles Darwin's Manuscript of "Pangenesis"', Brit. J. Hist. Sci., 1 (1963), 251-63.

3 Cf. L & L, I, 121; NY, I, 99.

4 See addendum dated 1867, on fol. 67 of ch. 4.

5 Variation, I, 5.

6 Variation, I, 8.


seems to fit best with Darwin's prospectus for the concluding part of his full-scale Species Book: 'In a third work I shall try the principle of natural selection by seeing how far it will give a fair explanation of. ..several large and independent classes of facts; such as the geological succession of organic beings, their distribution in past and present times, and their mutual affinities and homologies.'1

This program, which Darwin outlined in the introduction to Variation under Domestication, he never completed. His letter of July 6, 1868, to Alphonse de Candolle explains: 'You ask me when I shall publish on the "Variation of Species in a State of Nature." I have had the MS. for another volume almost ready during several years, but I was so much fatigued by my last book that I determined to amuse myself by publishing a short essay on the "Descent of Man".. .Now this essay has branched out into some collateral subjects, and I suppose will take me more than a year to complete. I shall then begin on "Species", but my health makes me a very slow workman.'2

For the Descent of Man (1st ed. 1871), Darwin again evidently quarried in his Natural Selection manuscript. On folio 13 of the manuscript for chapter x on instinct he scrawled in the margin: "Used Man Book", and the textual gaps created when he sheared off portions of folios 11 and 12 of that chapter can be filled from the corresponding passages in the Descent. As Dr Alice Guimond discovered when she was my research assistant, Darwin published more material from the manuscript in 1868 in an article on specific differences in Primula, later incorporated in his book on The Different Forms of Flowers in Plants of the same Species (1877) and other material in his book on The Effects of Cross and Self Fertilisation in the Vegetable Kingdom (1876).3

Besides Darwin's own use of the materials in the Natural Selection manuscript, its history also includes loans of sections to scientist friends, and some authorized posthumous publication. In November, 1859, Huxley had begun to consult Darwin in preparation for the lecture 'On Species and Races, and their Origin', which he gave at the Royal Institution on February 10, 1860,4 and Darwin soon loaned him the manuscript of chapter IX on hybridism and of his

1 Variation, I, 9.

2 L & L, III, 100; NY, II, 280.

3 For example, compare his discussion on the primrose and the cowslip on folios 68-79 of ch. IV with his article in Linn. Soc. J (Botany) 10 (1868) 437-54 and with ch. II of The Different Forms of Flowers.. .(London, 1877) and cf. fols. 27, 27 v, and 30 of ch. 3 with Cross Fertilisation, pp. 378-9, 395.

4 'On Species and Races, and their Origin', Roy. Inst. G. B., Proc., 3 (1860), 195-200.


TABLE. Stages of Darwin's Organized Writing
on the Origin of Species

Dates of Writing
Version Short Title (from Pocket Diary) Estimated Length
I 1842 Sketch May, June, 1842 15,000 words
II 1844 Essay Finished July 5, 1844 52,000 words
III Natural Selection July 1856-June 1858 225.000 words extant
IV Origin of Species July 1858-March 1859 155,000 words
(c. 80,000 in parts
corresponding to
(6th ed., present text of
June-Oct. 1871) Natural Selection)
V Variation under March. 1860- 315.000 words
Domestication January, 1867

discussion of pigeons, presumably from chapter II.1 The manuscript materials on instinct which Darwin loaned G. J. Romanes, who published portions of them, will be discussed in the editorial introduction to chapter X. After Darwin's death, his son Francis loaned some of the manuscript to Wallace, and allowed him to publish excerpts, particularly about variation among wild species, in his book on Darwinism.2

In reviewing the history of Darwin's organized writing on evolution we can see that the Natural Selection manuscript forms part of a sequence of versions which can be summarized in the table above.

1 L & L, ii, 251, 281; NY, 46, 75; ML, nos. 84, 85.

2 1st ed. (London, 1889), pp. viii, 46, 69, 79-80. These quotations are from Natural Selection ch. IV, fols. 25 to 33.

[page] 15


Since Darwin painstakingly wrote and revised his manuscript with publication in view, the first aim of this edition is to print the book Darwin had in mind.

The text is so long that I believe readability should take precedence over the inclusion of minor details of the manuscript such as insignificant cancellations. For such details, the original manuscript is available in the Anderson Room of the University Library, Cambridge, and a microfilm is available in the library of the University of Wisconsin at Madison. Examination of the accompanying facsimiles of manuscript passages and comparison with the printed text will reveal some of the problems and illustrate the editorial procedure followed.

The first edition of Variation of Animals and Plants under Domestication offers a model of format, including the setting of subordinate material in reduced type in the text. Today, however, long footnotes even covering more than a full page of text (e.g. Variation, II, pp. 375‐6), which did not discourage thousands of Victorian book buyers, now do seem extreme; and in the present work, where long notes could be smoothly incorporated into the main text, this has been done (e.g. chapter III, fol. 64 v).

Occasional gaps occur in the manuscript where Darwin later apparently used passages in preparing his published books. In such cases the continuity has been supplied by quotations on the same subject matter from his other works.1 For example on folio 105 of chapter VII, shown in the accompanying facsimile, the content of the surviving top and bottom portion of the cut‐up manuscript sheet corresponds closely to the text on page 163 of the first edition of the Origin, which thus supplies 'a double shoulder stripe' as the missing subject of the incomplete sentence fragment at the top of 105A. The information in the cancelled passage on the common donkey Darwin repeated in Variation under Domestication (I, 3) where he added a source reference to Martin's The Horse. Perhaps Darwin cut off the missing middle third of this folio in order to attach this reference to the expanded manuscript he was making up for chapter II of Variation, while reserving the information on Hemionus at the foot of the sheet (now 105 A) for

1 E.g. ch. IX, fol. 36 v, where the surviving MS. note corresponds to note 12, Variation II, p. 105.


use much farther on in chapter XIII, note 36 (Variation, II, 43). This probably explains Darwin's pencil scrawl: 'All used' on 105 A. In chapter IX, the present folio 21 (which was renumbered, since it was folio 6 of the earlier draft) has been cut up, and part is gone. (See the accompanying facsimiles.) The missing portion of the quotation from Herbert's Amaryllidaceae can be restored from the text quoted in Variation under Domestication as well as from Herbert's original text. Again in chapter IX the note surviving on the lower part of a sheared‐off folio now numbered 36 V corresponds to note 12, p. 105 of the second volume of Variation and the text for the missing upper part of the folio can be restored from the published text.

Besides the gaps left by these selective excisions, Darwin left occasional blank spaces to be filled in when he might later find the appropriate names, numbers, or citations, and these have been filled in from the sources Darwin used, where this is feasible.


Before the text reached its present form, Darwin had worked it over in ways which leave many traces in the manuscript. He customarily wrote in ink on one side of folios of paper measuring about 8_ by 12_ inches. In revising, he cancelled by making horizontal lines through words or lines or by vertical lines through longer passages so that the earlier wording is usually readable. (See the facsimile of folio 9 of chapter V.) Some revising he did immediately by cancelling an incomplete sentence and starting anew, as in the middle of folio 94, chapter IX, shown in the facsimile. Similarly in chapter III, in the long note following folio 32, Darwin made three false starts: '<It is almost superfluous, but I may state> that <Yet somewhere [?] I have observed instances quite off [?] Although I have seen quite enough to convince me that this claim is quite fanciful, yet> Nevertheless some facts could be given <to> in favour of <it> such a view:' Some cancellations merely show that Darwin alternated in his mind between equivalent wordings so that a complete reproduction of the manuscript would read: '<no doubt in all probability> no doubt' (ch. IV, fol. 7), '<identical absolutely similar> identical' (ch. VII, fol. 66), 'makes <us one> us feel' (ch. IX, fol. 71), and 'instinctive actions, wondrous though they <are be are> be' (ch. X, fol. 3). For the sake of readability such minor variants and cancelled passages which were rephrased with essentially the same content have been omitted


from the printed text. In other clearly important instances, such as alternatives for the phrase 'struggle for existence' the worked-over original text has been printed in the appendix. Other cancelled words, phrases, and passages which seem to amplify or clarify Darwin's thought have been printed within angle brackets in the regular text wherever feasible and otherwise have been placed in the appendix.

Besides cancellations, Darwin's revision led to additions of new material. Sometimes he wrote words or phrases in between the lines. Sometimes he wrote additions on the blank versos of his folios. He wrote some additions on separate slips of paper pinned or pasted on to the manuscript. (See the facsimile of folio 94, chapter IX, where he signalled an interpolation by adding 'a text') With rare exceptions I have found no useful clues such as water-marks to help date these additions. On a few occasions differences in the colour of the ink reveal a lapse of time between the writing of text and of revisions, but the length of the interval is uncertain. For a very few addenda or notes, Darwin supplied dates; these range from October 10, 1856 (ch. XI, fol. 6 V), through June, 1858 (ch. VI, fol. 53A) to 1867 (ch. IV, fol. 67). Addenda longer than a line or two written on the backs of folios have been designated by the folio number followed by 'V' for verso and this same sign designates many of the additions made on slips apparently later pinned to the backs of manuscript sheets.

If we can generalize from two instances where Darwin turned over a sheet after cancelling a false start of a few lines on a folio he had already numbered, he numbered his folios as he wrote. Interpolations he often designated by lettering; for example in chapter VI he interpolated a sequence running from folio 26a to 26nn, discarded folio 27, and designated the following sheet '27 & 28'. Some chapters he reorganized drastically, cancelling the original folio numbers and supplying new ones. In chapter IX, the folios originally numbered 5 and 6 were renumbered 20 and 21 (see facsimile); folios 13 to 16 were numbered 30 to 33, and the folio originally numbered 20 became 38.

In the printed text the end of each piece of paper is marked by a slant sign and the numbering and lettering of the new manuscript folios and slips is given followed by another slant sign, and thus the reader can recognize Darwin's additions where they amount to more than a line or two of writing. Where Darwin destroyed the continuity of the text by shearing off parts of folios, the location of the cut is signalled by double slant signs.

Besides cancellations and additions applying to the text as it was to be printed, Darwin wrote an occasional instruction to the


copyist or to the printer such as 'Lead', (ch. VI, fol. 28), 'Small Type - notes run into text', (ch. X, fol. 69), and 'Large type again', (ch. X, fol. 78). (See top of facsimile of folio 105, chapter VII.) These have been taken account of without special editorial comment.

On the manuscript Darwin sometimes scrawled pencil memoranda to himself such as: 'Get Huxley to read over for this.' (ch. VII, fol. 16). Part of these have since been rubbed out, but where they are intelligible they have been printed, usually as footnotes.

The Darwin papers also contain some reading notes and letters directly related to the manuscript. These have been printed in connection with the associated portions of the text. They have been cited according to the volume or item numbers in the Handlist of Darwin Papers or in Mrs Martineau's catalogue.

Finally, on the manuscript there are some signs of its later use. In the margin of folio 13 of chapter X, Darwin pencilled 'Used [ ] Man Book', and Lonsdale's anecdote about snails was published in 1871 on page 325 of volume one of The Descent of Man. Similarly, there are other jottings whose meaning is more or less obvious. On the verso of folio 136 of chapter IX, where one would expect the pinned-on slips with the reference previously cited, instead one finds that Darwin wrote in ink: 'Note used in Domestic Animals Chapter 15, Crossing.' Since note 9 of chapter 15 (Variation, II, p. 88) fits the context of the Natural Selection manuscript, this jotting is crystal clear. Elsewhere in chapters III, VI, VII, and IX we find the jottings 'all used', 'used' or an encircled U alone or with light vertical cancel lines down the page.1 (See the facsimile of folio 105 of chapter VII and of folio 21 of chapter IX.) Many of these jottings are easily connected with passages in Variation where Darwin used material from the Natural Selection manuscript. In chapter IV, 27 and elsewhere we find Francis Darwin's initials. With a few special exceptions, vertical cancellings and jottings of this sort have been ignored in my editing.


As Darwin described it 'My handwriting, I know, is dreadfully bad.'2 This often forces the reader to guess at words, and even his family had difficulty in reading it.3 In the fair copies made of

1 In regard to a similar use of vertical cancel lines, see Francis Darwin's introduction to the Foundations p. xxi.

2 ML no. 636.

3 H. E. Litchfield, ed., Emma Darwin, Wife of Charles Darwin. A Century of Family Letters. Cambridge, 1904. I, p. 436, and ML no. 2.


portions of this work, the copyist misread enough words which Darwin did not correct so that we must go back to Darwin's holograph for the basic text.

Darwin himself sometimes misread his own writing. For example, in chapter IX folio 19 he correctly quoted Herbert's 'in cases of natural impregnation', but later when he reviewed this passage he could not read the final 's' in 'cases' and so he added a 'the' above the line to make the phrase read 'in the case of natural impregnation'. In chapter X, folio 112 he quoted Kirby and Spence's 'utmost activity' then later misread his 'utmost' as 'almost' and added 'incessant' to restore meaning by saying 'almost incessant activity'.

Such handwriting makes it practically impossible to reproduce every detail of the text exactly as Darwin intended it to be printed. I have given the best reading I could, but only in the cases where my best interpretation seems to make doubtful sense or where several different words–often proper nouns–would fit the handwriting equally well and I have found no clues as to which Darwin probably meant, have I specifically warned the reader of a particular uncertainty in the manuscript by adding a question mark within square brackets.

The reader should be generally warned about certain specific difficulties in the handwriting. The following pairs of words are frequently indistinguishable: 'to' and 'the', 'when' and 'where', 'could' and 'would', 'than' and 'then', 'man' and 'men'. Considerable uncertainty often arises in the choice of alternative readings between 'that' and 'the', and, unfortunately, between 'probable' and 'possible'. As the last example suggests, Darwin's long 's' is a particularly obscure letter. The final 's' in plurals must usually be determined from the context. The pair of letters 'r' and 'o' are sometimes indistinguishable as in the case of 'grow' (ch. V, fol. 42 V). In an unusual proper name such as Gouan this can be troublesome. Of course other commonly indistinguishable letter pairs such as 'e' and 'i' also occur, so that a choice between possible proper names such as Marten's, Martens', Martin's, and Martins' must depend upon the context.

In some words, letters instead of being merely uncertain seem to be entirely missing, so that a reading such as 'Gret Britain' seems clear. Such omission seems specially frequent for letters before 'y': 'may' for 'many', 'thy' for 'they',' and 'vey' for 'very'. Particularly for the cases of proper names, students of Darwin manuscripts should remember the possibility that the

1 Cf. L & L I, 119; NY, I, 07.


correct form may be other than that which Darwin seems to have written; in chapter IV, folio 55, the reading seems to be Magillvray instead of Macgillivray, for example. When writing words such as Gaertner, he apparently intended to ligature the 'a' and the 'e', but the 'e' is usually undetectable so that the word seems to be written as Gartner. Such apparent lapses of the pen are rarely quite clear-cut, however, and the normal spelling has ordinarily been used in the text without special editorial comment.


A reading such as 'chesnut' might appear to represent either a lapse of the pen or an error in spelling, but it is one of a group of unexpected spellings including 'plaister', 'owzel', and 'Feroe Islands' for which Darwin had had reasonable precedents in his sources.1

He was inconsistent in spelling as is illustrated in chapter VII where on folio 37 he wrote both 'connection' and 'connexion' and where on folio 113 he clearly wrote 'organization' and on folio 118 'organisation'. He also made clear-cut errors such as 'thoroughily'.2 Where the handwriting is clear, the spelling of the manuscript is followed without any particular editorial comment. Where Darwin's spelling is uncertain, the normal English form is used.

Some quite clearly written words still puzzle me. In chapter V, 40, he mentions Inverorum, and this spelling also appears on page 295 of volume one of the Life and Letters, but I have not found exactly this place name in any of the numerous gazetteers I could consult. Perhaps Inveroran is the closest. In the same chapter on folio 29, the reading Colinsay seems quite certain. This exact place name I have not located either, and Colonsay does not seem to fit the context. Particularly in cases such as this, it seems best not to change Darwin's apparent spelling lest an essential clue to Darwin's meaning be discarded.


Just as about his spelling so about his punctuation Darwin's handwriting leaves many uncertainties. Clearly he often used colons where we would use semicolons. This suggests a system of

1 Or an unreasonable precedent. See the introduction to the Catalogue of Charles Darwin's Library (Cambridge, 1908), p. x, where Francis Darwin comments on his father's copying the spelling 'ciliae' from Robert Grant.

2 Ch. VII, fol. 117; cf. Nora Barlow's discussion of his spelling in her preface to Charles Darwin's Diary of the Voyage of H.M.S. 'Beagle'. (Cambridge, 1933), p. xix.


punctuation similar to that set forth in Lindley Murray's English Grammar which was so widely used that the book averaged an edition a year during the first half of the nineteenth century. In presenting this manuscript, Darwin's punctuation is retained in so far as it is clear. The many doubtful points such as distinctions between colons and semicolons have been interpreted to conform with present-day usage,

I have retained the parentheses Darwin used, but I have discarded his square brackets. Most often these simply set off material to be printed as a footnote. Occasionally Darwin used square brackets to mark the beginning or the end of a paragraph. This is most clear on folio 78 V of chapter VIII and on folio 64 of chapter X, where he added an ordinary paragraph sign as well. (See also facsimile of folio 94 of chapter IX.) In Darwin's text as here printed and in the related footnotes, square brackets will have the customary function of indicating material added by the editor.

Similarly in regard to a more frequent use of capital letters, Darwin's practice seems to have been different from ours; but, here again the handwriting often leaves his intentions uncertain. Such uncertainties have been resolved in favour of present-day practice. Where Darwin in revising changed the beginning or ending of sentences without completing a corresponding change in capitalization, I have changed this without special comment, e.g. chapter X folio 187 V.

I have also silently expanded Darwin's contractions for words such as should, island, and reverend, and have spelled out 'Natural Selection' where Darwin used the abbreviation 'Nat. Sel.' in his pencilled addition to folio 51 in chapter VI. I have omitted words which were unintentionally written twice.

The preceding editorial discussion applies to the work as a whole. Comments about points concerning single chapters and their history will appear in the separate introduction immediately preceding the individual chapters in a form similar to the following comment on Darwin's own tabulation of the contents of his manuscript.


The following extensive table of contents, which Darwin himself wrote out, merits some special consideration for what it tells us about the detailed history of the Natural Selection manuscript. First of all it supplies a full outline for the two initial chapters, now missing except for one single stray survivor, folio 40 from chapter


one. Presumably Darwin's manuscript for these chapters was incorporated and used up in the course of writing the two later volumes on The Variations of Animals and Plants under Domestication. The first folio of the manuscript for the Table of Contents has been cancelled by a single diagonal pencil line in the same manner as Darwin employed to mark passages farther on in the manuscript which he had used in later publications. The contents for these two missing chapters can be compared with those for the two volumes on Variation.

The fact that the very first entry for the table is for folio 16 of the first chapter raises the question, what about the preceding fifteen folios? I believe these formed Darwin's preface, which we know that he wrote because he referred to it in the postscript of a letter to Baden Powell on January 18, 1860: 'I have just bethought me of a Preface which I wrote to my larger work, before I broke down & was persuaded to write the now published abstract. In this Preface I find the following passage, which on my honour I had completely forgotten as if I had never written it. 'The "Philosophy of Creation" has lately been treated in an admirable manner by the Rev. Baden Powell in his Essay &c 1855. Nothing can be more striking than the manner in which he shows that the introduction of new species is a "regular not a casual phenomenon" Or as Sir John Herschel expresses it "a natural in contradistinction to a miraculous process"''. 1 To my particular regret I have as yet been unable to find any further trace of this Preface in the surviving Darwin manuscripts.

Darwin seems to have written his table of contents chapter by chapter and not very long after each part or chapter was finished. In the case of chapter VI, for example, Darwin completed his original table at the end of folio 4 of his table of contents section and continued straight on with the contents for chapter VII on folio 5 of his table. Then a year later in the spring of 1858 he returned to chapter VI to revise and expand it, particularly by interpolating a new discussion of divergence some forty folios long. Thereupon he cancelled the original table of contents for the latter part of the chapter, which had been notably transformed, and wrote out for this second draft a new table on a folio which he had to number '4 bis' to fit it into its place in the table as a whole. Here in my edition, in the following table only the new version is included; the older cancelled portion is to be found later, in my introduction for chapter VI, which immediately

1 Gavin de Beer, ed., 'Some unpublished Letters of Charles Darwin', Roy. Soc. London, Notes and Records, 14 (1959), 54.


precedes Darwin's text and discusses the specific history of that chapter. Similarly for chapter IV, about a year after he finished his original draft he wrote a long additional section. For this he wrote out a table of contents on a folio he had to number '3 bis' to fit it into its proper place. As is evident in the case of chapter IX and therefore probably for the other chapters he did wait until he had finished the chapter before he wrote out the table of contents for it. In the case of this chapter after finishing his original draft he revised it drastically. The table of contents for this chapter starts on folio 6 of his manuscript table immediately after the end of the table for chapter VIII, yet it fits the revised form of chapter IX and has no cancelled references relating to the earlier draft, so that it could only have been written out after the second version of chapter IX had been completed. In the case of the two portions of the manuscript of which Darwin had fair copies made, namely the addition to chapter IV and the discussion of geographical distribution which I have called chapter XI, he waited at least until the copyist had finished because the folio reference numbers in his tables fit only the fair copies and not the drafts he himself wrote out.

Most exceptionally, considering the Natural Selection manuscript as a whole, watermarks in five out of the eleven foolscap sheets used for the table of contents supply dates; these are compatible with the previous assumptions about the different times of writing of the different parts of the table of contents. The sheet for chapter XI, written first of all, bears the date 1856, as do folios 1, 3, and 5 of the table. Folio 3 bis giving the table for the addition written in 1858 for chapter IV bears the watermark date 1858. The complete manuscript for the table now consists of folios 1 to 3, 3 bis, 4, 4 bis, and 5 to 8. These are mounted at the very beginning of volume 8 of the Darwin Papers. The unnumbered folio with the contents of chapter XI is now item 58 in volume 72.

On the manuscript Darwin pencilled certain notes as memoranda or agenda which should be recorded for consideration. After the entry for folio 3 of chapter II, he scrawled: 'N.B. Why not animals domesticated We do not want them Goats & Asses.' He bracketed the next two entries, for 'The Cabbage' and 'Dog' adding: 'Sports in Plants? isolation' [?] On the verso of folio 1 opposite the last lines of the table for chapter II on folio 2 he wrote: 'I think here all naturalised animals—P. Santo Rabbits—Mice in different countries—Naturalisation Cardoon—Naturalised plants in N. America.' At the beginning of chapter IV: 'Antiquity of variation under nature of shells in Madeira must be included in this chapter',


and following the entry 'conspicuous and useful plants not cultivated' for folio 44 of chapter IV, Darwin scribbled: '??Feral animals & plants not domesticated.' For chapter VII, Darwin seems to have cancelled the entry for folio 41 on Brullé's law, presumably after he had received from Huxley the adverse comment on Brulle which is given in the introduction to chapter VII.

For Darwin's table of contents, there has been no attempt to reproduce the apparently insignificant exact details of his holograph. His abbreviations, such as 'Var. under domest.' for the title of chapter I, have been expanded without the use of square brackets. Similarly for folio 25 of chapter I, Darwin's 'avitism' has been corrected to 'atavism'. The numbers in the table, which precede each topic, are folio references to Darwin's holograph manuscript, largely as Darwin himself supplied them. Thus any topic indicated in the table of contents can be located in the printed text by referring to the folio numbers given between slant signs in the text. For the addition to chapter IV and for chapter XI Darwin's reference numbers correlated with the sometimes inaccurate contemporary fair copy, and in these cases I have replaced them by new numbers (not signalled by square brackets) which correlated with the folio numbers of the holograph manuscripts used as the basis for the published text.

[page] 25



16 What is domestication—changed conditions 19 Anyhow some slight variation must be admitted 20 Changes in the individual 21 Congenital variation 22 Hereditariness—why one thing in herited & not another—Mutilations, effect a need, not hereditary 25 Atavism 27 Peculiarities attached to sex 27 Variation ap pearing at same age 29 Causes of variation—Immediate causes 31 Habit 32 Variation from simple organisation 34 Indirect effects, making organs plastic 35 Laws regulating variation—balancement—nisus formativus 37 Effect of homologies 38 Mechanical relations—arrests 39 Cohesions 40 Multiple organs varying in number 41 Variations analogous to other species & to other varieties 41 Correlation of growth & constitution—obscure relations, as colour & constitution 43 Effects of crossing in obliterating & forming races; selection required 47 Selection, produces effect of adding up small changes 50 Breeds are true 51 Antiquity of selection 54 Unconscious selection 55 Consti tutional or natural selection 56 Acclimatisation of plants 58 Circumstances favourable & unfavourable to selection; non-varying 62 Facility in preventing crosses 64 Effects of selection on natural results 67 Effects of selection as shown in flowers & fruit compared.


3 General argument whether plants so much altered by domestication as not to be recognised; steps in cultivation by savages—why certain countries have not produced useful products 10 The Cabbage 13 Dog 18 Changes within historical times 20 Cat 22 Horse 26 Pig 30 Cattle 38 Sheep & goats 40 Rabbit 43 Fowl 46 Ducks 50 Pigeons 52 Pouter 54 Fantail 56 Jacobin 57 Tumbler 59 Turbit 60 Barb 61 Carrier 62 Runts 65 Dovecots &c 68 Imperfectly known breeds 69 Amount of important differences 70 Individual variability /2/74 History of changes of principal breeds 76 Discussion on origin 77 How many Rock Pigeons <81 Columba livia parent race> 83 Not feral: fertility 85 Reversion to blue breeds 90 C. livia parent race,


reasons for 92 Circumstances: antiquity &c, favourable to most variation, in Pigeons 94 Probable steps in the variation 96 Probable sports in Pigeons 103 Summary; varieties like species: man can not select internal changes, only conspicuous 107 Summary that multiple origins not general; for breeds abnormal, & no reason why such selected & such extinct 109 Argument from peculiar races, when no peculiar endemic species 110 Argument from general fertility of mongrels 111 Antiquity of animals & plants 114 Summary on ditto 118 Summary on breeds resumed; differ in variability 120 Objections, from assumed force nature 123 Especially plastic 124 Variation limited 126 Summary on objections & close of chapter


1 Ill‐effects of interbreeding 3 Opposed facts: in islands 5 Good from crossing in plants 7 Even in hybrids 11 Good from slight changed conditions 13 Opposed facts 14 Good in animals 16 No land animal real hermaphrodite 18 Flowers are fertilised sub jove 19 Hermaphrodites & unisexual in same groups 20 All animals can cross. 26 All plants can cross 26 bis Dichogamy 27 Varieties of plants do cross freely 28 Great care taken by hybridisers to prevent crosses 31 On pollen carried by wind & insects 32 Insects necessary to fertilisation How bees find flowers 37 Bees frequent visits 39 Pollen brought to Thyme & Rhododendrons 41 Opposed facts to crossing of plants; fertilisation in buds, Campanula &c 44 In wheat, varieties not crossing 46 Trees with separated sexes 47 bis Subularia, Podostemon 48 Goodenia 49 Aristolochia & Arum /3/ 50 Hollyocks 52 Orchidaceae 56 Ophrys 58 Leguminosae 62 Opposed facts 64 Summary on probability of crossing occasional in animals & plants 66 Superfluity of pollen 68 Crossed offspring better chance of surviving 70 Final cause of crossing 72 Changes of condition causing lessened fertility 74 In animals 78 In birds 83 Effects on secondary male characters & instincts 84 Effects in domesticated animals: not explained by disease &c 86 In plants 89 Can stand great changes in temperature & domestic plants 90 Contabescence 93 Double flowers, great fruit & tubers 95 Parallelism in hybrids 97 Domestication & cultivation itself not stop fertility: Wild flowers not seeding in own country 100 Conclusion 101 Bears on variability & on early period of cause of.



1 Definition of species 6 Different kinds of variation 8 Subspecies—gradation from species to individual differences 10 Impossible to tell how far variations inherited 12 Naturalists practically are guided by having intermediate forms 14 Naturalists' forms between good species 15 Watson's categories 16 Individual differences 17 Graba's 18 Are they of the same class as differences constituting recorded varieties? 21 Genera variable in one country, variable everywhere? 24 Cases of individual variability 35 Difficulty in deciding what are species not due to want of knowledge: British plants: numbers estimated 37 Brehm 38 Difficulty increases when specimens brought from many districts 39 Difficulty a climax when specimens taken from separated regions: but what distance is enough to make 2 regions distinct 41 Conspicuous animals & plants generally are perceived to vary 44 Conspicuous & useful plants not cultivated 48 Marked varieties, Elephant 50 Foxes 53 Raven, Carrion & Hooded Crow 56 Tetrao Scoticus 59 Guillemots—Madeira Black'cap 60 Carp 61 Wollaston's insects 63 Plants: Centaurea, Taraxicum, Isatis 64 Festuca, Polygonum 65 Myosotis Azores Tolpis 66 Digitalis, Iris 67 Anagallis 68 Primula 80 Conclusion: my definition of "species".—/


1 Wide ranging and common and much diffused species tend most to vary. 8 D'Archiac's rule not really opposed. 10 Geographical range of varieties themselves 14 Relation of common species to size of genera Average range of all the species in the larger genera not greater than in the smaller genera. Highly organised beings have acquired narrow ranges. 20 Species having recorded varieties are more frequent in larger than in smaller genera. 23 The species in large genera are themselves closely allied. 30 Reason why Prodromus not good. 35 Why I disregard other difficulties. 41 Summary on discussion./


2'5 Mere variability will not account for wonderful structures 9 Struggle for existence 10 Rate of increase 12 In feral animals & wild under good circumstances 15 Cannot be explained by altered fertility 17 All animals breed 18 Checks to increase in animals 24 Odd checks 25 In birds 29 Colonies destroyed 30


Definition of term Struggle for Existence 31 Facts in distribution explained by struggle in animals 33 Complicated 34 Rate of decrease, compensation 37 Checks, Animals & Plants 41 Checks Plants & Plants 44 Not limited absolutely by temperature 47 Herbert on Struggle 48 Maer Heath case 50 Rotation, influence of 53 Slight causes determining victory 55 Struggle severest between individuals of same species 55e Facts opposed to struggle: local species common: Madeira Helix cases; explained 58 Plants social on limits of range 59 Explained by mutual protection 61 Fertilising each other 63 Monstrosities in nature, appearing to show no struggle 65 Numbers not dependent on number of eggs or seed 66 Summary brief 70 Corollary, that structure is related most intimately in organism to organism./


4 Very slight differences can be selected 7 Selection may act on larvae or sexes 7b Sexual selection 7k In plants 8 Effect of change of climate on continent 11 Places open shown by naturalised productions 12 Imaginary illustrations: Catskill wolves—flowers & bees—dioicous 19 Comparison of man's and nature's selection /4 bis/ 26 Extinction 26c Principle of Divergence: in domestic animals (d) Amount of life due to diversity shown by culture & by natural distribution (h) Shown by naturalisation (1) Shown by physiological division of labour (p) Divergence acting in nature on large genera (s) Diagram of (w) Varieties how supplant parents (aa) Varietal differences become specific (cc) Bears on classification (gg) Limit to total number of species in any country (mm) Classification compared to Tree of Life 28 Crowd of difficulties 33 Causes favourable & unfavourable to selection 35 Isolation 38 Intercrossing 41 Illustration of isolation & intercrossing in Madeira & Galapagos 43 Varieties keeping separate 46 In plants 49 Large number of individuals favourable for selection 51 Slowness of selection 53 Absence of intermediate links; nature of such links 54 Links chiefly extinct 57 Links generally absent now when species mingle in border regions 58 Areas perhaps not formerly continuous 60 Conditions of continuous areas do not really graduate insensibly 62 Intermediate varieties rare 64 Intermediate varieties tend to be easily exterminated 68 Summary—illustrated by Malay Archipelago./



4 Immediate action of external causes: heat, brackish water, depth, proximity to sea 8 Analogous variations under different conditions 10 Acclimatisation of species 13 Species naturally extending range 19 Effects of use & disuse—flight of birds 22 Insects 26 Tarsi 28 Blindness in moles 32 In fishes 34 Correlation of growth—multiple parts varying 35 Homologous parts varying in like manner—& to unite—teeth & hair—seeds of Umbelliferae & Compositae 39 Early monstrosities & change must affect subsequent parts <41 Brullé's law, that most metamorphosed parts are developed first. Bears on M. Edwards> 45 Balancement 52 Extraordinarily developed parts are variable 59 Cases 70 Summary on: explanation of 77 Rudimentary parts variable 79 Monstrosities: arrests of development—do not produce new species 86 Distinct species present analogous varieties: varieties of one resemble normal character of other species. Reversions 90 Analogous varieties in nature 95 Cases of varieties mocking species 105 Horse case 114 Distinctive characters of varieties more variable than specific, & specific characters more variable than generic: sexual & specific differences allied. /


2. How could bat be formed? Birds' gradation in flight 9 Diversified habits in same species, & different species of genus 12 Change of habits in introduced organisms, & in nature in different regions 17 Organisms with intermediate habits likely to be rare 19 Habits not conforming to structure: Upland Geese 22 Does not appear like creation 25 Could so perfect organ as eye be formed? 33 Must admit any nerve could be made acoustic or optic 35 Do quite new organs appear? "Natura non facit saltum". Kinds of transitions 37 Changes in function 39 Two organs with same function 41 Same organ with two functions 42 Functions changing with age or condition, or always in two states 45 Cases of difficult transitions, generally only apparent: poison glands of snakes 50 Separation of sexes 53 Chemical compounds definite 55 Neuter insects 58 How sterile 61 How acquire their structure 76 Summary on transitions—natura non facit saltum explicable on our theory 77 Similar organs in remote animals, as electric organs in fish 81 Could organs of little importance be formed by


selection? Causes of error, attribute to selection what not due to it 83 Fly-flapper—eye-lashes; how selection may act on them 85 When more important to ancestors, or concurrently with other advantages /6/86 Some have protested against utilitarian view of all organs 88 Selection will produce nothing injurious 89 Organisms are not perfect 93 No structure can be modified for good of other species 97 Result of whole accords with Cuvier's principle of conditions of existence 97 Difficulty from supposing bird, fish, & mammal have had common ancestor 99 Summary


4. Final cause of sterility of hybrids & first crosses: distinction between which 9 Plants: species when first crossed generally sterile; Gaertner's method 10 Ill-effects of processes 10 Sterility of hybrids, only in function: sterility during reduction 14 When self-fertilised 16 Sterility in successive generations partly due to interbreeding, evidence of 18 W. Herbert's cases of fertile hybrids & first crosses 21 Even excess of fertility; & Gaertner's cases of ditto 24 Florists' crosses very fertile 27 Conclusion is that sterility not universal attribute of species 28 Difficulty of distinguishing species & varieties of plants by fertility, compared with other evidence.—Cases 36 Infertility of varieties (plants) when crossed 42 Explanation & difficulties in getting evidence of this 44 Laws & circumstances governing infertility: gradations of sterility 46 Affected by conditions & innately variable 48 Related to systematic affinity 50 Not governed by external form or constitution 51 Reciprocal crosses 55 No absolute relation between fertility of first crosses & hybrids 57 Special odd cases; mother's pollen gives fertility quickest: male organs fail first 58 Exceptional & decided types sterile 60 Prepotency & fertility do not go together 64 Conclusion that sterility is an incidental quality like 65 Grafting 70 Causes of Sterility, in first crosses, from distinct causes 72 In hybrids compared with that from changed conditions 78 Animals: instinct comes into play; aversion rare 80 Fertility of first crosses & of hybrids compared: reciprocity doubtful 83 Male sex fails easiest, gradation in sterility 83 Affected much by all conditions 85 Not governed by systematic affinity 88 Table of Rasores 89 Cases of the most fertile hybrid animals 89 Summary: reason for expecting much sterility in animals, from confinement & interbreeding 99 Domestication increasing fertility of first crosses & of hybrids 101 Fertility of crossed varieties & of their mongrels 104 Difficulties


in subject 105 In pigeons 107 Explanation: owing to difference in natural & artificial selection /7/ 110 Hybrids & mongrels compared, independently of fertility.—Plants: hybrids are not monsters 112 Gaertner's difference of mongrels more variable in first generation 115 Variability of hybrids in successive generations 116 Neither hybrids or mongrels universally variable 117 Reversions 118 Alike from reciprocal crosses. Rules of resemblance 119 Prepotency 120 Reduction—Old varieties not prepotent over new 121 Two close species crossed with third, compared with two varieties ditto 123 Animals: difficulties from complex laws of inheritance 124 I. St. Hilaire's law, hybrids fixed, mongrels heterogeneous & like either parent 126 Hybrids heterogeneous 127 Hybrids heterogeneous in successive generations 127 Mongrels usually intermediate, exceptions 128 Mongrels sometimes, though rarely, have fixed character 129 Laws of resemblance, part after father false, but same for species & varieties 130 Characters, as colours, suddenly appearing are transmitted either whole or not at all 132 Prepotency of species over species & race over race 133 Trumpeters 134 In individuals 134 bis Prepotency of old race over new, doubtful 135 Reduction in birth, rate of 136 Quagga case 137 Summary of rules for Animals & Plants, for hybrids, mongrels, & individuals 138 Summary for whole chapter.


3 Characteristics of 7 None universal 10 Intellect of animals 14 Degree to which end of instinct is recognised 15 Performed by the young 18 Imitation 22 Comparable with habits 31 Tricks & habits inherited by domestic animals 37 Not strictly acquired by habit but by selection of self-originating peculiarities & (f. 41) by training 46 Lost 49 Instincts of animals in a state of nature 50 Migration 58 Instinctive fear 62 Feigning death 64 Nidification 69 Variations of 78 Habitations of animals 80 Parasitism 82 Rhea 84 Hymenoptera 86 Social insects 88 Slave makers 92 Bees' combs 98 Miscellaneous remarks— variation under different circumstances or times of life /8/ 100 Occasional odd habits 102 Reversions 102 Special difficulty, same instinct in remote classes 103 Not for good of others 104 Instinct performed only once 105 Apparently no gradation 106 Trifling instincts: mere tricks 108 imperfect instincts 113 Conclusion. /



vol. 72, fol. 58 1. Alpine productions of Europe: Prof. E. Forbes' explanation 5 Of North America 7 Northern range of the organisms in common to Europe & N. America 9 Passage by arctic land & shore, under former warmer climate, before Glacial Epoch 11 Colonisation of arctic islands; species all identical 12 Alps with more representative species—Representative species of Europe, N. America & Japan or E. Asia 16 Extension of Forbes' theory to whole world: evidence of former "cold period" 23 Effects of "cold period" on migration of plants & on representative species 24 Applied to America 28 To Africa 29 To India 31 Australia 32 v New Zealand 33 Effects of "cold period"on migration of animals, birds, insects 34 Fish—marine mollusca 36 Crustacea 38 Sea-weeds 38 Difficulties on "cold period" migration-theory; viz. rapidity of travelling; non-modification of whole body of tropical productions; may erroneously attribute too many cases to modification 42 More migration from north to south than in reverse 44 Specially difficult case of supposed migration, viz. in Antarctic islands Kerguelen Land 45 Tristan d'Acunha 46 Probable transport by icebergs 49 New Zealand very difficult case 53 Summary on migration during cold period.

[page] 33




On October 3, 1856, Darwin wrote to his second cousin, W. D. Fox, that 'I… am now drawing up my work as perfect as my materials of nineteen years' collecting suffice, but do not intend to stop to perfect any line of investigation beyond current work… I find to my sorrow it will run to quite a big book.'1 Ten days later, according to his Pocket Diary, he finished his second chapter, and presumably he then proceeded to this chapter three and wrote rather fluently, for the manuscript is less laboured than for many of the later chapters. After several mentions of the theme of crossing in letters to Hooker2 he wrote to him on December 10. 'It is a most tiresome drawback to my satisfaction in writing, that though I leave out a good deal & try to condense, every chapter runs to such an inordinate length: my present chapter on the causes of fertility & sterility & on natural crossing has actually run out to 100 pages M.S., & yet I do not think I have put in anything superfluous.—'3 The completion date for this third chapter was December 16, 1856, according to the Pocket Diary.

Although for this chapter Darwin made very few revisions, one involving terminology is worth comment. On folio 20 of this chapter Darwin states: 'All the vertebrata are bisexual', here clearly meaning not hermaphrodite but having two separate and distinct sexes. This same usage occurs in the 1842 Sketch4 where he wrote 'All bisexual animals must cross, hermaphrodite plants do cross, it seems very possible that hermaphrodite animals do cross.' Similarly in the Monograph on the Fossil Lepadidae, published in 1851, he wrote: 'Ibla cumingii…is bisexual; one or two males being parasitic near the bottom of the sack of the female…hence Ibla cumingii is exactly analogous to Scalpellum ornatum. On the other hand, the closely allied Australian Ibla Cuvierii, like Scalpellum vulgare, is hermaphrodite… 5 In print later in this same year of 1851 he changed his usage completely around in his monograph

1 L & L, II, 84; NY, I, 442.

2 Letters of Nov. 15, ML no. 334; Deo. 1, C.D. MSS. vol. 114, no. 185.

3 C.D. MSS., vol. 114, no. 186; ML no. 337.

4 Foundations, p. 2‐3, and also in 1844 Essay, p. 92.

5 (London, 1851), p. 16, n. 1. Regarding this work Darwin wrote to Hooker in June, 1849: 'I am going to press….' L & L, II, 37; NY, 1, 397.



on recent Lepadidae to equate bisexual with hermaphrodite when he stated that: 'Ibla, though externally very different in appearance from Scalpellum, is more nearly related to that genus than to any other; in both genera some species have the sexes separate, the imperfect males being parasitic on the female, and other species are bisexual or hermaphrodite,' (p. 182). Discussing the affinities of the species which I. E. Gray had named Ibla Cuvierana and which Darwin named Ibla quadrivalvis, he wrote: 'Considering these so slight differences, it is highly remarkable that this species should be hermaphrodite, whilst I. cumingii is unisexual' (p. 207), and farther on in another general discussion of a genus stated that 'Scalpellum ornatum and perhaps S. rutilum, are unisexual; the other species are hermaphrodite', (p. 221). This same complete reversal of intended denotation regarding the term 'bisexual' appears in the manuscript of this chapter. Darwin clearly made the change well before he had finished the original version although he had some difficulty in keeping this change of usage in mind. Evidence of the standard usage occurs as early as folio 21, but even on folio 68 he started to write 'bisexual' to denote the opposite of hermaphrodite but caught himself before he had finished the word, drew a line through his error, and continued writing on the same line to make the phrase now read: 'in closely allied groups of hermaphrodite & <bisex> unisexual plants.' Study of folios 21 to 26 reveals a confusing oscillation of usage in Darwin's original drafting of this portion of the manuscript. (Since my editorial system does not signal interlinear interpolations, I should assure the critical reader that, for these folios, the cancellations in every case occur only on the original lines of writing, not in interpolations added between the original lines at some later but uncertain time.) In revising these passages Darwin cancelled and changed all but the first occurrence of his earlier usage of 'bisexual'; and he often tried to make his terminology clearer by substituting 'hermaphrodite' for 'bisexual'.


Evidences of Darwin's later uses of the text and material of this chapter abound in the manuscript. Dr Alice Guimond, while working as my research assistant, discovered direct quotations and close paraphrases from the manuscript in the published text of Darwin's Variation of Animals and Plants under Domestication (London, 1868), and also that use of information to be found in the manuscript was published in his Effects of Cross and Self-Fertilisation in the Vegetable Kingdom (London, 1876). At the top of folio 72 Darwin wrote: used to p. 102', that is to the end of the manuscript for this chapter, and in his Variation under Domestication volume II, pp. 148–72, starting with the section headed 'Sterility from changed Conditions of Life', the use is clearly evident. This folio 72 was also cancelled with a vertical line to indicate use and so were the following sheets to the top of folio 97. The lower part of folio 97 was marked with an encircled 'U' as were folios 98–102. One or both of these symbols, or occasionally a criss-cross cancel, also makes folios 1 to 44, 27, 37 to 40, 52 to 54, and 56 and 57 as used.

More drastic signs of use also occur. A quotation from William Herbert's Amaryllidaceae at the foot of folio 5 has been sheared off completely and is now missing, although the end of the quotation appears at the top of the next sheet of the manuscript. The missing text can be restored from p. 127 of volume II of Variation under Domestication which gives the full quotation.


Similarly the bottom fourth of folio 74 and the top fourth of folio 75 have been removed. The remaining manuscript text for the second paragraph of folio 74 is closely paralleled by the opening wording of the paragraph beginning towards the top of page 150 of volume II of Variation. After the gap the manuscript proceeds: 'Four wild species of the Horse genus have been bred in Europe', which the printed text in this same paragraph repeats as 'Four wild species of the horse genus have bred in Europe', and then paraphrases the rest of the manuscript remaining for this folio. It seems almost certain that while writing the manuscript for Variation under Domestication Darwin cut off these parts of the older Natural Selection manuscript and simply pinned or pasted them on to his new manuscript in the same way that certain existing sheets of the Natural Selection manuscript are pieced together from passages selected and sheared from earlier drafts (e.g ch. 9, fol. 84). The full list of cut off folios for this chapter is: 5, 6, 9, 13, 36, 39, 74, 75, 78 and 79. Similarly certain note slips are now missing from the Natural Selection manuscript, and were probably transferred to the later manuscript.

Much, both of the short passages of missing text and the missing notes can be restored with confidence from the first edition of Variation under Domestication.


[Completed Dec. 16, 1856]

I/The subject of the present chapter is related to some points discussed in the previous chapters as to breeds being kept constant by the blending of slight & individual differences & to several questions which follow, & may therefore be as well intercalated here as elsewhere.1

On the ill effects of close breeding in & in. That evil arises from this process carried to an extreme has been a general opinion in various countries & times, is universally known.2 That general beliefs of

1 [As Dr Alice Guimond, my research assistant, discovered, Darwin used material from this chapter in his Variation of Animals and Plants Under Domestication (London, 1868) and in his Cross and Self-Fertilisation (London, 1876). The first fourteen MS. folios he cancelled with a vertical line, marked with a large 'U' or both, presumably to indicate he had used the material written there. Moreover, he cut up his manuscript, and apparently removed some reference slips which I have not found, so that to restore the gist of these missing parts of the MS. I have borrowed from his published text on the same topics in the Variation.]

2 Sir G. Grey in his most interesting Journal of Expeditions into Australia Vol 2, p. 243 says that anything approaching to the crime of incest is held in abhorrence


this nature have often no foundation is very true; but in this case it may perhaps [be] more readily trusted as the breeder is often most unwilling to act on his beliefs, as it must seriously interfere with his process of continued selection of some peculiarity in his own stock./lv/Independently of the undoubted evil of matching animals having the same infirmity, which must always tend to be the case when relations unite/1/the general belief seems to be that decrease in size, & of general vigour is the first result of close interbreeding, & then lessened fertility.

I have never met a Pigeon Fancier who did not believe in the evil of close interbreeding; & he has the best/2/opportunity of judging, from pigeons being paired for life, & many generations raised in a short period: when size is an object as in the Pouter, it is asserted1 that the ill effects are very soon perceived, not so when small birds are wanted as in the Almond Tumbler; but in such cases many of the birds become shy breeders.—The high price of many fancy dogs, which have long been closely selected & interbred, I have been assured is, due more to the difficulty in getting them to breed freely, than in their throwing inferior animals; I have known the female requiring to be held, exactly as in the production of some Hybrids2 & indeed if no such difficulty existed the high price of such dogs would be quite inexplicable. The particulars have been given me of one gentleman who long had kept a small family of blood-hounds, & from being very unwilling to cross his breed, he almost lost them, so infertile had they become, until he was obliged to resort to a cross when his breed became fertile./

2 bis/The evidence of an acute observer like Sir John Sebright, who bred all sorts of animals during his whole life, & who boasted that he could produce any feather in [three] years & any form in [six] years;3 & who always worked by crossing & thereby closely interbreeding, is very good; & he was a most firm believer4 in the ill effects of this process carried on too long. I was assured by

by the Australians. So it is with the aborigines of N. America, & Dobrizhoffer makes the same in regard to the Abipones of S. America, [r, 71] It is singular that this feeling does not appear to have been felt by the Kingly class of the Polynesians; but Ellis [Tour through Hawaii (London, 1826), pp. 414–15.] does not doubt that evil followed from their incestuous marriages. Prescott [See William H. Prescott, Conquest of Peru, book 1, ch. 3, passage associated with note 47.]—Ohio [cf. Variation, ch. 17, n. 21.] Chinese of same name [?].

1 A Treatise on Fancy Pigeons by J. M. Eaton, p. 56 [citation clearly to 1st. ed.]

2 Hunter's Animal Economy in regard to a she wolf too [1837 ed., p. 323.]

3 [Darwin left blank spaces for these two numbers. They are attributed to Sebright a without source reference in Eaton's Fancy Pigeons, 1852, p. iv.]

4 [The Art of Improving the Breeds of Domestic Animals (London, 1809), pp. 8, 11–12.]


Mr Yarrell that Sir John had for so long interbred his Owl-Pigeons, that he nearly lost his whole stock by their extreme infertility: I have seen some silver Bantams, bred from Sir John Sebrights', which were nearly as sterile as Hybrids for they had laid in that season two full nests of eggs, "not one of which produced a <single>-chicken. The cock, also, seemed to have lost its secondary male characters, for it had not saddle-hackles, & was scarcely more brilliant <plumage> than the hen./

3/On the other hand some competent judges have doubted the ill effects of interbreeding. The case of Bakewells cattle has often been quoted, & it shows that a man with a large flock may continue the process for a considerable time; but Youatt1 speaking of the subsequent deterioration of this breed says 'it had acquired a delicacy of constitution inconsistent with common management' & 'many of them had been bred to that degree of refinement that the propagation of the species was not always certain.'—In most of the cases of closely selected cattle & sheep there has been much mystery, & crosses have been suspected. The English Race horse & Mr. Meynell's hounds2 have also been advanced as instances of pretty close interbreeding without any ill effect. In these cases it may be suspected from what we shall presently see, that individuals being taken to different parts of the country & differently treated, & then occasionally brought together & matched, would lessen the ill effects of interbreeding. Again the case of the half-wild cattle in Chillingham/4/which have gone on interbreeding for the last 400 or 500 years3 seems a strong case; but Lord Tankerville, the owner, expressly states that 'they are bad breeders'.4 Those in the Duke of Hamilton's Park, are believed to have degenerated in size; I am informed by Mr. D. Gairdner that the stock kept, in the park of 200 acres, varies from 65 to 80, & that only about 8 or 10 are yearly killed, which seems to show no great fertility.

In the closely analogous case of Fallow Deer in parks, I find that the owners go to the trouble of occasionally obtaining bucks from other parks to cross the breed. In the case of the aurochs of Lithuania, which have a much wider range than the cattle of the British park, some authors believe that they have become considerably reduced in size. So it certainly is5 with the Red Deer of Scotland; but in the latter & indeed in the other cases it seems

1 Cattle p. 199.

2 Karkeek, Veterinary Journal Vol. 4. p. 4 & Mr. Appleby [actually Apperley] in Encyclopedia of Rural Sports p. 280

3 Culley on Live Stock [Introd. pp. x–xi.]

4 British Association Zoolog Sect. 1888

5 Scrope [Deerstalking, cf. pp. 10–11, 170 of 1839 ed.]


impossible to decide how much of the decrease of size to attribute to less varied food, & in the case of the Red Deer to sportsmen having picked out for many generations the finest Bucks; the less fine having been thus allowed to propagate their kind./

5/Good effects of crossing. However difficult it may be to obtain quite satisfactory evidence of the ill effects of close interbreeding, the converse of the proposition, namely that good arises as far as increased size, vigour & fertility comes from crossing distinct families & breeds, I think admits of no doubt. I have never met any breeder of animals who doubted it; & it seems useless to adduce authorities or facts. But in regard to plants [,] as varieties have been much more seldom crossed than in animals, I will go into some details to show that the same rule holds with them.

Gaertner, whose accuracy & caution seem most trustworthy, believes in the good effect of taking the pollen from another individual of the same species; he states1 that he observed this many times, especially in exotic genera, as in Passiflora, Lobelia or Fuchsia.

<Herbert2 says>//['I am inclined to think that I have derived advantage from impregnating the flower from which I wished to obtain seed with pollen from another individual of the]/6/same variety, or at least from another flower, rather than with its own.'

In these cases we have referred to crossing individuals of the same variety; we now come to crosses of distinct varieties.//

6A/Andrew Knight3 found that the offspring of crossed varieties of Peas were remarkably tall & vigorous; & that crossed wheat resisted blight better than the pure kinds./

7/We have seen in crossing varieties, that the offspring gains in size, vigour & fertility; in crossing distinct species it would appear that size & vigour is gained in an equal or apparently greater degree, but fertility is greatly impaired or very often wholly lost. Every traveller has been struck with the vigour & health of the common mule & this holds good with the hybrid Yak in the Himalayas; in the almost quite sterile hybrid from the fowl & pheasant, marked increase of size has been often noticed. In plants every single experimentiser Kolreuter, Gaertner, Sageret, Lecoq Herbert &c have been struck with the wonderful height,

1 Beitrage zur Kenntniss der Befruchtung 1844, S. 366

2 Amaryllidaceae p. 371 [The bottom of MS. fol. 5 is sheared off; the missing part of the quotation ending at the top of fol. 6 is supplied from Variation, II, 127, ch. 17 n. 41.]

3 Philosoph. Transact. 1799 p. 200 [At end of this sentence Darwin pencilled a memo for addendum: 'Loudon's Gardener's Magazine for grapes & other cases.']


size vigour, tenacity of life, precocity number of flowers, power of resisting cold &c of most of their Hybrid productions. Kolreuter1 is astonished at the portentous size of some of his hybrids & gives numerous precise measurements in comparison with both parents. Gaertner2 sums up his conviction on this subject in the strongest manner. Kolreuter attributed these facts to the sterility of hybrids, owing, I presume, a sort of compensation, in the same manner that capons, emasculated cats, some breeds of oxen are larger than unmutilated males. But Gaertner (p. 394 & 526) has shown that there is much difficulty/8/in admitting this explanation to its full extent; for there is no parallelism between the degree of sterility & the increase of size or luxuriance of growth; indeed the most striking cases have been observed in not very sterile hybrids. It deserves notice that the mass [?]<luxuriance> & enormous size of the roots in a crossed Mirabilis of unusual fertility for a hybrid3 was found to be inherited. It seems probable that the result is due both to nutriment which ought to have gone to the sexual function being applied to general growth, & secondly to that same general law which as we have seen gives to mongrels, animals & plants not only increased fertility but greater constitutional vigour & size. It is not a little remarkable thus to see under such opposite contingencies as increased & decreased fertility, an accession of size & vigour.

It is well ascertained4 that hybrids will always breed more easily with one of their parents, & indeed not rarely with a third distinct species, than when self-fertilised or crossed inter se.—Herbert would [have] explained even this fact by the advantage of a fresh cross, but Gaertner far more justly accounts for it, by the pollen of the hybrid plant, being in itself in some degree vitiated, whereas the pollen/9/of either parent species or of a third distinct species is sound. Nevertheless there are some facts on record, which seem to show that even in hybrids a fresh cross does do some good in respect to their fertility. Herbert states 5 that having in flower at same time nine hybrid6 //

1 ['"Fortsetzung", 1763, s. 29; "Dritte Fortsetzung", S. 44, 96; "Act. Acad. St. Petersburg", 1782, part ii, p. 251; "Nova Acta", 1793, pp. 391, 394; "Nova Acta", 1795, pp. 316, 323.' as cited in Variation, II, 130, ch. 17 n. 53.]

2 ['"Bastarderzeugung", s. 259, 518, 526 et seq.' as cited in Variation, II, 130, ch. 17 n. 52.]

3 [Kolreuter] Nova Acta 1795. p. 316.

4 Gaertner Bastarderzeugung p. 430 [cf. Herbert Amaryllidaceae, p. 352.]

5 Amaryllidaceae 1837 p. 371. The statement is confirmed after experiments tried during several years in Horticultural Soc. Journal. Vol. 2. p. 19

6 [Fol. 9 is sheared oft at this point and the rest of the paragraph is quoted from Variation, II, 138-9, ch. 17.]


[Hippeastrums, of complicated origin, descended from several species, he found that "almost every flower touched with pollen from another cross produced seed abundantly, and those which were touched with their own pollen either failed entirely, or formed slowly a pod of inferior size, with fewer seeds." In the 'Horticultural Journal' he adds that, "the admission of the pollen of another cross-bred Hippeastrum (however complicated the cross) to any one flower of the number, is almost sure to check the fructification of the others." In a letter written to me in 1839, Dr. Herbert says that he had already tried these experiments during five consecutive years, and he subsequently repeated them, with the same invariable result. He was thus led to make an analogous trial on a pure species, namely, on the Hippeastrum aulicum, which he had lately imported from Brazil: this bulb produced four flowers, three of which were fertilised by their own pollen, and the fourth by the pollen of a triple cross between H. bulbulosum, reginae, and; the result was, that "the ovaries of the three first flowers soon ceased to grow, and after a few days perished entirely: whereas the pod impregnated by the hybrid made vigorous and rapid progress to maturity, and bore good seed, which vegetated freely." This is, indeed, as Herbert remarks, "a strange truth," but not so strange as it then appeared.]

9A/Now considering how many crossed Hippeastrums were experimentised on, & that they were crossed in all sorts of ways, & that the pollen in each case applied to the stigma of one plant was from some other hybrid, & therefore not sound, I can understand the strong &/10/overpowering <marked> good effect of its application, only on the abstract good from crossing, as seen in crossing varieties. Moreover this case of the hybrid Hippeastrums is confirmed as we shall hereafter see in the chapter on Hybridism in some degree by some extraordinary cases, well ascertained by Gaertner, Kolreuter, & Herbert, in which pure species of Lobelia, Passiflora, Hippeastrum, Verbascum, had both pollen & germ in proper condition as shown by their fertilising, & being fertilised by, other species, but yet were incapable of, self-fertilisation, when their own pollen was placed on their own stigmas. These facts seem to show that in hybrids from distinct species, independently of the greater vigour & luxuriance often acquired, that even in regard to fertility, which is undoubtedly almost universally diminished or quite annihilated, there is some slight counter-balancing good in the act of crossing which occasionally appears in the intercrossing of hybrid with hybrid./


11/ Good from slight changed conditions—I think some little light can be thrown on the good resulting from crossing the breed, from considering the effects on the individual of slightly changed conditions. It has been a very general belief from ancient times to the present day, in many countries that <decided> good results from taking the seed, tuber or bulb of a plant grown in one kind of soil or situation & planting it in another; the most opposite kinds of soil being chosen, seeds, tubers &c being often interchanged between residents thus situated. I should have thought less of this belief, if it had been confined to cottagers or common farmers, but I find on enquiring from some [who] attend especially to raising seed-corn, & whose success is testified by their obtaining the highest prices in the market, that they find it indispensable to change their seed every few years.—One eminent gentleman in this line has two farms at different heights & on very different soils, so that he is able to exchange his own seed, but even with this advantage, he yet finds it advantageous to purchase occasionally fresh seed grown on other land.1 Mr. Robson, a practical gardener,2 /12/positively states that he has seen himself decided advantage in obtaining bulbs of the onion, tubers of potatoes & seed from different soils, & from distant parts of England. Oberlin3 attributed in great part the surprising good he effected amongst the poor of the Vosges in the cultivation of the potato, (the yield having been reduced in between 50 & 60 years from 120–150 to only 30 or 40 bushels in the year 1767) to changing the sets. In the cases of good resulting from the exchanging of seeds, I should think it could not be explained on the same chemical principles as in the rotation <of crops> of different species, namely by the seed obtaining some <chemical> element in one soil good for use with say for wheat not found in sufficient abundance in another soil also good for wheat for how small a difference in a single grain could the excess be, & this one grain has to influence the whole yield of the plant. Such a chemical view has more probability, & yet not much I think, when applied to the exchange of tubers of potatoes; but even in this case the slice planted bears but a small proportion to the yield of tubers./

13/As animals are less fixed to one spot & the same conditions it is less easy to get evidence of the good of change. But with invalids,/14/no medical man doubts of such good being most

1 The Rev. D. Walker in his Prize Essay of Highland Agricult. Soo. Vol. 2. p. 200, expresses a strong opinion on this subject. See also Marshall's 'Minutes of Agriculture' Nov. 1775. <Mr. Loiseleur Deslongchamps in his 'Considerations sur les Cáreales 1843. p. 200' gives numerous references on this subject.>

2 Cottage Gardener 1856 p. 186

3 Memoirs of p. 73


evident. Small farmers again find their cattle prosper best when they can occasionally change their pasture. It seems very doubtful whether in these cases the good can simply be accounted for by some fresh element in their food, which was before wanting. It would rather appear as if the marvellous & complicated play of affinities & constant change by which life is kept up, was somehow stimulated by almost any sort of slight change in the conditions to which the individual is exposed. Judging from plants, as both those which are useful from the number & quality of their seeds, & those which are useful from their organs of vegetation seem to be benefitted by a change, we may infer, that as in the case of crossing, both general luxuriance & fertility are increased.

If the facts here just given can be trusted, I think we can in some degree understand the good of crossing, for the individual with a blended constitution, derived from the union of the male & female from two varieties, differing in/15/structure or constitution or even two individuals of different families will be exposed during its life whatever the conditions of its existence may be, to a somewhat different relation with external things to what either of its simple parents can have been;— <for I presume it will be admitted that every part of the structure is related either to the external conditions or to other portions of its own structure..>

Considering the various cases now discussed,—obscure as many of the facts are & doubtful the evidence—namely the apparent ill effects of close interbreeding, the good from crossing individuals of distinct families or varieties, & even of species in this latter case with the great exception of fertility, considering what little light is thrown on the subject from the good of changed conditions to the individual, I should be strongly tempted to believe with Mr. Andrew Knight,1 that it was an essential part of the great laws of propagation that occasionally there should be the concourse of two separate individuals in the act of reproduction. But instantly it will occur to everyone that there are very many hermaphrodite organisms, with/16/the two sexes united in one individual. How it may be asked can in such cases two individuals occasionally cross? If an organism can from the day of its creation go on most strictly interbreeding, that is self-fertilising itself from the day of its creation to its extinction, one may well doubt all the foregoing

1 Philosophical Transactions 1799. p 202. Mr. Knight argues "that nature intended a sexual intercourse should take place between neighbouring plants of the same species."—Kolreuter in Mem. de 1'Acad. St. Petersbourg Vol. 3. p. 197 makes striking similar remarks: I think in the Portfolio on "Dichogamy".—[Darwin's notes on dichogamy are now in vol. 49 of his papers in the Cambridge University Library. His notes on this article of Kelreuter's are on fol. 163 of vol. 49.]


facts & put them all down to popular prejudices. I can hardly believe this. The subject has sufficient importance for us, in relation to crossing of slight varieties being a powerful means of keeping a breed or species true,—in relation to some points in geographical distribution,—perhaps to the extinction of species when become very rare,—& to some other points, that I must discuss it at some little length.1

First for some general considerations, which seem to me to have considerable weight. In land animals, after attending to the subject for several years, I have not been able to find any one case, in which the concourse of two individuals is not requisite;2 yet there are a good many hermaphrodite animals/17/as land-shells, certain annelids, as earthworms, land-leeches & planariae, but these all unite in pairs for propagation. In aquatic animals there are numerous cases of hermaphrodite <bisexual> animals which can certainly propagate by self-fertilisation; but in these forms the fluid medium in which they live, & from the fluid nature of the liquor seminis there is a possibility of an occasional cross, & we shall presently see that this is favoured by their structure.

In land animals, on the other hand from the nature of the liquor seminis it is obvious there never could be a cross between two individuals, without their close contact or union; & this, as far as I can find out, is the universal rule in land bisexual animals. This fact is the more striking, when we contrast land animals & land plants; in these latter hermaphroditism <bisexuality> & self fertilisation is the rule &/18/unisexuality <monoecious & dioecious plants> the exception; but in plants the fertilising element or pollen is not liquid & can easily, as is well known, be carried through the air from individual to individual by insects & the wind.

Secondly, in plants it is known3 that damp winds & rain are very injurious to their fertilisation; yet the general rule is that flowers are open & fertilisation takes place <sub jove> under the open sky. Such cases as the snap-dragon & papilionaceous flowers cannot be considered as exceptions, but rather as confirming the remark, for though they protect the stigma & anthers from rain, as do drooping tubular & bell-shaped flowers, yet they are not sealed up, but frequently opened & visited by insects. The few cases in which fertilisation appears to take place in really closed flowers will be presently discussed. I am far from pretending that

1 [Here Darwin pencilled the following memorandum: 'Get Huxley to read over for this.']

2 The Acarus mentioned by Owen See the authority [?].

3 Gaertner Bastarderzeugung p. 11.


there may not be some other additional & quite different explanation of the generality of the fact of the fertilisation of plants taking place, exposed to the injurious effects of climate & to an enormous loss of pollen/19/by the consumption of insects, but yet if an occasional cross with another individual is a law of nature, we have an explanation of these facts.

Thirdly, in animals & plants there are many instances of hermaphrodite <bisexual> & unisexual species in the same group & even frequently in the same genus; that is, we have the two sexes united in the same individual, or in two separate individuals in organisms, in all other respects very closely allied. Now if there be no such thing in nature, as an hermaphrodite fertilising itself throughout its whole existence;—if the only difference be indegree, the hermaphrodite occasionally crossing with another individual, the unisexual at every act of propagation, then the concurrence of bisexual & unisexual organisms in the same groups is less surprising & Nature in this case, as in other cases, has not moved per saltum./

20/Now for some details showing that in all animals the occasional crossing of two individuals seems to be possible: if it could be demonstrated that the structure of any animal was at all times such that the access of the liquor seminis from another individual was impossible, then the conclusion towards which I am tending that an occasional cross is a law of nature would be proved to be erroneous.—I shall pass over those low animals, the protozoa,barely distinguishable from plants, for I believe true sexual generation has not been observed in them; but the steady progress of knowledge of late years should make us very cautious in assuming that they have not sexes. In the lower plants as mosses & lichens there are many cases of species for long periods & in certain districts which have here at most rarely been seen to fructify, being propagated by generation but which are known in other districts & at other times to follow the ordinary law;1 & so it may be with some of the lower animals.

All the vertebrata are bisexual, except as it would appear some fish of the genus Serranus2 but from what we know of the habits of fish, an/21/occasional cross seems far from improbable. In the enormous Kingdom of true articulata (excluding annelids) all are <bisexual> unisexual, except the acarus previously alluded to, & the order of Cirripedia. In Cirripedes I have shown that a very

1 [Durieu in] Silliman's Journal vol. 21. p. 171. Several instances are here given taken from the Transactions of the Linnean Soc. of Bordeaux.

2 Quatrefages. Revue [des deux mondes, 1856, tome 4, p. 80, n. 2.] I have not seen the memoir of M. [Dufosse?].


few are <bisexual> unisexual; & that the fertilisation of some other very few which are hermaphrodite are aided by what I have called complements! males, which are distinct individuals; these few species, therefore, can be crossed. But by a piece of good fortune I met with some monstrous specimens1 of Balanus balanoides, a truly hermaphrodite form, in which the male organs were rudimentary, & the channel absolutely imperforate, nevertheless three of these specimens included developed larvae; proving (without we admit lucina sine concubitu) that the liquor seminis from other individuals had gained access to the open sack of these monstrous individuals.

In the other <two> great animal Kingdoms, there are many <bisexual> hermaphrodite forms; but it deserves notice that during the last 20 or 30 <twenty> years a surprising number of these lower animals, which were formerly thought to be <bisexual> hermaphrodite are now known to be unisexual.—Of the <bisexual> hermaphrodite animals/22/many, as2 the gasteropod univalve shells, & marine worms or annelids require the concourse of two individuals. Until lately all acephalous mollusca, or bivalveshells, were thought to be hermaphrodite, but now many as the common mussel & cockle3 are known to be <bisexual> unisexual, & their fertilisation is probably (must be) effected by the spermatozoa being drawn in by the same ciliary currents by which food is obtained; & this same method could facilitate an occasional cross in the hermaphrodite bivalves. I long thought from the description which I had read that the common oyster was a case of perpetual self-fertilisation, but it now seems as I am informed by Prof. Huxley, from the observation of M. Devaine that the male & female products are matured at different periods & therefore that the oyster though in structure an hermaphrodite, in function would appear to be <bisexual> unisexual. /22 v/This likewise, according to Prof. Huxley's own observations is the case with the <bisexual> hermaphrodite ascidians./22/From the analogy of plants, I should expect that this maturity at different periods would prove to be of frequent occurrence with/23/animals. In parasitic worms or Entozoa, many are <bisexual> unisexual, but some which are hermaphrodite4 mutually unite; & Dr. Creplin remarks that in

1 Monograph on the Cirripedia, published by the Ray Soc. 1854. p. 102.

2 [Here Darwin pencilled 'all?' before 'the gasteropod', a line under and a question mark after 'gasteropod' and two question marks after 'annelids'. In the margin he pencilled: 'V. Owen & Huxley.']

3 Von Siebold in Wiegmann's Archiv fur Naturgesch 1837. p. 51

4 Dr. Creplin in appendix to Steenstrup's Untersuchungen iiber das Vorkommen des Hermaphroditismus tr. Hornschuch 1846. I have seen a translation of this owing to the kindness of Mr. Busk


those in which Von Siebold discovered an internal passage from the male to the female organs, apparently insuring perpetual self-fertilisation, the so-called cirrus exists, which would lead from analogy to the conclusion that there must be at least occasional mutual fertilisation.

Distrusting my own knowledge I applied to Professor Huxley, whose knowledge of the invertebrate animals is well known to be profound, whether he knew of any animals whose structure was such that an occasional cross was physically impossible. He informs me that some of the jelly-fish (Beroidae) seem to offer the greatest difficulty, but even in them it is not positively known whether or not the eggs are discharged fertilised;/24/& that as these animals derive their food from indrawn currents of water, which bathe the ovaria, it is certainly quite possible that the spermatozoa of other individuals might come into action.1 Again Prof. Huxley informs me that he should have thought that the hermaphrodite Bryozoa or Polyzoa (certain corallines) would have offered insuperable difficulties to an occasional cross, had it not been for Mr. Hinck's observations, who saw in some species the spermatozoa pouring out from pores between the tentacula; & as Prof. Huxley remarks what could this be for, except to fertilise some other individual. Moreover there are some <unisexual Polyzoa> Bryozoa, with the sexes distinct,2 which proves that fertilisation can be effected between the separated, & yet fixed polyps.3

In all these cases of aquatic animals it is well [to] remember Spallanzani's curious experiment,4 namely that three grains of the liquor seminis of a frog thoroughly diffused in a /25/pound & a half of water retained its full power, & that the same quantity when diffused in 22 pounds sufficed to vivify some of the eggs. The weight of the liquor seminis serving to fertilise a single egg was calculated to be only 1/199,468,7500 of a grain! Finally as far as I can discover, under our present state of knowledge, no animal is known, the structure of which would prevent an occasional cross; & this fact, considering the astounding diversity of nature, seems to me an improbable coincidence, without the capacity of such occasional crossing be one of the laws of reproduction <propagation>.6

1 [Pencilled memorandum:] Nordmann & Owen on sexes separate in Flustra. [Nordmann] L'Institut 1839 p. 95—on sexes in coralline allied to Flustra & on zoospermatic animalcules!

2 Hincks Brit. Assoc. 1852 Proc. of Sect. p. 75.

3 [Pencilled memorandum:] I might put in notes other examples of Pollen.

4 Dissertation II relative to Natural History. English Translat. [See vol. II, dissertation II, § CXLIII pp. 140–50.]

5 In Bee v. Siebold shows semen keeps power for 4 or 5 years. [Parthenogenesis in Moths and Bees, p. 14].


26/Crossing of Plants. To show that all plants are capable of being occasionally crossed by another individual of the same species, is more difficult than with animals. Hermaphroditism <Bisexuality> with self-fertilisation is here the rule, & the separation of the sexes the exception.' The mere proximity of the male & female organs in the same flower,—the apparent frequency of the pollen & stigma being ready at the same time,—the explosion of the anthers close to the stigma & the lightness of the pollen,—the movement of the stamens to the pistil & of the pistil towards the stamens, would all at first lead to the conclusion that self-fertilisation would be almost invariable. But I think we shall see that such a conclusion would be hasty. Besides the comparatively few monoicous & dioicous <mono- and dioecious> plants, C. C. Sprengel1 has shown that many hermaphrodite plants are what he calls dichogamous,—namely that either the pollen is mature & has been shed in one flower before its stigma is ready to receive it, or on the contrary (which is a less frequent case) the stigma is ready before the anthers have burst;2 hence in these cases, the plants are essentially <bisexual> unisexual, being fertilised by the pollen of an older or younger flower <or at least an occasional cross is greatly facilitated; I cannot doubt from the observations of others, & even from my own that these cases are frequent.> I may state that I have tested during several years many of Sprengel's observations, in those cases in which I could judge by the clefts of the stigma, opening &c, & am convinced of his general accuracy.3 / 26 bis/It would be useless to give examples of dichogamy from Sprengel; they are so numerous, for instance in many Scrophulariaceae, & in all or in most Umbelliferae;4 in many Onagraceae as I have myself observed in genera not noticed by Sprengel. So again Kölreuter observed similar facts long ago5 in many Mal-

1 The curious work containing these observations is entitled "Das Entdeckte Geheimniss der Natur 1793." The greatest living Botanist, Robert Brown, thinks highly of Sprengel's power of observation, as I <know from conversation> have heard from him. I mention this because Gaertner in his two admirable works does not seem to think highly of Sprengel; He gives however, (Bastarderzeugung p. 65) one strong case of inevitable dichogamy & admits (p. 659) that in many plants, as in whole families, that the pollen and stigma do not come to maturity at the same time, but in many of these cases, probably in most, the pollen is retained close at hand so that it may easily fertilise the pistil in the same flower. A. F. Wiegmann (Uber die Bastardzeugung 1828. S. x) says after careful observation he is convinced that most of the statements of Sprengel are correct, & that he could write a commentary on his work.

2 [Pencilled memorandum] Mr Schubert [?] add half-dichogamous

4 Das Entdeckte &c p. 28, 154, 50, 322 &c.

5 Mem. de l'Acad. St. Petersburg. Tom 3. p. 197


vaceae. This seems to be the case from Cassini's observations1 "nearly throughout the Compositae."; & the pollen in this great Family was observed by Kolreuter to be aculeate, & specially adapted to adhere to insects. In Lobelia, judging from my own examination of a few species, the pollen is swept clean out of the united anthers, in the same manner as in the Compositae, by the fringe on the style, some time before the stigma is ready for its reception. That the growth of the pistil in these cases is really adapted to sweep the pollen out of the anthers, before the stigma is mature, I must think from having observed the same process effected by very different means in the Crucianella stylosa [;] here the mouth of the corolla is much contracted, so that the anthers, which open whilst the flower is in bud, instead of being united together as in Lobelia & the Compositae, are pressed close round the pistil. The style is of remarkable length, & lies zig-zag in the bud; as soon as the flower opens it is rather quickly & sometimes suddenly protruded by its elasticity; & in this movement owing to the largely knobbed & rugose stigma, it pushes out the pollen; & not till some time afterward does the stigma open & becoming humid is apparently ready for fertilisation./

27/It is known that many cultivated varieties of plants, not only are capable of occasionally crossing, but without great care are actually crossed very frequently. The Cruciferae are particularly apt to be adulterated, a single cabbage*a2 plant sufficing to contaminate whole beds of other varieties. I had a radish plant which flowered in the same bed with several other varieties: I saved a few seeds from one plant, & out of the 22 plants which I raised only 12 came true to their kind. /27v/So, again, with turnips [.]3

1 quoted in Linn. Transact Vol XIII. p. 595

2 [Scholars should be warned that for fols. 27 & 30 Darwin's note slips seem dis arranged or interchanged, and their proper insertion points seem uncertain. Darwin used reference marks such as '*a' to identify notes and to indicate their insertion points. These will be given along with the present folio reference numbers of the note slips, so that the reader may rearrange them if he does not accept the order adopted here.]

*a [30 v] Gaertner (Bastardzeugung p. 566) gives an experiment on 4 plants of Matthiola annua, the flowers of which he castrated, & kept two in his room unfertilised & they produced no seed; the other two he placed in his garden, 100 yards distant from some other plants; both of them produced some poor pods, containing 68 apparently good seed, from which, however, only 20 seedlings were raised. See p. 573 for an analogous fact with Nicotiana. *a [30 v] <Wiegmann's experiments (Über die Bastardzeugung s. 32, 33) on cabbages show the extraordinary degree in which they cross without artificial aid. A most intelligent foreman in a nursery garden assured me that he had known seed of a plot of Brussels-sprouts spoiled by a bed of Drum-head cabbages about 200 yards off. Mr. Masters of Canterbury has known a single Red Cabbage spoil the seed of Savoys, Cabbage & Broccoli in neighbouring gardens.—Many other instances could be given>

3 Gardener's Chronicle 1856 p.729


In the Cruciferae according to Gaertner,1 the pollen & stigma are not ready at the same time; but I doubt whether this alone will account for the extent to which they blend, & I suspect that the pollen of another variety must have a prepotent effect over the pollen of the stigma's own flower, in the same way as it is known that the pollen of one species is prepotent over & obliterates the effect of the pollen of another species, previously placed on a stigma.—/27/Gallesio in his treatise <on oranges> does not doubt that oranges very commonly cross.2 It is impossible to prevent the different varieties of Rhubarb (as I have known myself) from crossing, if grown near each other./27 v/The various species of Crinum sent by W. Herbert3 to Calcutta cross so freely in the garden, that true seed cannot be saved./27/And many other instances in Rhododendron, Berberis, Poppies (in which latter I know of case in which not one seedling came true) could be given./27 v/The mere circumstance of great beds of one variety being cultivated in any one place is alone a considerable protection that seeds shall not be adulterated; & hence certain villages4 have become famous for pure seed of certain varieties, owing to the masses of the same variety there cultivated & to the exclusion of other kinds./

27/But by far the strongest proof, as it seems to me, of the extent to which the pollen from one flower is carried to other flowers of the same species;s incidentally offered by hybridisers. Without a single exception, all these naturalists, several of whom have devoted their lives to the subject, insist in the strongest manner on the absolute necessity of perfect isolation of the castrated flower,*b5 so as to preclude the possibility of access of its own pollen. Herbert6 positively states that it is not always sufficient to enclose/28/a flower in gauze; so subtle are the means by which pollen can be introduced: from some experiments in hybridising which I have made myself, I suspect that it is the minute Thrips which in these cases brings pollen, as I have found that it crawls into flowers protected by gauze & I have often found this insect

1 Bastardzeugung s. 659 [Traité du Citrus, pp. 40-1.]

3 Amaryllidaceae p. 32. [cf. Darwin's Effects of Cross and Self Fertilisation (London, 1876) ch. 10, note on p. 395. Another note slip marked 27 v, having no indication of insertion point, reads:] Kolreuter Dritte Fortsetzung. s. 56. Hybrid Pinks often arise naturally in gardens.

4 Lindley's Horticulture p. 319.

*(b)5 [30 v] See Gaertner's, the most admirable of all observers on this subject, strong expression on this subject in his Bastardzeugung s. 670. Experiments made in the open air, he says, must be absolutely rejected. (Beitrage zur Kenntniss s. 510, 573), See also Lecoq De la Fecondation &c. 1845. p. 27. [cf. Darwin's Cross Fertilisation, ch. 10, note on pp. 378–9.]

6 Amaryllidaceae p. 349


dusted with pollen. In the first season of Gaertners grand series of observations he crossed after castration 20 distinct genera, & obtained, as he thought hybrids from <nearly> all; but Herbert, who had been in the field before, at once published his entire disbelief of these experiments, & asserted that the isolation had not been sufficient; which was subsequently acknowledged1 with perfect candour by Gaertner. Prof. Henschel's experiments, worthless in all other respect2 are interesting as showing the extent to which crossing goes on without they be completely isolated; he castrated flowers of 37 species (belonging to about 22 genera,) & either put on no pollen or pollen of other genera &c, & yet obtained seedlings from all.—Other parallel cases of experiments made by Dr. Mauz might have been given. No doubt in many of these cases the fertilisation has been effected by pollen carelessly left in the castrated flowers. But a most curious table published by Gaertner3 /29/shows I think conclusively to what a wonderful extent pollen is.carried from flower to flower. In 1825 he castrated 520 flowers & placed in them pollen of other species & genera; & as he says he thought it laug[h]able to suppose4 that pollen could be brought to his castrated flowers from other flowers of the same species growing between 500 & 600 yards distant, he did not isolate the plants more perfectly. The result was5


19. Which produced seed that did not germinate, & therefore have no bearing whatever on the result, & may be eliminated.

29. which produced true hybrids, & therefore the pollen, intentionally placed on them produced its effect.—

270 which remained unimpregnated, & therefore on which the foreign pollen had no effect, & on which the pollen of its own kind had not been brought by any agency.

202. produced seed, which yielded pure plants, & therefore on which the foreign pollen had produced no effect, but pollen of its own kind had somehow been introduced.

520 total number of flowers experimentised on in 1825.

Now one's first impression is that in the 202 castrated flowers, attempted to be impregnated with other pollen, but which produced their own kind, is that their own pollen must have been carelessly left in; but Gaertners tables9 shows that this explanation is not

1 Bastardzeugung. s. 128

2 given by Gaertner in his Kentniss s. 574

3 Kenntniss s 550

4 Kenntniss. s. 539, & 575

5 Kenntniss s. 576

6 Kenntniss &c s. 555 & S76


sufficient, for during the 18 subsequent years he castrated no less than 8042/30/flowers, & always kept them in a closed room, so that they could not possibly get pollen from other individuals of the same species, & then during these many years & out of the 8042 flowers he had only 70 cases.of seed producing pure plants, showing, that the castration had been imperfect; whereas in 1825 when he experimentised in the open air we have seen that out of 520 flowers 202 produced pure seedlings! Yet plants of the same species in several of the cases did not grow within 500 or 600 yards distance !*1 It should, however, be not overlooked that when a flower is castrated, the stigma retains its capacity for fertilisation2 for a considerable time, so that a castrated flower would have a much better chance of being fertilised by pollen from another individual, than would a plant having pollen of its own, the action of which would fertilise the pistil.

Now considering that there are some monoecious & dioicious plants,—that there are many dichogamous plants of C. C. Sprengel, which are in fact monoecious—considering the many cases of the intercrossing of varieties in our gardens—& especially considering the astonishing care which all hybridisers have found absolutely essential to prevent the pollen of/31/its own kind being brought to the castrated flowers—& considering the hybridisers have indiscriminately from varied motives worked on nearly all kinds of flowers,/31 v/—lastly considering the many cases, in our gardens & in a state of nature, of hybrids spontaneously being formed—/31/ I must conclude that the transmission of the pollen from individual to individual is not only very generally possible, but that it actually its so transmitted. Some further facts will be presently given.

Before considering the many grave cases of difficulty opposed to the foregoing conclusion being made universal, it may be interesting briefly to consider the means of transmission. It is known that in many plants with the sexes in separate flowers the pollen is carried by the wind, & hence has to be produced in such astonishing quantities, that many buckets full of the pollen of various fir-trees have been swept off the decks of ships on the shores of N. America.3 In some associated hermaphrodite plants, as in Gramineae, in which the stigma is large, branched & at the period of fertilisation fully exposed, in which the pollen is but/32/

1 *[27 v] But even in the largest nurseries, it is surprising the trouble which the owners are compelled to take to keep their seed crops unadulterated; thus Messrs. Sharp "have land engaged in the growth of seed in no less than eight parishes."(Gardeners' Chronicle 1856. p. 823). [Cf. Cross Fertilisation, ch. 10, p.395.]

2 Kenntnis S. 145 [Anon.] Silliman's Journal Jan. 1842.


little coherent, & the long slender filaments seem formed to scatter the pollen, I do not doubt that crosses must be often effected by the wind./31/But in most hermaphrodite <bisexual> flowers, owing to their structure, or to the small quantity of the pollen, or to its coherence or to the small size of the stigma, I think it may safely be concluded that the wind can but seldom bring sufficient pollen (for several grains are almost always required for the act of fertilisation) from one flower to the other so as to effect a cross between two individuals./32/Insects of various orders, more especially Bees,1 are the great agents. Many flowers cannot be fertilised without their agency, as is admitted, though very unwillingly, by Gaertner:2 it is impossible to read C. C. Sprengels details & then examine many flowers, as most Irideae, Passiflora, Viola, most Orchideae &c &c & doubt this: in regard to all the Ascelpiadeae which have been carefully examined, Robert Brown, says the * absolute necessity' of the assistance of insects is manifest,3 & Sprengel <Gaertner> believes that their agency is rendered more effectual by their extraordinary activity, due to the intoxicating effect of the nectar.—It would be tedious to give other examples. All those who have personally attended to the subject have become strongly impressed with the efficiency of insect agency in the fertilisation of flowers./

32 n/There can be little doubt that C. C. Sprengel has pushed his views to a quite fanciful degree; as for instance, when he accounts for all the streaks of colour on the petals, as serving <formed> to guide insects to the nectary. Nevertheless some facts could be given in favour of such a view: Thus in a patch of the little blue Lobelia, which was incessantly visited by Hive Bees, I found that the flowers from which the corolla, or the lower streaked petal alone had been cut off, were no longer visited./ <Whether the Bees were then led to think that these flowers were withered, or whether the absence of this convenient alighting place on the lower petal was the cause, I know not. But I feel sure that> Bees seem to work against each other with excessive competition <industry>, so that they grudge the least loss of time: thus when visiting flowers with several nectaries if one be dry, they do not try the others; again when visiting flowers which have been bored, if one has accidentally not been bored I have seen Bee after Bee pass over it & not stop to bite a hole nor will they enter the open tubular flower, though having to crawl over it, but will dash on to another bored flower. By the way, if proof were wanted how little Bees require any guide to the nectary, their habit of biting holes in the lower part of the corolla or through the calyx, so as to reach the nectary without the loss of time of crawling in at the mouth of the flower would prove it. These holes when once formed are known to

1 Bees are found in all parts of the world; even in the extreme Arctic regions they have been seen sucking the flowers. But I must add that on the little coral islets, called the Keeling Islands, in the Indian Ocean, I found no Bees; but there were other insects.

2 Beitrage zur Kenntniss. s. 335

3 Linnean Transact. vol. 16, p. 731


& used by Bees/32 n. 1/of various species & genera: when as in Kidney beans, the hole has been bored on the lower side of the calyx, bee after bee flies to the under side with unerring precision. Bees, as far as I can judge are guided by various senses to flowers, & more especially by knowledge of the position of each tuft of flowers in a garden. It is well known that the same Bee keeps as much as it can to the same species, when getting nectar; & I have repeatedly seen them flying in a <direct line> clearly determined course from plant to plant of the same species, when round a corner & so out of sight.—They are good Botanists, & know well that plants of the same species may have brilliantly different colours; but they know that they are only varieties & visit them indiscriminately. I think they recognise a plant by its general habit; I have seen Humble bees after visiting a tall blue Larkspur fly to another plant, of which the buds were so little open, that they were hardly tinged with blue. They seem often to be aware if another Bee has almost instantly before visited a flower, & will then not try; but I have seen one blunder & itself visit the same flower twice./

33/It is, I think, impossible to doubt that the structure of very many flowers has been formed in direct relation to the part which insects play in their fertilisation. What can be a more beautiful adaptation than that shown by R. Brown to exist in the Asclepiadeae & Orchideae, between the stickiness of the gland of the pollen masses, of their separate grains one to another & to the surface of the stigma, by which it follows that the instant an insect touches the gland it draws out the whole pollen mass out of its case, & then the sticky stigmas of the several flowers, as the insect crawls from one to the other, each take a few grains of pollen from the coherent mass. It is worth anyones while to watch a Bee visiting a Salvia, or to push some thin body like the Bees head down the tube of this flower, & notice how the anthers & stigma are protruded & rubbed on the Bee's back; then let him cut open the flower & see the cause of this is two projections near the base of the stamens, closing the passage, & the movements of which by the/34/Bees proboscis, causes the protrusion of the anthers & stigma from beneath their hood: I can no more doubt the final cause of this structure than I can of a certain mouse-trap. I have seen a Bee enter the flower of a Mimulus & in doing this the two-lipped stigma fairly licked the back of the insect which was thickly dusted with the pollen from another flower, & then the two lips of the stigma slowly closed on the pollen which it had thus obtained. It is pretty to compare in those species of Fumaria, in which either one or both nectaries secrete honey, the different movements of the parts of the flower as a Bee enters. Even in trifling details as in the position of the stamens <anthers> & pistil, in relation to the nectary, I believe that there is very generally a distinct relation to the action of insects: thus in the


Dictamnus Fraxinella, I noticed during several days that the stamens & pistil were placed so that a Bee visiting the nectary would not touch them; but then came a hot day & the anthers all burst & stigma was humid, & I found their positions all changed & their tips now stood in the direct gangway to the nectary, & were/ 35/brushed by every Bee which entered. I could fill pages full of other instances from C. C. Sprengel & from my own observations.

I will only allude to the case of the Berberis/35 v/in which the stamens move to the pistil, & in which consequently, it might have been thought there would be seldom any chance of a cross with another individual: but Kolreuter has shown1 that they never move till touched by some insect; so that insects are necessary to their fertilisation; & their flying from flower to flower could hardly fail to bring pollen from individual to individual. Indeed the extent to which the American evergreen Barberries (Mahonia) have been hybridised together, so that it is almost difficult in our nursery gardens, as I have found, to get a pure plant, shows that this has occurred not only with the individuals of the same species, but of different species. Similar remarks are applicable to some other plants, of which either the stamen or pistils move on being touched./35 v/Thus in regard to the pistil of a Goldfussia, it is scarcely possible to doubt from Ch. Morrens2 remarks & curious observations that the movement of the stigma when touched towards the lower side of the corolla, where the fallen pollen is collected, stands in direct relation to the action of insects. Again in Stylidium, as described by Ch. Morren3 I can see no difficulty, from the proportions in the parts, of a Bee carrying pollen from flowers, when by sucking at the nectary it causes the sudden & remarkable movement of the column; though Ch. Morren may be quite correct that this movement, also, aids the fertilisation of the flower by its own pollen.—/35 v/In Parnassia palustris the stamens slowly move one after the other over the pistil; but Sprengel4 positively asserts that the pistil at this period is not fit for fertilisation, & <therefore that the plant is strictly dichogamous> he supposes that it is fertilised by pollen from a younger flower brought by some nocturnal insect. Allium5 is in nearly the same case./

35/Bees & other insects visit flowers both for the pollen & nectar. The nectar cannot be supposed to be formed, any more than the pollen, for the sole purpose of attracting insects; for nectar is sometimes secreted outside flowers, as by the bracts of various Legumi-

1 Nova Acta Petrop. 1788 p. 214

2 Nouv. Mém. de l'Acad. Roy. de Bruxelles. Tome XII 1839.

3 Ib. Tome xn. 1838

4 Geheimniss &c 167

4 ib. p. 186


nosae,1 but nature has utilised this secretion for the very distinct purpose of facilitating fertilisation, & as I believe occasional crossing.

When Kolreuter first discovered2 that the Malvaceae, owing to the adhesive pollen & stigma not being ready at the same time in the same flower, can be fertilised only by the agency of insects, he says he was astonished that so important a function should have been left, as he then thought to accident,—to a mere happy chance; but he adds that further observation convinced him that the wise Creator has thus used the most/36/sure means. Hardly any means, I am convinced, could be surer; & in regard to our present discussion, it should be borne in mind that in every case in which insect agency is essential to fertilisation, & indeed in every case in which insects habitually visit flowers during this period, it is hardly possible to doubt that pollen is often brought from flower to flower of distinct individuals, & thus a cross between <separate individuals of the same species> effected. I have repeatedly seen many minute beetles, dusted with pollen, fly from flower to flower: some flowers, which are very rarely visited by bees as the Phloxes (which I have never seen visited except during one year) are frequently visited by butterflies:3 //36 v/I have remarked this particularly with the Rhingia rostrata on the Lychnis dioica, on an Ajuga & on many others, I have seen the same thing with Volucella plumosa on a Myosotis,—I may add that I have never seen a Bee visiting a Daisy but I have seen the Rhingia, Scaeva iris (?) & Hilara globulipes all thickly dusted with the pollen of this plant/37/flight, dusted like millers with pollen. I have seen several times the same thing with Thrips, an insect hardly larger than a bit of chopped bristle: one day I watched with a lens, one in the flower of a convovulus having four grains of pollen on its head, & these I saw left on the stigma, as it crawled over it. The crossing of the great flowers of foreign lands, may well be aided by Humming & other birds: I remember shooting in S. America, a mocking thrush, which had its head of so bright an orange from the pollen of a Cassia that I at first thought it was a new species.—

But Bees are the most important of all insects for this end. Until I watched I was not at all aware how quickly they work. In exactly one minute I saw one Humble-Bee visit 24 of the closed flowers of a toad-flax (Linaria cymbalaria); another 22 flowers of the Snow-berry tree (Symphoricarpos <Chiococca> racemosa); another 17 flowers of a Larkspur on two separate plants &c.—

1 I called attention to this fact in the Gardeners Chronicle, 1855 July 21; & had at the time quite forgotten that it had been previously noticed by Sprengel

2 Vorlaufige Nachricht 1761 p. 22

3 [The rest of fol. 36 is sheared off.]


The top flower of an Oenothera was visited eight times by Humble Bees in 15 minutes/37 v/& I noticed that one Bee visited in the course of a few minutes every single plant of Oenothera in a large flower garden; passing over, without regard, other plants having large yellow flowers, like Escholtzia:/37/in 19 minutes each flower of a tuft of Nemophila insignis was visited twice: in a large plant of Dictamnus Fraxinella with 280 flowers, from the rate at which Bees visited it, as observed during/38/several days, each flower at lowest computation must have been visited daily 30 times. It is no wonder that the beauty of many flowers, as I have noticed in some Mimuli & Lathyrus grandiflora, is greatly destroyed by the scratching of the hooked tarsi of the bees. Some flowers seem never visited by Bees; but with the exception of the Gramineae in all other cases of indigenous plants to which I have attended I have found that they were visited by other insects. Night-blooming flowers, which are often sweet-scented & of a white colour, I have reason to believe are visited by moths. One must be very cautious before assuming that any flower is not visited by Bees: in the first summer of my observations on this subject I watched many times daily for 14 days the Linaria cymbalaria & never saw a Bee look at it, when suddenly after a hot day Bees were most industriously at work. So again for a fortnight I saw Bees visiting White & Red clovers, but never looking at the little yellow Trifolium minus; & as the flowers were so minute I doubted whether they would ever visit them; when suddenly I one day found innumerable bees hard at work at this species over the whole country, & neglecting the other kinds. In <all these> most cases I believe that the secretion of the nectar, which determines the visits of the Bees is coincident/39/with the flowers being ready for fertilisation. The secretion of the nectar seems in close relation to temperature: I have observed in a little blue Lobelia, that if the sun went behind a cloud for even half an hour, the visits of the Bees immediately slackened & soon ceased.1 /39*/I may give one more instance in regard to the action of insects. In Viola odorata Sprengel has shown that the pollen cannot escape owing to the manner in which the anthers with their scales close round the pistil, till disturbed by the proboscis of an insect: he proved this by covering up some flowers & leaving others uncovered,2 & finding pollen shed in latter, but never in the protected flowers.

1 [Fol. 39 is sheared off just after an asterisk at this point. The lower part of fol. 39 is now numbered 40A. In the CD. MSS., fol. 27 of vol. III is a full-sized sheet of gray foolscap marked: '* to p. 39', and is inserted here where it seems to belong.]

2 Das Entdeckte &c p. 390.


Now in 1841 I watched almost daily & many times a day several patches of the V. tricolor or Heartsease for seven weeks & never saw an insect of any kind visit them; when suddenly on two successive days I saw several small Humble-bees visiting all the flowers. In the next year after a fortnight watching in vain I again saw two or three species of Bees (& a fly dusted with pollen) visiting most of the flowers & I found pollen profusely shed on the lower petals, all around the stigma; & I noticed the same fact on the same day with some plants of wild V. tricolor. Now in both these years, I noticed a few days after the visit of the Bees & of the Fly (I marked the flowers visited by the Fly) a great number of the flowers on the several clumps suddenly withered as if the germens had been set. Hence I cannot in the least doubt that I saw in these Humble Bees, the priests who celebrated the marriage ceremony of the Heartsease./

40A./In a flower garden containing some plants of Oenothera, the pollen of which can easily be recognised from its great size & shape, I found not only single grains, but whole masses within many flowers, of Mimulus, Digitalis, Antirrhinum, & Linaria. Other kinds of pollen were likewise distributed in the same flowers. A large part of the stigmas of a plant of Thyme in which the anthers were completely [?] aborted were likewise examined & their stigmas, though scarcely larger than a split needle, were covered not only with the pollen of Thyme brought by the bees from other plants, but with several other kinds of pollen. {but I was not/40/surprised at this, seeing how much Thyme is frequented by Bees & flies>/40 v/Those who have not attended to the subject of Hybridism; may feel inclined to exclaim that if pollen is carried from distinct species to species, so freely as these facts show in the cases, an endless number of hybrids would be formed. But nature has provided a most efficient check to this, namely in the prepotent effect of each species own pollen; so that all effect from the pollen of another species is obliterated by the previous or subsequent action of its own.—/40/I found a hybrid Rhododendron which [was] quite destitute of pollen, & which was so seldom visited by Bees, that after long watching the branch [?] for many days I never saw but four Bees visit it: yet on one morning I found from 50 to 100 grains of pollen of Azalea or Rhododendron on the stigmas of these flowers: another day I examined the stigmas of 19 flowers & on 13 of them there was the same <I found some> pollen. Kolreuter relates1 a curious experiment bearing on this subject: in an Hibiscus, which is necessarily fertilised by insects,

1 Fortsetzung &c 1763 p. 69


because its pollen is shed before the stigmas are ready, he marked 310 flowers & daily put pollen on their stigmas & left the same number of other flowers to the agency of insects which did not work during some days as the weather was cold with continued rain. He then counted the seeds of both lots; the flowers which he fertilised with such astonishing care produced 11,237 seeds & those left to the insects 10,886—that is only 351 fewer seeds./

41/From the facts now given, at too great length, though I could have given many more, I think it can hardly be doubted that insects play a very important part in the fertilisation of flowers; & furthermore that in those cases in which their agency may be not at all necessary, yet that they can hardly fail occasionally to bring pollen from one individual to another. Nor must the action of the wind be quite overlooked, which probably is highly efficient for an occasional cross in some hermaphrodite flowers, as it undoubtedly is in some mono-oecious & dioecious plants for their ordinary fertilisation. I should, indeed, have been inclined'boldly to affirm the proposition that all plants are not only capable, but do actually receive an occasional cross, had it not been for the following cases of serious difficulty.

Facts opposed to the doctrine that in plants an occasional cross is necessary. Very many statements may be found in the works of Botanists not only that the pollen is often matured & the anthers burst before the bud is opened, which admits of no doubt, but that in certain plants the stigma is regularly fertilised in the unopened flowers.1 which would render an occasional cross a physical impossibility. But there are many difficulties in the way of/42/ascertaining this; & observations made on only a few flowers during one season cannot avail much; for Gaertner has shown2 that the bursting of the anthers & relative maturity of the stigmas depends much on the weather & varies in the same species; & there seems to be no doubt that a plant may be occasionally fertilised in the unopened bud, of which the pollen is ordinarily ready only when the flower is fully expanded. Again Gaertner has shown3 that an abnormal precocity not rarely affects many flowers & that in this abnormal state it can be fertilised in the bud. But Gaertner was a firm believer that in many plants, even in whole Families,4 as the Leguminosae, Cruciferae, Onagraceae, Campanu-

1 Thus Aug. de Saint. Hilaire in his admirable Lecons de Botanique 1841. p. 572 says, without entering in details, "Chez une foule de plantes c'est dans le bouton ….que la fecondation s'opère."

2 Beitrage &c. s. 104

3 Beitrage zur Kenntniss &c. s. 571

4 Bastardzeugung s. 655


laceae &c, fertilisation takes place, not only some hours, but even from one to two days before the corolla opens. Now I am quite unable to reconcile this statement with others: of the Leguminosae I shall speak afterwards: in regard to the Campanulaceae there has been much discussion on this very point, & notwithstanding Gaertner's statement1 that the stigma can be fertilised before the clefts are fully marked, I can hardly doubt that Sprengel formerly & Wilson lately2 are correct/43/in believing that the fertilisation takes place after the flower is fully opened; if Gaertner is correct that the fertilisation takes place in the bud there is an inconceivable waste of pollen on the curiously organised, & retractile collecting hairs of the pistil; & the manner in which Bees, as I have often watched, frequent the flowers is admirably adapted to bring the pollen from the collecting hairs of one flower on to the stigma of another./43 v/In Phyteuma, one of the Campanulaceae, Sprengel found3 plenty of the coloured pollen on the open stigma; but if a branch, with unopened flowers was put into a glass of water in a room where there were no insects, not a grain could be discovered on the stigmas./43/So again in regard to the Onagraceae, I must think the weightiest evidence would be required to overthrow Sprengels statements4 in regard to Epilobium & Oenothera (which as far I can judge from repeated observation seem strictly true) that far from being fertilised in the bud, they are dichogamous, & invariably fertilised by the pollen of younger, which he saw effected by Humble Bees: Gaertner himself, elsewhere5 admits that in some Fuchsias, the pollen is not shed for some days after the flower is fully expanded, as is well known to Hybridisers.6 Lastly with respect to the Cruciferae, Gaertner's statement that they are fertilised in the bud seems to me quite extraordinary, considering the everyday experience of gardeners with cabbages, turnips, Radishes, &c. & to my mind throws doubt on/44/his other statements in regard to habitual fertilisation in the bud.

M. Loiseleur-Deslongchamps7 believes, though confessedly on imperfect observations & in opposition to some other authors that Wheat is fertilised within the closed flowers. This surprises me much, for I have repeatedly seen the florets widely open, with the feathery stigma protruded on one side, with the dangling anthers not fully discharged, & with the grains of pollen sticking over all parts of the florets: in most grasses, all the florets open at the

1 Beitrage s. 338

2 Hookers. [Lond. J. Bot. 7 (1848) 92–7.]

3 Das Entdeckte. s. 117

4 Das Entdeckte &c. s. 225

5 Beitrage &c. s. 104

6 Lecoq de la Fecondation &c. p. 129

7 Consid. surles Céréales 1842. p. 80


same time & with the protruded stigma, the plant for the time, as every one must have observed has a very different appearance: in wheat each floret opens separately & keeps open for only 3 or 4 hours, leaving the empty anthers dangling outside so that the whole phenomenon is far less conspicuous than in most grasses; & if the Chinese are at all to be trusted some varieties as Huc states flower in the night.1 The structure is such that I can hardly understand how an occasional cross from another individual can be avoided. A. Knight2 asserts that by sowing different varieties together,/45/"I obtained as many varieties as I wished." Col Le Couteur whose great experience makes his opinion valuable, though he gives no precise facts believes3 that wheats cross. Puvis4 asserts that nearly all the varieties which were grown near each other in an Agricultural Garden under his charge were each year modified; but his evidence seems to me of little value, as he attributes to the action of the pollen on the grain itself that kind of change which it is known results from climate & culture.—Opposed to these statements we have a much more precise one from M. Loiseleur Deslongchamps.5 Namely that during eight years he cultivated from 100 to 200 varieties very near each other, & that he never saw a hybrid appear. Making some allowance for different varieties, as noted in this very respect by Col. Le Couteur, flowering at different times, & even from the positions of the beds with respect to the wind, this statement is very remarkable, & at first seems almost conclusive against occasional crosses. But I do not think the experiment has been fairly tried, until the different varieties are sown close together, as Knight sowed/46/them; for wheat is not, as far as I can observe visited by insects & a cross could take place only by the wind, & as the pollen though pretty plentiful bears no sort of comparison, to the quantity in those dioecious plants, in which the wind is the fertilising agent, crosses could very rarely, as several grains of pollen are probably required, take place without the two individual grew quite close together. This remark is probably applicable to most Graminea; but the social habits of most of the species in the Family makes the difficulty of an occasional cross less than it would be in less social plants./46 v/Water plants are very apt, I think, in proportion to their numbers, to have their sexes in separate flowers: these, also, are very social plants, as remarked by M. Alph. De Candolle. According to all analogy, the division of labour, or in this instance separation

1 [Huc/ Chinese Empire (London, 1855), 312-13.]

2 Philosophical Transactions 1799 p. 200

3 On the Varieties of Wheat. p. 66 &c.

4 De la Dégénération 1837. p. 77

5 Ib. p. 81


of sexes, is advantageous to all living beings, & therefore it may be that water plants can safely partake of this advantage, because they grow nearer each other, & therefore can be more easily fertilised by pollen brought by the wind or insects.—/

46/It may, perhaps, be objected, that large trees with thousands or tens of thousands of flowers, (like a large bed of the same variety of a plant in a garden with respect to another variety) could hardly ever be crossed with the pollen of another distinct individual;—that crossing between the several flowers on the same tree at best would be like the crossing of near relations in animals,— & this, I think is a valid objection. But on the other hand it is a curious fact, <which I have heard remarked on by Botanists, & which will strike> that if any one will turn over a Synopsis of the Vegetable Kingdom/47/on the Linnean system, he will find that the Monoecious, Dioecious, & Polygamous classes include a surprising number of trees,—that is that trees are apt to have their sexes separated. Now it is obvious that in flowers, which can be fertilised only by the pollen from another flower, there will be a better chance, (whether the pollen be habitually brought by wind or insects) that pollen should be brought from a quite distinct individual, than in the case of a hermaphrodite flower having its own pollen close at hand.—/47 v/<Let any one run over in his mind the trees even in our own small island, & he will find many in this predicament; & even some that are hermaphrodite, I have reason to believe are according to Sprengel dichogamous./47/Moreover trees are very apt to grow together or to be social as may be inferred from the much greater frequency of forest-clad-land, than of single scattered trees: This relation of sociability may not be so fanciful as it at first seems:—single trees would interbreed & would produce seedlings not so well able to struggle with surrounding vegetation, as the crossed offspring of the same species, & therefore the species might be able to take root & grow only where several individuals existed. I am aware that there are very numerous exceptions to the above remark that trees have their sexes in separate flowers; but yet the above coincidence of trees being so often mono or dioicous under our present point of view seems worth notice./

47 a/To test the foregoing remark a little further, I find that in Great Britain there are 82 indigenous trees1 of these 19 or more than half (5.93) have their sexes separated,—an enormous proportion compared with the remainder of the British Flora: nor

1 I have taken the 4th Edit, of the London Catalogue as my guide for the indigenous trees, & Loudon's Encyclop. to distinguish trees from bushes.


is this wholly owing to a chance coincidence in some one Family having many trees & having a tendency to separated sexes: for the 32 trees belong to nine Families, & the trees with separate sexes belong to five Families. This result, as far as the number of species of trees with separated sexes would have been greater had I included all the tall dioicous willows, but I have counted only half-a-dozen willows in the thirty-two.1 Remembering that Dr. Hooker2 had observed that the very peculiar Flora of New Zealand was characterised by the number of its trees, & by the number of the plants with more or less separated sexes; I thought the foregoing relation might here be thus well tested: hence I applied to Dr. Hooker, who, not remembering his former results, & as this/47 b/subject is open to doubt under several points of view, has gone over his materials & thinks the following a fair result. There are about 756 phanerogamous plants; & of them no less than 108 are trees. Of the 108, fifty-two or very nearly half, have the sexes separated: of shrubs, there are 149, & of these 61 or considerably less than a third have the sexes separated: of herbaceous plants there are 500, & of these only 121, or not one-fourth have sexes separated. So that we have here the same relation as in Great Britain, with Shrubs shown to be in an intermediate condition. In this case, also, the trees are not confined to some one or two Families, which chanced to have their sexes separated, for these 108 trees belong to no less than 38 Families, & the 52 trees with sexes separated belong to 18 Families, or exactly half. Whether or not, in the above record the trees which have not their sexes separated may be dichogamous in C. C. Sprengel's sense, I do not here consider./

47 bis/Some water plants seem to flower always under water with their corolla perfectly closed: if this could be shown to be invariably the case in any species, it would demonstrate that a cross with another individual could never take place. All British Botanists describe the rare Subularia aquatica as flowering under water with the corolla perfectly closed: Prof. Dickie is the only Botanist, whom I know to have examined it often, & he informs me that he has invariably found it near Aberdeen submerged, with the corolla closed, with fully developed anthers & plenty of

1 [Here Darwin pencilled: 'Dr Asa Gray', presumably in connection with the statistics on separation of sexes in trees in a letter by Asa Gray dated 1 June 1857, part of which is mounted as ULC vol. 8, fol. 47bA. This letter amplifies statistics given in: Asa Gray 'Statistics of the Flora of the Northern United States', Amer. Journ. Sci. and Arts, vol. XXIII no. 69, ART XXXVII, p. 400, May 1857. Darwin's copy is ULC, vol. 135(3).]

2 Introduction to the Flora of New Zealand. p. xxviii


seed in Autumn: but in Germany Koch1 expressly states that 'sub aqua clandestine floret, extra aquam flores parvi albi explicantur.'—The same thing happens with several other marsh plants; thus Limosella aquatica which in this country generally flowers in the open air, was seen by Dr. Hooker in Kerguelen land flowering with closed corolla under the ice.—The Menyanthes trifoliata is hardly a parallel case, for it is not said to flower under water, but on account of the very humid situations in which it grows, it has been asserted2 /47 tres/to shed in Russia its pollen & be fertilised with the flower closed; but in Staffordshire I found that this was by no means the case. A more curious instance is offered by Podostemon, some species of which Dr. Hooker3 informs me flower under water with their corollas closed, carpeting the rocky beds of the torrents of the Khasia mountains in Bengal. The species referred to are annual, & appear only in the rainy season when the torrents are swollen, & Dr. Hooker has never seen them flowering in the open air; but he will not assert that this may not sometimes occur, when the torrents sink. Some Podostemaceous species raise their caulescent stems above water, when they flower; & some few species are monoicous or dioicous, & it is not known whether the pollen in these latter species is carried under water from flower to flower, or whether they are fertilised above water. So that until the natural history of the Family is more thoroughly worked out, this case is not quite so fatal to the views here advocated as it at first appears. There are several other water plants, belonging to the Naiadaceae & allied Families,/47(4)/which seem to offer much difficulty to an occasional cross; but in most of them, the manner of fertilisation is imperfectly known, & several of them are monoicous or dioicous, & therefore it would seem that there must be some means of conveying the pollen under water, from flower to flower.4 /

48/The following appears a strong case against my doctrine: M. Auguste Saint-Hilaire5 states that in Goodenia the pollen is shed in the bud, & then becomes enclosed in a cup surrounding

1 Synopsis Florae Germ. Edit. 2. p. 73. I am indebted for this reference to Mr Babington & to Mr H. C. Watson

2 M. Gillibert. Act. Acad. St. Petersburg. 1777. p. 45. [on verso of fol. 47 bis, Darwin scribbled (& later cancelled) the following: 'Subularia: one of the strongest cases. See J. E. Smith & others; I long thought a case of clear self impregnation. Babington like Limosella— Menyanthes Podostemon. Zostera. Lindley Veg. Kingdom.']

3 Himalayan Journal vol. 2. p. 314

4 This is the conclusion of P. Cavolini in regard to Zostera oceanica, & of Willdenow in regard to Najas etc., see Annals of Botany vol. 2. p. 43. 1806.—It seems to be now made out that Ruppia maritima rises to the surface to flower.—

5 Lecons de Botanique p. 572


the stigma & is then hermetically sealed; so that here a cross would appear physically impossible. But I observe that R. Brown1 speaks of the cup enclosing the pollen till the stigma is ready. & Ch. Morren2 speaks of the cup as being excitable, & 'qui se ferme apres avoir recu quelques grains de pollen;' therefore I infer it may open itself again—As the cup seems to be analogous with the collecting hairs in the Campanulaceae & Lobeliaceae (to which Families the Goodenia is allied) one must doubt whether the cup would act in so opposite a manner as the collecting hairs./

49/The following is a somewhat different case: Fabricius & Sprengel3 have shown that Flies are necessary for the fertilisation of Aristolochia clematitis; but they believe, that when a Fly once enters the tubular flower, it is imprisoned for life by the thick set hairs on the inside of the corolla: if this be so a cross with another individual could never be effected. But having been myself deceived in a somewhat parallel case I am sceptical on this subject: in the common Arum maculatum, I found in some flowers from 30 to 60 midges & minute Diptera of three species, & as many were lying dead at the bottom, & as the filaments on the spadix above the anthers seemed to offer some difficulty to their escape, I concluded that after once entering a flower they probably never left it. To try this I quietly tied gauze over a flower & came back in an hour's time, when I found that several had crawled out of the spathe & were in the gauze: I then gathered a flower & breathed hard into it several times, soon several very minute Flies crawled out/4 50/dusted all over, even to their wings, with pollen, & flew away; three of them I distinctly saw fly to another arum about a yard off; they alighted on the spathe & then suddenly flew down into the flower. I opened this flower & found that not a single anther had burst, but at the bottom of the spathe, near to but not on the stigmas, I found a few grains of pollen, which must unquestionably have been brought by the above or other midges from another individual arum. I may mention that in some other arums which had their anthers burst I saw these midges crawling over the stigmas & leave pollen on them.

<I have given all the facts, which I have been able to collect, which seem to be opposed to the doctrine of occasional crossing

1 Appendix to Flinders Voyage p. 560. [Brown's statement refers to the order Goodenoviae, to which Goodenia belongs; seep. 561.]

2 Nouveaux Mem. de l'Acad. Roy. de Bruxelles Tom XI. 1838 p. 4

3 Das Entdeckte &c. p. 418 [See cols. 424-5.]

4 [Darwin revised the top of fol. 50 extensively, but the resulting text is fragmentary and obscure. The original unrevised version is given here. For the cancellations and scribbled additions of his revision see appendix.]


in more detail, than those which seem to favour it. And there still remain, three cases, viz Hollyocks, certain Orchideae & the Leguminosae.>

Hollyocks (Alcea). Loudon, Herbert & others have stated that the several differently coloured varieties come true from seed. As from the observation of Kolreuter & Sprengel there can be no doubt that the stigmata/51/are fertilised by the coherent pollen of younger flowers, by the agency of Bees, which I have actually witnessed myself in a carefully castrated flower; so this asserted trueness of the many varieties seemed to me very surprising. Hence I brought 18 packets of the best German seed, & raised 18 little beds of plants; but though generally very true, there were seven beds with one or more plants false; altogether out [of] 111 plants.85 came up quite true & 26 not true to their colour. Now if the seed-beds were, as is probable, large, from which it would follow that generally each flower would have pollen brought to it from the same variety, there is nothing in this proportion (even if we attribute, as we ought, some of the false plants to variability,) to cast doubt on the crossing of Hollyocks./

52/Orchidaceae: that in very many genera of this Family, the agency of insects is necessary for their fertilisation cannot be doubted, & therefore an occasional cross from another individual is probable.—Mr. R. Brown believes in this necessity, but adds that all the capsules of a dense spike not infrequently producing seeds, seems hardly reconcileable with impregnation by insects1 I will therefore give a few facts to show how efficient insects are in the Family. It is known that in Orchis, Gymnadenia, <Habenaria> & Listera the pollen-masses cannot be shaken out of their pouches, & can be drawn out only by something touching the sticky gland; yet in a plant of Orchis maculata with 44 flowers open, twelve beneath the buds had neither pollen-masses removed, but everyone of the 32 lower flowers had one or generally both removed: in a stem of Listera ovata, every one of the 17 lowest flowers had pollen on the stigmatic surface: in Gymnadenia conopsea with 54 open flowers, 52 flowers had their pollen masses removed; in another plant with 45 open flowers, 41 had been visited by insects; in another individual I found three pollen masses on one stigma. Four small plants of Orchis Morio grew in my orchard; I covered one with bell-glass;/53/the other three plants had 23 quite or partially opened flowers & day after day I found some of the pollen masses disappearing till all were gone with the exception of one single flower which withered with the pollen-masses in their

1 Linnean Transacts. vol. 16, p. 704


pouch: but one or two terminal flowers, in each plant not included in the 23, & which opened subsequently never had their pollen masses removed. I then looked at the plants under the bell-glass & found not one single pollen-mass removed; & though then left uncovered every flower withered in the course of six days with all pollen-masses in their pouches & the germens did not swell. From this fact, I infer that whatever nocturnal insect (for I never saw an insect visit the plants by day) haunts this orchis had ceased its visits, as indeed might be inferred from the extreme terminal flowers of the three plants which had never been covered, retaining their pollen-masses.

I have repeatedly seen in Listera ovata, Gymnadenia conopsea, Habenaria bifolia & Orchis morio, plenty of pollen on the stigmatic surface, but with pollen-masses of the same flower in their pouches; & still oftener the reverse case, namely the pollen-masses removed, but no pollen on the stigmas, — which clearly shows that each flower in these species is very generally fertilised not by its own pollen, but by that of another flower or individual. After having/ 54/attended to this subject at intervals during several years I have seen no insect visit an Orchid, except once a Butterfly sucking an Orchis pyramidalis & once a Gymnadenia conopsea; but Sprengel1 has been more fortunate for several times he saw a Hymenopterous insect visiting Listera ovata, & he saw the pollen-masses removed & the pollen left on the stigma by these insects: on Epipactis latifolia, also, he saw a Fly with the pollen adhering to its back. I do not doubt that usually moths are the agents for fertilisation; & I must think in the Butterfly orchis2 (Habenaria) the white-coloured flower, the sweet smell at night, the abundant nectar contained in a nectary with which only a tube as fine as needle can be inserted all stand in direct & beautiful relation to the visits of nocturnal Lepidoptera.—3

It is well known that in certain exotic Orchidaceous plants, parts of the flower have the power of movement, when irritated. In Mormodes the pollen-masses are jerked out with such force as sometimes to hit a person's face; & I was told by Mr. Loddiges that he thought not one in a hundred would miss hitting the stigmatic surface: but I am not able to say what the result would be on the chances of two individuals crossing in this case, & in that of those Australian genera4 /55/in which the labellum when

1 Das Entdeckte &c. s. 409, 415

2 Das Entdeckte s. 405

3 [Scribbled addition:] Entomologists have been often puzzled by finding their glass sticking to & flower feeding Beetles.

4 Lindley, Vegetable Kingdom. 1853. p. 179


touched by an insect suddenly turns round & shutting up a box-like cavity, imprisons the insect./

56/We now come to a case in which it appears, though the flower is open, that there is a direct mechanical provision for perpetual self-fertilisation: in certain species of Ophrys, R. Brown1 has shown that the pollen-masses readily fall out of their pouches, but being retained by their glands, & the stalk being of the proper length, they swing downwards, strike on & adhere to the stigmatic surface: hence insects as Mr. Brown remarks are not at all necessary for their fertilisation: to test this I covered up under a case of gauze some plants of Ophrys apifera, so that no insect could visit them or the wind agitate them, yet in every flower I found the pollen masses fallen on the stigmas. Again during three years I have examined many plants, one day looking at every flower in 18 plants, of this Ophrys, & I have never found the pollen-masses removed or pollen on the stigma of a flower excepting its own proper pollen. Hence I should have concluded that this was/57/ certainly a case of perpetual self-fertilisation; had it not been, firstly, that the sticky glands are here present, & if insects did ever visit this flower2 a cross might readily be effected, & if they never do why are the glands sticky? Secondly in Ophrys muscifera, the pollen-masses cannot be shaken out, as I have repeatedly tried; & therefore the agency of insects is required as in the other Orchidaceous genera for their removal & apposition of the stigma; but upon examining 102 fully expanded flowers, during different years, I found in this number that only in 13 flowers had one or both the pollen masses been removed; in the other 89 flowers (most of them withered) the pollen-masses were still in their pouches. Hence we see that in Ophrys muscifera in the district in which I live3 the agency for the ordinary fertilisation of the plant is far less effective than in other orchids & it may be that in Ophrys apifera the less important agency for an occasional cross is here likewise highly defective;—consequently that both species are here living under conditions unfavourable in one respect, but so favourable in some other, that they are able to survive;—nor need we be surprised at this, as there are many cases of plants

1 Linnean Transact. vol. 16, p. 739

2 Mr. Brown suspects that the flowers of Ophrys resemble insects in order to deter other insects visiting them. But I cannot avoid feeling very sceptical on this head. As we shall immediately see in Ophrys muscifera the agency of insects seems requisite.

3 In Spandow in Germany, Sprengel found that in Orchis militaris, of 138 flowers, only 31 had seed-capsules; & he attributes this to deficient fertilisation, & contrasts it with the case of Gymnadenia conopsea in which nearly all the germens had set.


living in a country, in which they seldom or never are known to seed. But as seeding is the normal condition with these very species in other countries or times; so may an occasional cross possibly be the normal condition with Ophrys apifera./

58/Leguminosae. We now come to our last & <perhaps> most difficult case. The stamens & pistil are here beautifully enclosed within the keel shut up as in a bivalve shell; & as the pollen is shed in profusion at an early period, I am not surprised that Pallas & some other authors have advanced this great order as an instance in which a hybrid, could never be naturally formed. Yet if I trusted only to Sprengel's observations on the action of insects & to my own after having attended especially to these flowers during several years, I should have inferred that they could not have escaped frequent crosses, between individuals of the same or another variety. The flowers in this Family are especially frequented by Bees, & I have seen on them certain flies, butterflies & the minute winged Thrips, all covered by pollen. It is really beautiful to see what takes place, when a large bee alights on the wing-petals of we will say a common bean; how its weight depresses the wing-petals & with them the keel, by which the rectangularly bent pistil & already shed pollen are forced out & rubbed against the hairy body of the Bee, as it visits flower after flower. In many Leguminosae the hairs beneath the stigma act in the prettiest manner to brush out the pollen in/59/masses against the bee. Even such very minute flowers as those of the yellow1 clover (Trifolium minus) are visited by Bees & the keel in them is generally split open: in Coronella after a hot day, I have seen the keel open of itself. But before anyone comes to a conclusion on the part which insects play in the fertilisation of the Leguminosae, long observation is required: for weeks together a Bee will not be seen even to look at a certain species, & then that species will suddenly be visited by thousands: Bees can suck the nectar as I have seen in the common Pea, without moving in the least the stamen & pistil; but then again I have seen at another time a Bee whilst sucking this flower force out the pollen in profusion & get its under surface well dusted against which the stigma was rubbed. Other Bees will visit the already fertilised flower & collect the old pollen. Other Bees frequently bite holes at the bottom of the calyx & corolla & so get the nectar, without aiding in any way its fertilisation <performing what I believe is their proper function> whilst Humble Bees are thus robbing the flower of the nectar hive-bees may be

1 [Here between the lines on the MS. there occurs the pencilled comment: 'No'.]


collecting its pollen./60/But the case which convinces me that there is a direct relation between the structure of papilionaceous flowers & the agency of insects, is that of the Kidney Bean, (Phaseolus) which it is worth any one's while to notice: the tubular keel, with the included pistil & stamens, is here curled like a french-horn & has its little open end directed to the right side: when a Humble-bee alights on the wing-petals, the tubular keel is so acted on that the pistil is protruded & the hairs on it brush out quantities of pollen, & the pollen & stigma are rubbed against the bee's side. Now I have noticed (which was overlooked by Sprengel) that the nectary is so placed as to induce both humble & hive bees invariably to alight on that side towards which the pistil is protruded. And that this is not a mere chance relation may be inferred from the structure of Lathyrus grandiflorus,1 in which the keel, though not actually spiral is distorted towards one side, & again it is on this side that bees are induced invariably to alight, & in so alighting they cause the pollen to be protruded against them.

But now let us see what direct evidence we have of the crossing of our many cultivated leguminous varieties. A. F. Wiegmann2 asserts that by merely planting together varieties of Phaseolus/61/Vicia, Pisum & Ervum, he procured various hybrids, & that the seeds in the pure <female> parent were affected by the pollen of the other varieties; as some of these crosses were bigeneric, I should not have even alluded to these statements; had not the accurate Gaertner, a most hostile witness,3 after most careful experiments in artificially crossing the varieties of Peas, come unwillingly to the conclusion that the pollen of one variety does sometime affect the seed of the castrated female plant, in the same way as happened with Wiegmann's plant, when left spontaneously to cross with each other. The only possible error in Gaertners experiments, which I can see is that it might have been the act of castration & not the pollen of the other varieties which affected the colour of the peas in the artificially fertilised pods. Certainly in some varieties, as I have witnessed (but Gaertner selected the most constant) the colour of the pea is extremely apt

1 A writer in Loudon's Gardener's Magazine (vol. 8. 1832 p. 50) [Letter signed G.C.] says that having observed that this plant never set its pods, by moving the keel & so causing the stigma & anther to protrude, he found that the greater number of the flowers, thus treated, formed pods. But this does not always succeed ([Godsall] Ib. p. 733), as may perhaps be accounted for by our climate being unfavourable to this plant. I should mention that Bees seem to visit this exotic plant only during certain seasons.

2 Ueber die Bastarderzeugung 1828

3 Bastarderzeugung s. 89


to vary.—I was led by their statements to apply to Mr. Masters of Canterbury, a great raiser of pea-seed & the author of an article on this subject, & he/62/answered me that undoubtedly some varieties of Peas & Beans occasionally become crossed with other varieties, but that he had never known a whole crop deteriorated./ 62 v/Again in Mr. Sharps great seed nurseries1 it is said that Peas are grown very extensively & as they are considered liable to be adulterated, 'considerable precautions are employed to secure separation.'/62/But in these cases, I must remark that it must always be very difficult to distinguish in close varieties between the effects of a cross & of simple variation. Lastly it is incidentally asserted in the Memoirs of the Board of Agriculture of New York2 that the varieties of the Kidney-bean easily cross with each other when grown together.

But now let us look to the evidence on the other side. A. Knight3 castrated several pea-flowers; on some he put pollen of other varieties, & some he left without any; & these latter did not set, showing that no pollen was brought to them by bees. Secondly I applied to Messrs. [ ] great raisers of seed-peas & they do not believe that their varieties cross, & they take no especial precautions to prevent it; & this seems to be the general practice of gardeners. Thirdly a friend had planted during two generations, three varieties of Peas & three of Beans in rows close together all in flower at the same time, & I saw their produce or third/63/ generation & they seemed to run all true; but most of these varieties were closely allied, & between some of them a cross would not easily have been detected. Lastly, (& this case has struck me most) Mr. Cattell of Westerhaven regularly has beds of five varieties of the Sweet Pea, (Lathyrus odoratus) for seed grown close together; these varieties differ in no respect whatever except in colour, they flower at the same time & are frequented by Bees; yet each variety comes up, as I know from experience, true. Here certainly there can be hardly any crossing; probably none whatever; but it would be rash to conclude positively that there was none, for I have noticed sometimes a plant of one variety growing amongst the others, which I have attributed to a stray pea having got into the wrong packet; but possibly such might be the result of a cross4 ;

1 [Anon.] Gardener's Chronicle 1856 p. 823.

2 [J. Armstrong] vol. 2. p. 100

3 Philosoph. Transact. 1799 p. 196

a I have failed in my endeavours to test this, for all the flowers which I castrated, both those on which I put pollen of other varieties & those which I left without any pollen [,] fell off unimpregnated. This difficulty in manipulation is well known to hybridisers, & I presume explains the reason so few hybrids have been formed in this Family; I have heard of only three <two> viz one in the genera Erythrina & two in Cytisus.—


for it is known that in very close varieties differing only in colour, the offspring sometimes are not intermediate but take after either parent: thus Kolreuter1 crossed red Hollyock with the pollen of yellow & the two seedlings were yellow; I crossed a dull purple Hollyock with the pollen of a bright yellow & the seedlings was red. Kolreuter crossed a white one with pollen of red, & the several offspring were red, with one purple./

64/With respect, then, to the Leguminosae, bearing in mind the facts given on their structure in relation to insects; bearing in mind Wiegmann & Mr. Masters & Messrs. Sharp's statements; & on the other hand the opposed facts just given, more especially the case of the Sweet Peas, it is difficult to come to any sure conclusion. But, I think, we may conclude that crosses between individual & individual, if such do occur, can take place but rarely in the Leguminosae; & the facts here given seem to me more strongly opposed to the law, which I am attempting to establish, than any others, at present known to me.—/64 v/We have seen in a former part of this discussion, that forest-trees, when hermaphrodite, offer a difficulty to my notion of general crossing from the simple occurrence of very numerous flowers on the same individual close together. Therefore as the papilionaceous structure alone offers a difficulty, this is much aggravated in forest-trees belonging to the papilionaceous division of the Leguminosae; of which, as I am informed by Mr. Bentham, there [are] a good many in Tropical countries of gigantic size; & of which the Robinia, pseudacacia offers a well-known example./

64/I will now sum up the discussion in this chapter, on the question whether it be a subordinate law in the mysterious act of reproduction that occasionally the concurrence of two distinct individuals is necessary. First for plants, the numerous cases of varieties which are known to cross freely if grown near each other;—the extraordinary precautions which hybridisers unanimously agree are necessary to prevent a castrated plant receiving pollen from another individual, thus obliterating the action of the foreign pollen;—the many cases of dichogamous/65/plants, or those in which the pollen is shed when the stigma is mature <at different times>;—the many cases in which insects are necessary for the fertilisation of plants; & the other cases in which they are not necessary, but in which they are frequently visited by insects, & in which there seems an obvious relation in their structure to the visits of insects,—all tend to show that crosses between

1 Acta. Acad. Petropol. 1782 p 256


individual & individual must at least, be frequent. A camel-hair brush which may be aptly compared with the hairy body of an insect is found useful by hybridisers to bring pollen from flower to flower; but ask any one, if he were to remove the pollen out of one flower with a brush, & use the same brush to bring foreign pollen, whether he could thus make a hybrid, & he will tell you that there would not be slightest chance of success.—

As it is known that protection from rain & damp is favourable to the fertilisation of flowers, it is remarkable how extremely general it is that the act takes place fully exposed. The reported cases of habitual fertilisation within <the bud or any closed chamber> the closed corolla are comparatively very few; & as has been shown are mostly open to some doubt [.] I cannot but suspect that such cases as that of/66/Subularia, Podostemon, Goodenia1 of Ophrys apifera, & even of the Sweet Pea & of other papilionaceous flowers will be modified & explained with the progress of knowledge. How comes it, with the almost infinite modifications of structure in the vegetable kingdom, that no case, as far as I can find out, is known of the anthers bursting actually on the stigma: in Stylidium <Goldfussia> there is a near approach to it, but here there is a wonderful contrivance of self movement & of collecting hairs, of nectariferous organs which I can hardly doubt would favour by the agency of insects an occasional cross: in several <many> of those cases in which the anthers move to the stigma or the stigma to the anthers, insects are requisite to excite the movement, & not only would favour a cross, but, in Mahonias <barberries> at least, crosses do frequently take place. What again is the meaning of the superfluity of pollen in many hermaphrodite flowers? Kolreuter has shown that in Hibiscus2 sixty/67/grains of pollen are sufficient to fertilise all the seeds in a flower, the anthers of which he calculated had 4863 grains of pollen; but Hibiscus though hermaphrodite is a dichogamous plant, & therefore might require a very great excess: in Geum urbanum the pollen is only ten times in excess3 : Gaertner thinks4

1 [In the MS., a question mark within parentheses is pencilled after Goodenia.]

2 Vorläufige Nachricht 1761. s 9. .& the statement is confirmed by Gaertner in his Beitrage zur Kenntnis. s. 346

3 Gaertner Beitrage &c s. 346. [After this note came the words 'in the' followed by a blank space practically large enough for the following slip now numbered fol. 67a reading:] Gardeners' Chron. Nov. 21. 1845. Article on there being 7000 pollen grains to every ovule or seed in Glycine—I mention because <good as> Papilionaccae, as argument for cross impregnation. We recognise use of numerous pollen grains in Zea, why not here? ['Wrong reference' is scribbled in the margin of this slip; the citation should be 1846 vol., p. 771.]

Beitrage zur Kennt. s. 440


that this superfluity of pollen is simply for ensuring the fertilisation of the plant: but on this view it must be admitted that generally flowers have been formed, without any object which we can see, with a structure rendering self-fertilisation so far difficult, that this difficulty is compensated by a great superfluity of so highly wrought an organic product as pollen! On the other hand we can understand the act of fertilisation taking place so generally in open flowers,—the maturity of the pollen & stigma being at different times—the many & very curious relations of structure to the visits of insects—the superfluity of pollen—/68/the presence in closely allied groups of hermaphrodite & <bisex> unisexual plants,—if the occasional concourse of two individuals be a law of nature. From the well-known elective power between various kinds of pollen specifically different & the stigmatic surface, it seems to me not improbable that the pollen of a distinct individual or slight variety may be prepotent over the flowers own pollen; & from the facts given in regard to the greater vigour of the crossed offspring of varieties, I believe that such crosses would have a better chance of surviving in the severe struggle for existence to which all living beings are subjected, than the offspring of self-fertilisation.—Although I believe good results from crossing & that probably the occasional concourse of two individuals is even a law of Nature, yet I come very far from supposing that such is the sole good of the separation of the sexes, (which necessitates a cross each time); for analogy leads to the belief that division of labour, to use Milne Edwards expression, tends to the perfection of every function.1 /

69/Turning to animals, although many are hermaphrodite, we have the remarkable fact that not one single land animal, in which a fortuitous cross is obviously impossible by the same agencies, viz insects & wind that are so efficient with plants, is hermaphrodite in the strict sense of the word or self sufficient. Again amongst aquatic animals, not one case is positively known, in which a perfectly enclosed structure would render a fortuitous cross impossible; & that in aquatic animals fortuitous crosses are not improbable we may infer from the fact that many fixed aquatic animals have separate sexes. These facts, & others identical with those just referred to under plants, as to maturity of the ova & spermatozoa at different times, we can understand if the occasional concourse of two individuals be a necessity; & I think it would be difficult to offer any other explanation.—/

1 [Milne-Edwards, Introduction à la zoologie générale, Paris, 1851, introd., pp. 35, 56–.]


70/If it be asked why the occasional concourse of two individuals should be a law, I think the facts given showing that the crossed offspring of two varieties, & even of two individuals in hermaphrodite plants, have their vigour & fertility increased, afford a sufficient answer. Even hybrids from between distinct species gain in stature & vigour compared with their pure parents; & in some strange cases their fertility which is always deteriorated seems somewhat improved by further crossing./70 v/On the other hand close interbreeding, even in animals with separated sexes in which a cross, between two individuals, is a necessary accompaniment, seems injurious./70/It would appear as if the good from crossing was like that felt by the individual from some slight change in the conditions of its existence. But if it be further asked, why changed conditions should do good to the individual, & why a slight cross should add to the vigour & fertility of the offspring, no answer can be given, or can be expected seeing how utterly ignorant we are in regard to Life & its Reproduction.—

Finally weighing all the evidence as well as I can, I certainly think that it will hereafter/71/be found, that the occasional con-course of two individuals, & these individuals not very closely related, is a subordinate Law in Reproduction.—I have stated in full all the facts opposed to this view, which are known to me, but have not given all those in favour of it. The difficulties many of which as we have seen are grave enough, I must leave to the judgment of the reader./

72/On changes of condition causing lessened fertility or complete sterility.—1 As we have in this chapter so largely discussed the good apparently derived from crossing varieties & individuals, & from slight changes in the conditions of existence, it will be convenient here, also, to discuss the effects of those changes which lessen or quite destroy the fertility of organic beings; though the subject is, I think it will be seen, more intimately related to hybridism than to the points hitherto treated of—

There is a wide difference, as strongly insisted on by Isidore Geoffroy St. Hilaire2 between taming an animal, & getting it to breed in captivity, which alone can be called domesticating it. The one is very easy, but domestication, as the experience of all ages shows, is very difficult. One's first impulse is to attribute

1 [Here in the MS. is pencilled the note: 'All used to p. 102' (i.e. to the end of this chapter.). See Variation, II, 148–72, ch. 18, section headed 'Sterility from changed Conditions of Life' to the end of the chapter.]

[Space for citation left blank. In Variation ch. 18, note 9, Darwin cited: 'Essais de Zoologie Générale, 1841, p. 256.']


the whole difficulty to the sexual instinct being affected, as has often been the explanation with respect to the Elephant in India; & in the case of birds in some instances to a proper place or materials for nidification. This in some instances may be a sufficient explanation,/73/but in very many cases, animals couple but very rarely or even never conceive; & here it cannot be an instinct which fails: moreover we shall find in plants a large parallel series of facts.—

Why many animals taken young, perfectly tamed, quite healthy & living long, should not breed, it is impossible to explain. One must attribute it to some change in the conditions of its existence. Sometimes one may infer that it is not owing to any change of climate, as when captive animals will not breed in their native country; in other cases it would appear not to be caused by want of exercise; in others not by change of food. Perhaps it may be due to these several slight changes combined. Some orders are far more affected than others, without any assignable reason; but it often happens that certain species in the orders usually least affected will not breed; & on the other hand that some species in those orders which are generally most affected will breed. In some cases the animals are never known to couple; in others they do couple but never or most rarely produce young. An apparently very slight change in the condition of existence has sometimes caused an animal to breed, which had never done so before./

74/I will now give some facts. My materials are derived from scattered notices; from an M.S. report from the Zoological Garden, between the years 1838 & 1846 inclusive, of all the animals which were seen to couple & of those which produced young; from subsequently published Reports, & from inquiries which I made from the keeper of the Birds at the Surrey Zoological Gardens, I should premise that I have no doubt that under very slightly different management, in other menageries, the results would be somewhat different; & that in the long course of years individuals of the least fruitful species would be found to produce young under the same treatment which rendered all other individuals sterile.1

First for the most notorious case of the Elephant, in its native

1 [The following dubiously legible pencilled comment occurs on the verso of fol. 74:] I lay particular stress on animals not breeding when thoroughly tamed & left considerable liberty in their own country—In menageries very many do not breed, or breed rarely & produce few young.—There it must be at part [?] through not ranging & attributable to ill-health, but some [?] live long & others suddenly double away like sheep as highly fertile—We shall now see that the lessened fertility runs in classes without any apparent rule—Instincts.


country of India, though kept in great numbers in perfect health, has with one or two exceptions, been never known to couple; but if we go1 [a little eastward to Ava, we hear from Mr. Crawfurd2 that their -breeding in the domestic state, or at least in the half-domestic state in which the female elephants are generally kept, is of everyday occurrence;"and Mr. Crawfurd informs me that he believes that the difference must be attributed solely to the females being allowed to roam the forests with some degree of freedom. The captive rhinoceros, on the other hand, seems from Bishop Heber's account3 to breed in India far more readily than the elephant]/75/In captivity. Four wild species of the Horse genus, have been bred in Europe, but generally one species with another already [?] hybrid here; though the conditions of their existence must be very different from those of their native desert home.— Most wild species of the Pig breed readily; & the Peccary [Dicotyles torquatus] has bred in the Zoological Gardens; but this animal, in its // [species, the D. labiatus, though rendered so tame as to be half-domesticated, breeds so rarely in its native country of Paraguay, that according to Rengger4 the fact requires confirmation.] 76/The carnivora generally breed nearly, or quite as freely, as the Ruminants in captivity, but the plantigrade division must be excepted. Bears of several species couple most freely in the Zoological Gardens but <with the exception of the cinnamon bear, have never bred> have bred only thrice. I have heard of the Badgers having bred twice, once in Germany5 & once in the Zoological Gardens; I suppose it must be very rare in Germany, as the fact was published. The Cuati or Nasua in its native country of Paraguay, though kept in pairs for many years, & perfectly tamed has never been known to breed there, or to show any sexual passion.6 So according to this same author it has been thus with two other plantigrades, Procyon or Raccoon, & the Gulo: these three genera, have been kept in the Zoological gardens, & the two former have been known to couple, but have never bred.— In the Dog-Family of the Carnivora, it is very different, as most breed, but it has very rarely taken place with Foxes & Jackalls.—In the Cat Family, breeding is likewise very general; but even here they couple far more freely than conceive; in the M.S. return

1 [The MS. is cut up here. The missing portions of text are supplied from Variation, II, ch. 150, 18, portion relating to notes 13–15, which are quoted as notes 2–4.]

2 ['"Embassy to the Court of Ava", vol. i. p. 534.']

3 ["'Journal", vol. i. p. 213.' (i.e. Journey through India)]

4 ["'Saugethiere", s. 327.']

5 [Sicmuszowa-Pietruski] Wiegmann's Archiv. fur Naturgesch. 1837. p. 162

6 Rengger. ib. p. 106


from the Zoological Gardens for eight years the coupling was noticed between various species 73 times, but young were produced only 15 times. It is remarkable that on a change of treatment with the Carnivora at these Gardens, & when they were freely exposed in open cages to a much colder temperature, they were found to breed very much/77/more freely. I have never been able to hear of the Tiger, though known to couple, breeding in India: nor does the hunting Leopard or Chetah; but [in] this latter case pains may have been taken to prevent their breeding, as animals which have hunted in a state of nature are alone worth taming.1 Every one knows under what unnatural conditions, shut up in a small cage, the Ferret breeds; & even the otter has once bred in the Zoological Gardens; whereas the Herpestes griseus, though many have been kept in the gardens, & some species of Viverra & Paradoxurus have never bred there.

In regard to Rodents, the Rabbit breeds most freely in wretched little hutches, <as does the Guinea Pig,> where the common Hare, though it has many times been tamed, most rarely will breed. Some few Rodents as the Chinchilla, some mice, a porcupine, a Lemming have bred in the Gardens; some have coupled & never bred & some have done neither. To give one example no Squirrel, has ever bred, though the Sciurus cinereus has been known to couple, & as many as fourteen of the S. palmarum have been kept together. Nor have I ever heard of the English squirrel breeding in captivity. What a strange contrast to the free breeding of the rabbit, guinea-pig & white mice !

Lastly in regard to the many species of Monkeys; most couple freely, but during the eight years, of which I had a return, there2 … [Monkeys, in the nine-year Report from the Zoological Gardens, are stated to unite most freely, but during this period, though many individuals were kept, there were only seven bir ths.]…

78/Birds. We have seen that the Carnivora, with the exception of the plantigrades, breed pretty freely in captivity; but the case is very different with Hawks. It is said//3 [that as many as eighteen species have been used in Europe for hawking, and several others in Persia and India;4 they have been kept in their native country

1 Sleemans Rambles in India Vol. 2. p. 10

2 [MS. fol. 78 has been cut up to leave only a narrow remnant. The preceding and following text is pieced out from Variation, II, 153, ch. 18, whose notes 28–30 follow as the next three notes.]

3 [' "Encyclop. of Rural Sports", p. 691.']

4 ['According to Sir A. Burnes ("Cabool", &c, p. 51), eight species are used for hawking in Scinde.']


in the finest condition, and have been flown during six, eight, or nine years;1 yet there is no record of their having ever produced young.]…//79/African, American & Australian Ostriches have often bred in confinement: yet what a change in habits, climate & nature of food they must have suffered!—Most Gallinaceous birds brought from all quarters of the world, breed very freely. We see what an astonishing change the Guinea-fowl, from the dry deserts of Africa; & the Peacock from the jungle of India have undergone, & yet breed freely. At Lord Derby's some Ortyges, Grouse, & even Partridges have bred. The Capercailzie has bred in the Regents Park; but in Sweden it has been found2 [ ] that the [ ] grouse would not breed without the birds were kept in a space, though small one, of enclosed wood. On the other hand it is well known that Partridges will not breed in captivity; but one case is recorded of the red-legged partridge having bred3 when kept in a large court with other birds./

80/Pigeons, again, breed much more readily than most birds in confinement: in the return from the Regents Park for the eight years, thirteen species bred, & only two were seen to couple with no result.—Both the magnificent crowned Pigeons have bred in the Gardens; but Mr. Crawfurd informs me that nearly fifty birds were kept in a pleasure ground for several years in Prince Edward Island, in a climate one would have thought admirably adapted to them, & that they never bred.

Parrots, of which such numbers are kept & which have often lived to such extraordinary ages, showing that they are healthy, breed so rarely that paragraphs in the newspapers4 are sometimes inserted when such occurs: in the Regents Park, & in the Surrey Zoological Gardens some few species couple, but I believe the Australian Euphema pulchella is the only species which has ever produced fertile eggs: Sir R. Schomburgk says5 that Parrots kept tame & loose in Guyana do not breed.—What a singular & inexplicable contrast is thus presented by Parrots with Pigeons.—

Of the small birds or insessores, several as the linnet, Goldfinch, Siskin &c are known freely to breed with the canary bred in confinement; but very many others, as the Bull-finch have with the exception of one or two crosses with the Canary, have never

1 [Hoy, 'Loudon's "Mag. of Nat. Hist." , vol. vi., 1833, p. 110.']

2 [Darwin left an unfilled space for the reference here.]

3 [Defay] Journal de Physique Tom. 25. p. 294.

4 [Denny] Athenaeum 1843. p. 829. Brit. Assoc. Report [for 1843, part 2, p. 71.]

5 [Darwin left an unfilled space for the citation here. Variation, ch. 18 at note 42 indicates this information was given in a personal communication from Schomburgk to Darwin.]


been known/81/to breed. Though Larks, (Alauda) of four species are kept in numbers, & I have known of some which lived in a large aviary for seven years; yet none, as I have been assured by a great Bird Fancier, here in their native country have ever been known to breed. In the 8 year returns from the Zoological Gardens, I have particulars of 24 confined species which have never bred, & of which only four have been known to couple.

Waders or Grallatores, as a class, seem eminently sterile in captivity; but many of them are short-lived in this state, so that the fact is not so remarkable as it would otherwise be.—I have heard only of three breeding: namely a Water-Hen (Gallinula chloropus) in the Regents Park; a Crane (Scops paradisea) at Lord Derby, & Grus antigone at Calcutta.1

The great Duck Family, Anatidae, seems the most fertile of all, apparently more so than even the Gallinaceous birds or Pigeons; yet one would have thought that their conditions of existence, considering their aquatic & generally wandering habits & insect food, would have been singularly affected by confinement. Between 20 & 30 species have bred in the Zoological Gardens. On the other hand, Sir R. Schomburgk2 says/82/that he has never heard of the Dendrocygna viduata, though easily tamed & frequently kept by the Indians of Guyana, breeding. Lastly with respect to Gulls (Larus) <& Pelicans>, though kept in numbers in their native country, in the Regents Park & Surrey gardens, are never known to couple or to breed, with the exception of the Herring Gull in the 1850–51, in the Regents Park. But their condition of existence & food, it might have been thought, would have been not more unnatural than with marine Ducks in confinement. Insects seem to suffer in their fertility like the larger animals. [The bottom half of this folio is blank.]

83/I have been informed in the <Regents Park> Zoological Gardens, that even those Mammals & Birds, which do breed in confinement very rarely breed for the first year or two. The secondary male characters seem sometimes to be affected, as in the <case of the crimson breast of the Cock Linnet> loss of the brilliant colours of many cock birds under confinement.3 The young are apt to be born dead or to die immediately,—of whichfact Rengger gives several instances in Paraguay: the flow of milk is often checked;, which all shows disturbance in the repro-

1 [J. E. Gray] The Knowsley Menageries 1846. PI. xiv.—Mr. Blyth Report Asiatic Soc. of Bengal, May. 1855.

2 Geograph. Journal vol. XIII. 1844. p. 32

3 Bronn Ges[ch]ichte der Natur B.2. s. 96: [Barrington] Phil. Transact. 1722. p. 278.


ductive functions. I have fancied that even the strangely perverted maternal instinct, so frequently leading animals in confinement to devour their new-born young, may likewise be connected with the same general disturbance.

Considering all the facts which I have been able to collect, most of which I have given, it seems impossible to come [to] any more definite conclusion, than that captivity has an especially injurious influence on the reproductive system; & more injurious in some orders than in others, but with many exceptions in every case. Generally, the cause can hardly be change of food, for the difference in the effect produced by captivity is vast, when we compare/84/ carnivorous mammals & birds; nor can it be generally want of exercise, when we consider for instance the case of Ostrich tribe, so cooped up in confinement, & ranging so widely in their natural state: nor can it be generally change of climate, when we see captive animals so frequently sterile in their own climate.

The case of domestic animals, perhaps, is hardly appropriate with respect to climate, as it may be said that their constitutions are enured to change; but it is remarkable that those Dogs, as the Bull-Dog, which degenerate in India, yet breed freely there as I am informed by Dr Falconer, as do likewise, according to Dr Daniel dogs imported from Britain into Sierra Leone. From the latter country, I have received owing to the kindness of Dr. Daniell, Poultry & Pigeons, & though brought here in Autumn & so exposed to a great change of climate the males were ready at once to procreate their kind. Rabbits breed pretty well in India. The only instance of the fertility of domestic animals having been affected of which I have heard, that of Geese & Poultry given by Roulin when first imported into Bolivia:1 Dr. Falconer, also, informs me that the eggs of Turkeys in the hot & dry province of Delhi are extremely apt to be infertile: Geese,/85/as I am informed by Mr. Crawfurd, do not lay at Manilla.2 Lastly we cannot generally account for the infertility of animals in captivity by the want of health, for many of them live to old age; & in the case Hawks, used for Hawking, must have been in robust health. Moreover the diseases of which animals die in menageries, (& numerous postmortem examinations of the cases in the Zoological gardens have been published in the Veterinary Journal), are chiefly inflammations of the internal viscera & membranes, & tubercular cases. Such diseases are known in mankind not to affect the

1 Bronn Ges[ch]ichte B.2. p. 100.

2 [See A descriptive dictionary of the Indian islands and adjacent countries, London, 1856, p. 145.]


reproductive system. Of all domestic animals, the sheep, perhaps, is the most subject to disease, yet it is very fertile. In captive animals, the reproductive organs, do not appear to be diseased; but their proper function is often most gravely interfered with. The case seems quite an especial one: I do not know if there are instances of any other organs, not diseased, yet not performing their function. We can attribute this deficient action only to general constitutional derangement./

86/Plants In the vegetable kingdom there is a large class of facts in regard to sterility analogous with those in the animal kingdom. But the subject is here much obscured by several considerations. It is notorious that very many plants in hot-houses & in our gardens, though living in apparently the most perfect health, & often more vigorous than in their native habitat, never produce seed. I do not allude to the cases in which the seed-pod, for want of heat or other causes does not ripen, (though this may be analogous to the frequent births of dead offspring in menageries) but to those cases in which the ovules, as far as we can judge, are not fertilised. Many productions of the temperate region, for instance most of our fruit trees, when grown in tropical countries do not flower; so it sometimes is with plants in our own country when treated with an excess of manure or kept too hot & damp in greenhouses: but it seems very doubtful whether such cases come under our present subject, for here the reproductive individual is not produced, & therefore cannot be classed as sterile./86 v/To check over luxuriance, gardeners in India mutilate in the oddest way European plants which they cultivate./86/But there are many foreign plants in our gardens, which do not seem injured by our climate, in which the pollen seems perfectly good, & in which the pistil seems perfectly formed, which nevertheless never or most rarely set their seeds. These cases seem analogous to those/ 87/of captive animals, in which the reproductive system seems far more sensitive to change than any other part of the organisation. Linnaeus long ago remarked1 that alpine plants when cultivated in gardens, though in their natural site loaded with seed, produce very few or entirely abort; but with care, & planted in favourable situations some will produce abundant seed, as in the case of Draba sylvestris, "one of our most thoroughly alpine plants"2

1 Swedish Acts. vol. I 1739 p. 3. [See Linné Versüch von Pflanzung der Gewächse…'

Abhandlungen aus der Naturlehre 1 (1749), 13.]—Pallas makes the same remark in his travels vol. i. p. 292.

2 Cybele Britannica vol 1. p. 131 [Watson's statement here is about D. rupestris.]


which multiplies itself by seed in Mr. H. C. Watson's garden. Zuccarini has remarked1 that scarcely any of the genus Oxalis from the Cape of Good Hope will seed in Europe./

87 v/In the genus Syringa, which seems perfectly hardy in our climate, I cannot hear that the Persian or Chinese Lilac ever set their seed; & I find that their pollen in water does not swell like that of common Lilac, which does produce (I do not know whether always) seed, which I have found to germinate: whereas in Germany, Gaertner instances2 the common Lilac, as never producing seed, though having well‐formed seed‐capsules, in the same manner as many quite sterile hybrids here./87/Many hardy liliaceous plants are quite sterile in our gardens; as are many Bog plants: Numberless instances could be given: but in some of these cases the subject is much obscured by what Gaertner has called contabescence, namely the abortion of the anthers, which in some cases at least seems to be in no way connected with any change of conditions; but I shall have to return to this subject.

There are several cases on record, as in Lobelia, Passiflora, Gladiolus, Lilium candidum &c, of plants having good pollen, as known by its fertilising/87 bis/other plants, but in which the female organ either cannot be fertilised anyway, or only by pollen of another individual or other species:3 some of these may be special cases, like those of the contabescent anthers, but as they generally occur in exotic genera, they are probably due to some‐thing unfavourable in the conditions of <existence> the cultivated plants. Pollen, when once in process of formation does not appear easily injured; a plant may be transplanted or a branch may [be] gathered with flowers in early bud, & <if> placed in water the pollen will be perfectly matured. But the female organs seem much more sensitive, for Gaertner found that generally with dicotyledons, previous transplanting, even if the plant did not flag at all, prevented the act of fertilisation; & this resulted even with plants in pots, if the root had grown out of the hole at the bottom but in some few cases as in Digitalis the transplanting did not prevent fertilisation. According to the testimony of Mauz, Brassica rapa ripened its seed, with the plant pulled up & placed with its roots in water, as have several monocotyledons when cut from their roots. But I do not know whether in these cases the flower had previously been fertilised, for this, judging from W. Herbert

1 [Darwin left an unfilled space for the reference here.]

2 Beitrage zur Kenntnis. s. 560, 564

3 Gaertner Bastarderzeugung s. 333, 356, [3]66

4 Beitrage zur Kenntniss s. 252, 333 [Mauz' work is described here.]


experiment on Crocus makes a great difference; for he found that after the act of fertilisation, neither transplantation or mutilation prevented the seed from being perfected, but that "no application of its own pollen would fertilise the flower after transplantation."1 /

88/In accordance to the nature of the species acted on, excess of food or manure, & some believe especially ammoniacal manures, will produce sterility. Nothing is easier, as I have tried to produce on some plants, as the common primrose, absolute sterility by manuring it too much. Plenty of perfect flowers are produced, but these produce no seed, or seed which will not grow: Gaertner also2 alludes to the excessive flowering of some sterile species, & compares the fact to the excessive flowering of sterile hybrids: in other cases too much manure, especially if accompanied by too much heat, as before alluded to, prevents flowering. The effect of much manure depends on the nature of the plant; in some cases it is hardly possible to give too much; & Gaertner enumerates3 Gramineae, Cruciferae & Leguminosae as standing much manure, whereas succulent, & bulbous-rooted plants &c are thus easily rendered sterile. Hence in some case potting by checking the supply of food increases the fertility of hybrid plants, & in other cases lessens them.4 The extreme poverty of soil seems to have much less effect than too much richness on causing sterility, although of course the number of seeds is lessened, owing to the lesser size of the plants: but in/89/some plants of Trifolium minus & repens, flowering on an old lawn never manured, not one seed seemed to be produced: some other plants produced very few. I have tried starving kitchen garden plants & very small & few Pods can be produced.5

The period of growth during which the plant is watered often seems to affect greatly the fertility of a plant; so also does bottom heat. Many pelargoniums are extremely sterile (many of them no doubt owing to their being hybrids) but seeds have been obtained from some by extremely slight changes in treatment. So Kolreuter6 after comparing the manner in which some pure species of Mirabilis shed their flowers like hybrids, says that some were rendered more fertile by being kept dryer in pots. Very slight changes in position as on a slight bank, inste[a]d of at its foot, will sometimes make the difference, of a plant which appeared equally healthy in both positions, setting its seed or not producing one.—

1 Journal of the Horticultural Soc. Vol. 2. 1847. p. 83

2 Bastarderzeugung s. 370

3 Beitrage zur Kenntnis s. 333

4 a Gaertner Bastard, s. 378 s. 519: Kolreuter Act. Acad. St. Petersburgh 1781 Part II p. 303.

5 Herbert in Hort. Journal on Crocus

6 Nova Acta Petrop 1793. p 391


No doubt temperature has a very important influence on the fertility of plants: but it is surprising what changes, in this respect some species will bear to which they are not naturally subjected. To give one example; Dean Herbert showed me in his garden Zephyranthes Candida seeding well after having been just covered by/90/snow; but this plant, he informs me is a native of La Plata, where snow does not fall; & it runs wild & spreads itself in the dry & hot climate of Lima.—

Several cultivated plants, like domesticated animals, will endure the greatest change of climate & yet retain their fertility; & what makes the case far more remarkable, have their natures so far changed that their chemical composition is sensibly modified: thus Dr. Falconer informs me that Hemp seeds well on the plains & on the mountains of India, but its fibre is brittle; Linum does the same, but its seeds contain 25 per cent more oil: the poppy contains on the plains much more narcotin in proportion to morphine; & in wheat there is a similar difference in the proportions of starch & gluten; yet these plants in both situations seed well.—I suspect cultivation allows a plant to undergo change without sterility. I have alluded to the more or less complete abortion of the anthers, called by Gaertner, contabescence: until I read Gaertners able discussion on this subject,1 I attributed all these causes to sterility from changed conditions. The cases are very numerous: Kolreuter gives many2 in Dianthus & Verbascum: Herber3 adduces the N. America Azaleas,/91/which anyone may compare (as I have often done) with the most sterile hybrids, & the anthers will be found to be in exactly the same aborted condition. Gaertner has shown, that contabescence varies in different plants in intensity;— that it occasionally affects very many species in all classes but is most apt to occur in certain orders, as in Caryophyllaceae, Liliaceae (& Ericaceae may, I think be added);—that when one flower is affected generally all are affected;—<that whatever the degree of contabescence may be plants propagated by cuttings, layers etc retain4 the same degree of contabescence> & that it comes on at a very early period in the bud.

These facts alone, would not have convinced me that contabescence was due to some cause distinct to exposure to unnatural conditions; for in plants, very differently from in animals, we may I think infer that the fertility of the reproductive individual or flower is fully as much affected by the conditions to which the

1 Beitrage zur Kenntniss &c s 117 et seq.

2 Zweite Fortsetzung p. 10, p. 121—Dritte F. p 57.

3 Amaryllidaceae p 355

4 [Beiträge] s. 119.


whole plant, or vegetative individuals have been exposed, as by those to which the reproductive individual itself is exposed; we see this in the effect of previous treatment on the bearing of fruit trees, & this perhaps would account for contabescence coming on very early in life, & for all the flowers on the same plant being affected. But Gaertner further/92 /shows that contabescence, when it once comes on, is permanent (with one exception) in degree for life;—that it can be propagated by layers cuttings &c;—that no change in treatment, as potting &c affect the degree;—that it is doubtfully hereditary in hybrids from a contabescent plant;—& lastly that the female organs generally not affected or only rendered precocious, & that in some instances in which after artificial fertilisation the seeds were counted, the full normal number were produced. These facts more especially the last one seem quite incompatible with the view that contabescence can be caused by unnatural conditions of existence; for it seems incredible that the female organs should not be at all affected whilst the male were rendered completely sterile: some degree of inequality of affection would be not at all improbable, from the frequent production of hybrids in those captive animals which very rarely produce pure [?] young in confinement. Moreover many endemic plants are contabescent, which seems equally incompatible with the above view. One potent cause of contabescence probably is a tendency to become dioicous, as indicated by Gaertner in the case of Silene; & that may have nothing to do with external conditions. On the other hand, as exotic plants seem very often affected; & as Kolreuter1 seems to think that it is most apt to affect indigenous plants, when transplanted into a garden; & as Wiegmann2 /93/states that the contabescent wild plants of Dianthus & Verbascum which he found, grew on a dry, sunny sterile bank, the affection may in some instances be due to exposure to unnatural conditions.3

Double flowers: seedless fruit.—Flowers are often made, (as commonly expressed) nearly or even quite infertile by doubling. The male organs are much more often affected than the female, as everyone may see.4 The tendency to double depends on the nature of the species; for we have some species extremely double, as the Gorze, in classes which very rarely have double flowers. It depends, also, on the structure, as flowers with many stamens & petals are most apt to become double. Luxuriant growth & rich soil no

1 Dritte Fortsetzung s. 57

2 Über die Bastarderzeugung s. 27

3 [Here Darwin scribbled in pencil: 'Anthemis nobilis.']

4 Gaertner Bastardzeugung s. 363 s 569


doubt are highly favourable to doubling; & Prof Lehmann1 found several wild plants double near a hot spring: on the other hand I may mentioned that I found many stunted wild plants of Gentiana campestris,2 growing on a very poor chalky bank very double; I have also noticed a Staphylea & Aesculus pavia, & some other plants growing very poorly under favourable conditions, with a distinct tendency to become double: therefore luxuriant growth & good soil are not absolutely necessary concomitants./

94/Again when the fruit is largely developed seeds are rarely perfected:3 we see this in our best pears: the Enville pineapple which is a poor one is the only kind having seeds: this is notoriously the case with the Banana & Bread-fruit; it being extremely rare to find even a single good seed, except in some poor varieties. So again it is generally believed that a great development of tubers or roots often (certainly not always as in carrots, turnips &c) causes infertility; as does a great tendency to propagate by runners, & suckers.

These several affections have always been considered as the causes of the lessened or destroyed fertility, owing to an antagonism or compensation in growth. I strongly suspect the effect has been here held for the cause. I do not doubt that if any cause whatever produced a great development, especially if in the proximity to the reproductive organs, this would tend to produce infertility: but we have to consider what so frequently gives in cultivated plants the first tendency to such development often in connection with lessened fertility. There can be no doubt that the first tendency having been given, selection, taking advantage of the hereditary principle has played a most important part in nearly every case, & as we know/95/in the history of several double flowers, in which the work commenced in the seed of a flower having one or two stamens converted into petals. I believe that the first cause is lessened fertility from the plant being exposed to unnatural conditions, more especially to excess of food; & that the doubling of the flowers, the great size & succulence of the fruit, of the roots, & the tendency to form suckers &c is the result of, <or is compensation of> organic matter not being consumed in the formation of seeds, together with generally an excess of food <the process having been perfected by man's selection>. I have come to this conclusion, from finding an exactly parallel series of facts, but not perfected & added to by continual selection in a case in

1 Quoted by Gaertner Bastard. s. 567

2 Gardener's Chronicle.

3 See Prof. Lindleys excellent remarks on this subject in Theory of Horticulture p 175–179


which lessened fertility or entire sterility has supervened from an entirely independent cause; namely from hybridity. Gaertner has shown1 that hybrid plants are more inclined to produce double flowers than pure species; & the tendency is hereditary; in hybrids & in double flowers the male organs are first affected; in both there is a strong tendency to yield innumerable flowers. Again Gaertner insists2 /96/most strongly on the very general tendency of hybrids, even utterly sterile kinds, to produce the perfect receptacles of the seed or fruit: thus, Sabine on Passion Flower.

With respect to the development of roots, Kolreuter expresses his unbounded astonishment at the size of those of hybrid Mirabilis. All hybridisers, also,3 are unanimous in the strong tendency in hybrids to increase by their roots, & throw up suckers &c.—Considering this strictly parallel series of facts, & that it can hardly be disputed that unnatural conditions have a special action in lessening the fertility of organic beings, it seems to me, that the view here adopted, that the lessened fertility is the first cause aided by excess of food & selection, & that double flowers, fine fruit, large roots, &c is the result. Therefore the enormous class of facts here alluded to, come, I think, fairly under the present discussion, & support the conclusion that considerable changes of condition have an especial action on the reproductive system. I may add that horticulturists have often/97/spoken of infertility as the bane of horticulture; but on the views here advocated they ought to confess that though this may be so, they owe to it, their choicest productions.—

96 v/How far the several known & extraordinary cases of plants never flowering or never seeding in their native country, when they are abundant, come under our present subject, I am doubtful. Certain plants ascend mountains to a height, & in the arctic regions to a latitude in which they do not produce seed. In such cases I presume that there can be no doubt that their infertility is owing to the climate to which they are exposed, but that they have some other advantage over their few competitors in these sterile regions, which allows them to hold their own. We may suppose this to be the case in the curious instance mentioned by Kalm that the coniferous trees which cover in an impenetrable mass the swamps on the shores of N. America, never seed there; but only when growing in the higher country. Certain water-plants in our own country rarely or never seed. Dr Bromfield4 gives a still more curious

1 Bastard. s. 565

2 Bastard. "Fruchtungsvermögen der Bastarde: diese Eigenschaft ist Behr ausgebreitet bei den Bastarden." s. 537

3 Gärtner Bastard. s. 527

4 [Phytologist, 1848, 3.i, p. 376]


instance, namely in the common ivy which abounds in Russia & over the North of Europe but never flowers.

97/Although we have seen so many animals in captivity & so many plants under cultivation are rendered more or less infertile: yet those animals which do submit to the particular changes of conditions implied by domestication, are far from having their fertility checked; on the contrary the more abundant & regular supplies of food which domestic animals probably receive in comparison with wild ones, appears, as might have been expected, to increase their fertility./97 v/If it be denied that domestic animals which are often fattened & which are protected from famine, do receive more food on average than wild ones then I know not how to test the dictum/97/I have compared the produce of nearly all our domestic animals,1 with their wild prototypes, when known or with the most nearly allied animals. Of course there is often doubt about the rate of increase of wild animals, but as far as known all domestic animals, without it be the Peacock, bear either a greater number of young at a birth or at shorter intervals, probably at a younger age, than wild. In some domestic animals selection/98/may have increased their fertility, by the most fertile individuals, but in others as in cat, Pigeons &c I do not suppose this point has ever been attended to. In regularly cultivated plants, some as we have seen are nearly sterile; but these are such as can be propagated by cuttings, grafting &c; & in most of these the infertility, in accordance with the views just advocated has been of use, as causing greater development of some useful product, & therefore here infertility has been selected. In many plants, cultivated for their seed, selection probably will have increased their fertility: but there are many other plants propagated by seed, but yet which would never have been selected for this advantage; as the carrot, parsnip, cabbage, asparagus. As in these instances the wild prototype is known, I have taken the finest wild plants which I could find, & ordinarily fine cultivated plants, & I find that the cabbage has about…2 [Seeds vary so much in number that it is difficult to estimate them; but on comparing beds of carrots saved for seed in a nursery garden with wild plants, the former seemed to produce about twice as much seed. Cultivated cabbages yielded thrice as many pods by measure as wild cabbages from the rocks of South Wales. The excess of

1 [Memorandum pencilled on verso: 'Bechstein—Ferret, Rabbit—Wild Pigeon Poultry'.]

2 [Here well before the end of the page Darwin stopped in the middle of his sentence. The text is pieced out from the last two paragraphs of ch. 16 in Variation, II, 112–13.]


berries, produced by the cultivated Asparagus in comparison with the wild plant is enormous… with plants like carrots, cabbages, and asparagus, which are not valued for their prolificacy, selection can have played only a subordinate part; and their increased fertility must be attributed to the more favourable conditions of life under which they have long existed.]1

99/I have alluded to this last subject more particularly on account of Mr Doubleday's2 theory, which is that an abundance of food checks fertility & poverty increases it or "that prolificness is in the ratio of the state of depletion ". Independently of mankind, in regard to whom, I should have thought that the Malthusian explanations of restrained or reckless marriages, would have accounted for the asserted facts, the only evidence appears to me the undoubted fact that you can fatten individual animals to such an excess, as to check their fertility;3 & that in plants the same can be easily done by excess of manure.4 If indeed it could be proved that the most flourishing wild animals & plants, which exist in the greatest numbers in any country, from this very cause of their flourishing so much, had their fertility checked, it would be a most serious objection to the principles hereafter to be elucidated in the chapter on selection. <But to me, all the facts seem to point in an opposite direction.>/

1 [At the foot of the blank portion of fol. 98, Darwin pencilled the following dubiously legible memorandum: 'In carrot [?] I did not measure but after selecting the finest wild plant compared it with—The wild one grew in cultivated ground & had more than those growing in natural ground.']

2 [The following note, now to be found in ULC vol. 46.1, fol. 24, appears to belong here:] The True Law of Population. I have read this work, an article by Mr. Hickson in the Westminster & Foreign Quarterly Review Oct. 1849,' Godwin on Population' & various other Essays written against Malthus' great work, with all the attention of which I am capable, but I cannot say that they have had any weight with me, in opposition to the few facts given in this chapter, & which could have been largely added to. I am bound to add that so eminent an authority as Dr Carpenter (Principles of Comparative Physiology 1854 p 122) seems to admit Mr Doubleday's doctrine; so again that shrewd observer Hugh Miller (Schools and Schoolmasters p. 266) seems of same opinion, & remarks, that "when hardship presses on the life of the individual, so as to threaten its extinction, it is rendered more fruitful."—

3 [The following note on the verso of fol. 99 seems to belong here:] Gaertner in his Bastardzeugung s. 378, gives references to Henschel & Girou de Buzareingues, that domestic animals produce more in fruitful years, than when food fails. But in plants we have seen that there can be no question that by poverty of soil the number of seeds can be lessened. No one can doubt that few[er] ears of any corn will be produced on very poor land than on rich. In the case of the most wonderful increase on record, namely that of the domestic animals become feral & rapidly spreading over America, can it be believed, that this astonishing increase was owing to lessened fertility, for want of food: if there was any change whatever in fertility, which may be doubted, in all probability this would be increased.

4 [A note slip seems lost here. Note 76 in ch. 18 of Variation, which may well represent the missing reference, cites Gaertner: "Beitrage zur Kenntniss der Befruchtung", 1844, s. 333.]


100/In concluding this Chapter, it must be admitted that the evidence on the several points discussed in it, has been often very dubious & partly rests on the weakest possible grounds [,] general belief. Yet to my mind the evidence does seem to weigh in favour of the following conclusions; that slight changes in the condition of existence are favourable to the life of both animals & plants;— that in both, close interbreeding between the nearest relations is unfavourable to vigour & fertility, & that, on the converse hand, crossing with a distinct individual or variety (& even distinct species in some respects) is favourable in all respects; & further that there is some probability, though many of the gravest difficulties at present stand in the way, that it is a fundamental principle in the act of reproduction that there should be, perhaps at very wide intervals, the occasional concourse of two distinct individuals.—On the other hand, I think it must be admitted that greater changes of condition, or more strictly changes of a particular nature with respect to each species, have an special tendency, in both animals & plants, to cause infertility, that the cause seems to us to act most capriciously, affecting/101/one order far more than another; but with numerous exceptions in each order. That as slight changes of condition <& slight crosses> are good to the individual & as the offspring of the crossing of closely allied forms are more vigorous & fertile so we have a parallel series, in greater changes of condition causing more or less sterility in the individual & in the notorious fact of the lessened fertility or utter sterility in the hybrids produced by the crossing of distinct species or unlike forms. Neither in hybrids, or in an individual species placed out of its natural conditions, can we tell, till we try, whether the fertility will be greatly or slightly affected, so ignorant are we of the exact cause. But to the subject of Hybridity we shall hereafter to return.—

Hence I cannot doubt the truth of the propositions that in all living beings the reproductive system is acted on in an especial manner, unlike any other part of the organisation, by the conditions of existence; that both male & female element is acted on, the action appearing to us most capricious either for good or evil. This proposition seems to me important, for it brings into connection all the facts in this Chapter with the variability of organic beings when placed out of their natural conditions under domestication. If the reproductive system is so easily acted on, that/102/ changes of condition, which do not in the least affect the health of the individual, yet seriously affect or entirely stop its function; surely it is not surprising that the product of the reproductive

[page] 92




Darwin wrote the original draft of chapter IV during the period from mid December 1856 to late January 1857 according to his Pocket Diary. A year later he wrote a fifty sheet section on the contrasts in variation in genera with large and small numbers of species, which he intended for insertion towards the end of his original chapter. The history of these two parts of the chapter is best considered separately. In comparison with some chapters, the original section of chapter four shows relatively little revision. On folio 67, pencilled additions by Darwin dated 1861 and 1867 indicate his return to this section of the manuscript. From folio 68 to 72 cancellations signal that Darwin made later use of the material as Dr Guimond discovered. But here in contrast to chapter three Darwin did not take passages directly from the Natural Selection manuscript but in 1868 he only used some of his earlier examples and references incorporated with new material including experiments of his own.1 At some time after the final revision of the section on variation among Indian elephants, on folios 48 and 49, Darwin had a fair copy made of these two sheets.2

The large section written later on the commonness, range, and variation of species in large and small genera has a history rather separate from the rest of the fourth chapter. In an earlier memorandum dated January 4, 1855, Darwin indicated one theme of this section:

it may be concluded, as Mr. Watson remarks (Cybele Brit. vol. I, p. 18) that "those most widely & generally distributed, even in large spaces, being usually also the most common species."… Hence we may rudely conclude, that wide-ranging species are commonest: this harmonises with fact that they range far & are numerous, from same cause, viz successfully struggling with the organic & Physical conditions of area.—

The number of individuals must especially depend on struggle with other individuals.3

In regard to extensive numerical analyses of catalogues of regional flora, including helpful volumes borrowed from Hooker, all to provide quantitative evidence for his view of varieties as incipient species, Darwin later wrote Hooker:

1 'On Specific Differences in Primula', Linn. Soc. J. (Botany) 10 (1868), 441–2. The Different Forms of Flowers on Plants of the Same Species (London, 1877), ch. II, pp. 60–2.

2 Fair copy now in CD. MSS. vol. 45, fols. 18–19.

3 CD. MSS. vol. 15.1, fols. 36–7 of 2nd no. sequence. The last sentence was added in pencil along the margin.


I was led to all this work by a remark of Fries, that the species in large genera were more closely related to each other than in small genera; and if this were so, seeing that varieties and species are so hardly distinguishable, I concluded that I should find more varieties in the large genera than in the small…1

Fries' statement appears in Darwin's reading notes on the Botanical Gazette, where Darwin wrote:

p. 188 "In genera containing many species, the individual species stand much closer together than in poor genera; hence it is well in the former case to collect them around certain types or principal species, about which, as around a centre, the others arrange themselves, as satellites."This very important, it shows that extinction has not been at work in the large genera.— But some of the small growing genera ought to have close species.—2

In regard to the Fries quotation, Darwin later added in pencil the note that: "Bentham, Hooker & Thompson say Hieracium not large genus only forms. All three greatly doubted truth of statement & quoted case of Senecio & others where species very distinct."He also added on a pinned on note slip: "Waterhouse does not in least believe in Fries statement that large genera have closer species.—"

Darwin saved about 300 foolscap sheets of tabulations of genera & species from standard catalogues and calculations of relative proportions of species to genera and varieties to species.3 At the beginning, tabulations and calculations on Boreau's Flore du centre de la France are marked 'wrong & useless'. In July, 1857, John Lubbock pointed out some fundamental error in procedure which Darwin had been making in his calculations, thus vitiating his initial labours on statistics from Boreau and Fürnrohr.4

Having appealed to Hooker for a loan of these Floras so that he could rework them Darwin continued his tabulations and calculations of ratios of variation and speciation in parallel with his writing of successive chapters of Natural Selection, and he frequently mentioned this statistical work in his letters to Hooker. On August 22, 1857, he wrote:

I am very glad to hear that you have been tabulating some Floras about varieties. Will you just tell me roughly the result?—Do you not find it takes much time? I am employing a laboriously careful Schoolmaster who does the tabulating & dividing part into two great cohorts more carefully than I can. This being so, I shd be

1 L & L, II, 102–3; NY, I, 460.

2 CD. MSS. vol. 73, fol. 118, notes regarding Fries article, 'A Monograph of the Hieracia', Bot. Gaz. 2 (1850), 85–92, 185–8, 203–19.

3 C.D. MSS. vols. 15.2, & 16.1, 16.2.4 L & L, II, 103–4; NY,I, 461. Fred. Somkin,'The Contribution of Sir John Lubbuck, F.R.S., to the Origin of Species: Some Annotations to Darwin.' Roy. Soc. London, Notes & Rec 17 (1962) 185.


very glad some time to have Koch-Webb's Canaries—& Ledebour, & Grisebach….

On September 11: 'The magnificent & awful Box of Books arrived quite safely this morning I shall not, of course, try to do all, but will invest a handsome sum with our Schoolmaster…."Then on September 30th: "I hope you are not getting impatient for your books back: for I have done only a few of them which I shd like to do; for it is very slow work, & our Schoolmaster has only his evenings to spare.'1

The following spring, on March 10, 1858, the day after finishing chapter x, on instinct, Darwin mentioned to Hooker that he was putting notes together on large and small genera, and the next day he warned Hooker he would want him to read his draft when it was finished.2

A month later, on April 10, Darwin wrote Hooker:

I have almost finished my discussion; but it will take some little time to have it copied; & as my health has been lately wretched, I start in 9 days for a fortnight of Hydropathy & rest. On my return I will send it, & most grateful I am to you being willing to take the trouble to read it. I enclose a memorandum on way which I want you to consider my M.S. which please keep & read, when I send the M.S.—3


'Is the whole worth publishing? I do not promise to be guided by your judgment, but it will have great weight when in some <year> months time I reconsider subject.

Have I fairly stated the more important objections in abstract: to have given all in full would have made my now tedious discussion intolerably tedious.

I shd be very glad to hear any criticisms in detail; & you & Watson have done me an enormous service in drawing my attention to & enumerating the numerous objections but what I want you to do now is, in as candid a frame as you can, to balance all the vague probabilities on both sides of question.—

Remember that my book is written for geologists & zoologists, so that on some points I daresay my remarks may appear to you trivial.

I have discussed some extra hypothetical points chiefly for sake,

1 CD. MSS. vol. 114, Ltr. no. 208 (cf. ML, no. 53), no. 211, 210 (sic i.e. order inverted here). For instructions from Darwin to the schoolmaster, Mr Norman, see vol. 16.1, fols. 133A, 136A.

2 Ltr. 228, L & L, II, 103; NY, I, 460.

3 CD. MSS. vol. 114, Ltr. 231.


here & in other places to show what points ought to be considered in theory of the descent of species, rather than in hopes of throwing light on the many points of present inextricable confusion.—1


'My pencil <notes> alterations were intended to make passages clear to myself not for corrections or hints to you so do not mind them.' J. D. Hooker"1

In the Pocket Diary, the first two lines entered for 1858: 'March 9th Finished Instinct Chapter April 14 Discussion on Large genera &small' together with the letter of April 10 strongly suggest to me that the April 14 entry was intended to record the completion date of this additional section. Then upon his return from Dr Lang's hydropathic establishment at Moor Park, Farnham, Surrey, on May 6 Darwin sent off the fair copy to Hooker.2

Early in June Hooker sent Darwin an encouraging note about the manuscript to which Darwin replied most gratefully on June 8.3 Then came the startling interruption of Wallace's letter on natural selection, and Hooker's joint efforts with Lyell to secure fair recognition for both Wallace and Darwin. Apparently only on July 13, could Hooker complete his examination and send Darwin his verdict of considered approval:

I went deep into your MS. on variable species in big and small genera and tabulated Bentham after a fashion, but not very carefully. After very full deliberation I cordially concur in your view and accept it with all its consequences.4

Hooker's immediate comments on Darwin's draft are recorded on the fair copy and these appear as notes in this portion of chapter IV.


1/In this Chapter we have to discuss the variability of species in a state of nature. The first & obvious thing to do would be to give a clear & simple definition of what is meant by a species; but this has been found hopelessly difficult by naturalists, if we may judge by scarcely two having given the same.

I will copy the latest & most laboured definition by Alph. De Candolle5 who has carefully discussed the subject in relation to plants: he says species are "collections d'individus qui se ressemblent assez pour 1 ° avoir en commun des caracteres nombreux et

1 CD. MSS. vol. 15.1, fol. 0.

2 L & L, II, 107; NY, I, 465.

3 ML, no. 64.

4 Hooker, L & L, I, 458.

5 Geographic Botanique p. 1072


importants, qui se continuent pendant plusieurs générations, sous 1'empire de circonstances variées; 2° s'ils ont des fleurs, se féconder avec facilité les uns les autres et donner des graines presque toujours fertiles; 3° se comporter à l' égard de la température et des autres agents extérieurs dune maniére semblable ou presque semblable; 4° en un mot, se ressembler comme les plantes analogues de structure, que nous savons positivement étre sorties d'une souche commune, depuis un nombre considérable de générations."M. De Candolle lays stress on making the element of descent subordinate to that of resemblance, so that the definition may be less hypothetical. But as animals & plants must be here equally considered, I agree with Dr. Carpenter who gave at Glasgow to the British Association an interesting lecture on this subject,1 that descent does come in as a prominent idea. Although when speaking of the resemblance of two forms, the comparison should of course extend to all ages & sexes, yet as zoologists/2/have often described these stages as specifically distinct, an error instantly corrected when their descent was known, it is very natural that they should bring this idea prominently forward. Thus if the development of <Trichoda lynceus> had not been known, the stages through which it passes, as M. Quatrefages2 has remarked, would have been considered as forming eight distinct genera: I am convinced that in the cirripede Ibla without knowledge of its descent, the male & female & its two larval stages would have formed four distinct Families in the eyes of most systematic naturalists. Again the most ill-shapen monster is rendered home to its species the instant we know its parents.

Let us test M. de Candolles definition with a plant. Assuming for the moment that it was demonstrated (& we shall presently see that the evidence is very strong) that the primrose & cowslip can be produced from the same stock; would they be called by any Botanist distinct species in the ordinary acceptation of the word? Yet/3/the individuals of the cowslip & the individuals of the primrose accord in every single respect with M. Decandolle's definition of two distinct species: for (1st) the individuals in each agree in many important characters, which are constant during many generations under different conditions, for they are found in distinct parts of Europe: (2nd) they do not fertilise each other with facility, as the best experimentiser, who ever lived, Gaertner, found after repeated trials during many years: (3rd) they do not behave in the same way in regard to temperature & soil, for they

1 [See The Athenaeum (1855), 1090.]

2 Revue des deux Mondes [(1856 tome 3), 871.]


have different ranges & inhabit different situations (4th they cannot be said to resemble each other as much as analogous plants do, which we positively & habitually know to have descended from a common source. Hence I conclude, that descent is a prominent idea under the word species as commonly accepted.

The idea of descent almost inevitably leads the mind to the first parent, & consequently to its first appearance, or creation. We see this in Morton's pithy definition of "primordial forms",1 adopted by Agassiz. The same idea is supreme, & resemblance goes for nothing, with those zoologists, who consider two forms, absolutely similar as far as our senses serve, when inhabiting distant countries, or distant geological/4/times, as specifically distinct. Having the idea of the first appearance of a form prominently in their minds, they argue logically that as most of the forms in the two countries or times are distinct, the distinction being in some great, in others less & less, they naturally ask, why forms apparently absolutely identical should not have been separately created, & which they in consequence would call distinct species.—As we have to discuss in this work whether forms called by all naturalists distinct species are not lineal descendants of other forms, this minor question will fall or rise with the greater question; & is here only alluded to in connection with the definition of the word species.—

Some authors, as Kölreuter, take the fertility of the offspring of two forms as the sole <or leading> test of what to consider as species; & however unlike two forms may be, if they produce quite fertile offspring, they consider them as specifically the same. The great importance of this difference in fertility in what are ordinarily called varieties/5/& species, has in my opinion of late years been much undervalued by some authors. In the chapter on Hybridism we shall fully consider this subject & we shall find that there are great difficulties (I do not mean merely practical ones in its application) in taking lessened fertility in the offspring as an unerring guide what forms to call species. I will here only remark, that perfect fertility & utter sterility glide into each other, in so insensible a manner that it is hardly possible to draw any line; hence the two most laborious experimentisers who ever lived, Kolreuter & Gaertner after numerous experiments in regard to certain forms, have come to diametrically opposite conclusions; the one concluding that certain forms are varieties, & the other that they are undoubted species.—

Short as this discussion has been it suffices, I think, to show

1 [See J. C. Nott and G. R. Gliddon, Types of Mankind, p. 375.]


how various are the ideas, that enter into the minds of naturalists when speaking of species./5 v/With some, resemblance is the reigning idea & descent goes for little; with others descent is the infallible criterion; with others resemblance goes for almost nothing, & Creation is everything; with others sterility in crossed forms is an unfailing test, whilst with others it is regarded of no value./5/ At the end of this chapter, it will be seen that according to the views, which we have to discuss in this volume, it is no wonder that there should be difficulty in/6/defining the difference between a species & a variety;—there being no essential, only an arbitrary difference. In the following pages I mean by species, those collections of individuals, which have commonly been so designated by naturalists. Everyone loosely understands what is meant when one speaks of the cabbage, Radish & sea-kale as species; or of the Broccoli, & cauliflowers as varieties: between such extremes there is often a wide neutral territory in which the term species & varieties are bandied about according to the state of our knowledge & our ideas of the term species. —

Botanists in discussing the subject of variation have usually included <together> that variation which occurs under domestication & that under natural conditions; & this is probably the best plan, though not for our particular object. They have divided1 varieties into "variations" in which the varying characters are not fixed even in the individual plant, all the buds produced on the same plant being here considered as one individual. In animals we have very few instances of this class; but as the black colour in cage birds produced by hemp-seed goes off with change of food; & slight changes in the/7/hairy covering of animals when transported into a different climate2 have been observed. The term 'Variety' is applied to forms often offering considerable differences, & which can be securely propagated by buds, grafts, cuttings, suckers &c, but which are believed not to be inheritable by seed. This class nearly corresponds with "abanderungen" in Bernhardi's classification3 in which the form is not hereditary or only so in certain soils; & likewise in a lesser degree with his "Spielarten" in which the form tends to go back in one or more generations to the parent type. As we know scarcely anything of the variation of those lower animals which can be propagated by division, the class "Variety" in the above strictest sense is not applicable to the animal Kingdom; though no doubt, in the less strict sense of

1 M. Alp. De Candolle has given a full discussion on this subject. Geograph. Bot. p. 1078

2 The cat in West Africa

3 Ueber den Begriff der Pflanzenart 1834. p. 5


being hereditary in only a slight degree, there are very many cases amongst animals, & some even in a state of nature. Lastly we have the class "Race", corresponding with "Abarten" of Bernhardi/ 8/& with subspecies of some authors, in which the form is strictly inherited, often even under changed conditions; of this class we know there are plenty under domestication, some known, & more suspected in a state of nature, as in the geographical races of some Zoologists. But the term subspecies is used by some authors, to define (& corresponds in this sense with "unterart" of Bernhardi) very close species, in which they cannot determine whether to consider them as species or varieties. The existence of these doubtful forms has lately been explicitly admitted by M. Alp. Decandolle in regard to plants, & by implication by Mr. Wollaston1 in regard to insects: M. Decaisne & Dr. Hooker use the term without expressing more than that the difference between such subspecies is slight, yet permanent. As these authors are of the highest authority, this admission is important as sub-species fill up a gap, between species, admitted by everyone & varieties admitted by everyone. Between varieties & individual differences there seems a gradual passage but to this subject we shall recur. In species we should remember how extremely close some undoubtedly distinct forms are, as many plants, & as in some of the willow wrens, which are so close that the most experienced ornithologists can hardly distinguish them except by their voice, & the materials with which they line their nests; yet as these wrens inhabit the same country [? county] & always exhibit the same/9/ difference, no one can doubt that they are good species. So that between individual differences & undoubted species naturalists have made various short steps.

In the above classification of several varieties the main difference rests on the hereditariness of the characters. Though the classes blend insensibly into each other, this classification is of some use when applied to domestic productions; & no doubt it holds good in varieties in a state of nature, which we are here considering. But it seems to me that we are far too ignorant to apply it to varieties under natural conditions, more especially in regard to animals. We have seen in our first chapter that the same character is inherited in very different degrees by different species, & even in different individuals of the same species; we have reason to suspect that a character becomes more fixed by long continued generation; although on the other hand, a character suddenly appearing is sometimes strongly inherited. Who can tell how much

1 On the Variation of Species. p. 185 99


the dwarfed character of a plant or the dark colour of an insect on a mountain, or of a shell in brackish water/9 v/or of the improved character of the fur of Beavers Martins &c the further we go north1 /9is due to inheritance & how much to the exposure of the individual from its earliest days to the condition in question. Probably in all such cases, the/10/form would change when placed under other circumstances; & some in fewer generations than others; but then it might be argued that this was not a fair test, as many races or strongly hereditary varieties change in some degree under new conditions. <I am inclined to believe that with the rarest exceptions every changed structure is in some degree inheritable.> In animals perfectly black individuals are not very rarely born, even in the same litter with ordinary coloured individuals: & in some places these appear much more frequently than in others, thus I am informed by Mr. Crawfurd that black Leopards are far more commonly produced in Java, than elsewhere; & in such cases I know not whether to attribute this to a strange hereditary principle, or to some unknown conditions acting on the parents. Fish of the same species are well known to present distinguishable differences in different lakes: Sir H. Davy2 states that red-fleshed dark-banded trout were taken from one Scotch lake & put into another, where the trout were white-fleshed; the young here produced had their flesh less red, & in 20 years the variety was lost. Laying on one side the probability of crosses having taken place, we see here that the red flesh was in some degree inherited; & some would assert that if the red trout in their own lake had/11/transmitted their character for some additional hundred-thousand generations, the character would have kept truer. From these & similar considerations I have thought it advisable to use only the term "variety", & where it is known or almost known to be strictly inherited "race": and I use the term variety loosely, simply in accordance with common acceptation, as I do the term species. <for the same reason in both cases> If the distinction could be drawn between hereditary & temporary variation in a state of nature it would be of great importance for our object; for variations in a state of nature which are not inherited are of little signification, & deserve notice, (perhaps) only as showing the possibility of change in structure.—

Practically the systematic naturalist, without troubling himself more than he can help about descent & creation, considers those forms as one species which he can unite by other intermediate & graduated forms. It is his golden rule. But those who have not

1 Kalms Travels vol 3. p. 58

2 Salmonia p 53


themselves worked, can form little idea of the irksome labour required in its application. For example look at the case of Aqilegia vulgaris, as worked out by/12/Dr. Hooker in his Flora Indica [I, 44], who devoted weeks to the examination of specimens from all parts of Asia & Europe, & who ends in uniting about 16 species of other authors into one. I may state, as I know that similar cases have occurred with others, that in Lepas anatifera & Alan's tintinnabulum. I at first wrote out full descriptions of several supposed species; then after getting more specimens from various parts of the world, I thought that I ought to run them all into one, & tore up my separate descriptions: after an interval of some months I looked over my specimens & could not persuade myself to call such different forms one species & rewrote separate descriptions; but lastly having got still more specimens, I had again to tear up those & finally concluded that it was impossible to separate them! When the Naturalist has got the intermediate forms between two supposed species, the work though laborious is generally simple; but he is very often obliged to judge by analogy. And here springs up an endless source of doubt. On how widely distinct groups may he draw for his analogies?/13/it is a remark repeated in almost every systematic work, that the very same organ whether or not of physiological importance will be constant in one group & so afford good specific characters, & will be <highly> variable in another. His power of drawing analogies will not only obviously depend on his amount of knowledge, but on the frame of his mind. Is it then surprising that naturalists should differ in the extreme degree in which they do, in determining what forms to call by the various <defined &> recognised term species? I have remarked that generally when the naturalist has got intermediate stages he unites with confidence two forms distinct in appearance. But here, also, he sometimes has cause to doubt. The intermediate forms may be hybrids; these he may often recognise by their sterility, but by no means always, at least without counting their seeds & comparing them in number with those of both presumed pure parents; but Gaertner thinks that a hybrid should be artificially made for comparison; or he may discover that they are not hybrids by one of the supposed parent forms not growing in the neighbourhood./14/But independently of this source of doubt, which perhaps has been over-rated by some authors, there is another & more important one, namely the probability of one of two forms, or of two forms which deserve in every sense to be called species, both varying greatly & running so closely together that the extreme varieties become undistin-


guishable. This is the more probable, as we shall afterwards see that, certainly varieties of one form tend to mock the characters of other species in the same group. To give a very few examples: Drs. Torrey & Gray, in speaking of the N. American Asters say "that several species, which we cannot but consider as distinct do frequently present very puzzling intermediate forms; & that an apparent transition is not always real."1 /14 v/—Such cases more or less striking do not seem to be very rare, for even in the small British Flora, Mr. Hewett C. Watson has marked for me 15 cases, (not including the protean forms in Rubus, Hieracium &c) in which two species & in some cases three species apparently distinct are/15/united, more or less perfectly, by intermediate forms:2 to give a single example,—Geum urbanum & rivale are universally thought to be distinct; but between them we have the var[iety] G. intermedium (considered a distinct species by some authors) & several intermediate forms, breaking down every character between the two types: in this instance Dr. T. Bell Salter has stated that he produced G. intermedium by crossing the above two species; but from observations in the Flora 1848 p. 42 [Hornschuch], in regard to the absence of the two parents in a place where G. intermedium was found, we perhaps have here two distinct origins of the connecting links, making the confusion doubly confounded.

Mr. Watson, who has paid the closest attention to the subject under discussion,3 & to whose assistance I am under great obligation, in a letter, <which he has permitted me to publish>, has pointed out in a very clear manner the following four categories in our British plants.

Mr. Watson's note./

15 A-l/<Categories of Species>

1. Plants distinguishable from each other by positive characters, & generally received as true Species.

2. Same as No 1; but so closely resembling each other as to be frequently mistaken one for the other, & by botanists even of some experience.

1 Silliman's American Journal of Science vol. 40. p 280 [actually vol. 41 (1841), p. 281.]

2 Mr. H. C. Watson has given me a list of examples divided into three groups.

(1) of two species actually passing into each other by intermediate varieties.

(2) of two species closely approximating to each other by intermediate varieties.

(3) & more commonly of one two species varying & its varieties assuming some of the characters of the other species, either positive or negative, but without actually passing into that other species.

3 An admirable paper <entitled on the Theory of Progressive Development) from which I have largely borrowed views & facts by Mr. H. C. Watson on the relations of species to each other & the varieties is given in the Phytologist 1845 p. 140 & 161.


3. Same as No 1; & not liable to be mistaken in their typical forms; but accompanied by intermediate or transition forms, approximating so much to each or both, as not to be quite satisfactorily assigned to either. <(N.B. The primrose & cowslip would be in this category, but it has been there proved that the intermediate produces both the alleged species from the same year's seed)>/

15A-2/4. Plants deemed true species where their typical & most general forms only are looked at; but the limit of the species is rendered uncertain by the existence of forms closely allied, deemed varieties of the type by some botanists, distinct species by other botanists. As is the case with the intermediates of no. 3, so these varieties or sub-species of No 4 are usually much more rare or local than the type species. They differ from the intermediates of No 3 only as varieties or quasi species clustering around one, instead of uniting together two supposed genuine species./

15 A-3/Altho' four such categories are easily defined on paper, & illustrated by selected examples, they glide together by other examples; & thus, as groups, they are different in degree rather than in kind. To give examples of the four categories,

1. The Apricot, plum, & Cherry are commonly placed under one genus, Prunus; & as species these are very readily distinguished by any body. 2. But there are two Cherries spontaneous in England, an arborescent & a fruticose, which by most botanists are deemed two real though very similar species, & between which in a wild state we can hardly point out any connecting links./

15A-4/3. Many botanists deem the wild sloe of England to be quite a distinct species from the cultivated & probably imported plum-tree of the gardens. Nevertheless, between the plum-tree of the garden & the sloe-bush of the hedges, there exist numerous intermediate forms or links, which render it highly difficult to say, 'here ends the sloe & its varieties, there begins the plum & its varieties I If we hold the intermediate Bullace a good species, this also passes insensibly down to the sloe, & improves almost as insensibly into the plum, so numerous & fine are the steps or links either way. 4. Linne described the fruticose bramble as a species, under name of Rubus fruticosus; but various modern botanists make out 50 to 100 supposed species of Bramble which others call varieties of R. fruticosus, & others again group into a small number of species, say half a dozen./

16/To the naturalist who looks at species as not essentially differing from varieties, being only more permanent, with the connecting links extinct, the occasional blending by intermedial forms of two or more apparently' distinct species, will not be wonderful; indeed the wonder is to us, with our restricted notions of the lapse of time, that many more cases are not on record.—

Individual differences.—Besides the varieties recorded by naturalists we have individual differences, which are not thought worthy of separate notice, either from being so slight, or from being believed to be so little permanent or forms graduating or blending into each other so that they cannot be divided even into distinct


varieties./16 v/Nothing can be looser than this distinction; no doubt a multitude of what perhaps should be called individual variations, with no degree of permanence figure as recognised varieties; more-over it is quite a common practice with naturalists to pick out of a graduated & inextricable mass of forms, a few leading types & designate them as varieties as does Mr. Wollaston1 when speaking of his "technical" varieties. In other cases, when this has not been done, it might be easily effected, especially if a few of the intermediate forms were to become lost; as remarked to me by Mr. Watson in regard to Polygonum aviculare. But on the other hand if we take the extreme case of well marked & permanent varieties, & the difference, just perceptible though hereditary, between a brother & sister organism, some such distinction does exist, as no one would put these differences into the same class. M. Boreau, who has so carefully studied the Flora of France has called attention to this distinction & says "les varieties proprement dites sont plus tranchees"./16/Individual differences from being generally very slight compared with the difference between species have not I think always been sufficiently noticed by naturalists. When discussing the subject of varieties one is apt, except in very variable forms [,] after a short preliminary study to forget them; but let any one collect specimens in almost any group of beings, about which he is profoundly ignorant, & he/17/will be for a short time, at least I have been, utterly perplexed to tell what are individual & what specific characters. This indeed is tacitly acknowledged by every cautious naturalist, by their dislike to define a new species, without it be some strongly marked form, if he possesses only a single specimen. I have been in the habit during many years of marking in all careful monographs & works, in which measurements have been given of several individuals, with care taken to note sexes & age, & I cannot doubt that individual differences are very often considerable; & no one doubts that this is the case with plants [.] It is impossible to give instances: many cases might be selected from Mr. Waterhouse's excellent work on two great orders of Mammalia, & likewise in Macgillivray's elaborate work on British Birds. I will refer only to one other instance, as it, also, relates to birds, generally considered, & I believe truly, as very fixed in form: Graba3 who particularly attended to this subject, says that he shot hundreds of seabirds

1 Variation of Species. p. 5.

2 Flore du centre de la France 1840 p 101

3 Tagebuch auf eine Reise nach Färo 1830 s. 103: he gives details of measurement of beaks &c of Anthus s. 56 & 67.-of beak & tarsi in Larus s 65 & 80.-& in Colymbus s. 118 etc


at Faroe & that he seldom omitted to measure very one, & the result was that rarely did two individuals of the same species agree throughout in their measurements./

18/These individual differences differ in amount to a surprising degree in various species & in various groups of species, one part or organ being affected in one species or group, & the same part being very constant in another set of species [.] Some forms are extremely constant in their whole organisation others as variable, causing to the naturalist an odious amount of perplexity. Generally the characters which individually vary, are of slight physiological importance, but this is not always the case; & I will immediately give a table of some of the more important & curious cases of variation (the slighter ones not being worth notice) which do not seem to be characteristic of any breed or variety, & therefore are not marked as separate varieties by Naturalists.

But here arises a perplexing question; are these individual differences of the same order & have they the same origin as those other differences, either greater, more permanent, or less closely linked together, which separate recognised varieties. <Many authors seem to consider that each species was created with a certain fixed amount of variability, or to use an expression in a letter of Prof. Dana, with "its system of librations under the influences of nature to which it may be subject", & this would include both recognised varieties & individual variations.> No one will pretend that any clear line of demarcation can be drawn between these two classes of facts; but some authors as Dr Prosper Lucas/19/ think that the production of slight differences is the normal & invariable function of the reproductive system in all organisms, independently of their conditions of existence; & the universality of some slight individual differences countenance this conclusion; but this view I presume no one would extend to marked varieties, thus even a fundamental difference between individual differences varieties seem to be indicated. But to me it seems a simpler view to account for all individual differences, which cannot be explained by differences in the parents or more remote ancestors, by the effects of varied <external> conditions acting on the parents ancestral forms & thus affecting indirectly (as we have seen in the last chapter) the reproductive system & consequently its products. According to this view if we could start with quite similar organisms & bred them for many generations during their whole lives under absolutely similar conditions, the offspring would be absolutely similar; & consequently we should look at all individual differences (independently of those produced by


crossing) as having the same nature & origin with those marked by naturalists as varieties.

In favour of this view we have the broad facts that there is much more individual variability as well as distinct varieties in domestic/20/productions, than in those under their natural & unchanged conditions. M. Boreau thinks that it is the very common plants, which vegetate in all places & under all exposures, which offer innumerable slight differences. It is certain that some species which are extremely constant in one area are extremely variable in another: thus the Helix aspersa one of our most constant land-shells in the South of France, as I am informed by Sir C. Lyell is very variable; & many instances might be given. On the other hand the general impression which I have taken, is that a variable species is in all places & all times variable; but I have not met with careful observation on this head. Variable sea-shells seem to be variable everywhere, but these in most cases are attached shells, as Limpets & oysters & cirripedes & they would everywhere be modified by the surfaces of attachment. In Beetles Coccinella seems everywhere variable in its spotted colouring. I applied to Dr. Hooker on this subject & he went through the Tasmanian & New Zealand Flora with this idea, & he found that those genera which were very variable in Europe were there also very variable; but in the Himalaya, the species of Willows, Rubus, /21/Senecio Gnaphalium, which are so eminently variable in Europe & in N. America were there not so./

21 a/I have applied, also, to Mr. Davidson, whose vast experience in Brachiopodous shells, makes his opinion of the highest value & I find he has specially attended to this subject & is puzzled by it equally with myself: he says that certainly many fossil shells, as Spirifer rostratus of which he has examined vast numbers of specimens from various places & periods, present everywhere the same quite extraordinary amount of variability: on the other hand some other shells of this same order vary but little either in time or space: innumerable examples could be given of the foregoing cases & this was all that I could learn on this subject from the late Prof. E. Forbes & from Mr. Woodward. Under certain conditions the same species, of which Mr. Davidson has given me examples, will be very variable in one space & constant in another: thus, also, Mr. Searles Wood, who is so intimately acquainted with the Crag fossil shells, informs me that several species, from the Mammaliferous stage are remarkably variable more so than the same shells at the present day, & which he is inclined to attribute to the former estuary conditions of the site:


on the other hand Mr Wood has not found the same degree of variability in the Eocene estuary shells of Hampshire/

21/These facts, & more especially the existence in every great class of organisms of groups of species adapted to varied conditions & growing in different countries eminently variable, as the genera of plants just mentioned & many others & as in the Brachiopods in various geological formations seem to indicate that the variability is here innate & independent of the conditions of existence: or that according to the common view, that they have been created with this tendency, each having its own system of libration to use an expression of Prof. Dana in a letter to me. But this tendency can seldom be predicated of every species in the variable group; thus even in Rubus, the R. [ ] is a very fixed form: in the eminently variable genus of shells, Pleurotomaria M. Eudes-Deslongchamps1 states that some vary hardly at all, some, so to speak without any limit. How variable are the species of Squirrels, yet Dr Bachman who has so carefully studied the N. American species, informed me that some are very true to their characters. As under cultivation forms are often produced which are characterised by being variable, it/22/is perhaps possible, according to the views we are examining in this work, to account for groups of variable species by their inheriting this tendency from a common parent; but I am not satisfied with this conjecture.

If it could be rendered probable that in the course of time some one or two of the forms of a species individually very variable might become fixed, then with the extinction of the intermediate forms we should see the stages & in some cases better understand the origin of the more permanent varieties. The occurrence of certain constant species in the most variable groups harmonises with such a view.—M. Lecoq, I presume is of opinion that this would happen, & likewise that the more fixed varieties would be converted into & deserve to be called species, for he speaks of such genera as Rubus; as being genera in process of formation. But I must leave the case of these many Protean groups an open question; not doubting, however, that in very many instances there is no real distinction in nature or origin between individual differences & more strongly marked & permanent varieties./

23/I will now give a few <selected> examples of individual variation or differences, not known to characterise a recognised variety; & I shall select them from various motives, some from the physiological importance of the organ affected, or from such part being in the group in question generally constant &c.

1 Mem. de la Soc. Linn. de Normandie Tom. 8. 1849. p. 23 107


Several other cases might have been added, & will be subsequently given, illustrating the variability of rudimentary organs, of greatly developed parts, & of sexual characters &c. One chief object in the following list, is to show that the common remark that organs called important by naturalists never vary is not quite correct, but anyone, unacquainted with Natural History, who might infer that because this or that part varies in certain species given as examples, it would likewise vary in other groups, would err greatly.—/

24/In Utricularia nelumbifolia,1 in the perfect (sexual) flower, especially where only one stamen is antheriferous the anther is commonly found to be one celled. The lobes of the style are variable in number, as are the scales of corolla & calyx.

In Zannichellia palustris2 "the form of the stigma the length of the style, the number of anther-cells… the fruits more or less stipitate are very variable."

In the common Beech Fagus sylvestris3 Persoon has described a wild individual with extraordinary large leaves & fruit, & another with the bark & manner of branching so precisely like an oak, that the country people consider it a cross.

Prof. Vaucher says that he has found the kind of gemmation4 with one exception always the same in the same species of tree, & that it generally is a generic character; but that in the common Lilac, Syringa vulgaris, he has observed two forms, "bourgeon terminal" & "presentant ruptures"./

25/Papaver bracteatum & orientale5 present indifferently two sepals & four petals or three sepals & six petals, which is sufficiently rare with the other species of the genus.

In the Primulaceae, & in the great class to which this Family belongs6 the unilocular ovarium is free, but M. Duby has often found individuals in Cyclamen hederaefolium "ou la base de l'ovaire etait soudee jusque a un tiers de la longeur avec la partie inferieure un peu charnue et dilatee du calice."

M. Aug. St. Hilaire7 speaking of some bushes of the Gomphia oleaefolia, which he at first thought formed a quite distinct species, says, "Voila done dans un meme individu des loges et un style

1 Dr. Asa Gray, Silliman's American Journal vol. 45. p 215 [where reference is to Oakesia conradii not Utricularia.]

2 Sir W. J. Hooker & Arnott's British Flora 1855 p 486

3 Linnaean Transactions vol 5. p. 232

4 Mem. Soc. Phys. de Geneve Tom. [1 (1822)] p. 300

5 Decandolle, Mém. Soc. Phys de Genève. Tom 2. Part 2. p 127

6 Duby Mém Soc. Phys de Genève Tom. x. p 406

7 Sur le Gynobase, Mem. du Mus. d'Hist. Nat. Tom x. (1823) p 134


qui se rattachent tantot a un axe vertical, et tantôt a un gynobase; donc celui-ci n'est qu'un axe véritable; mais cet axe est deprimé au lieu d'etre vertical."He adds (p 151) "Tout ce qui prec´dè1 n'/26/indiqueroit-il pas que la nature s'est en quelque sorte essayé dans la famille des Rutacees a former d'un seul ovaire multiloculaire, monostylé et symetrique, plusieurs ovaires uniloculaires munis chacun d'un style." And he subsequently shows (p. 364) that in Zanthoxylum monogynum "il arrive sou vent que sur le même pied, sur la même panicule [text seems actually 'le même panicle'] on trouve des fleurs à un ou deux ovaires." And that this [is] an important character, from the Rutaceae, to which Zanthoxylum belongs being placed "dans la cohorte (Tom. XI. p. 48) a ovaire solitaire."—The same author (Tom XI. p. 49) referring to this same character differing in the different species of Helianthemum, states that in the H. mutabile "une lame, plus ou moins large, s'etend entre le pericarpe et le placenta."/

27/De Candolle has divided the Cruciferae into five sub-orders in accordance with position of radicle & cotyledons, yet M. <Monnard> J. Gay2 found in 16 seeds of Petrocallis Pyrenaica the form of the embryo so uncertain that he could not tell whether it ought to be placed in "pleurorhi[z]ée" or "notorhizée": so again (p 400) in Cochlearia saxatilis M. Gay examined 29 embryos & of these 16 were rigorously "pleurorhizées" 9 had character intermediate between pleuro- & Notorhizees & 4 were pure notorhizées: a few other examples are given of variability in a character of great importance in this large Family.

In the Cruciferae it is well known, that Bracteae are generally absent, but these have been observed3 in certain individuals of Cardamine pratensis, in Erucastrum Pollichii & in (cultivated) Wall-flowers.—In regard to bracts, I may add that W. Herbert4 says that there are varieties natural & arising from cultivation of Crocus aureus, with & without bracts.5 /

28/The insertion of petals & stamens is a character of high generality; but M. J. Gay6 found in Arenaria tetraquetra, that in var. uniflora, which is polygamous, that in the hermaphrodite flowers the insertion was ambiguous neither visibly perigynous or

1 [Pencil note at bottom of fol. 25: 'I forget & I am not sure that this has bearing'.]

2 Annales des Scien. Naturelles 1. S. Tom 7. p. 389 [citation should be p. 391.]

3 [Anon.], Henfrey's. Botanical Gazette vol. 3. p. 82. & vol 1 p. 307

4 Journal of Hort. Soc. vol. 2 p. 283

5 [Here at the foot of the sheet Darwin added in pencil: 'Hooker says there are species of crocus with Bracts.' as if this were a memorandum later producing the previous sentence]

6 Ann des Sci. nat. Tom. 3 (1 series) p. 27 [citation should be p. 35.]


hypogynous, whereas in the female individuals, the insertion was perigynous: in var. aggregata (thought by some to be a distinct species) the insertion was ambiguous in all the individuals.

M. Raspail asserts1 that a grass Nastus Borbonicus is so eminently variable in its floral organization, that the varieties might serve to make a Family with sufficiently numerous genera & tribes,— a remark which shows that important organs must be here variable.

In Globularia nudicaulis2 the upper lip of the corolla varies remarkably, being sometimes entirely wanting, sometimes very small & divided to the base./

29/In some species of Hern[i]aria3 on the same individual, the divisions of the calyx are regular or irregular with four or five sepals.

In Suaeda, the vertical or horizontal position of the seeds in the pericarp has been thought a character of some importance, but M. A. Moquin4 found that S. altissima "presente des grain[e]s tantot droit[e]s, tantôt obliques et quelquefois couchées." With the different position of the seeds the point of attachment of the umbilicus varies./

30/M. Milne Edwards5 has given a curious table of measurements of 14 specimens of Lacerta, & taking the length of the head of standard, he finds, neck, trunk, tail, front & hind legs, second toes of posterior legs, colour & femoral pores all varying wonderfully; & so it is more or less with eleven other species. So apparently trifling a character, as the scales on the head, affording almost the only constant character.

Mr. Couch6 has seen the common ling Gadus molva with two cirri on the throat & G. mustela with five barbs.

The eggs of many Birds, especially of the Crow genus, of Shrikes, & Gulls vary in tint of colour, in spotting & size, even sometimes in the same nest.7

The Beak of birds, though generally so constant in character that most of the systematic divisions are founded on it, varies sometimes considerably in length; & I was shown in the British Museum by Mr. G. R. Gray three examples of/31/a Nutcracker

1 Annal. des Sci. Nat. 1 ser. Tom. 5. p. 440

2 Cambessèdes in Annal. Nat. Scien. 1 ser. Tom. 9. p. 15 [see p. 17.]

3 Decaisne, in Annal. des Sc. Nat. 1 ser. Tom. 22. p. 97

4 Annal. des Soc. Nat. 1 Ser. Tom. 23 p. 274

5 Annal. des Scienc. Nat. 1 Series. Tom. 16, p. 50

6 Linn. Transacts. vol. XIV p. 73

7 [Sheppard,] Linn. Transact. vol XV Part I. p. 9. See, also, for numerous cases W. C. Hewitsons British Oology where the variations are shown by coloured Plates.


(Nucifraga) shot in some forest, with beaks of remarkably different length: he showed me, also, a Himalayan Nuthatch (Sitta) with beaks similarly varying. I observed the same fact in two S. American birds1 the Uppucerthia & Opetiorhynchus. The conspicuous character of the tooth on the upper mandible, varies in some Hawks, as in the Jer Falcon.2 In whole Families of Birds the number of tail feathers is constant; but in some, as in Swans & in some Gallinaceae the number is variable; & this is the case according to [ ] (Isis [ ]) in many short‐tailed Birds as the King‐ fisher: in the N. American coot3 the number varies from 10 to 16. In some Hawks & Owls, the proportional lengths of the primaries, a character perpetually used to separate species, varies.4 I have already quoted from Graba instances of variations in length of the/32/tarsi in several sea‐birds & so it is with Anser Canadensis.5

Is. Geoffroy St. Hilaire6 has mentioned the case of a Monkey with an extra pair of molar teeth.7 Such cases, I may remark, are often called monstrosities; but if the teeth are well formed, I hardly see that they should be so called without every deviation from the normal structure be so designated. <Mr. Bellamy exhibited to Brit. Association in 1841, the head of Arvicola agrestis with fangs to its teeth, a character known to separate two groups of Mice.8 ) Dr. J. E. Gray has found considerable variability9 in the molars of certain seals. The form of the lower jaw seems also to vary1 considerably in Sloths. So according to M. De Blainville it is with the lower jaws of the Hippopotamus/

33/Dr. Andrew Smith2 in speaking of the antelope Cephalopus Natalensis, "the females are almost always found without horns, yet individuals are occasionally killed in which they exist; hence it would appear that their presence or absence ought not to be highly considered in establishing the generic characters."

In some species of Shrews (Sorex) & in some field‐mice Arvicolae, the Revd L. Jenyns3 found the proportional length of the intestines to vary considerably. He found the same variability in the number

1 Zoology of Voyage of Beagle: Birds p. 66, 67

2 Sir J. Richardson & Swainson, Fauna Boreali‐Americana. p. 27

3 Fulica Americana, in Richardson's Fauna Bor. Americ. p. 404

4 Ib. Birds. p. 58, 60, 80, 90

5 Ib. p. 469

6 [Here Darwin left an unfilled blank space for the citation.] See Histoire des anomalies, I, p. 660.]

7 [Here Darwin later added: 'Owen Ourang Outang'.]

8 [See Br. Ass. Rep. for 1841 (1842) part 1, p. 68.]

9 Proceed Zoolog. Soc June 12 1849 [see 'On the variation in the teeth of the crested seal...']

1 Ib. May 8 1849 [see 'On the genus Bradypus of Linnaeus'].

2 Illust. Zoolog. of S. Africa. 1849 P 1 . 32

3 Annals of Nat. Hist. vol 7. 1841. p. 267. 272.


of the caudal vertebrae. In three specimens of an Arvicola,1 he found the Gall‐Bladder having a very different degree of development, & there is reason to believe it is sometimes absent. Prof. Owen has shown2 that this is the case with the gall‐bladder of the Giraffe/

34/It has been long known that the presence of nails on the posterior thumbs of the <Borneo> Ourang3 is variable; & Prof. Owen has shown that with the nail there is an additional joint & bone. Prof. Owen informs me that he has seen a specimen having that muscle of the index‐finger, which has been thought characteristic of man; but in another specimen it ran to the second finger as well as to the index.

In Spiders, from six cases recorded by Mr. Blackwall4 the more or less complete absence of pairs of the eyes, & even the presence of a symmetrical superpernumerary one does not seem to be so rare a variation, as might have been anticipated in so important an organ.

In the sea‐urchins (Clypeastroida) the position of the anal orifice is highly variable, being even in the same undoubted species, sometimes above, sometimes below, & sometimes on the border of the shell.5 /

34 v/In many insects of several widely different classes, the presence of wings, is extremely variable within the limits of the same undoubted species; as in one British beetle Calathus mollis, in some Hymenoptera, & in several aquatic hemiptera.6 In a rare case described by Mr. Wollaston (p. 96) the connateness of the elytra varied.—/

35/It has been remarked by some authors, that the difficulty in determining what forms are really species, is due simply to want of knowledge. Undoubtedly this is often true, more especially in regard to the different stages of growth & sex of animals. But I suppose the Flora of Great Britain may be considered well‐known, & yet how differently is the number of species estimated by different authors! Mr. Hewett C. Watson informs me that after examining the London Catalogue (4th Edit) for this object, he finds that there are about 1800 names which have been considered by some Botanists as Species, but that out of this number, about 450 are

1 Ib. p. 272

2 [Darwin left space for citation. See Zool. Soc. London, Proc. 6 (1838), 10.]

3 Sir J. Brooke in Annals of Nat. Hist. vol. 9. 1842 p 58

4 Annals of Nat. Hist Vol XI. 1843. p 166

5 Agassiz & Desor in Annal. des Scienc. Nat. 3 series. Tom. 6. p. 318

6 Westwood Modern Classification of Insects Vol. 2. p. 431. & Wollaston Variation of Species. p. 43. 101


considered by other Botanists as mere varieties: moreover he has given me curious details, showing how opinions have alternated in successive periods two forms having been considered varieties, then species then varieties & lastly species again; these opinions being probably at no time unanimous. In certain Protean British genera the following table, published by Mr. Watson1 shows at a glance how unfixed is the criterion of a species.—And it particularly deserves notice that most of these genera in our own country/36/have been the subject of special monographs, sometimes by successive authors, who have devoted the closest attention to these genera./

36 v/ Salix. Mentha. Rosa. Rubus. Saxifraga.
Hudson (1791) 18 6 5 5 9
Smith (1824-8) 64 13 22 14 25
Lindley (1835) 29 9 17 21 24
Hooker (1842) 70 13 19 14 16
Babington (1843) 57 8 19 24 20
London Catalogue (1844) 38 8 7 34 16

["The table is intended to show the number of indigenous species in some of these genera, varying according to the author who describes and catalogues them." H. C. Watson, loc. cit.]/36 v/ Atriplex is another protean genus. The Rev. Leighton told me that he had some seeds of several species collected in various places in his garden, & that a mass of plants came up, which defied the powers of the two botanists most skilful in this tribe, to classify.—/

36/So again M. Ch. Des Moulins2 in his discourse on the well‐known Flora of central France, says that in 2332 phanerograms, there are still 250 forms under litigation.

I suppose no two land‐shells are better known than Helix hortensis & nemoralis. Mr. Bean3 of Scarborough has collected 152 vars. of H. hortensis 58 of H. pullata of some authors or the white-mouthed var. of this species; 236 vars. of H. nemoralis, & 21 of its variety or supposed species H. notabilis. Notwithstanding all this attention, & notwithstanding the fact, as I am informed by Sir C. Lyell, that H. hortensis ranges further north than H. nemoralis & is alone found in Canada, yet some great conchologists, as Deshayes doubt whether H. hortensis & nemoralis are not the same species./

1 Phytologist. May 1845. p. 143

2 Actes de la Soc. Linn. Tom. 16. 1849 p. 56

3 As quoted in Forbes Report Brit. Assoc. 1839 p. 136


37/To give another example, not so much to show that there is difficulty in deciding what form to call species & what varieties, but that even in a class, generally having such fixed characters as Birds, there is some appreciable amount of variation. In Germany, according to common authors, there are about 282 Birds, but Brehm1 by dividing species, adds to this number 576 species, making a total of 856 species: thus he divides the tit‐lark (Anthus pratensis) into 12 species & the Nightingale into 6 etc.—Now I have never met an ornithologist who thought these species worthy of consideration, & it has been asserted in Germany that many have been formed on single specimens.—On the other hand Brehm was a laborious observer: he collected2 more than 4000 skins, & he positively asserts that his new species are often found paired together, that they can be found on the same spot in successive years, & that they can often be distinguished by their voices & habits; & lastly that Bird catchers practically make similar distinctions. He grounds his distinctions chiefly on slight differences in the shape of the skull, beak, tail & feet. Though it may be very proper to/38/ignore these fine differences as specific, I can hardly doubt but that they exist. I believe this the more as our great ornithologist Mr. Gould has lately shown me some of our commonest birds from different districts, certainly presenting an appreciable difference.—

Lamarck long since remarked that there was not much difficulty in distinguishing species <from varieties> as long as specimens were brought from a single country,—not that this can be considered, as we have just seen, as always quite correct—but that the real difficulty begins when specimens pour in from every region inhabited by the genus.—Though this may be very true, yet with cautious & sound naturalists, how often do these numerous specimens if collected from continuous regions clear away doubts;3

1 Vögel Deutschlands 1831

2 Ib. Introduct p. xix

3 I am far from wishing to assert that this always the case: on the contrary I was formerly much struck, when witnessing Mr. Waterhouse (than whom a more <cautious> accurate naturalist can hardly be found) examining the large collection of Mice, which I made in S. America: when the specimens came all from the same exact locality, or from very distant localities, there was seldom much difficulty in distinguishing the species; but when a specimen or two had been collected at a moderate distance from any other locality, then I repeatedly observed there was very great doubt & difficulty. Exactly the same thing was noticeable in the difficult genus of Birds, Synallaxis, of which I collected many specimens.—Probably if I had collected still more numerous specimens, from every inter mediate station, there would have been less difficulty, but the difficulty would have been removed only by admitting considerable variations, or by designating every infinetesimal difference as specific.


but the doubts are generally dispelled by admitting considerable variation;—intermediate forms connecting others which might have been classed as specifically distinct. Hence apparently it arises that those who study local floras are apt to admit more forms as species, than those who take‐a wider field. But the/39/difficulty rises to a climax & indeed seems insuperable where very closely similar forms are compared coming from islands & from countries apparently now quite separated:1 I was much struck how entirely arbitrary the distinction is between varieties & species, when I witnessed different naturalists comparing the organic productions which I brought home from the islands, off the coast of S. America. In such cases there is no intermediate territory for the existence of intermediate forms; & the naturalist must rely wholly on analogy. North America & Europe offer the most striking example of this difficulty: let it be observed to what different conclusions the best naturalists have come to in regard to many quadrupeds, birds, insects & plants2 of these two quarters of the world; some

1 Instances innumerable could be given in regard to the islands of several great archipelagoes; & even from so small a one as the Galapagos group. Mr. G. R. Gray showed me some small pigeons (Peristera Macro‐dactylus, Brasiliensis, brevipennis &c) from the W. Indian Islands & mainland, which certainly differed <slightly> sensibly in length of wings, toes & plumage; & yet so cautious a naturalist as Mr. Gray, is strongly inclined to believe that they are only local races. I have quoted this instance, because Mr. Blyth has instanced in a letter to me another genus Treron in this family, as offering in the East the very same cause of doubt; but he leans to considering these slight differences as specific; for he remarks if these be given up, where can we stop; & well may he ask this: the answer in future years, will be, as I believe, no where.—It is known that large rivers in S. America form an impassible barrier to monkeys, & on their opposite sides the monkeys often differ slightly; these forms have been described by many authors as distinct species; but Dr. Natterer, a most careful observer who resided many years in Brazil (Note by Mr. Waterhouse in Annals of Nat. History 1844. vol. 13. p. 48) was convinced that these forms were only races of the same species See Mr. Wollastons works in regard to the insects of Madeira. See Mr. Layards & Blyths remarks in regard to the Birds of Ceylon.—

2 Dr. Asa Gray has lately published a truly admirable paper on the Statistics of the Flora of the Northern United States (<American> Journal of Science 2 ser. vol 23. p 80) & he gives a list of 15 varieties of plants common to Europe "which not only have been, but are not unlikely to be again distinguished as species", & another list of 42 N. American species, "almost all of which are more or less liable to be reduced to geographical varieties", of European plants. Had the United States been worked as carefully by local botanists as have the different parts of Europe, there can be no question, that a number of forms, which Dr. Asa Gray considers identical with European plants, would have been cut off by Botanists having less widely extended knowledge, as distinct North American species.

For Birds compare Prince Napoleons list with that by Sir J. Richardson & Swainson & other work. For Coleoptera, compare Mr. Murrays remarks on (Proc. Phys. Soc. Edinburgh in Zoologist. vols. 11 & 12. p. 3894) the differences, which he (as Kirby likewise) considers too slight to be specific, with M. Leconte in Agassiz Lake Superior p. [blank—see pp. 239*‐240*] who seems to consider that all? [sic] or nearly all, should be specifically separated.


calling the slight differences which can undoubtedly be observed in nearly all the animal productions from the old & new world, varieties, & some calling them species.

At present a considerable number of naturalists cut the knot by calling all forms from distinct regions, distinct species, even if the differences are excessively slight & even if apparently they are/40/identical. To those who rest on the hypothesis of distinct creation as the criterion of a species, this may be logical; but who can say what regions should be called distinct? Can we say we know all the means of distribution; past & present; as what part was land & what sea, & what was the exact temperature of either, within comparatively recent geological times? In regard to distance, as Mr. S. Haldeman & Wollaston1 have well remarked where shall we draw the line; if N. America & Europe are so distant from each other, that we may call their most closely allied inhabitants distinct species; are the Azores or Madeira sufficiently distant in regard to Europe to justify the same distinction. Must we extend the same view to Madeira & Porto Santo, within [ ] miles of each other, but with so many shells & insects quite distinct, & so many forms presenting marked varieties? Lastly must we extend it to Ireland & England, with only extremely few species distinct, but with some few, as generally considered, well marked varieties? Practically each naturalist arbitrarily decides the question for himself, in accordance/41/with his hypothetical idea of the term species, in accordance with what he knows of the amount of variation witnessed during the present time, & according to his tendency to trust in analogy.

We have seen that in the best known countries there is much uncertainty in deciding what to call species & what varieties. And further it seems to me that very generally if an animal or plant inhabits different districts or even if very common in one district, if it be conspicuous for any quality, or if it be valuable or in any way attracts man's notice, so as to be thoroughly well studied varieties will have been observed, & the more striking varieties will often have been considered as distinct species. Look to the King of beasts, as popularly called, how naturalists have doubted whether or not the Maneless Lion of Persia2 is a distinct species: Some few think that of Nubia also distinct; & the great lion‐slaughterer Mr. Gordon/42/Cumming is convinced that there is

1 Boston Journ. of Nat Hist vol 4. p 480. Wollaston Variation of Species p. 38.

2 Capt. Smee in Zoolog. Transacts. vol III [actually vol. I] p. 165 concludes that the Maneless lion of Guzerat is only a variety; I believe many naturalists now think it distinct. The Hyaena of Persia (Harlan's Researches p. 535) is, also, said to differ from that of Morocco only in wanting a mane.


more than one even in the Cape district.1 or look to the Elephant in India, but the variation in this animal is so curious that I shall presently enter into some little detail on the subject; as I shall on the well‐known & persecuted Fox of Europe. What disputes there have been in regard to the Bears of Scandinavia, there so ardently hunted, whether these there be one or more species. How many moles may a person casually examine without perceiving the slightest difference, yet being a thoroughly well known, animal, we hear from Mr. Bell, in his excellent history of British Quadrupeds (p. 106) that there are several remarkable varieties.2 The Sportsman3 can distinguish the Red Deer (Cervus elaphus)/43/of the different Scotch forests; "the Braemar deer are allowed to be quite different from those of Atholl, they stand higher & are in general of greater weight": those of Corrichebar are again different & have larger head than those of Atholl: the red deer of the outer Hebrides are very small4 So in Germany three varieties of this deer are distinguished & inhabit different localities.5 Other instances could be given as with the common Hare. So with Fish, it is certain that the salmon of many different rivers can be distinguished by fisher‐men; & the Herring which has been so closely studied, is found to present a vast range of variation.6 To descend lower in the scale; Fishmongers can distinguish whence their oysters come, & so they can on the coast of N. America with the clam, of which they distinguish five varieties:7 /

44/In plants most of those useful or much noticed by man are cultivated, & therefore do not come in here, as their variations may be all due to cultivation. To begin with a humble example; varieties of the water‐cress (Nasturtium officinale) are hardly noticed by botanists, but those who cultivate acres of this plant (not seedlings raised under cultivation) for the London market distinguish three varieties, which are not caused by any difference in the quality of the water, for they may be seen growing together; they differ in hardiness & other qualities; & the large brown‐leaved variety is the only one which will grow well, when the water is

1 Lichtenstein in his Travels vol 2. p. 31. says the country people distinguish three different sorts of Lions at the Cape.

2 The Rev R. Sheppard in Linn. Transact. vol. XIV. p 587, describes a remarkable variety with a white snout, & white line on the head, belly orange, forming a line on the chest; tail covered with long white hairs, & with the tip quite white.

3 See W. Scrope's Art of Deer Stalking,—a most interesting work. [c. p. 408.]

4 [Here Darwin left a blank space for a reference.]

5 Bechstein Naturgesch. Deutschlands. 1801. p 458.

6 Wilson's Voyage round Scotland vol 2. p. 206. The Herring Fishery was one of the points especially attended to in this voyage.

7 Venus mercenaria. Dr. Mitchill in Sillimans Journal of Sciences. vol 10. p. 287.


not very shallow.1 What is the tree, which ought to be best known in Britain? assuredly the Oak; yet I see that Mr. Babington, Hooker & Arnott with Dr Greville in their last Edition, treat Quercus robur & sessiliflora as varieties, whereas Dr Lindley in the Gardener's Chronicle speaks decisively of them as distinct species, & Sir James Smith seems to entertain no doubt on this subject. Every forester can distinguish the two forms: it is asserted that they come true to seed2 /45/though this has been denied: the quality of their timber is said to be different3 & Quercus sessiliflora is hardier & ascends the Scotch mountains higher than Q. robur.4 On the other hand the existence of a perfect gradation of inter‐mediate forms is admitted by everyone & Dr. Bromfield quotes5 with approval the remarks of another most careful observer Mr. Bree that 'though there are sessile oaks bearing fruit on peduncles & pedunculated oaks bearing almost sessile fruit, there is yet a certain indescribable something about the trees, by means of which I can always distinguish each, without minutely examining either the acorns or the leaf‐stalks.' So that according to these two excellent observers the distinction of the two varieties or two species, (& the highest possible authority can be quoted for either term) of our one most conspicuous tree can be best recognised, like a man's face, by 'a certain indescribable something.'

It would be superfluous to give other examples; but parallel ones could be given in regard to our Elms, to our Birches, & most striking ones/46/in regard to the Scotch Fir, in which the varieties or species, call them which you please & you will have high authority for doing so, are adapted to different situations, produce different kinds of timber & are hereditary in their quality.6 In the Yew, the highest authority Dr. Asa Gray thinks the Canadian form perhaps only a variety; & this seems the general opinion of Botanists in regard to that most remarkable bush, the Irish yew, found growing wild in Ireland with its upright dwarf habit, & large scattered leaves; but I presume, if this plant had been found of both sexes, growing abundantly in some distant region, no

1 Mr. Bradbery account of the cultivation of the water‐cress. [See Ker,] Transact. Hort. Soc. Vol IV. p. 537.

2 Gardener's Chronicle [1855, p. 776; 1856, 191‐2, 405.]

3 Sir J Smith English Flora. vol IV. p 149 & Gardener's Chronicle [1856, pp. 191‐2]

4 Mr. Farquarson in Hooker's Bot. Misc. vol 3. p 127.

5 Phytologist Vol. 3. p. 883.

6 In regard to the Elm, see Dr. Bromfields remarks in Phytologist. vol 3. p. 837. Mr. H. C. Watson exhibited before the Bot. Soc. of London (Annals of Nat History. vol 12. 1843 p 450) specimens showing that Betula alba, pendula, glutinosa, & pubescens are all mere fleeting varieties of the common Birch. For the Pinus sylvestris, see Loudon's Arboretum p. 2189 & 2150 & Gardeners Chronicle [ ].


botanist would have hesitated to name it as a distinct species. The last example which I will give is that of the noble Cedar of Lebanon: it appears in our gardens most/47/distinct from the Deodar, yet when old, Botanists cannot point out any good character between these two forms & the Cedar of the Atlas, & as the seedlings vary hence are inclined to consider them as varieties, a conclusion indignantly repudiated by other Botanists.1 The question in these several cases, is not whether these forms deserve a name, popular usage has settled that point, but whether they should be designated by the undefined title of Species.—

Incidentally several cases of variation in a state of nature have now been given, & incidentally others will be hereafter given. It would be as easy as useless to quote the almost numberless instances of forms, which have been considered on good authorities as permanent varieties having much of the character of species; & I will conclude this chapter by giving from various motives, a few additional instances of variation, in which the evidence is rather better than in most cases./

48/Indian Elephants. Dr Falconer who has had great experience in Elephants, & who has seen as many as 1200 at a fair, informs me that they differ considerably, more than horses of the same breed, in size, general proportions, manner of carrying the head, form of tusks, shape of feet & in the absence of the nail on one toe: Mr Corse has given a nearly similar account2 & says that the different castes have their proper names. In the Ayeen Akbery, written about the year 1600, four kinds of Elephants are specified. Most of these differences probably come under our class of merely individual differences; but both Dr. Falconer & Mr. Corse believe that some of the breeds inhabit different, adjoining districts; & animals which are thought to be cross‐bred, are occasionally caught. As far as size is concerned, climate appears influential; at least, as I am informed by Mr. Crawfurd, elephants northward of a certain latitude are excluded by the government contracts. Dr. Falconer tells me that there are two marked breeds, one thicker in its general proportions, more courageous, & with short tusks directed downwards; in the other breed, the tusks are upturned, & the/49/animal when attacked by a tiger tries to pitch his opponent into the air; whereas the breed with the downward directed tusks, when attacked, falls as if instinctively on its knees, & endeavours to crush & pin the tiger to the ground; this breed is consequently more dangerous to ride, as sometimes even experienced hunters

1 [Darwin here added in pencil: 'Hooker—Gardeners Chronicle'.]

2 Philosoph. Transacts 1799 p. 206.


are thrown on to the tiger. Now such differences in structure & habits, I think all zoologists, will agree, would in most cases be thought of specific value; but I believe no one has even suspected that there are two species in India. In Ceylon, there is, also, a distinct breed, but this has by some1 been thought to form another species. Until quite lately the Elephant of Sumatra, was thought to be the same,2 but now from differences in its skeleton it is thought to be a distinct species./

50/Foxes. These are well known to be variable animals & all over the world the species are discriminated with difficulty. British sportsmen speak3 of three kinds, but it is doubtful whether these are anything but individual indifferences [sic]. In Scotland the accurate Macgillivray4 describes four kinds, but he uses besides general proportions the tail being tipped with white, which Bechstein5 has shown is a quite variable point. But the Highland or mountain Fox of Scotland seems certainly to form a distinct race: Mr Colquhoun6 a very good observer says any one can distinguish this animal even at a distance from the small fox of the low grounds; he stands higher, his head broad, nose not so pointed, his coat more shaggy & mixed with white hairs: he is much more powerful & preys on young sheep, & rears his young, not in holes, but in clefts in the rocks; is less nocturnal in his habits,/51/& altogether, as Mr. St. John remarks, is more like a wolf, than a lowland fox. In Scandinavia it has been a question disputed both by naturalists 6 hunters, whether the common red, the black & crucigerous Foxes are distinct species or only varieties. So in N. America a parallel series occurs & it has been disputed whether the red Fox, (ranked as a different species from that of Europe) the black & silver & crucigerous (ie with a dorsal stripe & a transverse one on the shoulders) foxes are distinct or not: Sir J. Richardson7 inclines to consider them all as varieties. So much interest has this question excited in Scandinavia as the differences are said not to be confined to colour alone that a fox colony was established by some gentlemen near Stockholm8 & in it two crucigerous foxes produced in the course of four years 19 cubs; of these 9 were crucigerous; 8 were black (including those with white tipped tails),

1 Mr. Hodgson in Asi[a]t. Soc. of Bengal vol I (1832) p. 345.

2 Crawfurd Descriptive Diet, of Indian Islands. 1856 p. 136.

3 Encyclop. of Rural Sports. p. 448.

4 Transacts of Wernerian Soc. vol VII. p. 481.

5 Naturgesch. Deutschlands. B I. s. 627.

6 The Moor & the Loch. p. 97. Ch. St. John, Wild Sports & Nat. History of the Highlands p 232.

7 Frauna Boreali‐america p 93.

8 L.Lloyd Scandinavian Adventures vol II. (1854) p. 52.


& 2 red: two of the black cubs, also, produced young & these, six in number, were all black. Mr. Lloyd infers from these experiments that the crucigerous fox is a cross from the black & red, which he seems to consider, with/52/many of the inhabitants, as distinct species, producing, as it thus seems perfectly fertile hybrids, but Prof Nillson's conclusion that they are proved by these experiments to deserve only the name of varieties, seems to me the most probable. It is clear that these variations are in some degree hereditary, & the whole case is interesting as showing how difficult it is to decide what to call species & what varieties. The occurrence, also, of strictly analogous varieties in N. America from the generally received distinct American & Arctic Foxes (C. fulvus & lagopus) is an interesting fact; & the more interesting from these forms, not being produced in Great Britain, though they are, according to Bechstein, in Germany./

53/Raven. It has long been known that pied Ravens are found at the little islands of Faroe. This bird is white somewhat symmetrically marked with black, & as the beak "is much larger being not only higher at the base, but more elongated, & in form more attenuated at the end" than that of the Ravens,1 it has been admitted by Brisson, Vie[i]llot, Wagler, Temminck, & others the most distinguished ornithologists, as a distinct species under the name of Corvus leucophaeus./53 v/As this particular race is known no where else, (though partial albino ravens do occur elsewhere) this fact has been used as an argument that it is a distinct species; but perhaps the argument might be reversed with equal force, as not one other bird or indeed other production is endemic in this small spot./53/When, however the ornithologist Graba visits these islands, & investigates the case he finds that the pied ravens (at first quite white, the black feathers appearing with age), are produced in the same nest with ordinary ravens; & that in one case when black & pied were mated either exclusively black birds or one white bird with the others black were produced.2 The fact of the black & pied ravens being sometimes/54/mated & producing either black or white young, is not, as we shall immediately see in the case of the Hooded crow, so conclusive as Graba seems to think; but combined with the white appearing in the nests of common ravens, & more especially with the fact of the

1 Magillvray [sic] History of British Birds. vol. 3. p 745.

2 Tagebuch auf einer Reise nach Färo 1830. p. 51. Graba's description of the beak a nearly agrees with that of Macgill[i]vray. I may add that Landt in 1810 in his Description of Feroe p 220, says that black & speckled ravens are sometimes seen paired & that both kinds are sometimes found in the same nest.


two birds described by Graba, the one by Macgillivray, & that by Temminck, differing very considerably in their colouring, even sometimes on opposite sides of the same individual, I think this can leave no doubt that the C. Ieucophaeus is only a variety. Graba says that they are not very rare, & he states the interesting fact of which he was a witness that the pied birds are persecuted & driven away by the common ravens (p. 51, 54.); & Macgillvray once saw on the Hebrides a bird of this kind, apparently a wanderer, which he describes as "a neglected & persecuted stranger". Now suppose whatever the cause may be, which gives rise to this variety in Faroe to act with rather more intensity, so that pied ravens alone were to hold possession of these islands, how utterly impossible it would be ever to ascertain whether it was right to call this form a variety or species./55/No doubt any chance wandering black raven would be persecuted & driven away by the pied majority, as these latter now are by the black birds; & crosses being thus prevented, it is probable that the pied colouring & other characters would become in the course of many generations more fixed & constant.

Now let us turn to the Carrion & Hooded crows (Corvus corone & comix): these birds are so much alike that as Magillvray observes1 "were the colours the same in both it would be almost impossible to distinguish them". "The extent & tint of the grey‐coloured space varies greatly in the Hooded crow ["]2 & Bechstein asserts that in Siberia some are quite black, but how these can be distinguished from carrion crows I know not. The eggs of the two species are undistinguishable as are their digestive organs & their general habits are alike. Numerous cases are on record in Germany, England, Scotland & Ireland of these two forms being seen paired, & the young are either/56/quite like one of the parents or intermediate in colour.3 Hence several respectable ornithologists have looked at these birds as varieties; yet, as their voice is slightly different & as different districts are often inhabited separately by either one or the other form; & as when occurring together they keep separate; & as the carrion crow seems to have a more southern range than the Hooded crow & more especially as ordinary specimens of both can be distinguished with the utmost facility, I must agree with Mr. Macgillvray that, in common parlance, "the two species are perfectly distinct".

1 History of British Birds vol I. p. 529.

2 Ib. p. 534.

3 Bechstein, refers to three cases in his Naturgeschichte B. 2. s. 1170; Mr. Slater informs me that he has known of a case in Hampshire W. Thompson gives cases [See Nat. Hist. Ireland, vol. I p. 309] Macgillivray vol 3 p. 721 gives cases in Yorkshire & Scotland.


Lastly let us consider one other case; we have in Britain one single well‐known bird, the red Grouse, (Tetrao Scoticus) which has been almost universally ranked as a distinct species, & is confined to the British islands. On the other hand the Tetrao saliceti of Scandinavia, is a bird which we might have expected to inhabit Great Britain, but is not found here./57/Gloger alone, as I believe, has argued at length1 that they are certainly only local varieties of the same species.—Mr. Gould after studying T. saliceti in Scandinavia tells me that they agree perfectly in eggs, in the immature plumage, in habits, in voice & in summer plumage, with the exception of the white primary feathers & that he cannot avoid the suspicion that they may possibly be varieties./ 57 v/The Red Grouse is very variable in plumage, & easily runs into sub‐local races.2 Macgillvray says that it differs from T. saliceti in having a lesser beak; but Nilson, as quoted by Gloger says he examined 30 specimens of T. saliceti, & the beak was scarcely alike in two./57/I apprehend if these birds had been found together, & it does not seem improbable that colonies of the one might now be established in the territory of the other; no ornithologist whatever would have thrown a suspicion on their specific distinctness; hence their geographical separation & consequent exposure to a different climate seems to have been the sole cause of their specific diversity having been suspected; & undoubtedly as Britain has no other endemic bird this is an argument of some apparent weight <in favour of the two forms being identical>: on the other hand if we had possessed a few more endemic species the argument might have been reversed, notwithstanding it might most truly be said that every gradation exists in the proportional number of endemic forms possessed/58/by a country, & why should not insular Britain possess its single endemic Bird?

I have entered into these three last cases at some little length in order to show how difficult it is to determine what to call a species & what a variety, even with using all sorts of collateral evidence in well‐known Birds, which are amongst the least varying animals. The series seems to me an interesting one, from the pied & black ravens which must be considered as varieties, though hitherto esteemed by most ornithologists as distinct & which inhabit the same little island, but with some tendency to keep separate—to the carrion & Hooded crow, considered by a vast majority of ornithologists as distinct, often inhabiting distinct regions, but when mingling, often crossing—to the red & willow grouse,

1 Das Abandern der Vögel 1833. p. 117.

2 Macgillivray British Birds. vol I p. 174 & p. 186.


almost universally considered as distinct & inhabiting quite distinct countries, but yet with a taint of suspicion hanging over them./

59/As it is so rare that varieties of Birds, sufficiently distinct to have been esteemed species by first rate naturalists can be proved to be not distinct that I will give one more case. The common & ring eyed Guillemot (Uria troile & U. ringvia or lacrymans) have been by about an equal number of ornithologists esteemed as good or doubtful species or as mere varieties. The ring‐eyed form inhabits the northern islands & is generally rare; but Graba found that the Faroe islands1 were its home, one out of about five existing as this form; at first Graba thought it was specifically distinct, for besides the conspicuous ring of white round the eyes & from the eye backwards, it differed in other respects; but these differences Graba found were not constant, & he subsequently himself twice saw it paired with the common Guillemot; & the inhabitants affirm that sometimes from the two eggs of the common Guillemot, one will be ring‐eyed.

In Madeira, there is only one endemic bird, but some of the European birds are slightly smaller, & some are slightly duskier; & the Redpole (Fringilla cannabina) retains/60/its crimson breast throughout the year.2 The black‐cap (Sylvia atricapilla) besides being sometimes duskier, presents a variety, in which the black colour extends from the cap to the shoulders & occasionally even over all the under parts of the body; this has been described by so good an ornithologist as Sir W. Jardine as a distinct species; but as the inhabitants believe that it is produced from the same nest as the common black cap, there cannot be much doubt that Dr. Heineken & Mr. Harcourt are right in esteeming it as a variety.

I will now give a single case in Fish taken from Bronn;3 the Cyprinus gibelio & carassius have generally been considered distinct species, for they differ in almost every part in proportion, as shown by the table given by Bronn; but Eckstrom narrates that the offspring of the C. carassius removed from a large lake into a small pond, assumed an intermediate form; & on the other hand the offspring from C. gibelio from a small pond turned into a large lake 40‐50 years before, had become changed into C. carassius./ 61/I have selected the foregoing instances, from being able to adduce some other evidence besides the mere existence of a graduated series of intermediate forms. But I will now give two instances from Mr. Wollaston's works of Variation deduced from

1 Tagebuch. p. 106, p 150.

2 E. Vernon Harcourt Annals & Mag. of Nat. History. June 1855 and Sketch of Island of Madeira 1851.

3 Ges[ch]ichte der Natur. B. 2. s. 106.


intermediate forms observed with sepecial care by this excellent entomologist in the confined locality of Madeira. Harpalus vividus1 is perhaps the best example; if very many specimens from many sites had not been collected, <clearly showing a perfectly graduating series,> the varieties would have been described as forming several species; those from the lowland & the wooded mountain slopes appearing "altogether distinct". It is an interesting fact, that it attains its maximum of sculpture & minimum of size at about the elevation of 3000 to 4000 feet; both above & below which height, "as it recedes from the upper & lower limits of the sylvan districts, it becomes gradually modified, & almost in a similar manner". It varies greatly in colour/62/shape, in puncturing & in striation & what is even more important in the degree to which the elytra are soldered together: the united elytra are found only rarely in the sylvan districts. This beetle, also offers an instance, of which very many could be cited in the most distinct genera, namely of the individuals inhabiting the rocky islet called the Deserta grande, attaining a larger size than elsewhere. To take another very different genus of beetles, namely Ptinus2 in which some species of which "do not attain half the bulk on many of the adjacent rocks, that they do in more sheltered districts; & so marvellously is this verified in a particular instance, that I have but little doubt that five or six species, so called, might have been recorded out of one". Ptinus albopictus has a separate radiating form on every islet of the group, but all merge together by innumerable intermediate links. Very many other examples might have been adduced of each islet & even rock of different altitude having its separate variety./

63/ Plants

Centaurea nigrescens has been separated by some botanists from C. niga, (the common Knap‐weed) by several characters, of which the most conspicuous is that the heads are rayed. The Rev. Prof. Henslow informs me that this form kept true for two generations in his garden, but that in the fourth year it was clearly reduced to C. nigra. I mention this case, because, the var. C. nigrescens, as I am informed by Prof. Henslow occupies nearly the whole of Hampshire to the exclusion of the common forms; and here we have the argument from range, on a small scale, as with the Red Grouse of Britain, which may be used on either side.

1 Insecta Maderensia p. 54: The Variation of Species. p. 67.

2 Insecta Maderensia p. 260, 267. For other cases see p. 11, 30, 36, & 78.—


Koch raised the ensuing year from seeds of a dandelion (Taraxicum palustri)1 T. palustri, T. officinale, T erectum, T nigricans, & T corniculatum,—forms which have been admitted by some Botanists as species, & two of which were first named by De Candolle. Prof. Henslow on the other hand, though not doubting that T. palustre, is a variety, has found it/64/come up true for three or four generations when self‐sown in his garden Koch has, also, raised from seed of one species of Isatis2 tinctoria, campestris, praecox, dasycarpa,—forms as species by De Candolle, Ledebour & other distinguished Botanists: most of these forms inhabit different parts of Europe & Siberia. From cultivating another cruciferous plant Sisymbrium austriacum, Koch concludes that S. eckart[s]bergense, Willd. & taraxacifolium & acutangulum, both of De Candolle, are only varieties.

Mr. Hewett C. Watson is one of the few British botanists who has experimentally tried to test species by cultivation; thus he has succeeded in raising on plants of Festuca loliacea "stems which a botanist would assuredly have assigned to F. pratensis";3 & he almost succeeded in running together the common & Italian Rye grass (Lolium perenne & multiflorum).4 But Mr. Watsons experiments on seeds & living plants which he collected at the Azores, are particularly interesting: thus plants raised from Azorean seed of the Polygonum maritimum "partook much of/65/ the physical characters of P. Raii from the shores of Great Britain". Seeds of the Tolpis crinita from the Azores, produced plants undistinguishable from T. umbellata; yet these plants differ in the pappus of the fruit, in a manner on which distinct genera have been founded by some authors.5 Again Mr. Watson has found6 that cultivation during four generations in England of the forms of Raphanus raphanistrum found in the Azores has partially obliterated a character in the pods which was at first obvious. The rich deep colour of Myosotis Azorica7 tends to fail in our country; & the seedlings have varied so much that Mr. Watson is unable to say which should be referred to M. Azorica & which to M. maritima; & some approximate to the Canary species, M. sylvatica: yet in their wild state they were as easily distinguished as any other/66/species of the genus.

1 Annal. des Scienc. Nat. 2 Series. Bot. Tom 2. p. 119.

2 Annal des Scienc. Nat. Bot. 2 series. Tom. 3. p. 375.

3 Phytologist, June 1845. p. 166.

4 I may add to these cases of conversion in Graminea that [of] Bernhardi (Ueber den Begriff der Pflanzenart. 1834. s. 30) that by repeated sowings Panicum ciliare was perfectly changed into P. sanguinale.

5 [Watson,] Phytologist. 1845. p. 167.

6 London Journal of Botany 2d ser. vol. 6. p. 385

7 Ib. p. 388.


The accurate Kölreuter1 asserts that he has seen the Digitalis thapsi, when cultivated in northern Europe, & when artificially fertilised, so as to preclude any possibility of a cross, after four or five generations assume the characters of D. purpurea, & at last was completely converted into it. The hybrid offspring from the reciprocal crosses of D. thapsi & purpurea were perfectly fertile. Dr Lindley in his Monograph on Digitalis expresses some doubt whether Kölreuter may not have taken a variety of D. purpurea for D. thapsi, but as he speaks of his D. thapsi, as that of Spain, he may probably be trusted. These two forms are considerably unlike in many respects, & have generally been received as good species.2

E. von Berg gives a curious account3 of the extreme variability of the seedlings of cultivated plants of Iris so that Dr. Horns[ch]uch4 asserts that he raised twenty reputed species from Iris sambucina or Germanica; I confess that owing to some other recorded experiments of E. von Berg/67/I should have thought that there had been some mistake here, had not his results in the case of the genus Iris been strongly corroborated by quite independent testimony. For M. C. Bouchés5 by sowing seeds of I. Germanica raised 13 reputed species; & what is important for us, three of these, namely I. florentina, Germanica & pallida are Linnean species & have been found growing in separate districts, & in their own native habitats remain unaltered.

The blue & red pimpernel (Anagallis arvensis & coerulea)6 have by a good many botanists been considered as distinct species, for besides in the colour of the flower, they differ in some other

1 Journal de Physique Tom 21. p 291.

2 In Loudon's Arboretum vol. 3. p 1374 it is stated that Mr. Masters of Canterbury a great raiser of Elms, & therefore one who ought to judge well, is convinced that the Ulmus Americana is identical with the Huntingdon Elms, a variety undoubtedly of English origin. In Bronn's Ges[ch]ichte de[r] Natur B. 2. p. 85, there is a marvellous account of the change of a plant of Lobelia lutea into L. bellidifolia; & by Link of Ziziphora intermedia from Z. dasyantha, & of a great change in Ribes alpinumi. Mr. Gordon of Birnie in his Flora of Moray‐shire [p.iv] says the 'Avena pratensis is confined to soils of this description (calcareous), changing its habits, as the proportion of their ingredients differs. Where there is a super abundance of limy matter, the plant often assumes a glaucous‐rigid appearance, which has probably originated the A. alpina and causes it still to hold a place as a distinct species.'—

3 Flora 1833 Beiblatter & 1835. B. 2. s. 564.

4 Flora 1848 p 55.

5 Flora 1833, Nachschrift, Horns[ch]uch. s. 44.

6 [Here Darwin later scribbled in pencil: '1861 a new var. Eugenia [?] I read [?] latterly [?] came red and blue

'In 1867 Red & Blue var[ietie]s of A. grandiflora produced both var[ietie]s & intermediate—See notes on crossing plants.

'1867 Both vars. extra fertile when crossed.']


respects./67 v/It is certain that each kind can be long perpetuated by seed & keep true.1 On the other hand/67/the Revd . Prof. Henslow's experiments,2 though nearly can hardly be considered absolutely decisive, in showing that one form can be raised from the other: Dr. Bromfield3 has seen bright blue & flesh‐coloured flowers on actually the same plant, when cultivated in a garden. Dr. Asa Gray says that in the United States whither this species has been introduced all the coloured varieties are met with, having flowers of variable size. Bernhardi4 /68/says that it is almost certain (& I have received corroborative evidence) that the allied Anagallis collina produces blue & red flowered varieties. Considering these several statements the probability seems to me strong that the A. coerulea & arvensis should be considered only as varieties. I have alluded to this case chiefly owing to the remarkable fact, that Gaertner with all his experience failed after repeated & reciprocal trials5 to raise a single hybrid between these two forms, whence he concludes that they are distinct: Herbert succeeded with Anagallis collina; & if Gaertner had shown that he could artificially fertilise either variety with its own pollen one would then have had more confidence in his result.—

The most interesting case on record is that of the Primrose, <common> oxlip Bardfield oxlip & cowslip (Primula vulgaris elatior & veris). These plants differ, as everyone knows, in their flowers foliage & habit; they all three differ in the forms of the capsule & seed:6 the primrose & cowslip have a/69/different scent: they flower at somewhat different times: they ordinarily inhabit different stations, the cowslip in open fields & the primrose on banks & in shaded woods, but they are sometimes mingled; they abound in different districts in different proportions;/69 v/& in Switzerland the P. vulgaris & elatior ascend to different heights the primrose being the more tender.7 They have, also, different geographical ranges; Dr. Bromfield has remarked8 "that the primrose is absent from all the interior regions of Northern Europe, where the cowslip

1 [Teesdale,] Linnean Transactions vol. 5. p. 44 & [Wiegmann] Flora 1821. B. 1. s 15.

2 Loudon's Mag. of Nat. Hist. vol. 3. 1830. p. 537, but compare with vol. 5. p. 493.

3 Phytologist vol. 3. p. 699.

Begriff der Pflanzenart p 9.

5 Bastarderzeugung s. 309. [On the verso of this folio Darwin asked: <"could Gaertner have by chance tried only a male plant or with female pollen??") and later cancelled this query.]

6 Revd . W. Leighton in Annals of Nat. History vol 2. 2 series. 1848. p. 164.

7 Annals of Nat. History, vol. IX. 1842. [P. J. Brown,] p. 156. & [E. Doubleday,] p. 515. See, also, Boreau Flore du centre de la France 1840. Tom 2. p 376. and Hooker's & Arnotts British Flora 1855, on the rarity of P. veris in Scotland.

8 [E. Doubleday,] p. 515. Phytologist vol: 3. p. 694.


is indigenous": & Messrs. Bentham & Hooker inform me that in the East, the primrose is found only in the Caucasus; that the oxlip ranges from the Caucasus to about the latitude of Moscow & the Cowslip from the Caucasus to four degrees northwards to the latitude of St. Petersburgh.1

Lastly Gaertner laboriously experimentised on these several forms during four years, & actually castrated & crossed no less than 170 flowers, & yet <strange to say> he only twice succeeded in getting any good yet scanty seed2 /

70/He expressly states that the Primulaceae offer no mechanical difficulties to crossing,3 but yet it would have been far more satisfactory if he had shown that he could artificially fertilise a Primula with its own pollen. On the supposition which seems to me most probable that this extreme infertility is not real, but only apparent, & caused by some want of skill or knowledge, we have, nevertheless, as good as, indeed far better evidence than is attainable, in most cases, of the infertility of these forms together,—seeing how perseveringly the experiment was tried by the most practised operator who ever lived.—

Considering these several statements, it seems to me difficult to imagine better evidence than in this case that the primrose & cowslip deserve to be called distinct species. But now let us look to the other side: it is universally acknowledged4 that in England there are so many intermediate forms found wild that it is most difficult to draw any strict line of demarcation between the two extremes of the primrose & cowslip. And what is the result of the many experiments/71/which have been made? Several years ago, the Hon. & Revd . W. Herbert5 raised from the seed of a highly manured red cowslip, a primrose, cowslip, oxlips of various colours, <a black polyanthus,> a hose‐in‐hose cowslip, & a natural primrose bearing its flowers on a polyanthus stalk: from the seedling hosein‐hose cowslip, he raised a hose‐in‐hose primrose. Subsequently the Revd . Prof. Henslow6 doubting Mr. Herbert's experiment

1 In Britain see Cybele Brit. W. C. Watson Cybele Britannica vol. 2. p. 293. says only on range of Primrose & Cowslip that "P. veris would seem to be an uncommon plant in the W. of Scotland".

2 Bastarderzeugung. s. 721; & s. 178; but the table is not quite correct for a cross is mentioned at p. 247 not introduced into the table.

3 Le[c]oq[?]—at Maer gardener assured me he had known whole bed of Polyanthus spoilt by P. veris growing near. [See Lecoq, De la Fécondation naturelle..., Paris, 1845.]

4 See Phytologist vol. 3. p 43. for some excellent observations on the intermediate states by Mr. Watson.

5 Transactions of the Horticult. Soc. vol. IV, p 19.

6 Loudon's Mag. of Nat. Hist. vol. 3. 1830. p. 409.


took the seed of some cowslips growing in a shady part of his garden, & raised seedlings which varied considerably, approaching more or less closely to certain wild oxlips <which Prof Henslow had observed>; "& one was a perfect primrose". These experiments were not thought sufficient;1 & that most critical observer Mr. H. C. Watson raised at several periods many seedlings, from the cowslip, (P. veris), from a Claygate oxlip, & from an oxlip, truly intermediate in most points, but with/72/the primrose predominating, & the conclusion at which he arrives2 is "that seeds of a cowslip can produce cowslips & oxlips; & that seeds of an oxlip can produce cowslips, oxlips & primroses."

The experiments of Mr. Sidebotham3 are, perhaps, the most important of all, for the plants from which he procured seed, were covered by bell‐glasses & so crossing was prevented. He performed all the operations with his own hands. Moreover he experimentised on the Bardfield oxlip (P. Jacquinii or P. elatior of Jacq.), which has very generally been received as a third distinct species; though in this case, as with the common oxlip, Mr. Watson & Dr. Bromfield4 have "seen exceptional instances to all the characters, taken singly, by which this plant is distinguished from P. vulgaris & P. veris"; but Dr. Bromfield admits that it certainly has much the air of a distinct species. /73/Mr. Sidebotham's experiments were as follows, & they are the more important as he was a hostile witness, & confesses that the experiments "disappointed me greatly & interfered very materially with my previous idea of specific identity".

These experiments bring out clearly the hereditary tendency in all five forms. Both here & in Mr. Watson's experiments there is no direct passage from a true cowslip to a primrose or reversely; but Mr. Herbert experimentised on a cultivated red cowslip, highly manured, & from it he raised "a natural primrose on a polyanthus

1 <Some nurserymen (as I have been myself informed) are convinced that such changes take place in their seed beds, others have strongly denied them, as in Gardener's Magazine vol. VII p. 123, 247>.

2 Phytologist Vol. 3. p 43, and vol 2. p. 217. p. 852.

3 Phytologist vol. 3. p. 703 [Darwin later rejected Sidebotham's claims. Note sheet 69v on H.C.Watson's Cybele Britannica states: 'Vol 3. p. 488—doubts Mr. Sidebotham experiments, so I had better not speak so enthusiastically—doubts them from want of general accuracy—& from his want of Botanical knowledge,—relates more especially to P. elatior from P. veris; & to P. vulgaris from P. elatior Allows they support P. vulgaris P. veris coming from an intermediate form, & reverse case.—Thinks he did not take sufficient precautions, what I know not—Express a doubt about P. elatior.—'. See also The Different Forms of Flowers on Plants of the Some Species. 1st ed., London, 1877. note on p. 60, where Darwin states that these experiments 'may be passed over as valueless.']

4 Phytologist vol 1. p. 1001 & vol. 3. p. 695.


Names of seedlings produced Seed from P. veris produced (cowslip) Seed from P. veris, var. major produced (oxlip) Seed from P. vulgaris an intermedia produced (Claygate oxlip)1 Seed from P. vulgaris produced (primrose) Seed from P. Jacquinii = P. elatior of Jacq. Produced (Bardfield oxlip)
P. veris (common cowslip) 412 9
P. veris var. major of Lond. 30 21
Cat. (oxlip) (3 being hose-in-hose)
P. veris approaching Polyanthus (oxlip) (some hose-in- hose) some dark-coloured 13 20
P. vulgaris var. intermedia (Claygate oxlip) 7 19
—var. caulescens 3 3 1 1
P. vulgaris (primrose) 2 1 15 1
P. Jacquinii (Bardfield oxlip) 1 24
Plants producing no flowers. 18 2 4 2 6
Total number of seedlings raised 473 65 27 18 32

[Table compiled from lists in Sidebotham]

stalk" & again on the succeeding year from his seedling hose‐in‐hose (calycantha) cowslip he raised a hose‐in‐hose/74/primrose. The Revd . Prof Henslow's cowslip, whence he raised "a perfect primrose", was a garden plant & grew in a shady place. It goes for nothing that some authors have planted seeds, especially if gathered from wild plants,2 & have found that all the seedlings, have come true to their kind; it only shows how true the kind is, when not disturbed by cultivation.—

No one, I believe, has disputed the accuracy of the statements of these four Botanists, Messrs. Herbert, Henslow, Watson & Sidebotham; three of whom, I may add, commenced their experiments in a sceptical frame of mind. But the results have been attempted to be explained away by the supposition of the intercrossing of the several forms. Now laying on one side Gaertner's laborious & careful experiments, (which nearly all failed,) &

1 From seeds of this form Mr. Watson (Phytologist vol. 2. p 218. & Dr. Bromfield remarks vol 3. 69 [695?]) raised 88 seedlings of which 63 were intermediate, 5 were genuine cowslips & 20 true primroses.

2 Phytologist vol. 3 p 180.


assuming that insects could effect, that which he could not; do the results agree with this view of crossing? It seems to me most decidedly not. Mr. Sidebotham expressly/75/states that he protected his flowers by glasses; & this having been done, it seems quite incredible that there should have been so much crossing in all his five cases. <indeed apparently as much variations in the offspring in most of the experiments.> Moreover on the mountains of Switzerland the P. elatior or supposed hybrid between P. veris & vulgaris grows "by thousands in places within many leagues of which the P. vulgaris is absolutely unknown".1 so it must be with the oxlip from its Northern range in Russia; so with the oxlip (or P. Jacquinii) of Bardfield, round which place "the primrose does not occur for some miles".2 Lastly, & I may venture to say that I speak after a careful study of all well ascertained facts on Hybridism, there is no known instance of one species fertilised by the pollen of another species producing pure forms of both or either parent as must have occurred on this view with Herbert's & Henslow's cowslips & with Mr. Sidebotham's P. Jacquinii, if they had been fertilised by the pollen of the primrose. Moreover the common oxlip, or supposed Hybrid between the primrose & cowslip, yielded, as we have seen in Mr. Watson's & Sidebotham's/76/ experiments, various oxlips & pure primroses & pure cowslips; whether we choose to imagine these hybrids were self‐fertilised, or were fertilised by either pure supposed parent, so sudden & absolute a reversion to either or both parent‐forms is <in the case of species> without any known analogy3 in carefully recorded experiments on the crossing of species. From these several & combined reasons I think we are justified in absolutely rejecting the view that all the forms produced in the foregoing several recorded experiments, & likewise existing in nature, can be accounted for by the crossing of two or three aboriginally distinct species; their origin I think must be attributed to variation, but I am far from wishing to assert that some or many of the graduated intermediate forms may not likewise be in large part due to their having at some time crossed, which no doubt would increase their variability & probably aid in their tendency to reversion to either one or both of the parent varieties./

77/In all the experiments, the common oxlip seems the most

1 [P. J. Brown,] Annals of Nat. History vol IX. 1842. p 156.

2 [Doubleday,] Annals of Nat. Hist. vol. IX. p. 515.

3 The well‐known & marvellous case of Cytisus adami would be analogous in the individual, (though not in seedlings) if it could be shown that this tree was really a Hybrid: some competent judges firmly believe that it was produced by the union of two buds of the two species.—


variable form; though the cowslip is sometimes little less so, for in Prof. Henslow's seedlings "not one had the decided characters of the common cowslip" [p. 409]. Unfortunately no one, except Mr. Sidebotham seems to have tried the seed of the pure primrose; & it would be very rash to draw any conclusions from the apparent greater trueness of the primrose; but if this one experiment were confirmed, the primrose probably should be looked at as the primordial form, whence has been derived through intermediate oxlip‐forms the cowslip, & the Bardfield oxlip. It is, perhaps, the most probable view that the common oxlips are varieties of the cowslip, easily reverting back towards the primrose; some of the forms having been complicated by crosses with either the primrose or cowslip. I have entered into this case with great detail because, considering the structure habitat, range in height & latitude, & apparent infertility of the two forms, & the many careful experiments made on them, this seems the most interesting case on record. An able Botanist has remarked1 that if the primrose & cowslip are proved to be specifically identical, "we may question 20,000 other/78/presumed species." If common descent is to enter into the definition of a species, as is almost universally admitted, then I think it is impossible to doubt that the primrose & cowslip are one species. But if, in accordance to the views which we are examining in this work, all the species of the same genus have a common descent; this case differs from ordinary cases, only in as much as the intermediate forms still exist in a state of nature, & that we are enabled to prove experimentally the common descent. <Hence common practice & common language is right in giving to the primrose & cowslip distinct names.>

I will end this long discussion by recalling attention to another statement by Mr. Herbert in regard to the species of Primula, which, though it may seem incredible I think ought not to be lightly rejected, as Mr. Herberts observations on the common cowslip & on various other subjects/79/have stood the test of subsequent observation. Mr. Herbert affirms2 that he raised a powdered Auricula (P. auricula) from P. nivalis; & that he likewise raised P. Helvetica (described as a species by Don, but treated as a variety of viscosa in Steudel) from P. nivalis; & that thirdly he raised P. Helvetica likewise from P. viscosa. Hence Mr. Herbert concludes that these Swiss Primulas are only local varieties.3

1 Phytologist vol. 2. p 875

2 Transact. Hort. Soc. vol. IV. p. 19.

3 [On the blank lower third of this folio, Darwin pencilled: 'Here discussion on large genera.']


A1/Wide ranging, common and much diffused species tend most to vary:—The elder De Candolle, & several other Botanists1 have insisted that it is the widely ranging, the common & vigorous plants which vary most./A1 v/Alph. De Candolle2 gives a list of 117 species which range over at least a third of the terrestrial surface, & he states that the greater part of these offer varieties. I have attempted to test this proposition conversely; that is by taking the species which present varieties, & seeing whether a large proportion of them are common & widely diffused in their own country./A1/Ledebour divides the enormous territory, included in his Flora Rossica into 16 Provinces; & to each species he appends the number of Provinces which it inhabits. There are 999 phanerogamic species which present varieties, marked by Greek letters, & these on an average range over 4.94 Provinces; whereas there are 5347 species which have no varieties, & these range over only 2.43 provinces; so that the varying species range over rather more than twice as large an area as the other species. The rule holds very nearly the same when each of the four volumes is tried separately. But we shall presently see & have to discuss the many difficulties which arise in considering the value of the varieties appended by Botanists to their species/

A2/In the London Catalogue of British Plants the number of the 18 provinces, in which each species has been found, is added from Mr H. C. Watson's Cybele Britannica. The number of varieties given in this Catalogue is not great, but Mr Watson has added for me in M.S. some others; the principle on which he has acted in doing this, & the reasons for omitting some varieties & some few whole genera, are given in the Supplement to this Chapter; but I may add that all the varieties here included have been ranked as species by some one or more botanists. Now there are 1053 species which have no such varieties appended to them, & these on an average range over 10.76 of the Provinces; whereas there are 169 species which have such varieties, & these range over an average of 14.55 provinces. I have, also, tried these species in another way, not by taking an average, but by seeing how many species range over all 18 provinces; & I find that of the 1053 non‐varying species, 216 occur in the whole 18 provinces, or in the proportion of 205/1000 whereas of the 169 species which present varieties, there are 70 which range over the 18 provinces, that is the proportion of 414/1000; so that proportionally twice as many of the varying species range throughout the eighteen provinces, as of the non‐varying species./

1 Boreau. Flore du Centre de la France. Tom. 1. p. 101.

2 Geographic Botanique <1855> p. 586. [Actually 564‐81.]


A3/With respect to 'commonness', it is evident that a species might, as indeed is the case with many aquatic plants, range over an enormous territory, & yet not be common or individually numerous anywhere. In a small area, like Britain, where a plant is found in every province, diffusion & commonness almost blend together. Boreau in his Flora of the Central part of France (See supplement to this chapter, for particulars on this & other works quoted) has marked by C. C the very common species; & I find he has 1280 species not presenting <any marked> variety, of which 240 are very common,—that is in the proportion of 187/1000; there are other 193 species with varieties recorded, & of these 78 are very common, or in the proportion of 404/1000; so that proportionally more than twice as many of the varying species are very common in comparison with the nonvarying. <I may here remark that Boreau draws a distinction between the polymorphic species, which vary almost indefinitely & are not included in the above number, & those species which present varieties sufficiently distinct to be marked by Greek letters.>/

A 4/ Miquel in his list of the plants of Holland, marks a very few species having varieties & marks all the very common species; but the recorded varieties are so few, & no particulars specified in regard to them, that the list is not satisfactory: there are 1133 non‐varying species, of which 201 are common or in proportion of 177/1000; & on [the] other hand there are 46 varying species of which 27 are common, or in proportion of 586/1000; hence more than thrice as many of the varying species are common than of the non‐varying species, but the proportion is probably here exaggerated.

Again Prof. Asa Gray in his Flora of the N. United States, appends the word common to many species, & I find that of the 1851 non‐varying species, 439 are marked as common, 237/1000; whereas there are other 202 species which present varieties (either marked in small or large type, see supplement to this chapter), of which 82 are marked as common,—i.e.405/1000, here then, not far from proportionally twice as many varying species are common as of the non‐varying.1

From the foregoing cases, we see, that such numerical evidence as can be obtained, subjected as it is [to] doubts on the value of

1 In Mr Wollaston's Insecta Maderensia (Introduct. p. XIII) 12 Coleoptera are mentioned as the most abundant in individuals in this group of islets, to which may be added, as I am informed by Mr Wollaston, a Ptinus and Oxytelus. Hence out of the 482 species, about one in 34 of all the species is very common. But of the 61 species, which present varieties, six are very common, i.e. one tenth of the varying species are very common.


the recorded varieties, supports the opinion of those botanists, who believe that the much diffused & common/A 5/species are most liable to vary, or to present varieties, which have been thought sufficiently distinct to be recorded. We can understand why wide‐ranging species, which live under various climates, & which come into contact with diverse groups of organic beings (a much more important consideration, as I think will be seen in a future chapter) should vary more than local species. Wide ranging species will also generally from/A 5A/the mere fact of their inhabiting many places, & from the vigour which they show in thus ranging far & coming into successful competition with many organic beings under different climates, will generally be common or individually numerous: indeed Dr. Asa Gray after examining this question says, "so true is it as a general rule that species of wide range in our country are species of frequent occurrence, that I have not noticed any strongly marked exceptions to it"1 Even in regard to species strictly confined to a moderately sized & uniform locality, which are not exposed to very different conditions, we may, I think, see why such species, when common & much diffused in their own country, should present more varieties than when rare. If we suppose varieties to be mere fleeting productions, like monstrosities, then, if originating in exactly the same proportional numbers in common & rare species say one in a million individuals, they would, within the life‐time of Botanists, be far oftener encountered amongst the common than the rare species; & so would be oftener /A6/recorded in botanical works. But of two species, if one were common & one rare during the whole or greater part of their existence on the earth, then a greater number of such fleeting varieties would, it is probable, actually originate in the common than in the rare species. Now I believe, though we are here fore‐stalling what we shall have hereafter to discuss, that by far the most effective origin of well marked varieties and of species, is the natural selection or preservation of those successive, slight, & accidental (as we in our ignorance must call them) variations, which are in any way advantageous to the individuals thus characterized: hence there would be a better chance of varieties & species being thus formed amongst common than amongst rare. I may add, to illustrate what I mean, that a nurseryman who raises seedlings of a plant by the hundreds of thousand far oftener succeeds in his life‐time in producing a new & valuable variety, than does a small amateur florist. So it would be with a common,

1 Statistics of the Flora of the N. United States, in American Journal of Science, 2nd. Series, 1857, Vol. 23, p. 393.


in comparison with a rare species, raised by the hand of nature in millions on millions during the incomparably longer period of its existence on the earth.

But botanists do not actually wish (though unintentionally it is often done) to record, & define as varieties, mere fleeting variations or monstrosities./A7/Boreau, for instance, & others have expressly stated that they record only the more strongly defined varieties: more than one‐third of the varieties marked by Asa Gray are considered by him as possibly deserving to be called species: in the London Catalogue, the greater number of the most trifling varieties have been removed for me by Mr. Watson & all those which are left (182 in number) have been ranked by some one botanist as species. Of the degree of permanence of varieties in plants we know hardly anything: but when a variety is the common form throughout any province or even quite small district, we must suppose that it is in some degree permanent. We have seen in the case of certain land‐shells of Madeira that some of the varieties are of extremely high antiquity. Now when a variety is in some degree permanent, whether it has originated in a single accidental variation, or by the addition of several such successive variations through natural selection, or through the direct & gradual action of external conditions, as of climate, its first origin is even of less importance to it, than its preservation; for in order to become in any degree permanent, it has to struggle with all other organic beings in its own country; & this shows that it has/ A 8/at least nearly equal, or has perhaps acquired even some greater, constitutional advantages, in comparison with its parent‐species. The mere fact of a species being very common or widely extended shows that it is advantageously situated in respect to the inorganic conditions of its life, & in respect to all the other organic beings, animal & vegetable, with which it has to come into competition; & the varieties produced from such common species, from differing little from them, will gradually partake of (or have in excess) their advantages, whatever they may be. Finally then, I suppose that common species present more varieties, when these are in some degree permanent, than do rare species, from partaking of the advantages which make the parent species common; and that varieties (not now considering those wholly due to the direct action of climate &c) originate more frequently amongst common species than amongst rare, owing to more accidental (as we must call them) variations arising during the whole existence of a species which abounds in individuals, than during the existence of a species which has presented much fewer individuals.


The law first enunciated by M. M. d'Archiac & Verneuil & since confirmed by several geologists, that the species which range over a very wide area, are those which have existed for the longest period, seems at first opposed to the/A 9/foregoing conclusion, taken in connexion with my view that closely allied species do not essentially differ from varieties; for it implies that the species which have ranged furthest have longest remained immutable. But if we reverse the proposition, which can be done with equal truth, it is not so discordant;—namely that species which have existed longest, have had, owing to geological & other changes, the best chance of spreading furthest. The majority of such species we may, without contradicting the law, suppose to have become modified either into varieties or into new species, but that a certain number having undergone no change (& it has never been pretended that wide ranging species universally vary) has given rise to the fore‐going palaeontological law./

A10/Geographical Range of Varieties themselves:—I have met with scarcely any observations on this head. When two varieties inhabit two distinct countries, as is often the case & as is very generally the case with the higher animals, it is obvious that the two varieties separately have a much narrower range than the parent species. A variety, for instance, inhabiting N. America & another variety of the same species inhabiting Europe will both have a very much more confined range than the parent form; so on a much smaller scale, the many varieties of endemic species, confined to the separate islets of the same small archipelago (for instance in the case of the insects of the small Madeira group described by Mr Wollaston) follow the same rule.* So again the numerous alpine, maritime, shade or moisture‐loving varieties of species, which commonly live in other and different habitats, have confined ranges compared with their parent‐Types. These considerations alone make it probable that the far greater number of varieties have narrower ranges than the species whence they have sprung. I have looked to many local Floras, & as far as I could judge, the recorded varieties seem usually to have restricted ranges. In the London Catalogue (1857) the range within Britain is given by Mr Watson of some, namely 53 varieties, & I find that on an average they range over 7.7 Provinces; whereas the/A 11/46 species, to which these varieties belong, range over 14.3 of the provinces;—

* All this depends on the arbitrary assumption of which is var. & which species. [J.D.H.] Begin with stating that it is a truism Probably not worth giving so much of a truism. [CD.]


or over nearly twice as wide an area. At my request Mr Watson was so kind as to append remarks on the nature of the habitats & of the ranges of those varieties of British plants with which he was personally acquainted; but as he stated to me, it was not possible to arrive at any definite conclusions from the numerous sources of error; but I may add that from this list it seems that a large number are alpine, maritime, &c forms; sometimes confined to one or to a few localities, but often pretty widely diffused: a good many varieties are, as far as known, strictly local, & some of them have become extinct since having been first noticed: in several cases the varieties, when not strictly confined to any particular locality, or habitat, seem to be rarer than the type‐forms:—

The only published observation which I have met with on the range of varieties is by Mr. C. B. Adams,—a competent judge in regard to the terrestrial mollusca on which he treats:1 he states that the several/A 12/varieties of a species seldom have the same range with it or with' each other; 'each variety has its own limits of distribution;' one variety will often have an 'extent of distribution equal to that of two or more other varieties' of the same species. He believes that varieties follow the same laws of geographical distribution with species; and hence he concludes that they have been aboriginally created as varieties. But it follows from his remarks that varieties generally have more confined ranges than their type‐species.*

In all cases, this latter remark, is to a large extent a mere truism; for when two forms are so closely similar, that one is called a species and the other a variety, the commoner of the two, is almost sure to be called the species, and the less common one, the variety: for we cannot tell which of the two has branched off from the other.2

As by our theory two closely allied species do not differ essentially from a species & its strongly defined variety, I was anxious to ascertain anything about the ranges of such closely allied species but I can advance only one single case, as follows: Mr Watson has marked for me in/A13/the London Catalogue (4th Edit:), which is a pretty well sifted list & does not include the most doubtful

* This is reasoning in a O [circle.] The idea of a var[iety] is founded on variety. [ J.D.H.]

1 Contributions to Conchology. No. 10. On the nature & origin of the species of Mollusca in Jamaica. p. 193.

2 See an excellent discussion on this point in Dr. Hooker's Introductory Essay to the Flora of New Zealand. p. XVII & note.—Dr. Asa Gray, also has remarked to me that mere priority of description has in some cases determined which form has been called the species & which the variety.


species, the forms therein admitted as species, which he considers as most like varieties: he has marked 63, & adds that most of these have been of late years, as it were, cut out of other species: they have all been considered by some few botanists as mere varieties, but by the large majority of local authors have been ranked as good species. Now I find that these 63 species in the London Catalogue range on an average over 6.9 provinces; so that, they have very nearly the same extent of range, with that (7.7) of the 53 printed varieties in this same catalogue.*/

A14/On the relation of the commonness and diffusion of species to the size of the orders and genera in which they are included:—My object in looking at this question regards Variation:—As we have seen that a large proportion of the common and widely diffused species present varieties, if these common species occur most frequently in the numerically large groups, it would be some indication that a greater number of varying species would occur in them—& this latter subject is an important one which we shall presently have to discuss./A14 v/There is, as it seems to me, some a priori probability that the species in the large groups would be generally common & more widely diffused than in the small groups; for the simple fact of many closely allied species inhabiting any country shows that there is something in its condition, organic or inorganic favourable to them; & this by itself would tend to make the species numerous in individuals & widely diffused within that country beyond the common average.1 /

A 14/Alph: De Candolle has shown2 that there is some but very slight evidence that the Orders numerically large in a country, include more common or "vulgar" species than do the smaller Orders; but that the species of such large orders generally have

* Very good remark. [J.D.H.]

1 Alph. De Candolle (Geograph. Bot. p. 562) takes a directly opposite view. <He supposes that when the conditions of life are most favourable to a group, many delicate species could live. "Or, les espèces les plus delicates doivent avoir l'aire la plus restrainte" But we have seen that there is generally a relation between the extent of range and commonness of a species: and how is a species known to be delicate except from being rare and having a narrow range? Is it not saying that certain species are delicate because they are rare & have a confined range, & therefore they are rare and have a confined range? The rarity and confined range of a species, depends, I believe, in the vast majority of cases, on its not being able to compete with or withstand other organic beings; and by no means on the conditions of its existence being favourable.> [Deleted in fair copy.] M. De Candolle throughout his admirable work seems to me very often to greatly underrate the predominant importance of the struggle for life on all organic beings,—a subject to be discussed in our next chapter—

2 Geograph. Bot. p. 465‐470. p. 562.


more confined ranges; & he concludes with some doubt that where only a few species of an order exist, these will be the more robust & the widest rangers. It has appeared to me, from reasons not worth giving, that if any such rule did hold good, it would be more likely to appear in smaller groups or genera rather than in orders.1 But whether in genera or orders/A 15/there are very many causes which would tend to conceal such a result. Namely, our best classifications are considered by many able botanists as still highly artificial. The species in large genera are as remarked to me by Mr H. C. Watson, more difficult to identify, & he believes that many species in such large genera, which are now ranked as, distinct in distant countries, would on close examination often be found to be identical; & consequently such species in the larger genera would really have wider ranges than they appear to have in books; moreover there would sometimes be the greatest difference in the range of a species, according to the value set on its specific characters; for instance a European species having a variety in N. America would have an enormous range, but if that variety were ranked as a species, the range of the European form would be immensely reduced. Aquatic & littoral plants generally have very wide ranges, quite independently of the question whether they form parts of large or small genera. Lowly organised plants as a general rule range further than the more highly organised, and lastly when two areas, separated by the sea or by other/A 16/ barriers, are considered, the capacity for dissemination in the species in common, would probably come into play.

(Some of these multiform causes of error may, I think, be in some degree eliminated by not considering the whole range of the species, but only the degree of diffusion & commonness of the species, described by a single botanist, within one continuous territory, more especially if not of vast size. And for my special object of finding out whether more varieties have originated in any country (or if originating elsewhere, are in this country enabled to subsist) amongst the larger or the smaller genera, it

1 Dr. Asa Gray (in American Journal of Science, 2nd Series, Vol. 23 p. 391) has distributed under their orders 430 species which are the widest rangers in the northern U. States & at the same time the most common species. I have had these orders so arranged that all the species (977) included in the larger orders in the U. States are nearly equal in number to all the species (939) included in the smaller orders. And I find that the number of the wide‐ranging & common species are more numerous in the smaller than in the larger orders, in the proportion of 233/1000 to 215/1000. If the species could have been arranged by genera instead of by orders, namely if all the larger genera had been put on one side & the smaller genera in the other: I hardly doubt from the following Table (Tab. A), that the larger genera would have included a larger proportion of these common & widely ranging species.



The numerator gives the number of the much diffused or the common species in each country.

The denominator gives the number of species in the left column in the larger genera & in the right hand column in smaller genera—See Supplement to this chapter for titles of Works etc.

Larger Genera Smaller Genera
Britain: London Catalogue (1857) H. C. Watson—Larger genera with 5 species and upwards, smaller with 4 species and downwards—The numerator expresses the number of species found in all the 18 Provinces, into which Britain is divided. 148 592 = 250 1000 138 629 = 219 1000
Russia: Ledebour (Dicotyledonae alone). Larger Genera with 10 species and upwards, smaller genera with 9 species and downwards. The numerator expresses the number of species found in at least 8 of his 16 Provinces. The species inhabiting 8 Provinces have about thrice the average range of all the phanerogamic plants:— 239/3385 = 70/1000 131/1937 = 67/1000
Centre France: Boreau—Larger genera with 5 species and upwards, smaller with 4 and downwards. The numerator expresses the species marked C.C. or very common. 163/732 = 222/1000 155/741 = 209/1000
Holland: Miquel—Larger Genera with 4 species and upwards, smaller with 3 species and downwards. The numerator expresses the number of common species. 120/622 = 192/1000 108/557 = 193/1000
Ratisbon: Furnrohr—Larger genera with 4 species and upwards, smaller with 3 species and downwards. The numerator expresses the number of species marked "sehr gemein". 102/533 = 191/1000 79/526 = 150/1000
N. United States: Asa Gray—Larger genera with 5 species and upwards, smaller with 4 species and downwards. The numerator expresses the number of species marked as "common". 326/1136 = 286/1000 195/917 = 212/1000

1 [Darwin's holograph draft of this table is ULC Darwin MSS. vol. 16.1, fol. 172.]


seems to me quite immaterial whether the same species in other countries have very wide or narrow ranges,—are very common or rare.)

(The following short table (Tab. A.) gives the proportions of the common & of the most widely diffused species, in the larger & in the smaller genera, in six countries.)

We here see a slight preponderance, in the larger genera in all the cases except in Holland, and Miquel's tables differ more or less, in every single respect, as far as I have tried them, from those of other Botanists. The slight preponderance would probably/A17/be somewhat increased, more especially in such large territories as those included in the Flora Rossica, if some of the many above‐specified causes of error could be removed: for instance the influence of peculiar stations on the range, which is independent of the size of the genera./A17 v/I may add, as supporting the table that Dr. Asa Gray finds that 75 per cent of the widest ranging species in N. America belong to genera having above the average number of species1 and in regard to "commonness", we see in the table that a greater number of species marked as "common" are included in the larger genera; <& indeed as already remarked Dr Asa Gray has shown that the common & widely ranging species are almost invariably the same.) Dr. Hooker also finds a similar result by tabulating the species common to Europe & N. America, which have a vast range & these usually belong to large genera. Conversely, in regard to commonness, Dr. Hooker has remarked to me in a letter* that in a general Herbarium, genera with single species are represented by a single specimen far oftener than large† genera, showing that the genera with a single species are usually rarer in individuals./

A17/In regard to the extent of diffusion, the preponderance small as it is in Table A, quite or almost disappears, if an average of the ranges of all the species in the larger & smaller genera be taken, instead of, as in the Table, the proportional numbers of the species having unusually wide ranges. Thus in the Flora Rossica,

* I cannot now find your letter on this subject, but I hope I shall & I quote now only vaguely from memory. [C.D.]

† Or more local [J.D.H.]

1 American Journal of Science, 2nd series. Vol. XXIII. 1857 p. 380, Dr. Gray remarks that the converse of the above proposition does not hold good for out of 33 species which have the narrowest range of all the species, 21 belong to large genera. But it is conformable with my views that many species in the large genera should like varieties be extremely local. The species with a wide but disjointed range (p. 387) seem to make a real exception: but with disjointed species, several interfering causes, as extinction, the action of the Glacial epoch, chance dissemination, may have come into play.


all the species (3955 in number) in the larger genera (for the size of the genera see the table) have an average range of 2.8 provinces: whereas the species (2407 in number) in the smaller genera have a slightly larger average range over 2.88 Provinces. Again in the London Catalogue of British plants (5th edit.), the species in the larger genera range on an average over 11.4 provinces, in the smaller over 11.2 provinces. Nor according to the views, which we are in this work discussing, is this surprising; for we here look at species as first branching off into varieties, & these then becoming modified (by means which it will hereafter be attempted to be explained) into closely allied, & ultimately into quite distinct species: now we have seen that varieties generally have narrow range, as have those closely allied forms which were marked for me by Mr Watson, &/A18/which are admitted in the London Catalogue as true species; & such forms, when a general average is struck, would greatly reduce the range of the widely diffused species,—including those species, of which the varieties had not as yet become converted into local species.

On our theory, however, another cause of doubt and difficulty here comes in. We have no reason to suppose that all forms, even within the same class, undergo modification at the same rate; indeed geology leads to the belief that the more highly organised forms,—as Vertebrata compared with most other animals— brachiopods in comparison with acephala, & these with gasteropoda —are replaced at a quicker rate than the more lowly organised. Hence of two sets of species, having originally exactly equal ranges, one set might become after a given period converted into a greater number of new specific forms having restricted ranges, whilst the other set remained unaltered with their original wide ranges. I suspect that, on our theory, this may be the explanation of the Compositae,/A19/for instance, which are considered by many botanists as very highly organised plants, having species on an average with very narrow ranges.1 This view may perhaps, also,

1 Mr Gould in his Introduction to the Birds of Australia (1848, p. 122) divides this country into five sections & adds one for a few outlying regions: he gives the range of each species in these six divisions. As Birds are very highly organised beings, & as Mr Gould admits extremely slight modifications of structure to be of specific value, I have thought it worth while to have the species (omitting the Natatores) tabulated in genera having four species & upwards & into genera with three species & downwards. The 300 species in the larger genera range over an average of 1.84 sections; whereas the 228 species in the smaller genera range over an average of 2.24 sections. Here we see that the closely allied species in Mr Gould's larger genera have narrower ranges than those species which have not according to my view, been converted into representative races & species in the several sections of the country.


throw light on the general rule1 of lowly organised plants having wider ranges than the more highly organised: though probably the greater facility of dissemination in most of the lowest plants has largely influenced the result. On this view, it is not that the more highly organised productions of nature have originally had narrower ranges, but that they soonest become changed into local & distinct species.*/

A19 A/The undoubted fact that not rarely species in the smallest genera in a country are extremely common & range very widely is not opposed to our view; for a species, before it can have become modified into several distinct species inhabiting distant localities, must have ranged, according to our theory, over the whole area, inhabited by the forms derived from it, either in its original unaltered specific state, or during its successively modified states. On the other hand, some cases are on record of groups, possessing numerous species, all of which are individually very rare & have very confined ranges, & yet with nothing special in the stations inhabited by them to account for this. Dr Hooker has given2 a most striking instance of this fact in the Coniferae of New Zealand & Tasmania; & whilst examining the fossil Lepadidae of the Chalk period, I was much struck with the number of the species of certain genera in comparison with those now living; & yet all were very scarce in individual specimens. We may, perhaps, hypothetically account for such cases, by supposing that such genera are on the road towards extinction: for E.Forbes & others have remarked that the first step in this road is marked by a reduction of the individuals of the species.†/

A 20 /On species with recorded varieties being more frequent in large than in small genera:—/3

/Fair copy 15 A/From looking at species as only strongly marked & well defined varieties, I was led to anticipate that the species of the larger genera in each country would oftener tend to present varieties, than the species of the smaller genera; for on this view wherever many closely related species, (i.e. species of the same genus)/A20/have been formed [,] many varieties, or as I look at them incipient species ought, as a general rule, to be now forming.

* Good [J.D.H.]

† how can it be otherwise? [J.D.H.]

by catastrophe it would be otherwise [C.D.]

1 Alph. De Candolle. Géographic Botanique. p. 499, 519.

2 Dr. Hooker in [Flora Novae‐Zelandiae, I, xxix.]

3 [See Appendix for Darwin's earlier version of the opening for this section.]


Where many large trees grow, we expect to find saplings. But if we look at each species as a special act of creation, there is no apparent reason why more varieties should occur in a group having many species, than in one having few. On the other hand, where many species of a genus have been formed through variation, circumstances have been favourable for variation; & hence we might expect that the circumstances should generally be still favourable to variation & that varieties should occur there at the present day in larger numbers than elsewhere./

A21/To explain my meaning further by a loose simile,—if a nation consisted of clans of very unequal sizes, & if we <knew that these clans in ancient times had been very different in size, some much larger, some much smaller & some not then existing, & yet imagine ourselves quite ignorant of the cause of the difference of size, whether due to immigration or some other influence; thengt; if we divided the population into two nearly equal halves, all the large clans on one side, & the many small clans on the other side; we should expect to find, on taking a census at a moderately long interval that the rate of births over deaths was greater in the larger clans than in the smaller; and we should expect to find it so, notwithstanding that we knew that some of the small clans were now rapidly increasing in size & some of the larger clans declining./A 21 v/If we found this to be the case in several nations composed of clans, we should conclude that the greater rate of births over deaths was the cause of the size of the larger clans: & not, for instance, the recent immigration of the large clans./A21/ What the rate of births over deaths is to our clans, I suppose the production of varieties to be to the number of species in a genus; but unfortu‐nately in looking to the varieties existing at any one time, we are acting as if we took a census of the clans at excessively short intervals. Each child does not grow up to man's estate, nor by any means do I suppose that each variety becomes converted into a species. What death is to the individual & ultimately to the clan, I suppose extinc‐tion to be to the varieties, to the species, & ultimately to the genus. I may add that if we found any trace of the breaking up of the larger clans into smaller clans, we should infer that this was the origin of any new clans, which, had arisen since ancient historical times./

A 22/I was strengthened in my expectation of finding more varieties in the larger genera by a remark of Fries,1 that, "in genera containing many species, the individual species stand much closer together than in poor genera: hence it is well in the former case to collect them around certain types or principal species,

1 Quoted in Henfrey's Bot. Gazette. Vol. I [actually vol. II], p. 188.


about which, as around a centre, the others arrange themselves as satellites." And according to our theory the closer two or more species stand together, the more nearly do they in so far approach the character of varieties; we should also bear in mind, as has been shown in the earlier parts of this chapter, with how much difficulty naturalists distinguish species from varieties, even in the best known countries, How many debateable forms there are amongst the plants of Great Britain, of France and of the United States, ranked confidently by one eminent botanist as a species, by another as only a variety. In regard to insects, Mr. Westwood has made1 nearly the same remark with Fries: he says 'in very extensive genera, the distinctions of the species are so minute, that it requires the most practised eye to separate them'. I consulted Dr. Hooker on Fries' remark, & though he at first dissented* he subsequently quite concurred in its substance; & indeed this I find is an extremely general impression with all good observers. I likewise consulted Mr. H. C. Watson, of whose caution & judgment I have the highest opinion: after some deliberation he wrote to me, that although the difficulty/A 23/in distinguishing in a genus of 50 species, each species from 49 others, is obviously much greater than in distinguishing one species from two others in a genus of three species; yet he believes that generally the extremes are more remote in the larger genera than in the smaller, & moreover that the species in the smaller genera are more distinct from each other.

He represented the difference in the following diagram. Larger genus with ten species.—1, 2, 3, 4, 5, 6, 7, 8, 9, 10 Smaller genus with four species, 2, 4, 6, 8.

No one will pretend that the rule is universal; some small genera having very closely related species; & some few large genera having very distinct species. Further, I feel sure that all these naturalists would allow that in very many genera, some few species stand out much more distinctly than the others; & that the remaining closely allied species are not all equally related to each other: this might have been represented by the figures in the above two rows being placed at unequal distances from each other; some being crowded, like satellites, as Fries would have called them, around certain figures.—

I have tried to test numerically this doctrine of large genera including many very closely related species. But numerous dif-

* Because Fries does not observe that all? [sic] large genera are made up of two sets of species, one set as distinct inter se as those of small genera—the other all inosculate. [J.D.H.]

1 Quoted in the Boston Journal of Nat. Hist. Vol. 4, p. 474. [In article by Haldeman.]


faculties interfere: thus all the genera with a single species have to be entirely removed, as such genera/A 24/could not include two closely related species; but one species is sometimes equally related closely to two or even three other species, & then one does not know what to do for a standard of comparison. Moreover in these very closely related forms, the difference of opinion between botanists, whether or not they have been rightly classed as species, is carried to an extreme. However, I may briefly state that Mr Watson marked for me in the London Catalogue 71 forms therein admitted as true species, but which are very closely related to other species, & have indeed all been ranked by at least some one botanist as only varieties: of these, 57 occur in genera having five species and upwards, & only 14 in genera having 4, 3 or 2 species; so that in proportion to the number of species in these two great bodies of genera, the very closely related species stand as .90 in the larger genera to .35 in the smaller. Dr. Asa Gray has kindly gone through his Flora of the N. United States & has marked for me all the closest‐allied forms, which he has classed as & believes to be nearly all, true species, but which he considers as the most likely hereafter to be ranked as varieties: he has marked these in couplets & sometimes in triplets: in the 996 species included in genera having six species & upwards, there are 296 close species: in the 696 species included in genera, having 5, 4, 3 & 2 species, there are 192 close species: so that the close species in the larger genera are as .297 to .275 in the smaller genera. Dr. Hooker also marked for me the closest allied species in his Flora of New Zealand (see supplement for certain omissions & for manner in which the genera are divided) & they occurred in the larger genera, in the proportion of .175 to .166 in the smaller genera./

A25/To return to our question whether a greater number of varieties occur in the larger genera, which, as we have just seen, appear to include a larger proportion of closely allied forms, distinguishable with difficulty, or indistinguishable with any approach to certainty, from varieties. At first, I thought it would be a simple affair to discover this by dividing all the species in a Flora into two nearly equal masses,—all those in the larger genera on one side, & all those in the smaller on the other side, & then count the number of species presenting varieties./A 25 v/ I chose Floras, because these are much better known than any considerable Faunas, & plants are highly variable. But I have taken two well‐worked out insect faunas./A 25/I soon found, however, owing to the kind suggestions of Mr Watson & Dr Hooker



For particulars on the works here tabulated and on the few corrections made, see the Supplement to this Chapter. The numerators in the columns give the number of species presenting varieties; the denominators the number of species in the larger and smaller genera: these fractions are all reduced to common denominators of a thousand for comparison, and are printed in larger type to catch the eye. The right hand rows of figures in the three columns, with decimals, show the average number of varieties which each varying species has, thus the number 1.50 shows that each two varying species have on average between them three varieties.
Larger Genera Smaller Genera (including those with single species) Genera with a single species
Great Britain. Bentham
Great Britain: Babington—Larger Genera with 5 species and upwards, smaller with 4 species and downwards [Pencilnote by C.D.: Write this column larger'.] 101/663 = 152/1000 1.40 89/745 = 119/1000 1.30 [Pencil note by CD.: 'Write this larger'.] 24/255 = 94/1000 1.50
Great Britain, Henslow—Larger Genera with 5 species and upwards, smaller with 4 species and downwards.
The Varieties are divided into two groups, the less strongly marked, and those which have been ranked by some eminent Botanists as species. Lesser Vars: 69/560 = 123/1000 1.55 67/692 = 96/1000 1.40
Stronger Vars: 33/560 = 58/1000 1.33 29/692 = 41/1000 1.20
Great Britain—London Catalogue (1853) (see Supplement for nature of Varieties)—Larger Genera with 5 species and upwards, smaller with 4 species and downwards Great Britain—London 97/616 = 157/1000 1.35 85/642 = 132/1000 1.27
Catalogue—forms ranked as species in this catalogue but which have been thought by some authors to be varieties. In this second line, larger genera with 5 species and upwards, smaller with 4, 3, and 2 species 57/559 = 101/1000 14 = 377 37/1000

1 [Darwin's holograph draft for this table is in ULC vol. 16.1, fol. 167.]


Table I cont.

Larger Genera Smaller Genera (including those with single species) Genera with a single species
Centre France: Boreau—Larger Genera with 5 species and upwards, smaller with 4 species and downwards. 113/732 = 154/1000 1.38 84/741 = 107/1000 1.47 19/267 = 721/1000 1.47
Holland: Miquel—Larger Genera with 4 species and upwards, smaller with 3 species and downwards. 22/622 = 35/1000 25/557 = 44/1000
Germany & Switzerland: Koch—Larger Genera with 7 species, and upwards, smaller with 6 species and downwards 390/2093 = 186/1000 1.72 162/1365 = 118/1000 1.79 32/345 = 92/1000 1.50
Dalmatia: Visiani—Larger Genera with 5 species and upwards, smaller with 4 species and downwards. 164/1007 = 162/1000 1.37 130/899 = 144/1000 1.31 46/290 = 158/1000 1.26
Rumelia: Grisebach—Larger Genera with 6 species and upwards, smaller with 5 species and downwards. 98/1136 = 86/1000 1.45 54/1083 = 49/1000 1.14 12/326 = 36/1000 1.16
Russia, Ledebour (All 4 vols together) Larger Genera with 10 species and upwards, smaller with 9 species and downwards 692/3955 = 174/1000 1.48 307/2407 = 127/1000 1.39 45/475 = 94/1000 1.26
Ledebour—Vol:I separately. 207/1237 = 167/1000 1.42 62/576 = 107/1000 1.32
—Vol: II— 192/1243 = 154/1000 1.56 94/767 = 122/1000 1.35
—Vol: III— 171/905 = 188/1000 1.49 94/595 = 157/1000 1.50
—Vol: IV— 122/570 = 214/1000 1.45 57/470 = 121/1000 1.36
N. United States. A. Gray‐Larger Genera with 5 species and upwards, smaller with 4 sp. and downwards. The two kinds of vars. marked in this work are here classed together. 112/1136 = 98/1000 1.40 65/917 = 70/1000 1.36 32/361 = 88/1000 1.37


Table I cont.

Larger Genera Smaller Genera (including those with single species) Genera with a single species
Canary Islands, Webb & Berthelot—Larger Genera with 4 species and upwards, smaller with 3 and downwards. 49/421 = 116/1000 42/551 = 76/1000
India (part of Flora) Hooker & Thomson—Larger Genera with 7 species and upwards, smaller with 6 species and downwards. 21/258 = 81/1000 1.01 13/165 = 78/1000 1.53
Tierra del Fuego: Hooker—Larger Genera with 3 species and upwards, smaller with 2 species and downwards 19/177 = 107/1000 1.57 16/163 = 98/1000 1.37
New Zealand: Hooker—Larger Genera with 4 species and upwards, smaller with 3 species and downwards 52/361 = 149/1000 1.82 37/323 = 114/1000 2.05 15/159 = 94/1000 2.00
Insecta: Coleoptera Madeira: Wollaston—Larger Genera with 4 species and upwards, smaller with 3 species and downwards 35/225 = 155/1000 1.71 26/257 = 101/1000 1.34
Sweden‐Gyllenhal—Larger Genera with 11 species and upwards, smaller with 10 species and downwards 512/1344 = 380/1000 1.85 151/485 = 311/1000 1.43 11/43 = 255/1000 1.54

that there were many great difficulties in the way. The subject is so highly important to us, as we shall see in a future chapter, that these difficulties must be discussed at tedious length; but it will be convenient first to give the tables./

A26/In Table 1, we have several of the best known local Floras, (some of which were selected for me by Dr. Hooker) with the species divided into two great groups, those in the larger & those in the smaller genera. On the extreme right hand we have the genera with only a single species, but these are likewise included amongst the smaller genera. Some of the smaller Floras have been selected simply from giving remote countries under different climates. I may premise that I have given every single Flora (&



Larger Genera Smaller Genera (with the smallest wholly removed)
Great Britain: Bentham Great Britain: Babington—Larger Genera with 8 species and upwards, smaller with 7‐4 species both included.— 79/55 = 173/1000 1.41 53/360 = 147/1000 1.24
Centre of France: Boreau—Larger Genera with 8 species and upwards, smaller with 7‐4 species both included 86/505 = 170/1000 1.40 41/343 = 119/1000 1.31
Germany & Switzerland: Koch—Larger Genera with 11 species and upwards, smaller with 10‐5 species both included 257/1216 = 211/1000 1.99 114/683 = 166/1000 1.95
Dalmatia: Visiani—Larger Genera with 8 species and upwards, smaller with 7‐4 species both included 120/707 = 169/1000 1.39 71/492 = 144/1000 1.36
Rumelia: Grisebach—Larger Genera with 8 species and upwards, smaller with 7‐4 species both included 78/917 = 85/1000 1.44 33/513 = 64/1000 1.33
Russia: Ledebour—Larger Genera with 16 species and upwards, smaller with 15‐6 species both included 573/3285 = 174/1000 1.48 234/1437 = 162/1000 1.42
N. United States: A Gray‐Larger Genera with 9 species and upwards, smaller with 8‐5 species both included. (The two kinds of varieties classed together.) 76/710 = 107/1000 1.36 34/426 = 79/1000 1.26

two Entomological Faunas) which I have had tabulated, & have not picked out those which favoured my views. Nor have I divided the genera first in one way & then in another; but before knowing what the result would be, I determined to divide the smaller Floras nearly equally, but in the larger floras to have a greater number of species on the side of the larger genera, & then reduce

1 [The holograph draft for this table is in ULC vol. 16.1, fol. 170.]


TABLE III.1 Decandolle Prodromus, Vols. 2, 10, 11, 12, 13, 14

Name of Orders The Numerator and Denominator as in the foregoing Tables. Genera with 11species and upwards Genera with 10 species and downwards
Leguminosae 223/2937 = 75/1000 1.34 38/619 = 61/1000 1.39
Rosaceae 103/562 = 183/1000 3.09 24/144 = 166/1000 2.20
Borragineae 59/480 = 122/1000 1.38 10/111 = 90/1000 1.40
Scrophulariaceae 118/1413 = 83/1000 1.15 24/347 = 68/1000 1.29
Acanthaceae 232/1088 = 213/1000 1.43 65/335= 194/1000 1.35
Verbenaceae 21/500 = 41/1000 1.00 5/82 = 60/1000 1.00
Labiatae 207/1999 = 103/1000 1.34 32/278 = 115/1000 1.62
Solanaceae 258/1419 = 181/1000 1.45 11/139 = 79/1000 1.72
Proteaceae 153/912 = 167/1000 1.41 6/72 = 83/1000 1.16
Polygonaceae 130/614 = 211/1000 1.63 8/62 = 129/1000 1.37
Nineteen Small Orders 112/1120 = 100/1000 1.50 48/406 = 118/1000 1.14
All the species in the 6 Vols. 1616/13050 = 123/1000 1.51 271/2595 = 104/1000 1.43
Urticaceae—Weddell—Tabulated by Dr. Hooker 65/334 = 194/1000 1.89 26/100 = 260/1000 1.57*
All six volumes together Genera with 17 species and upwards 1510/12103 = 124/1000 1.52 Genera with 16‐8 species both included 188/1472 = 127/1000 1.48

* Is Weddell's Urticaceae part of Deeandolle or a separate work? [C.D.]

1 [Darwin's draft of this table is in ULC vol. 16.1, fol. 169.]


Table III cont.

All six volumes Largest Genera (76 in number) including half the species Smaller Genera (1088 in number) including the other half of the species
959/7815 =    122   /1000 1.59 929/7830    =    118/1000    1.40
28 Largest Genera, each including on average 134 species, taken out of all the Orders in the Six Vols. 455/3772    =    120/1000    1.74

all to a common denominator: for if the larger Floras had been divided equally, from the great size of many of the genera, but comparatively few would have been included amongst the "Larger Genera": & as we cannot suppose that the larger genera go on varying or increasing in species for ever, it requires a considerable number of genera, as will presently be more fully explained, to strike a fair average. In the very large Flora Rossica, I have given in the table, the result for each volume separately, just to show that the excess of varieties in the larger genera is common to the whole/A 27/Flora: I did the same in some other cases with the same results. I have given Great Britain as worked out by several Botanists,/A27 v/not as being particularly well‐known, but in order to show that personal differences in estimating the value of species & varieties, makes no essential difference in the general result. /

A27/Now if we look to the two columns, under the larger & smaller genera, printed in larger type, in which the number of species, presenting varieties, are reduced to a common denominator, we see that with one single exception, the species in the large genera present decidedly more species having varieties, than do the species in the smaller genera. Moreover the average number of varieties to the varying species, with few exceptions, is larger in the larger than in the smaller genera: this is seen in the right hand columns of decimals,—the figures 1.50 for instance, showing that each two varying species have an average of three varieties. The one exception in the table, just alluded to, is Miquel's list of the plants of Holland: but so extremely few varieties are here marked, & as the results deduced from his list differ in several other respects from those obtained by other botanists, it may, I think, be disregarded.


In Table II, I have selected a few (& given all which I have selected) of the larger local Floras, & have entirely removed the smallest genera: & by looking at the columns printed in the larger type, & at the column with decimals we see the same rule throughout, namely of a greater number of varying species, & a greater average number of varieties, in the larger than in the smaller genera.—

If, then, local floras are to be trusted, & if the varieties recorded by various botanists (& two celebrated Entomologists) are worth anything, & if the varieties have been recorded fairly or nearly equally in the larger & smaller genera.—/A 28/all subjects presently to be discussed— we must conclude that there is a decided preponderance of varieties in the larger in comparison with the smaller genera.—

Table III gives the results of the tabulation of all the species (15,645 in number) in six volumes of De Candolle's Prodromus: selected for me by Dr. Hooker, & done at his suggestion. We here see a very different result from that deduced from the local Floras. In the genera having only 11 species & upwards there are more recorded varieties than in the genera with 10 species, & downwards; this holds good for the summary of the six volumes, & for most of the separate orders, but fails in some orders, especially in the great, natural & most carefully worked out (by Bentham) order of the Labiatae. The rule, however, does not hold good, (see Table) if all the genera with seven species & downwards be wholly excluded; so that all that can be said, is that the smallest genera usually present fewer recorded varieties. It deserves remark, how closely similar the result is when all the genera with 10 [11] species & upwards, with 17 species & upwards, when the 76 largest genera which include half the species, & when the 28 very largest genera are taken:—the proportion of the species having varieties in these several cases varying only from120/1000 to 124/1000. The larger the genera are, however, the average number of varieties to the varying species seems to increase being in the 28 gigantic genera, as much as 1.74: so that each two varying species has on an average more than three varieties./

A 29/Now what is the evidence from these three Tables worth? The first question to consider is, whether it is best to take local Floras, or parts of the whole vegetable kingdom. The latter though having some advantages, has, for my special purpose several most serious sources of error. Geology tells us that in the long course of time, small groups have increased, come to a maximum, then declined, & ultimately disappeared. Hence we may feel pretty sure that some groups of plants, now numerically large, have nearly or quite arrived at their maximum, or are now declining; & that


other small groups are now increasing more or less rapidly in numbers./A29 v/Greatly as genera differ in size, yet there is a limit <in number of species> beyond which they rarely pass; & therefore, on my view of varieties being incipient species, there must always come a period when the largest genera will cease to increase at least as a single genus; though it does not by any means follow that sections or portions of such genera may not go on increasing, & other sections decline & be lost./A29/It is idle to speculate what would be the precise effect on varieties of the declination, from less favourable conditions of life, of a group of species; but as the individual numbers of most of the species would probably decrease, from the relations lately pointed out, the amount of variation at any one time would probably be less: we do not even at all know, whether commencing extinction would generally first act on the species in the larger or smaller genera: though one may surmise on the latter: the ultimate result, we shall in a future chapter see, would probably be to leave in any group, those forms which are most distinct from each other. Now in a local Flora any genera, still large, which had come to/A30/vary in a less degree, or a small genus which was varying largely, would, supposing for the moment our rule to be true of the species in large genera varying more than those in small genera, be on an average compensated by the other genera of the same country: so it should be in a Prodromus of the whole vegetable kingdom, if such existed, & there were no other causes of error: but looking to each separate order we might expect, if there be any truth in my view, to find some orders in which the large genera varied little,1 & some in which the small genera varied greatly.

Secondly it is known2 that the same order or genus often has

1 I suspect that the Labiatae, viewed as a whole are now undergoing some great change in development. When divided in the three different ways shown in Tab III the smaller genera have a preponderance of varying species: yet there are two gigantic genera containing together no less than 653 species, & these contain fewer varying species (viz. 90/1000 & only 1.20 varieties to each varying species)

than the smaller genera however divided. If the sub‐order Satureieae, (including only between 1/5 and 1/6 of the Labiatae) be removed, the larger genera have

a preponderance of varying species. In the smaller genera of Labiatae the average number of varieties to the varying species is unusually large. Lastly looking to some of the local Floras, I find that in Boreau, Koch & Visiani the smaller genera in this order have more varying species than the larger: on the other hand in Babington & Ledebour, the large genera in this order, as generally throughout all these several Floras, have a preponderance of varieties.

2 Alph: De Candolle, Geographic Bot. p. 1237‐1245. In Hooker's Hot. Miscell: (Vol. 2 p. 257) there is given from Ledebour several curious cases of the great predominance of certain genera in the Altai: for instance there are 62 species of Personatae, & one‐third of these belong to the genus Pedicularis: of the 130 Leguminosae, three‐fourths belong to Astragalus, Oxytropis & Phaca.—


many more species in one country, than in another, either owing to differences of climate or other unknown conditions. Where many species of a genus exist, relatively to the other inhabitants of the country, we have seen that there is some evidence that, on an average, a large number of them are common & widely diffused; and that of such common & diffused species a large number present varieties. This at least is possible, but it could be hardly detected except in a local Flora; for when all the species of the genus were collected in a general Prodromus, the supposed greater amount of variation where the species were numerous, & the less amount, where thinly scattered & where the genus did not seem to flourish, would tend to counterbalance each other, & conceal the result. Again there are many moderately‐sized genera with all their species confined to one country, & which in that country would be a large or rich genus, & which, according to my general theory ought to be largely varying, as they have in that/A 31/ country become modified into many species; but the greater number of such moderately‐sized endemic genera would in a general Prodromus have to be tabulated amongst the smaller genera, & would vitiate the result. In fact such genera with absolutely few species in comparison with genera in the whole vegetable kingdom, but rich in species in their own country, are exactly those genera which we might expect would yield the best evidence on our view. Gigantic genera are often widely distributed over a large portion of the world; & we must believe (as Sir C. Lyell has remarked in his Principles in regard to the wide range of the same species) that owing to the slowness of geological changes, of climate, &c., this spreading of the species of the same genus (descendants from common parents according to our theory) must have taken an enormous length of time: hence, although in a very large widely‐spread genus there must have been, on our view, a great amount of modification, this modification may have been slow. On the other hand in local genera, we may believe from the very fact of their not having ranged widely, that they often are not of such ancient origin as the widely spread genera; & in taking a census of such comparatively fleeting objects as varieties, we ought to look as much as possible to those groups of species, which are undergoing the most rapid change; & it is just these very endemic genera/A32/rich in the species in their own country, which would be lost, or rather would give a directly false answer when tabulated in a general prodromus.

To take as a final illustration, the case alluded to in a previous note of the genera Pedicularis and Astragalus, so extraordinarily


rich in species in the region of the Altai. As so many species have been formed there, we ought to look to these two genera/A 32 v/in that quarter, in order to see the manufactory of species at work: that is, according to my view, we ought there to find in these two genera, a greater than average number of varieties. And if this rule were found generally to hold good in local Floras, namely that the genera which had many species had many varieties, it would throw much light on the origin of species. But what can it signify under this special point of view, whether or not other species of Pedicularis and Astragalus are varying in other quarters of the world?*1

A32/Hence I conclude from the several reasons just assigned, namely that some large genera must have arrived at their maxima and be now declining, & some small genera be rapidly increasing in number of species,—that some genera have been largely developed in certain countries, and elsewhere much more feebly,—that endemic genera probably have in many cases increased at a quicker rate than mundane genera, & yet would be ranked as small genera in a general Prodromus,—from these several reasons, I conclude/A 33/that a fragment of a Prodromus would be of little service, and an entire Prodromus of far less service for our special purpose than local floras. Nor should I have tabulated the six volumes of De Candolle, had it not been for Dr. Hooker's advice, nor should I have published the results, had not honesty compelled me, as they are on the whole unfavourable. Nevertheless I am bound to confess that from the wide diffusion of plants, and from genera largely dominant being generally everwhere numerous, I had expected more favourable results.

The best territories for my special object, would be those with all the species endemic, for all the species will probably have originated in such areas and where many species of the same genus have been formed, there as a general rule we ought now to find most variation in progress. Under this point of view, New Zealand & Madeira are the best areas in Tab. 1, but they would have been better, had they included a greater number of species. I can, however, see no valid objection to taking, as a representative of the whole, fragments of one natural area, as (in Tab. 1) the several kingdoms of Europe. Another advantage in local floras over a Prodromus, in which latter the orders are worked out by different men, is that there would be generally more uniformity

* Hence the smaller the area the better the result? [J.D.H.]

1 [From here until the middle of fol. A 41, the text of the draft is not in Darwin's handwriting.]


in the value attached to varieties & species; there must be a prodigious difference in the value of the species as given by Dunal in the Solanaceae and by Bentham in the Scrophulariaceae, & though it is quite immaterial for us whether a greater or less amount of difference causes two forms to be called species or varieties, it is of some consequence that there should/A 34/be some approach to uniformity in the relative value of the species & varieties when all are tabulated together.

Now comes the question, what is the value of the varieties recorded in Botanical works? Am I justified in hypothetically looking at them as incipient species? do they differ in the same manner, only less in degree, from their types, as one closely allied species differs from another? I do not doubt that mere monstrosities have been recorded sometimes as varieties, though I do not suppose that any botanist would intentionally do so, & some authors have expressly stated that they have endeavoured to avoid this. Some also have stated, for instance Boreau, Visiani & Wollaston, that they have endeavoured to record as varieties not mere fleeting differences, but those alone with some degree of permanence. So again I do not doubt that a good many varieties are merely nominal, & owe their origin to doubts & confusion; & as such would be more likely to arise in large genera, than in small, this would directly vitiate our tables. That varieties even in the most carefully worked out floras are of very unequal values must be admitted; but it would have been a serious objection to my view of varieties being incipient species in various stages of modification, had they been all equally like or unlike each other and their parental types. I may here repeat that I am far from supposing that all varieties become converted into what are called species; extinction may equally well annihilate varieties, as it has so infinitely many species. That many varieties have in some degree the character of species I cannot doubt, for so many have been ranked as species by one botanist or another. Thus in the small British Flora, we have in Mr. Watson's list (Tab I) 182 varieties, so ranked by the greater number of sound botanists, /A35/but which have all been considered as species by some one botanical author; & we have in addition 71 other forms called species in the well sifted London Catalogue, but which have been ranked as varieties by some one botanist. So again in Professor Henslow's list there are 62 forms considered by him as varieties, but which have been ranked by such eminent men as the elder De Candolle, Sir J. Smith, Sir W. Hooker & Lindly as true species.

Dr. Hooker objects to my whole manner of treating the present


subject because varieties are so ill defined; had he added that species were likewise ill defined, I should have entirely agreed with him; for my belief is that both are liable to this imputation; varieties more than closely allied species, & these more than strongly marked species.

Mr. Watson & Dr. Hooker have also objected that there are many species so highly variable, & with the varieties running so closely into each other, that botanists do not attempt to mark them as distinct; hence in my tables, some of the most variable species do not appear to have any varieties. Boreau & Mr Wollaston also state that such polymorphic forms are not included amongst their recorded varieties. In the former part of this chapter we have seen how difficult it is to decide whether Polymorphism is of the same nature with more defined variation,/A 36/so that I am inclined to think that it is an advantage that such polymorphic species are partly excluded from my tables. That they are not by any means wholly excluded I am aware; for botanists occasionally mark by Greek letters ideal types which cannot really be defined from an inextricable mass of varying forms. So again when only a few specimens have been collected of some rare polymorphic species, the varieties would necessarily appear far more defined than they really are, & so would be liable to be recorded as distinct. I do not suppose that polymorphism which is partly excluded from our tables is much commoner in small than in large genera, or conversely; if it were so, it would have seriously vitiated our tables,—that is, if we suppose Polymorphism to be essentially of the same nature with more definite variation. In some of the floras I have excluded the most notorious polymorphic genera, which abound with doubtful species & doubtful varieties; but this has never been done except with the larger genera; & the result has invariably been to make the preponderance of varieties in the larger genera, less than it would have been, had these genera been admitted.

Mr Watson & Dr Hooker likewise object that* our best classifications are very far from natural; but any great perfection on this head is not material for my purpose: I divide all the species in a country /A 37/in to two great bodies; all those in the larger genera on one side, all those in the smaller on the other side; & I presume it will not be disputed that the species in the larger genera taken together present a greater number of forms more closely allied together in little groups, than do the species in the smaller genera. I have however, found in tabulating the British

* remind me [J.D.H.]


Flora that the species of some few genera when split up into smaller genera, had to be placed among the smaller genera, whereas in other British floras they stood on the other side. But the several British floras in Tab. I show that this has not materially affected the result.

I cannot look at any of these causes of error as very important; they would, I think, to a large extent disappear when averages are taken; & the uniform result in Tab I & II bears out this con‐clusion. But now comes a far more serious cause of doubt, suggested to me by Dr. Hooker after seeking some of my tables; namely that botanists have recorded varieties more fully in the large than in the smaller genera. He believes this to have been the case from several reasons, but more‐especially from floras serving in part as mere dictionaries; & as it is obviously more difficult to name a species in a large than in a small genus, he thinks botanists have guarded against error by more carefully recording the varieties in the larger genera. I have consulted several other botanists, & though it does not appear that they had previously thought on this point, they generally/A 38/concur in this view. One botanist, however, Dr. A. Gray, whose opinion will be considered by all as of the greatest weight, after deliberation does not believe that he has himself so acted: he at first thought that he might have unfairly recorded a greater number of varieties in the smaller genera, which, from what little systematic work I have myself done, was my impression owing to the greater interest of mono‐typic genera. Now if Dr. Hooker & the others who concur with him be right, all the foregoing tables are utterly worthless;* for they do not show nature's work only the imperfect handiwork of botanists. It is presumptious in me to believe that botanists have worked more philosophically than they themselves think they have; but I can hardly avoid this conclusion.

For in the first place it is somewhat remarkable that so many botanists & two Entomologists should all unconsciously & un‐intentionally have produced so uniform a result, as may be seen in the first two tables: more especially as the varieties recorded by different authors are of such different values. To test Dr. Hooker's capital objection, I selected some of the principal local floras, & entirely removed the genera of least size; these are all given in Tab. II; here the larger genera (larger than in Tab. I) still show a marked preponderance in the proportional number of varying species over the smaller genera,† here not so small as in

* vitiated though perhaps not overturned [J.D.H.]

† give the case of Rubus [J.D.H.]


Tab. I. Dr. Hooker/A 39/would probably account for this fact by saying that the larger the genera & the more difficult the species were to identify, the greater the number of the recorded varieties would be; but as the difficulty goes on regularly increasing with the size of the genus the excess is not so great or so uniform as might have been expected on this view. The excess in the number of the varieties in the larger genera not regularly increasing with the size of the genera, may be explained on my hypothesis by some of the largest genera having reached their maxima. If we now look to the genera with a single species (right hand column in Tab. I) the difficulty in identifying the species is reduced to a minimum, yet we find that the number of species in these monotypic genera1 which have varieties, though proportionally less than in the next group of larger genera, is by no means diminished in an extreme degree, as might have been confidently expected on Dr. Hooker's view: in two instances, namely in the U. States & Dalmatia, the number is actually greater than in the next group of larger genera. All this may be seen by comparing the right hand & middle columns in Tab. I.

If we look to the rows of figures with decimals in Tab. I & II, which give the average numbers of varieties which the varying species include, we find a degree of uniformity, especially in Tab. II very remarkable as it seems to me on Dr. Hooker's view. For my own part I look at these rows of figures as shewing, that not only/A40/more species present varieties, but that the varying species generally present more varieties in the larger than in the smaller genera.

In the monotypic genera (right hand column in Tab. I) where the difficulty in naming species is reduced, as already remarked, to a minimum, we find the average number of varieties to the varying species, in five cases, either equal to, or actually greater, than in the next group of larger genera. <This fact, I think, if the average from the small number of species in the monotypic genera can be trusted, might be explained on my view, but the explanation is not worth giving.*> On Dr. Hooker's view that the species in the larger & smaller genera really have on an average an equal number of varieties; but that the varieties have not been fully

* Small genera being few in individuals do not present so many Herbarium varieties. [J.D.H.]

1 [ ] says p. 574 that some have thought that monotypic species do not vary. He does not give any authority except [Puvis] (De la Dégéneration p. 37) who refers only to varieties raised under [cultivation], and adduces the supposed fact in regard to all variations being due to intercrossing.


recorded by botanists in the smaller genera, we are driven to conclude (as may be seen by comparing the middle & left hand columns in Tab. I) that although Boreau in France, Koch in Germany, & Hooker in New Zealand, did not fully & fairly record all the species having varieties in the small genera: yet that in these very genera/A 41/they have recorded a greater than average number of the varieties themselves. This strikes me as improbable, & on the whole it seems to me far more probable that the tables make some approach to a fair representation of the manner in which species vary in nature. Any how I have endeavoured to give an abstract of the more important facts & arguments on both sides, & those few naturalists who are interested in the subject, can form their own judgement.

Finally, then, if we review our whole discussion on local Floras, which alone are well adapted for our purpose, it may I think be concluded, that on an average, a greater number of species in the large genera are common & widely diffused in their own country, than in the smaller genera; but that this greater number is (according to our theory) being slowly & steadily diminished by these species tending to vary, & thus being converted first into local varieties & then into local species. We can understand why a species which ranges widely & thus becomes exposed to somewhat different conditions of life is the most likely to vary; and a species numerous in individuals has a better chance, within any given/A 42/period, of breaking into varieties, which from possessing some advantage might be preserved & so become more or less permanent. Moreover common & widely diffused species must generally be better adapted to the conditions of life, to which they are exposed than the rarer & more local species, as will be more fully discussed in the next chapter when we treat of the severe competition to which every being is exposed; hence varieties from such favoured species will have the best chance of enduring for a long period & of increasing in numbers. It may be added that if a variety has ever increased so largely in individual numbers that it has come to exceed those of its parental type; it assuredly will have been called the species, & the original species the variety.

From these relations, & more especially from the actual facts given in the tables of the local Floras, I believe that the species in the larger genera, which as a general rule are very closely related to each other & in so far themselves approach in character to varieties, oftener present varieties (& a greater number of

† had [not] the means [J.D.H.]


varieties) than do the species in the smaller genera./A 42 v/It is not that the species of very small genera never vary, or that the species of large genera invariably present a great number of varieties; for if it were so, it would be fatal to my theory, as genera of all sizes have to increase & decline. Nor by any means is it, that all the species of a genus present varieties; for this is a very rare case;—it is only that more species have varieties clustered round them in the larger than in the smaller genera. And in regard to the close affinity of the species to each other in the large genera, it is not that all are equally related to each other; but, that some species are closely clustered round other species; causing the genus to consist of smaller & unequal sub‐groups. These/A 43/conclusions as far as they can be trusted, strengthen our general theory, that species do not essentially differ from varieties, & that varieties by further modification may be converted into species. But our tables more especially throw light on the origin of the species of a genus, where very many are endemic in a moderately sized territory, & where we may suspect that they have been formed within comparatively recent times; for it is in local floras alone, that we invariably find more recorded varieties in the large genera, than in the small; & I have given my reasons for putting some faith in the records of so many Botanists, whose works agree in this respect. Furthermore, I believe, that the rule of the species in the larger genera on an average varying more, & therefore as I look at it, increasing in the number of their species at a quicker rate, than the species in the smaller genera, when taken in connexion with a large amount of extinction & with a principle, hereafter to be explained, which may be called that of divergence—taken together throw a clear light on the affinities of all organic beings within the same great classes; for we invariably see organic beings related to each other in groups within groups—or somewhat like the branches of a tree sub‐dividing from a central trunk.

Conclusion. From the various facts now given in this chapter, & innumerable others might have been added, I cannot doubt that there is much variability in organic beings in a state of nature/80 v/ The widely‐ranging, the much diffused & common, in short the vigorous species are those which are the most apt to vary./80/The variation differs greatly in degree; in some it is scarcely perceptible, in others strongly marked; so that we have a graduated series from the finest shades of individual differences, to well defined races, distinguishable with great difficulty, if really distinguishable


at all, from sub‐species & closely allied species. In certain protean genera, the variability may in part be of a different nature; but on this point it seems difficult to arrive at any definite conclusion. From what we have seen of the effects of domestication or changed conditions on organisms of all kinds, & which beings, it has been shown in the second chapter, could not have been originally selected from the plasticity of their organisation, & knowing well that the history of the world is emphatically that/81/of Change, it would have been a discordant result if there had been no variability in a state of nature. Judging from the effects of domestication it is indeed surprising that we do not clearly see in nature more organic change, but if such greatly changed organisms do exist, they would be universally called species & not varieties.

According to the views discussed in this work, species do not differ essentially from varieties;—two closely allied species usually differing more from each other than two varieties, & being much more constant in all their characters. This greater constancy may be looked at as partly due to the several causes of variability having acted less energetically on the two species under comparison than on the one species yielding the two or more varieties; and partly to the characters of the two species having been long inherited, & by this very cause having become more/82/fixed. The greater amount of difference between the two species than between the two varieties, may be looked at as simply the result of a greater amount of variation; the intermediate varieties between the two species or between them & a common parent having become extinct. Hence as a general rule, species may be looked at as the result of variation at a former period; & varieties, as the result of contemporaneous variation.

But the forms generally considered as varieties & those considered as species differ in one other most important respect; namely in the perfect fertility of varieties together & the lessened fertility of the offspring of two species. This subject will be discussed in a separate chapter; & I will here only repeat that the infertility of species when crossed graduates away so insensibly/83/that the two most experienced observers who ever lived have come to diametrically opposite results when experimentising on the same forms;—that the infertility does not closely go with the general amount of difference between the two forms, but follows laws of its own;—that it is most powerfully affected by the sex in reciprocal or reversed crosses of the very same two species;—and finally that, as we have seen in the last chapter, the reproductive system is eminently subject to disturbance & that infertility of an analogous


kind to that resulting from hybridism supervenes from other & totally distinct causes. Hence, as it will be attempted to be shown in the chapter devoted to this subject, there is no valid reason, why the different "sexual affinity" (to use Gaertner's expression) of different species to each other should be thought a character of overpowering weight, in comparison with the other differences between species when contrasted with the difference between varieties. <with each other; as, for example, in the tendency to adhere when grafted together.>/

84/It seems to me that the term species is one arbit[r]arily given for convenience sake to a set of individuals closely like each other; &, that it is not essentially different from the term variety, which is given to less distinct & more fluctuating forms. The term variety, in comparison with mere individual differences, is applied, also, arbit[r]arily & for convenience. Practically if two forms are tolerably constant in their characters & are not known to be connected by a nearly perfect series of intermediate forms they are called species; & according to the views here given, even should the two distinct forms be thus connected, if the intermediate forms are comparatively rare, so as seldom to cause much difficulty in naming an individual specimen, there seems no good reason why they should not be called species; & in that case science & common language would accord in giving names of equal value, to the primrose & cowslip,—/85/to the deodar & cedar of Lebanon, —to the Durmast and common oak,—as well as to the many fine species distinguished by the naturalists on characters of little physiological importance.

As the only known cause of close similarity in two organic beings, is descent from a common parent, it is natural that the idea of descent should have entered into almost every definition of the term species. A monster may be abnormal in any degree, but the instant we know its parentage, we do not doubt about referring it to its species.—On the views here discussed, the idea of the common descent of all the individuals of the same species equally comes into play; but it is not confined, as in the ordinary definition, to the individuals of the same species, but is extended to the species themselves belonging to the same genus & family, or to whatever higher group our facts will lead us.—/

86/According to these views it is not surprising that naturalists should have found such extreme difficulty in defining to each other's satisfaction the term species <as distinct from variety.) It ceases to be surprising, indeed it is what might have been


expected, that there should exist the finest gradation in the differences between organic beings, from individual differences to quite distinct species;—that there should be often the gravest difficulty in knowing what to call species & what varieties in the best known countries, & amongst the most conspicuous & best known organic beings if ranging over a wide territory; & that the difficulty should be hopelessly great in two adjoining but now perfectly, or almost perfectly separated regions./86 v/We can understand why it is that the species in large genera are generally more closely related to each other & related in little clusters like satellites around certain other species, why they are apparently often confined in their distribution, & lastly why they oftener present varieties & a greater number of varieties, than do the species in small genera: for, on our views, where, in any country, many species of a genus have been formed there has been in such genus a greater than average amount of modification within the existing geological period; & hence we might expect that the resultant forms would tend to resemble varieties in closely resembling each other & in being grouped around certain species, like varieties around their parents & in being local. We might moreover, expect, on these views that where there has been lately much specific modification, there generally would be now most variation in progress.

The conclusion that there is no/86/essential difference, only one of degree & often in the period of variation, between Species & Varieties, seems to me at least as simple an explanation of the many/87/difficulties by which naturalists are beset, as that each species should have been, independently created with its own system of variability,—the varieties imitating the characters of other species, supposed to have also been independently created, so closely as to defy in many cases the labours of the most experienced Naturalists.


a/Phanerogamic plants alone have been tabulated out of the following works./a v/In the counting the number of varieties themselves, I have not except in a very few cases which are specified counted those marked a: for these seem generally to be the typeforms more fully described: or the type forms in an exaggerated degree. I would, however, here make no important difference for our object whether counted or not, as they would have been counted both for the large & small genera.—/


a/C. C. Babington. Manual of British Botany. 3 Edit. 1851 The naturalized & doubtful plants, included in brackets and marked by asterisks are all omitted. The genera Rubus, Rosa, Salix & Hieracium are, also, omitted; from the extreme doubts, almost universally entertained, which forms to consider varieties and which species: as these are large genera & have many varieties had they been admitted the proportional number of varieties would have been greater in the larger genera. Mr Babington is generally considered to admit very fine species.

The. Rev. Prof Henslow. A. Catalogue of British Plants. 2nd Edit. 1835. The species certainly not indigenous have been expunged; but those marked (4) as possibly introduced by man have been left in. The genera Rubus, Rosa, Salix & Hieracium have been excluded for reasons given above; if left in, the result would have been as above stated. The Varieties are marked by Greek letters, but certain varieties (62 in number) are preceded by (‐); & this signified that these forms have been considered by one of the four following eminent Botanists, as true species, namely De Candolle in Bot. Gallicum: Hooker in British Flora, Lindley in Synopsis of British Flora, & Sir J. E. Smith in English Flora. These varieties, as so ranked by Prof. Henslow, must have much of the character of species./

b/Mr H. C. Watson & J. T. Syme, London Catalogue of British Plants, 4th Edit. 1853. All the species printed in italics, thought to be naturalized, are expunged. Genera Rosa, Rubus, Hieracium & Salix for reasons & with results already assigned have been omitted. In this well sifted list, only few varieties are recorded: but Mr Watson has added for me some which have been ranked by at least one Botanist as a species: he has, also, expunged some few of the most trifling printed varieties, which have not been considered by any one botanist as a species. He has, also, marked for me some of the forms ranked in this catalogue as species, but which have been considered by some Botanists as varieties. If considered, as in the second line of the Tab. I, as varieties, the number of species is diminished both amongst the large & small genera; when considered as species as in another part of our discussion, such could not occur in genera having only a single species, so that these have been also removed in our calculation, though not strictly necessary; & their removal makes the result less striking than it would otherwise have been. For the calculation of the Ranges I have used the 5th Edition 1857. The number of the Pro-


vinces is not appended to some of the species, & these are wholly excluded: as are those confined to Ireland & the Channel Islands.—/

c/A. Boreau. Flore du Centre de la France, 1840. Cultivated plants omitted. Genera Rubus, Rosa, Salix omitted, for reasons & with the result before assigned. M. Boreau expressly states (Tom. 1, p. 101) that he distinguishes "les varieties [sic] proprement dites", which are "plus tranchées" from the endless variations, which many common & widely ranging plants display.

Ant. Miquel: Disquisitio Geographico—Botanica de Plantarum Regni Batavi. 1837. This list is unsatisfactory for our purpose so few varieties being indicated: owing to a mistake in the printed list, I am doubtful about one variety, but have admitted it. I should not have given the results from this list, had I not felt bound to do so from honesty, as the result differed from those in all the other Floras, in several respects. The certainly naturalized plants (marked with †) are omitted.

Koch. Synopsis Florae Germanicae et Helveticae Edit. 2. 1843. I have here made no omissions: in counting the number of the varieties themselves, I have counted those marked a as well as B &c. I have not counted the subvarieties of varieties. This is a very large Flora including 3458 species.

Rob. de Visiani: Flora Dalmatica 1842‐1852. I have excluded the cultivated plants. In counting the number of varieties to each species, I have not counted those marked (a). Visiani seems to have carefully distinguished varieties from variations./

d/A. Grisebach, Spicilegium Florae Rumelicae et Bithynicae 1843. Doubtful species excluded. Monstrosities marked 'lusus' not included: in counting the number of varieties, those marked (a) not included.

C. Ledebour Flora Rossica. 1842. I have made no exclusions, but I have not taken the trouble to add the species in the Addenda, I have counted as varieties, only those marked by Greek letters, & not those species which are merely said to be variable. In counting the number of varieties themselves, those marked (a) not counted: nor the sub‐varieties of varieties.

Asa Gray. Manual of the Botany of the Northern United States. 2nd. Edit. 1856. The naturalized plants are omitted & all in the


Larger Genera Smaller Genera
Small‐type varieties 74/1136 = 65/1000 1.43 34/917 = 37/1000 1.36
Large‐type varieties 38/1136 = 33/1000 1.34 31/917 = 33/1000 1.38

large genus Salix, according to Dr. Gray's advice. The varieties are divided into two classes, the ordinary ones which are less strongly marked, & others printed in full‐faced type, which have been thought to be species by some Botanists & about which Dr. Gray is doubtful: the two kinds of varieties are classed together in the tables, as the number of the more strongly marked varieties is so small: taken separately we have as follows.

e/Webb & Berthelot. Hist Nat. des Iles Canaries: Phytographie. I have not been able to exclude the many naturalised plants. I have here included not only the varieties marked by Greek‐letters, but those polymorphic species of which the variations are divided into groups.

Hooker & Thompson. Flora Indica. 1855. This is a mere fragment, including only 428 species; & was taken only because illustrating a tropical country. The species "dubiae" have been excluded. The variations marked "variants" not counted, only those marked by Greek letters.

Hooker. Flora Antartica 1844. I have taken only the portion including Tierra del Fuego, the Falkland Islands & Kerguelen Land. This Flora including only 340 species is too small for our purposes: & was taken only from giving so distant a locality.

Hooker. Flora of New Zealand 1853. This is an interesting Flora for our purpose from containing so many endemic species. The large genera Senecio, Coprosma & Veronica have been omitted, as Dr. Hooker informs me that the species are so variable, that it is difficult to say what are species & what varieties. Had these been included the proportion of species having varieties would have been larger in the larger genera.— In calculating the "close‐species" ,/ev/the above genera, & Carex & Uncinia have been omitted: these latter because Dr. Hooker had not himself described them: the monotypic genera have, also, of course been excluded for this latter purpose:—/


Wollaston—Catalogue of the Coleopterous Insects of Madeira. 1857. The certainly & probably naturalised species have been omitted. Several new species have been added since the publication of the Insecta Maderensia: I tabulated the insects in this latter work without removing the naturalised species, & the result is for the large genera 148/1000 & for the small 86/1000. The varieties have been most carefully attended to in these admirable works.—

Gyllenhal. Insecta Suecica 1808‐1827. I selected this work on the advice of Mr. Wollaston. The species given in the addenda have not been added in.—The numerous variations are mostly of a very trifling nature, being chiefly confined to colour.—

Furnrohr: Flora Ratisbonensis 1839 (in Naturhist. Topog von Regensburg) This list has been used only for the species marked "sehr gemein".

Alph: de Candolle Prodromus. Vols. 2, 10, 11, 12, 13 & 14. These volumes were kindly selected for me by Dr. Hooker for various reasons, as containing several large & well worked out Orders & several small Orders. The Proteaceae are remarkable for their confined range. These six volumes/g/include 15,645 species. Those "non satis notae" & "dubiae" have been excluded. In counting the number of varieties, the few cultivated ones have been excluded: those marked (a) have not been enumerated, as being generally only the typical forms in excess. I have experienced some doubt about some of the varieties marked by asterisks.

<I may here add that in Tables II & III, several of the works were selected for & kindly lent to me by Dr. Hooker.>

[page] 172 CHAPTER V



A fortunate change of ink after the first 18 folios of the manuscript of chapter five reveals significant details in its history. Darwin started writing this chapter under the title 'On Natural Selection' and only later decided to add 'The struggle for existence' as the main theme. The original ink, now brown, is clearly distinguishable from the black of the later additions, notably in the title of the chapter, the added last sentence on folio 8: 'This present chapter will be devoted to the Struggle for existence,' and the slip of paper with the revised beginning of the direct discussion of this theme (fol. 9 A).

Although in the original brown ink version Darwin placed 'War of nature' as an alternative to 'Struggle of nature' as a rubric for this section, and began it in the Hobbesian vein, 'all nature is at war,' and although, through Erasmus Darwin he knew the even harsher Linnaean image of 'One great slaughter‐house the warring world!'1 he later changed his rubric to 'struggle for existence'. This he could interpret more broadly than war between organisms to include the physical environment as well: 'A plant on the edge of a desert is often said to struggle for existence' (fol. 30A').

The latter statement is in an interpolated addition equalling about three to four full folios of text giving an extended definition of the term struggle for existence. This interpolation and another one of half a dozen pages mainly on intraspecific competition are the only extensive revisions of the manuscript for this chapter. I have found no sure evidence of their dates of writing, but it is tempting to relate them to the entries for this chapter in Darwin's Pocket Diary. Here he originally wrote under 1857: 'Feb. 27 Find Ch. 5 Struggle for Existence' then later cancelled the date to change it to March third. The earlier date might well be for the completion of the original draft and the later for the completion of Darwin's revising.

The further accumulation of notes and observations in the portfolio for this chapter continued, even long after the completion of this draft. In the spring of 1857 soon after he completed the chapter draft, he began a series of observations on the effects of enclosure on portions of the heath at Farnham, Surrey, near Moor Park, once the residence of Sir William Temple and occasionally of Jonathan Swift, and later of Dr. Lane, whose water‐cure Darwin took several times. These observations relate closely to those given in this chapter on folios 48 and 49 on the heath at Maer Hall, near Stoke‐on‐Trent, Staffordshire, the estate belonging to Josiah Wedgwood, Darwin's uncle and father‐in‐law. Some of these Farnham Heath observations are

1 The Temple of Nature (London, 1803) Canto IV, line 66. Cf. Linné: 'Ianienam & Bellum omnium in omnes', in Politia naturae, Cap. II, 15, but omitted in Brand's English translation, and Linné: 'bellum omnium perpetuum in omnes, et horrenda laniena', in Föreläsningar öfver Djurriket..., ed. E. Lönnberg (Uppsala, 1913), p.1.


included in the appendix to this chapter. The corresponding published passage is on pages 71 and 72 of the first edition of the Origin. Even in May 1862, a year after the publication of the third edition of the Origin, Darwin recorded observations on the struggle for existence on heath land, which, by labelling 'Ch. V, he still associated with the Natural Selection manuscript rather than with the Origin, for. which the chapter number would be three.1


[completed March 3, 1857]

1/In treating of the variation of our domestic productions it was shown that the changed conditions of their existence had some direct effect on them, as food on size, heat on their hair &c, but that indirectly the effect was more potent in tending to render their whole organisation plastic, or less true to the parental type. This view of the organisation being thus rendered plastic in various ways, as if, though of course not really, by mere chance, is, I think strongly supported by the many facts given in our third chapter, showing how sensitive the reproductive functions generally are to changed conditions. It was further shown in the first chapter, that Selection by man, whether intentional or un‐intentional, combined with the strong principle of inheritance, played a most important part in adding/2/up very slight variations in a given direction.

In the last chapter we have seen that in all organisms in a state of nature there are at least individual differences, & in some a considerable amount of variation. It would be strange, inasmuch as variability in main part is due to changed conditions, if this were not so, as Geology consists of the history of the many changes which the earth and its inhabitants have undergone. <& from these changes its inhabitants must suffer or profit. No one who has studied Lyell's Principles of Geology will dispute this. Look to our last epoch, within which the far greater proportion of the now living beings have existed, & reflect over how a vast an expanse of land in Europe & both Americas the sea flowed & left its shells & boulders: reflect on the prodigious changes of climate evidenced by the long intercalated glacial period:/3/all those organisms which were so situated that they could not emigrate must have suffered almost every possible change which their organization could withstand; indeed far more, & there must have

1 CD. MSS. vol. 46.1, item 58.


been much local extinction. Occasionally a living being must get into an island or other isolated site, where it would be exposed to new conditions & yet might survive, like the very many productions naturalised by man's intervention. Some reasons were given in the first chapter for supposing that abundant food might be one main cause of variation under domestication; & I think I shall be able to show hereafter that the species, which are now most vigorous, ranging furthest & abounding most in individuals are those which vary most; & thus we may believe are the best nurtured.

Let the cause be what it may, organisms/4/in a state of nature are in some degree variable;>1 But mere fluctuating variability, or any direct effect of external conditions (to which subject we shall return) are wholly inadequate to explain the infinitude of exquisitely correlated structures, which we see on all sides of us. Look at the Anteater with its great claws & wonderful tongue; or at the Woodpecker, or the Hawk which may swoop down on it, or at the wood‐boring beetle on which it preys, or at what we consider the humblest creature, the parasite so admirably formed to cling to its feathers./

5/The most credulous believer in the "fortuitous concourse of atoms " will surely be baffled when he thinks of those innumerable & complicated yet manifest correlations. In quite simple cases, as in seeds furnished with hooks so as to be transported by animals, the believers in such a doctrine might, perhaps, adduce the case of the cultivated Teazle, believed by many botanists to be a mere variety, & yet so well adapted, that it cannot be imitated by man's art, for a special purpose; & he might say as chance in this instance has favoured man, so in other cases it might favour the plant. But no one I should think could extend this doctrine of chance to the whole structure of an animal, in which there is the clearest relation of part to part, & at the same time to other wholly distinct beings. It is superfluous to give examples: every animal if we know it well, could suffice; but the/6/instances are more obvious in some cases than in others, as perhaps in those given, or as in those insects, which have their structures specially adapted to lay their eggs in the larvae of other particular species of insects; others again being adapted to lay their eggs in special plants together with a marvellous poison <which no chemist can understand or imitate>, which will cause the tissues of the plant in question to develop a gall of fixed form, serving as food to the insect, & appearing like a prison, but out of which the prisoner in due time knows full well how to escape.

1 [For the rest of this cancelled passage see appendix.]


No theory of the derivation of groups of species from a common parent can be thought satisfactory until it can be shown how these wondrous correlations1 of structure can arise. I believe such/ 7/means do exist in nature, analogous, but incomparably superior, to those by which man selects & adds up trifling changes, & thus brings his pigeon or canary‐bird or flower up to a preconceived standard;—or gets one breed of dog to point to his game & another to retrieve it, in a manner which no wild animal would follow;—or gets the wool of one breed of sheep to be good for blankets, & another for broad cloth. If those slight variations of structure, which we see occurring in beings in a state of nature & which from our ignorance we attribute to chance, or changed conditions, if these could be selected & added up, not for man's good, but for that of the being in question, in such case the structure of one part might be adapted to another part, or to some distinct organism/ 8/& the whole being might be harmoniously modified. And for myself I am fully convinced that there does exist, in Nature, means of Selection, always in action & of which the perfection cannot be exaggerated. I refer to that severe, though not continuous struggle for existence, to which as we shall immediately see all organic beings are subjected, & which would give to any individual with the slightest variation of service to it <at any period of its life> a better chance of surviving, & which would almost ensure the destruction of an individual varying in the slightest degree in an opposite direction. I can see no limit to the perfection of this means of Selection; & I will now discuss this subject,—the most important of all to our work. This present Chapter will be devoted to the Struggle for existence./

9A/The Struggle for existence.2 All Nature, as the elder Decandolle has declared with respect to plants, is at war.3 When one views the contented face of a bright landscape or a/9/tropical forest glowing with life, one may well doubt this; & at such periods most of the inhabitants are probably living with no great danger hanging over them & often with a superabundance of food. Nevertheless the doctrine that all nature is at war is most true. The struggle very often falls on the egg & seed, or on the seedling,

1 [Above the word correlations in the MS., Darwin pencilled in lsquo;co‐adaptation' as an alternative term.]

2 [Cf. appendix for Darwin's earlier versions regarding this key phrase and for discussion.]

3 [Quoted by Lyell sans cite: Principles, 9th ed. p. 670, in the chapter in Theories respecting the original introduction of species, where CD. marked his copy. See Augustin P. de Candolle, art. 'Géographie botanique', in Dictionnaire des sciences naturelles, vol. 18, p.384.]


larva & young; but fall it must sometime in the life of each individual, or more commonly at intervals on successive generations & then with extreme severity. This struggle & destruction follows inevitably in accordance with the law of increase so philosophically enunciated by Malthus.1 In a country undergoing no great change, on a long average the numbers of all the species cannot increase; & unlike man, other organisms/10/cannot artificially increase their means of support, which must determine the extreme limit of their numbers. Yet all living beings, if not destroyed, even the slowest breeders, tend to increase in geometrical proportion, & often at an enormous ratio.

Everyone must have seen statements of the number of eggs & seeds produced by many of the lower animals & plants.2 To illustrate geometrical progression one meets in works on arithmetic calculations such as, that a Herring in eight generations, each

1 Essay on the Principle of Population 1826. Franklin & many others have clearly seen & exemplified the great tendency to increase in all the lower animals & plants. [See Franklin, B., observations concerning the Increase of Mankind...Boston, 1755.]

2 I will copy out a few instances of numbers of eggs & seed. Mr. Harmer in Phil. Transact. 1767, p. 280, weighed the whole & portions of roe & counted in this portion the number of eggs. The number differed considerably in different individuals.

Carp 203.109 and 101,200 lowest number
Cod 3,681,760
Flounder 1,357,400 and 133,407 do
Herring 36,960 and 21,285 do
Smelt 38,278 and 14,411 do
Lobster 21,699 — 7,227 do
Prawn 3,806 3,479 do
Shrimp 6,807 4,090 do

(N.B. These observations on the F. Water fish are confirmed by independent calculations by C. F. Lund in Acts of Swedish Academy Vol. 4.)

Astacus gammarus   12,440 [Linné] Brand Amoen Acad. p. 343

Holothuria 5000 ova in one night Sir J. Dalyell

Doris, 600,000—counted by myself. Journ. of Researches p. 201. [Powers of the Creator, vol. I (London, 1851), p. 52]

Bombyx mori 500. [Anon.] Silliman's Journal. Vol. 18, p. 282.

Wasp the Rev. Prof. Henslow counted 300 females in one nest in Autumn Ascaris lumbricoides, sixty‐four million. Carpenter Comp. Phys. p. 590. This is the greatest number, I recollect to have seen; & it is almost inconceivable.—

Helenium 3000 seeds Linnaeus [p. 93] in Brands Amoen. Acad. vol. 2 p. 409
Zea mays 2000 Brands Amoen.
Papaver    3200 Acad. vol. 2.
Nicotiana 4300 p. 409

Wild carrot, (a very fine one) according to my calculations had 40,000 seeds Wild parsnip according to Rev. Prof. Henslow had 2250 seeds one which I gathered, had I fully believe 12,000 seeds.


fish laying 2000 eggs, would cover like a sheet the whole globe, land & water: Linnaeus in the Amoenitates Acad. says that an annual plant producing a single flower with only two seeds (& no plant nearly so barren exists) in twenty years would yield one million plants.1 The great‐engineer Vauban calculates that from one sow... [Sentence left incomplete]. Buffon ranks fifteen animals as less fertile than man (a statement which I rather doubt); &/11/ yet man in the United States, has doubled in 25 years. The Elephant is supposed to be the slowest breeder of all living creatures; & I have seen it stated that were this not so, elephants would overrun the world! The elephant is supposed not to breed till <20> perhaps 30 years old; its length of life is not known, but as one of unknown age when taken lived according to Dr. Falconer 120 years, I think it will not be an exaggerated statement to take <80> 90 years as the possible duration of life & that each pair produces >four> three pair of young: in this case from one pair there will be at the end of 500 years 5,111,514 elephants alive: or if we assume that the pair produced eight young there would be above fifteen millions alive. Hence we can plainly see that it is not from want of fertility that this animal, the least fertile of any, does not overrun the world.2

But we have far better evidence than calculations of the possible increase, namely the actual increase of many animals & plants under favourable circumstances. The marvellous increase of several of our domestic animals where run wild in different parts/12/of America have repeatedly been quoted:/12 v/for instance the great herds, some even of 8000, seen in Cuba only 27 years after the discovery of that island:3 /12/Nothing has astonished me so much in this respect as to find in Sarmiento's Voyage that in only 43 years after the horse was first imported4 into Buenos Ayres, where it immediately ran wild, it was in possession of the Indians at the Straits of Magellan, 1200 miles to the south. We have similar facts in New S. Wales: thus in 17885 29 sheep & 5 cattle were introduced;

1 ['On the Increase of the Habitable Earth' pp. 94‐5 in Select Dissertations from the Amoenitates Academicae.. .Trans. F. J. Brand, London, 1781.]

2 [See ULC vol. 46.1 f. 35 for calculations and comment re increase of elephants.]

3 Lyell's Principles of Geology 9th Edit. 1853, p. 685. Robertson [II, book 8, p. 394] quotes for increase of cattle in S. America Oviedo, ap. Ramusio III, 101 Hackluyt III, 466 & 511 Churchill Collect III, 47 & V, 680, 692 Feuille, I, 249 Acosta, Lib III c. 33 [Robertson's citation for Acosta should read 'Lib. iiii cap. 33'. To identify all these citations, the editions used by Robertson would have to be identified For example, his reference to 'Churchill Collect, III, 47' seems to fit pp. 48‐4 of the 1st ed., i.e. Ovalle, Chile, book 1, ch. 21.]

4 In 1587, see Rengger, Natur. der Sä[u]gethiere von Paraguay S. 334.

5 [Report given in Sydney Smith's Works, vol. I, p. 324.]


29 years afterwards the numbers were for sheep, 170,920, & for cattle 44,753; & no doubt many must have been slaughtered in the interval. In 1418 a single female rabbit was turned out in the island of Porto Santo1 in a few years 3000 were killed at one time; & 36 years afterwards Cada Mosto in his voyage/13/speaks of them as innumerable;2 nor is this wonderful as it has been calculated3 that one pair might produce 1,274,840 individuals in four years. Equally striking & well known are the many facts, showing the astonishing increase of many native animals, when two or three favourable seasons have followed each other consecutively: thus during the famous drought of 1826‐28 (inclusive) in La Plata the whole country literally swarmed with mice, which disappeared with the returning wet. In Germany a similar increase of field mice was accompanied by an astonishing increase in stoats &c. which preyed on them. It would be superfluous to give the cases amongst my notes of the enormous increase of Birds, fish, frogs, snails & insects, when turned out in new countries: the one island of Mauritius4 would afford striking instances in all these classes except fishes; & for fish we may turn to N. America. Bees & wasps taken from Mauritius have come to swarm, as I am informed by Capt. Moresby on the miserable /14/coral Chagos islets.

Of the rapid <& often overwhelming> increase of plants run wild, innumerable instances could be given. America over large districts has been peopled by plants from the old World & in La Plata to a quite overwhelming extent: on the other hand there is scarce a region of the world which has not got now widely extended colonists from America since the time of Columbus: in India, as I am informed by Dr. Falconer, three of the commonest plants from Cape Comorin to the Himalaya are of American origin. In the island of P. Santo.5 .../

15/In the foregoing cases, & innumerable others could have been added, we cannot account, at least in any great degree, for this wonderful <observed> rate of increase, by the law of fertility having been suddenly altered in each species. In the higher animals, the period of gestation & suckling, the number of young produced at a birth, the length of natural life, would almost certainly remain constant; probably the animals would breed at a little younger age & oftener when better fed than in their native country; but

1 By J. G. Zarco in Kerr's Collect, of Voyages. Vol. 2, p. 177.

2 [For Cada Mosto see Kerr, Voyages, 2: 205.]

3 Fothergill Philos. of Nat. Hist. p. 137

4 Sonnerat's Voyage aux Indes vol. 2. p. 83. I could add other instances.

5 [Darwin left this sentence incomplete and the rest of the folio blank except for the pencilled memorandum: 'St. Helena'.]


this could hardly apply in all cases as in short‐lived animals & annuals. No one will maintain that the American Parkinsonia has spread over all India, or that the European cardoon & thistle have overwhelmed the plains of La Plata, owing to their producing more seed than in their aboriginal land. Undoubtedly the great increase must almost exclusively be due to all, or nearly all the young surviving & breeding, with the old likewise still surviving & breeding. The result of geometrical progressions invariably strikes one with surprise. The observed rate/16/of increase in the foregoing instances could not possibly be continued for centuries, for neither earth nor ocean could hold the product:/16 A/Nor is it probable that the cessation of increase or actual decrease as with the mice of La Plata, would be in any high degree influenced by lessened fertility; for I think the young would perish, before the old were starved to the degree as not to breed; & in the case of the domestic animals run wild they would hardly spread into districts, already stocked with native animals, so unfavourable to render them in any marked degree sterile. Indeed according to Mr. Doubleday's theory, in which for reasons given in our third chapter, [See ch. 3, fol. 99] I do not believe, but/16/which has found several advocates, organic beings when pressed for food, breed the more freely, causing the struggle for life to be more fearful.

In a state of nature, all plants annually produce seed, excepting a few which propagate at a great rate by suckers &c, & still fewer which are just able to live in the extreme arctic regions & on high mountains, where they have to struggle not against other living beings but against cold. All or nearly all animals pair in a state of nature excepting apparently a few males in excess, & a few barren individuals. Had this not been so, it could/17/hardly fail to have been observed in our game‐birds & other carefully observed wild animals.1 The time of pairing, I believe, always falls at a

1 With respect to barren birds, which are not at least in the case of Solan Geese, young individuals, it seems that they are not very rare in sea‐fowl. See Wilson's Voyage round Scotland Vol. 2, p. 77. For the excess of males see the fact given in regard to Partridge by White of Selbourne in Letter XXIX. But there are other facts mentioned in the same letter in regard to both males & females of sparrows & owls, quickly getting a new mate, when one has been shot, which are of difficult explanation. This fact has been particularly observed in the case of the Magpie: Jenner (in Philosoph. Transact. 1824. p. 21) relates the case of a pair of Magpies with a nest, of which seven were successively shot, but the widow or widdower was again immediately paired: in another case given by Macgillivray (British Birds, vol. I, p. 571) six females were successively shot on the same nest of eggs. As many nests, especially conspicuous nests like that of the magpie, are annually taken, one may conjecture, that a bird having a nest, offers an irresistible attraction to either sex of a nestless pair, to break their marriage vow.


period when the animal is at full vigour; though no doubt it is of still more consequence that the young should be produced at a time when food is superabundant & the other conditions of life favourable: hence it is in itself highly probable that nearly all animals pair annually or biennially according to the period of gestation. We have seen how great has been the actual increase of horses & cattle, in short periods, though many must have been slaughtered or killed by accidents; & these animals, when compared to the great mass of living beings must be considered as extremely slow breeders: we know the actual rate of doubling of man, a still slower breeder; & we have seen the possible increase of the supposed slowest breeder, the elephant, if allowed to live & breed at its natural rate, even for a few centuries, whereas we have to consider hundreds of thousands of years. Therefore I consider nothing can be/18/more certain, than that every single species on the face of this earth would rapidly swarm to an incalculable degree, if many individuals were not continually destroyed at some period of their lives from the egg or seed upwards, either during each generation or at short intervals in the successive generations.

Checks to increase in animals. What are the checks to this <possible, & as we sometimes see the actual> tendency to a high rate of increase in every living thing? This is a most difficult & curious question, which cannot be completely answered in any single instance. This subject of the Police or economy of nature has been ably discussed by many authors from the time of Wilcke1 nearly a century ago to the present day when it has been ably handled by Sir Charles Lyell. A volume would be required to treat the subject properly, & I can give here only a few of the leading facts, which have most struck me. The checks are often of a very unexpected nature. Let us look first at our domestic animals/19/ become feral in America, about which we might expect to know most. Though both cattle & horses multiplied greatly in La Plata when left on the desertion of the colony in 1537 to themselves, & <though> subjected to the attacks of Indians; yet at no time have they run wild in Paraguay; & both Azara & Rengger2 clearly show that this is owing to the greater number of a certain fly, there, which lays its eggs in the navel of the newly born young. In parts of Brazil, cattle can hardly be kept even in a domestic state, whole

1 [Carl von Linné, praeses, H. C. D. Wilcke, respondent, 'On the Police of Nature', pp. 129‐64 in F. J. Brand's Select Dissertations from the Amoenitates Academieae.]

2 [Rengger] Naturgesch. der Säugethiere p. 335, & 360. [Felix d'Azara, Voyages dans I'Amérique méridionale. I, p. 215 and Azara Essais sur quadrupèdes, II, p. 368.]


herds perishing from exhaustion in the dry season from the multitude of ticks (Ixodes) with which they are infested:1 in another part they failed from the attacks of blood‐sucking bats on the calves.2 In La Plata, where these causes do not come into play, great droughts are almost periodical, & <horses &> cattle of all kinds perish actually by the million, more especially by rushing by thousands into the great rivers, & from drinking saline water.3 These droughts destroy myriads of wild/20/animals, & even birds, whereas we have seen that during these very same periods mice swarm to an incalculable degree.—I may add that everyone has heard of the terrible destruction of sheep in Australia from the droughts: so it is in India, & Dr. Falconer tells me in places where formerly one man could kill 30 or 40 Deer in a day, for some years after a great famine & drought, hardly a single deer could be got. But to return to the cattle, further south in the Falkland Islands, there are no droughts, or injurious flies, or ticks or bats, & the cattle are magnificent animals & have multiplied greatly; but, as I am informed by Capt. Sulivan,4 who has kept cattle in these islands, every few years a hard winter like the 1849 destroys numbers, & even those that survive in the following spring are so much weakened that many die of diseases & get lost in the bogs.—The Horses there do not suffer so much from the snow, as their instinct teaches them to scrape the ground with their hoofs; but oddly enough they have multiplied/21/far less than the cattle, & here were left to eastern end of the island; though the western is the more fertile:5 the Gauchos can account for this only from the stallions constantly roaming from place to place & compelling by kicks & bites the mares to desert their young: Capt. Sulivan can so far corroborate this statement that he has several times found young foals dead, whereas he has never found a dead calf.6 Horses here deteriorate in size, & they are apt to grow lame from the boggy soil, so climate here, no doubt, aids in checking their increase but the fact of their not spreading seems to show that

1 Gardner's Travels in Brazil p. 295, 388.

2 In parts of Demerara Fowls cannot be kept from the same cause, Waterton's Wanderings p. 163, 4th Edit.

3 Darwin Journal of Researches p. 134.

4 [Bartholomew James Sulivan, admiral and hydrographer was one of Darwin's shipmates on the 'Beagle'. See ULC vol. 46.1 fols. 17‐18 of second numbering, for Darwin's notes, dated March, 1856, on Sulivan's information.]

5 Journal of Researches p. 191.

6 [Suhvan MS. letters to Darwin. C.D. MSS. vol. 46.1 fols. 73 v‐74, (undated portion of letter) and fol. 81 v from letter dated Jan. 13, 1844.]

7 It is possible that in this case the Horses' fertility may be somewhat lessened: for in the Shetland Isl'd (Fleming's Philosophy of Zoology. Vol. 2. p. 10) the Pony does not leach maturity till its four year, is not vigorous beyond its twelfth; &


the check falls chiefly on the young. I may add that Rabbits, though very numerous in certain parts of the Falklands likewise have not spread: what the check is here, I have no idea; or what the check is in Jamaica1 where the Rabbit is feral but has not multiplied.

There can be no doubt that carnivorous animals keep down the numbers of the animals on which they prey. It is worth noticing the number of domestic animals destroyed in single Kingdom./ 21A/In the year 1823 in Livonia there were destroyed by the wolves 1800 horses, 1800 cattle, 15,000 sheep, 2500 goats, 4000 pigs, 1200 fowls, 673 geese &c &c.2 The number destroyed, however, must often depend on complex relations: to give a single instance, according to Nillsson3 wolves have of late increased in Halland & foxes decreased; & this it is believed/22/is chiefly owing to the wolves running down & devouring the foxes, as has often been witnessed; but they can do this only on open plains, so that the proportional increase & decrease of wolves & foxes here depends indirectly on the presence of trees.4 We are perhaps apt to lay too much stress on the amount of food as determining the numbers of any species; for it seems well ascertained that game in any district, even in this our highly cultivated country, where so few hawks or carnivorous animals are seen, can more certainly be increased by the trapping of vermin than any other means.—But there are some few animals which are probably never, either whilst young or old, destroyed by beasts of prey as the elephant; & yet they do not increase to the extent, which their degree of fertility would soon permit: in this case the check is no doubt periodical famines & droughts which we have seen occur in India; & when weakened they would be very apt to perish in morasses, as seems to have happened with the fossil mastodons of N. America. On the coast of Africa, Capt. Owen5 gives a curious account of the

breeds only biennially. The dampness of the climate probably is the deteriorating agency, for Wrangell (Expedition to the Polar sea p. 28) states that in the extreme climate of N. E. Siberia, the Horse is serviceable even at 30 years old. With respect to the wild stallions killing their foals, the same thing has been observed in Australia, see Haygarth's Bush Life, p. 76.

1 Gosse's Sojourn p. 441.

2 [Anon.] Silliman's Jour. v. 20, p. 177. Rev. encyclop. Sept. 1830.

3 Lloyd Field Sports of N. Europe Vol. I, p. 395.

4 A beast of prey must often prevent other animals from haunting districts in which they could live and might prefer.

5 Surveying Voyage Vol. 2, p. 274 [Contrast with the note on this same passage which Darwin gives in his Journal of Researches (1845) p. 133, where he correctly quotes Owen as writing: 'A number of these animals had some time since entered the town in a body, to possess themselves of the wells, not being able to procure any water in the country. The inhabitants mustered, when a desperate conflict


sufferings of the/23/elephants, which in a body fairly took possession of a town for the sake of the water & drove out the inhabitants who numbered about three thousand!

I will give a few other instances of checks to increase from apparently trivial causes. The ferret cannot be kept in W. Indies1 owing to the chigo or sort of flea, which burrows in their feet. In [ ] the half‐wild dogs invade each other's districts when pressed for food, fight & <wound each other> flies lay their eggs in the slight wounds & cause their death. Everyone has heard how Rein‐deer2 are forced to migrate in vast bodies & annually perish in multitudes owing to the mosquitoes. Dean Herbert seems often to have been perplexed3 why certain animals do not increase: he instances the toad, of which such myriads are often seen, showing that they do not perish in the egg‐state, & as no animal preys on the toad, he asks why they do not increase infinitely: I can adduce one check, namely a maggot of some fly, which breeds in their nostrils, & which destroys thousands in Surrey, as I have seen, & in parts of Kent, as I have been informed by Mr. Brent. But the Dean might have asked with still more force/24/why the natter‐jack, (Bufo rubeta), which lays eggs enough to people the earth in a few generations, is confined to a few spots in England, where, however, it is common as on Gamling‐gay Heath. What animals can seem less concerned with each other than a cat & Humble‐Bee; yet Mr. [H. W. Newman]4 shows that field mice are the most powerful enemies to the Bee, & the cats determine the number of mice, as everyone knows in his house, & hence he believes that Humble‐bees are apt to abound near villages, owing to the destruction of the mice. From the facts given in our third chapter, I cannot doubt that the number of seed produced by certain flowers will be determined by the part which Bees play in their fertilisation; & on the number of seed to a certain extent depends the number of the plants; & on them the number of certain other insects & on them certain birds ad infinitum. To attempt to follow the mutual action & reaction in any one case, would be as hopeless

ensued, which terminated in the ultimate discomfiture of the invaders, but not until they had killed one man and wounded several others.']

1 Gosse's Sojourn p. 447.

2 Wrangell's Travels p. 48. [i.e. Wrangel, Eocpedition to Polar Sea].

3 [Darwin left a space between brackets here in his manuscript for a reference to be supplied later.]

4 [Darwin left blank spaces here in his manuscript for the name of his authority and for the reference. In the published version he supplied the name H. Newman. See The Origin of Species, 1st ed, p. 74. H. W. Newman, 'On the Habits of the Bombmatrices', Entomol. Soc. London, Trans. N.S., Proceedings Section, p. 88, 1850‐51.]


as to throw up a handful of (sawdust) feathers on a gusty day & attempt to predict, where each particle (of sawdust) would fall./

25/This subject is so important for us, that I must be excused for making a few more remarks. Our British Birds are probably the best known wild animals. Take the case of the familiar Hedge sparrow (Accentor modularis), which that acute observer, Mr. Waterton,1 says will not increase in numbers, however carefully protected. If not killed it could probably live at least seven years:2 it generally has two broods of about five eggs, but let us suppose that only every other pair rears any young, we will say only two pair. We thus seem to allow a fair amount of destruction at an early age; yet if we suppose that in Mr. Waterton's grounds there were at one time eight pair, the above rate of increase would yield at the end of the seven years, when the eight old pair would die, 2048 birds; but we have just seen, that though carefully protected by man they do not increase at all. It cannot in this case be any difficulty in finding a place for a nest; & I sh'd think hardly more than three out of four nests would be taken by cats; & only one out of four nests are supposed to be preserved/26/in the above calculation. That in many other cases the loss of the nest is a most important check we may infer from the wonderful increase of Magpies & some other comparatively rare birds in Mr. Waterton's park,3 where in one year 34 pair of Magpies bred & reared 238 young ones. The Hedge sparrow in a garden near a house can hardly suffer much from Hawks & the smaller wild carnivores, which are so influential in checking the increase of game‐birds. I doubt whether the young birds, during the first few months suffer greatly; at least with the Robin everyone must have noticed their numbers in their mottled plumage. <& in our migratory birds, as White long ago observed in his letters, the check must fall on the young birds which leave us, for what we imagine to be a more favourable climate, for comparatively few of those which migrate return to us.>4 The domestic cat is I believe a potent enemy, which with other occasional causes of death must prevent any great increase in numbers; but I believe nearly all our Birds

1 Essays on Nat. Hist. 2 Series p. 95.

2 In the N. American Journal of Science vol. 30, p. 81. It is said [by J. Bachman] that the same pair of Saxicola sialis built its nest in one place for 10 successive years;—a Muscicapa fusca for 9 years; a Turdus for a longer period; Falco borealis for 12 winters. Eckmarck Amoen. Acad. noted the same home in starling for 8 years [see: Linne, 'On the Migration of Birds']; a Motacilla & Kestrel for 6 years. In Montagu's Ornith. Diet. [p. 217] it is said that a Goldfinch lived in confinement for 23 years.

3 Essays p. 269. [1st series, essay on Magpie.]

4 [Selborne, part 2, letter 16.]


do go on increasing/27/in numbers, till there comes a severe winter, which greatly reduces their numbers & sometimes exterminates them in certain districts.1 After the winter of 1854,2 judging from the number of nests in my shrubberies & from the number of birds on my lawn, I estimated the decrease at four‐fifths compared

with previous years. In the summer of 1855, butterflies & moths abounded in an extraordinary manner, which some naturalists at the Entomological Society attributed, I believe rightly, to the lessened destruction of the caterpillars by birds: the little Tomtit (Parus coeruleus) has been observed to feed its young with caterpillars 475 times in the day.3 With man we consider an epidemic which destroys ten percent as frightful; but in <this> above case with the birds it seemed to me that the destruction had been at least 80 per cent.

With the higher animals, as soon as the young can provide for themselves they are generally driven away by the old: in their forced wanderings many probably perish; but some no doubt find a home, in spots where the destruction has been above the average, more especially after any unusually fatal period. The Rev. L. Jenyns informs me that in Swaffham, during twenty years, sparrows/28/& Rooks were unmercifully killed <for a reward offered per head>, but the most careful observer could observe no diminution in their numbers during this period: no doubt the spare birds from the surrounding parishes flocked in; but what would have become of these birds had not there been room made for them in Swaffham? undoubtedly they would have wandered away, some few have found a home & the others have perished during the first severe winter. In all cases, probably, the destruction is unequal in different parts of the whole area inhabited by the species; but this does not alter the final result; Nor is it applicable to the endemic species of small insulated regions: we may go in imagination from spot to spot, <& everywhere the rate of increase is far higher than what can possibly be supported> & we may fancy that here & there the conditions <of life) are so favourable that all survive to their full term of life; but if this be so the destruction must be very heavy in other spots, for, as repeatedly remarked the rate of increase in every living being is so high that

1 [John L. Knapp] Journal of a Naturalist, p. 182.

2 Severe winters destroy not only the inhabitants of the land, but of the sea; both certain species on the coast, as described by Hugh Miller (Royal Physical Soc. of Edinburgh Feb 28th 1855), but likewise on banks under the water: thus in 1820‐30 Kröyer (Eding New Phil. Journal 1840 p. 25) says eight million oysters were computed to have been destroyed by the frost.

3 Macgillvray's British Birds Vol. 2, p. 438.


the earth could not hold the product. In animals capable of much locomotion, & inhabiting a continent or the ocean, it is likely that many wander to the/29/extreme confines of their natural range & there perish in larger numbers than elsewhere. But how rarely could this be ascertained !/29v/A pair of sparrows bred for the first time in 1833 in [the] island of Colinsay, one of the Hebrides, but in 1841, no descendants could beseen.1 /29/Richardson2 speaking of the extreme northern range of the American Antelope, says that almost every year a small herd lingers on a piece of rising ground not far from Carlton‐house; but few or none "survive until the spring, as they are persecuted by the wolves, during the whole winter." So again with Arctic Fox, he says "Most of those which travel far southward are destroyed by rapacious animals; & the few which survive to the spring, breed in their new quarters, instead of returning to the north. The colonies they found, are, however, soon extirpated by their numerous enemies."

In those animals which produce an astonishing number of eggs, the destruction probably chiefly falls on the eggs, as is known to be the case with Fish, from other fish, water‐beetles &c. But when the old can protect their young few are generally produced as with the larger carnivorous birds: the Lion, however, produces several young at a birth, but when the/30/Lioness is hunting for food, it is asserted the hyaenas prey on her young. In very many other cases the check falls not on‐the egg, but on the young: thus Smeathman3 thinks that "not a pair in many millions" of the Termes or white ant "lays the foundation of a new community," common ants being the chief destroyers. In other cases, of which instances have been given, the very young do not seem especially to suffer: thus White of Selbourne long since remarked in his sixteenth Letter [to Barrington] that in our migratory birds those returning yearly, from what we imagine to be a more favourable climate "bear no sort of proportion to the birds that retire."/

30/A/As in this chapter I repeatedly use the expression of struggle for existence; I may here remark that I employ it in a very large sense./30A'/Carnivorous animals prowling for their prey in a time of dearth may be truly said to be struggling for existence; so when seeds are sown so thickly that all cannot grow, they may

1 Wilson's Voyage round Scotland Vol. I p. 368. [Wilson here wrote of Stornoway not Colinsay.] I may add that Partridge hens have been turned out here but they became extinct. In 1841 Rooks bred for the first time in this island. Will they hold their own?

2 Fauna Boreali‐Americana [part 1] p. 88, 268.

3 Philosoph. Transact. 1781, p. 167.


be said to struggle, though not voluntarily against each other. A multitude of animals are directly dependent on other animals & on plants; & plants on the nature of the station inhabited by them; & here the idea of dependency seems quite distinct from a struggle. But a plant on the edge of a desert is often said to struggle for existence; this struggle consisting in the chance of a seed alighting in a somewhat damper spot, & then being just able to live; so it may metaphorically be said that carrion‐beetles struggle for existence, when fewer animals die than usual in any district. In many cases when an animal depends on another or on a plant/30a/it destroys or injures it to a certain degree; & here more strictly there may be said to be a struggle. Again another idea comes into play, for it may be said to be chance, which seeds in the capsule of any one plant shall be devoured by a bird or insect, but it may metaphorically be called a struggle which individual plant of the species shall produce most seed, & so have the best chance of leaving descendants;/30a'/& again it may be called a struggle whether the plant or the bird <or insect> which feeds on its seeds gets the upper hand. A minute parasite which is absolutely dependant on an animal, cannot be said to struggle with it; yet its numbers will generally be dependant on the vigour of the animal which it will sometimes injure, & with the increasing vigour of the animal the weaker parasites will perish; so that here there may be said to be a struggle between parasite, & parasite & the animal; as there likewise will be which parasite or which carrion feeding beetle shall lay most eggs & so have the best chance of getting into another animal's body or feeding on its carcass.

I hardly know any living being which is more dependent on others, & which seems less subject to a struggle in the strict sense of the word then the Misseltoe; for it depends on certain trees for support, on certain insects for fertilisation, & on certain birds for diffusion; yet even here, when several seeds are dropped close together there must be a struggle which shall grow; there may be said to be a struggle which plant/30 B/shall produce most seeds with most tempting pulp for the thrushes; & lastly there may be said to be a struggle between parasite & tree, for the latter will suffer severely from too many misseltoes. In many of these cases, the term used by Sir C. Lyell of "equilibrium in the number of species"1 is the more correct but to my mind it expresses far too much quiescence. Hence I shall employ the word struggle, which

1 [E.g. section heading, Principles, 9th ed., p. 670 and page heading 1st. ed. Vol. II, p. 134.]


has been used by Herbert & Hooker &c.,1 including in this term several ideas primarily distinct, but graduating into each other, as the dependency of one organic being on another,—the agency whether organic or inorganic of what may be called chance, as in the dispersal of seeds & eggs, & lastly what may be more strictly called a struggle, whether voluntary as in animals or involuntary as in plants./

30B'/To return to our subject, it is difficult to realise that every animal is kept down by a severe "struggle"; yet it accords with, & aids us in understanding, much that is passing around us. Lighten the pressure on any one organism in the slightest degree, quite inappreciable by us,. & its numbers will instantly increase. Why are some species rare or quite absent in one district, & abundant in another, under, as far as we can judge, similar/31/conditions. Innumerable instances could be given; & several even within the limits of England; as the absence of the Nightingale in Devonshire, water‐wagtails (Motacillae) & carrion‐crows in certain districts: during 15 years I have only twice seen a swift (Cypselus) in the parish in which I live; yet how common a bird over nearly all England. We can perceive why the sparrow & partridge have increased in numbers in some districts with extended cultivation; but who can explain why during the last 20‐80 years the Missel‐thrush (Turdus [ ]) has increased in Ireland, Scotland in England, as I have likewise myself noticed. Why did the Robin (Sylvia rubecula) decrease & finally disappear in the year in parts of Belgium. A small wading bird (Pelidna [ ]) has increased of late considerably on the shores of the United States. In New S. Wales as Mr. Sutton stated before the Geographical Society some parrots have greatly decreased, & some disappeared; others equally conspicuous as the white cockatoo have remained in about the same numbers, & others as the Blue Mountain parrot have increased. No doubt if we had accurate accounts in past centuries, we shd have endless cases of great changes/32/in proportional numbers: I will give only a single instance from Prof. Nilsson.2 A large Bat (Vespertilio noctula) is now common in Sweden, having appeared about the year 1825, & was quite unknown to Linnaeus; but it seems from the bones found in parts of the walls of the

1 [Herbert, 'Local Habitations and Wants of Plants', J. Hort. Soc, 1 (1840), 47. Hooker, Flora Indica, I, 41, 42. Aug. de Candolle, article 'geographie botanique', Did. sci. nat, vol. 18 (1820), pp. 384, 386. Alphonse de Candolle, Geographic botanique (1855), p. 453.]

2 Report Brit Assoc. 1847. p. 79. Prof. Nilsson gives other curious instances: the water wagtail, (Motacilla alba) was very numerous 30 years ago, then it vanished & now again has reappeared.


Cathedral, which it now again haunts that about 700 years ago it was also very common. Lastly it is the common rule, that a species is abundant within what has been called its metropolis, & towards the confines of its range both in longitude & latitude becomes, often rather abruptly, rarer & rarer, till it disappears; & there seems to be no difference in this rule, whether or not the beings be locomotive: yet as it can exist towards the confines of its range, & as its fertility certainly usually then lessened, how is this? In all these cases, namely of a species abundant in one district & rare or quite absent in an adjoining one,—in their increase or decrease in numbers,—we shall feel little surprise, if we steadily look at the average number of every single species in its most favoured site, as determined by a severe struggle, of which in no one case can we perceive/33/all the elements: the merest grain in the balance will then determine whether the range should be lessened or increased.

The manner in which the diverse checks act & react must be exceedingly complicated. When there is no compensation there will be a steady but slow decrease in numbers: thus "the fur‐trade even when best managed has always been a decaying trade," & post has to be pushed beyond post into the interior: so it has been with whaling; but how different our game. Neither partridges, or grouse or hares are fed, & yet how many hundred thousands are annually killed with no decrease in the stock: no doubt they could be exterminated as the capercailye has been: with our game man compensates by the destruction of vermin, & he kills many which would otherwise have perished during the winter. Let not a gun be fired or a trap set in England for the next 20 years, & I think it may safely be predicted that there would be less game, almost certainly not more. For instance/34/Bruce remarks1 that in Abyssinia Boars, <foxes> & Hares are held unclean & are not hunted, but yet they do not increase in numbers; & he accounts for this by the number of Hyaenas; but whether Hyaenas would destroy many hares may be doubted.

Whatever the number of a species in any country may be, the average being determined by a complex struggle, that number will steadily decrease, if we add without any compensation the least additional cause of destruction, until the species becomes extinct. But the rate of decrease will be very slow: if we have 1000 individuals & we destroy on an average ten per cent more every year at the period when the number is least than were heretofore destroyed, it will take 298 years to reduce the numbers

1 Travels, vol 5, p. 84.


to fifty. But often with the decreasing numbers of the organism destroyed, the numbers of the destroyer will be diminished, & the check thus lessened & its action almost infinitely prolonged. It may well happen that a large additional number of a species might be destroyed without in the least lessening/35/their average numbers; for the destruction may fall before an habitually recurrent period of dearth, which would have in any case thinned their numbers: it is even quite conceivable that such destruction might increase the minimum average, for more food might thus be preserved against the period of dearth, as for instance in dry countries, in which the herbage withers up & serves as natural hay. Many other considerations might have been added showing how complex the action & reaction of the checks to increase must be.

Besides the many & complex checks tending to cause a decrease in the numbers of a species; an inordinate increase, under the most favourable conditions, is prevented in some cases at least, as in our game, by mysterious epidemics, which seem connected we know not how, with the closer aggregation of many individuals of the same kind.—

The great difficulty, which at least I have experienced in fully realising the struggle for life covertly going on around us: I think is partly due to our familiarity with our domestic/36/animals. We see how easily they are reared, how long they live & how seldom they perish from accident; & we overlook our care of them whilst very young & that we artificially preserve food for them & so prevent recurrent famines; but the millions annually slaughtered over Europe, with the stock still kept up, ought clearly to show us what destruction there must be with the allied animals in a state of nature. Nor ought we to feel the least surprise at our not being able to point, how, when & where the check falls on any animal in a state of nature: for the case of man, incomparably the best known, (& in some respects more simple, though in others as in the moral <restraint> check of Malthus or as Laing1 more correctly calls it the prudential restraint, very much more compli‐cated) shows how ignorant we are. Without careful statistical tables: how little could we have judged of the different rates of increase, & expectancy of life amongst different ranks, at different times in different countries & even within the limits of the same town.2

1 [See Samuel Laing, Journal in Norway..., 2nd ed., 481. (Conclusions at very end of book), and his Notes of a Traveller. First Series, 158, (ch. x, section on 'Checks on Over‐population'.)]

2 Mr. Neison has shown (Statistical Soc. March 17th 1845) that in the same town the expectancy of life with mature men of different trades differs by 50 per cent.—


37/Mutual Checks of Animals & Plants. We have considered as yet almost exclusively the manner in which animals check the increase of other animals. But plants & animals are even more importantly related; as are plants with plants. This subject is so important for us, in several ways, that I must be excused for entering into some details, but they shall be few. All animals live on plants either directly or indirectly; & their breath is the plants' chief food; so that the relation of the two kingdoms on a grand scale is very obvious. But it is probably much more precise than it at first appears. One at first supposes that grass‐eating animals devour all plants nearly alike; but of Swedish plants it has been ascertained1 that oxen eat 276 kinds & refuse 218; goats eat 449 & refuse 126; swine eat 72 & refuse 271,2 &c. Southward of La Plata, I was astonished, as others have been,3 at the change effected in the appearance of the plains by the depasturing of the cattle; & could not for some time believe but that there must have been a change in the geological nature of the country. What plants the many small/38/rodents live upon is seldom known, but every one must have heard of the destruction of whole plantations by mice, & rabbits &c. I have heard it remarked that all, or nearly all our spinose & prickly plants are liked by the larger quadrupeds; the spines being an evident protection to them; & I have sometimes fancied that the very common prickliness of the bushes on desert plains was chiefly due to the greater protection from animals requisite for any bush to live, where the vegetation was scanty. It has, also, been shown in detail by Forskahl4 that those plants which are not eaten by cattle are attacked in an extraordinary degree by insects; from 30 to 50 species sometimes preying on a single plant: I presume a plant preyed on by both insects & quadrupeds would be exterminated.

I will not do more than allude to the enormous amount of injury, even to extermination, effected by insects on plants; on which subject copious details are given by Kirby & Spence.5 Land mollusca are, likewise, potent enemies to many plants, especially when young, as every gardener knows: and early on a dewey/39/ morning in what extraordinary numbers they sometimes swarm! In all these cases the relation is obviously mutual: the increase

1 Stillingfleet Tracts. 1762. p. 361, on authority of Hasselgren in Amoen. Acad. [See Linné, 'Swedish Pan', in Stillingfleet Tracts (where name is mis‐spelled.]

2 [Sic. StillingHeet Tracts p. 361 gives the figure 171.]

3 Journal of Researches p. 118.

4 [Linné, 'The Flora of Insects', in Brand's Select Dissertations, pp. 361, 366, 367.]

5 [An Introduction to Entomology, I Letter 6 is on injuries caused to living plants by insects.]


or decrease from any cause of plant & animal mutually affecting each other.

But animals serve plants, as well as destroying them; & in destroying some plants they invariably favour others. In how many ways do they transport their seeds! Even when they devour the seeds if one out of a thousand escapes, it may be of the utmost importance to the plant; of which I shall presently give a curious instance. Though Bees devour much pollen, they are indispensable to the fertilisation of some plants, & generally most useful: different plants are visited by different kinds of Bees; & some by none, but which absolutely require other insects in order to produce seed. Worms I believe1 play an important part for plants in turning up the ground, & in burying seeds. I have often thought when seeing the quantities of manure collected under the most shady tree in a field during hot weather that even this in the great war of nature/40/might make a sensible difference in the vigour & spreading of a tree: on the other hand, Lieut. Breton2 says he has known in Tasmania that trees which were flourishing have actually perished as soon as the land was depastured; & he suspects that this is caused by the ground being bared & thus dryed.

At St. Helena the upper plains, to an extent of 2000 acres were originally wooded, & it seems pretty well made out that the goats & swine which were introduced3 in 1502 & soon multiplied, destroyed all the young trees; & that by degrees the old ones perished of age; so that 220 [years] afterwards it is said "the old trees have mostly fallen"; & now the upper plains are covered with grass without a single tree. Some of the trees are known to be now absolutely extinct. In the surface soil, I collected eight kinds of land‐shells, now extinct; & their extermination & that of many insects has likewise been in all probability, indirectly due to the goats. To give one more example: near Inverorum [Inveroran ?] in Scotland, I saw a whole hill‐side covered with young birch‐trees so nearly of the same age, that I enquired why so useless a/41/tree had been planted; but was told that about ten years before the district had <been> converted from sheep‐pasture into a deer forest; & that sheep devour young birch‐trees, but that deer do not. The growth of the birch, would certainly greatly alter the vegetation on the whole bank; & with the plants, the insects would change; & with them, the birds, of which I shall presently give an instance. It is not too strong an expression to say that the

1 Geological Transactions Vol. [ ]. [Charles Darwin, 'On the Formation of Mould', Geol. Soc. London, Trans., 2nd ser. 5 (1840) 505‐9.]

2 Tasmanian Journal Vol. 2. 1843 p. 136.

3 Darwin, Journal of Researches p. 489.


introduction of a single mammal might change the whole aspect of a district, even to the minutest living details.

On the struggle between plant & plant: the struggle here is not so obvious, but not less certain. Plant does not actually prey on plant, excepting in a few root & branch parasites. Nearly all plants, however, are favoured by the decay of others; and this is indispensable to those which live in peaty earth. In very many cases, also, shade is indispensable or highly favourable: but in plants growing in the shade of others there is some, though perhaps slight reciprocal action, for such plants must rob their protectors of some nutriment;/42/as we see in the greater vigour of our orchard fruit‐trees,when the ground is kept .bare beneath them. Plants, also, often offer protection to the seedlings of others; & as Stillingfleetl has remarked how often do we see a young tree springing out of a furze or thorn bush on a common which has protected it from the attacks of cattle, ultimately to be overshadowed & destroyed by it.

Generally the struggle between plants is like that of those quadrupeds in the same country, which devour nearly the same kind of food. We have evidence of the struggle on a grand scale in the many thousand hardy plants which can be perfectly preserved by simple weeding in our Botanic & common gardens & shrubberies, but which never spread beyond our gardens or spread to perish./ 42 v/Long ago Gouan was in the habit of sowing near Montpellier many foreign seeds likely to grow, several of which succeeded for some years;2 but Mr. Bentham informs me that he searched in vain, & all are now extinct: the ground here is sterile & bare, & we must suppose the native plants in the long run beat the foreigners in the spots where both could grow./42/It is instructive to observe how frequently foreign plants spring up for a year or two in the rubbish thrown from a garden; but how certainly in a/43/few years, more or less, they are overwhelmed by our native weeds. The foreigners languish, perfect few seeds; & of these seeds, few germinate; & the seedlings are generally smothered./43v/ Rothof3 sowed 39 kinds of hardy garden & agricultural seeds on earth thrown out of a ditch in a bog in process of being reclaimed, & only seven came to maturity; eleven seemed capable of ripening their seed; twelve germinated but did not thrive & nine did not

1 . Tracts p. 74. [Carl von Linné, praeses, Isaac J. Biberg, resp., 'The Oeconomy of Nature'.]

2. [Gouan, Antoine, Herborisations des environs de Montpelher...(Montpellier, 1796). pp. ix‐x, 227‐42. cf. Candolle, G é ogr. Bot. pp. 799‐800.]

3. Acts of Swedish Academy Vol. 4.


germinate./43/In our uncultivated banks & woods, far more seedlings of our native plants spring from the ground, than can possibly come to perfection; this may be conspicuously observed with some of our trees. We see the same fact in our crops; for thin‐seeding requires good farming,—that is land with many weeds must be thickly sown, to give the right number a chance of succeeding. In our gardens we can raise common culinary plants with certainty; but sow the same seeds in any number on an adjoining grass field, where there would be nearly the same animal enemies, & you will not raise a plant. Preoccupation of the ground, no doubt, is most influential against chance seeds; but its power has been, I think, sometimes over‐rated: all plants in a state of nature undergo a kind of rotation of crops, exhausting one spot & springing up in another, being supplanted & supplanting others: in a coarse meadow the patches of Dactylis &c. which are not browsed, if marked, will be found to change their place; so that if/44/the seed of a plant fitted to overmaster the others, be annually sown it will at last find a proper site. And the many naturalised plants in every land from the even chance seedlings will not rarely intrude on a preoccupied surface. Seeing on what a nice balance of power a plant can become naturalised, it is no wonder that the most skilful Botanist cannot in the least predict, as was remarked to me by Dr. Hooker, what plant will become naturalised in a given country, though he may safely assert that some will not.

No one will question that there is a limit of heat & cold, dampness & dryness, beyond which a plant cannot survive; but it seems that few plants reach this extreme limit. This may, I think, be inferred from what they can‐endure in our gardens; but more especially as once or twice in a century we have a winter of extreme cold or a very chilly or dry or wet summer; & yet I have not seen any record of a zone of dead plants having been observed towards the confines of their natural range. But what havoc an extraordinary winter will make in our gardens & more especially in our shrubberies! It may be inferred from this, that owing to the struggle between plant & plant, hardly any species reaches/45/very near its extreme climatal limit. In arctic regions & on lofty mountains, where each plant has to struggle against few other living beings, but against severe conditions; zones of dead trees have been observed, as by Ledebour on the Altai, & by Hearne in N. America, who describes a band of dead and blasted stumps upwards of 20 miles in width beyond the living wood.1

1. Ledebour, in Hooker Bot. Miscell. vol. 2, p. 251. Hearne's Journey to the Northern Ocean p. 101.


In the arctic regions & on high mountains very many plants become much stunted; & though I have not met with any precise observations on this head, I think it would certainly have been noticed had this often happened with plants at their lower limits on mountains & at their southern limits in the lowlands: of this latter ease I have noticed only one instance,/46/namely the Sugarmaple which in the southern United States is said1 not to attain above the third of the height which it does in Canada: on mountains, also, I have met with only one instance, namely in the Beech, which is stated [ ] to be stunted below the level of [ ] on the [ ]. Again when the northern range of a plant does not fall near the Arctic regions, it seems seldom to become stunted at its northern limit: as several British plants do not range beyond North‐umberland & Durham, I asked Mr. Story to attend to this point for me, & he has sent me a list of 32 plants in this predicament observed by himself & friends & it appears that only three or four of these are at all dwarfish. Trees,2 however, seem more commonly to suffer I presume, from being more exposed to the winter temperatures: & several of our British trees become dwarf in Scotland; & so it is according to Kalm3 with the Sassafras & Tulip‐tree in the United States./

47/These several facts are explicable if we look at plants as not actually limited by climate, but by struggling with other plants under conditions beginning to be unfavourable; for the struggle would be severer in proportion to the number of enemies or opposed species, & these would be more numerous on the lower than on the higher slopes of a mountain, & in the southern than in the northern half of our colder temperate regions.

No one has written more forcibly on the struggle between plant & plant than the experienced horticulturist, the Dean of Manchester. Mr. Herbert shows4 most clearly that those plants which live in sterile & peculiar soils often do not live there or under this or that degree of moisture, because they prefer it, but because they can thus "get a poor livelihood in peace & quiet" and their "enemies cannot grow to choke them." Speaking of some Crocuses confined to sterile hills in the Ionian islands he says that when secured from their native soil & transported into a garden they acquired ten‐fold vigour. There are many cases on record5 <besides

1 Kalm Travels in N. America Vol. 1, p. 142.

2 Alph. De Candolle Geograph. Bot. p. 72.

3. Travels in N. America Vol. 1. p. 142‐Sir G. Grey in his Expedition Vol. 2, p. 262 says that the Xanthorroea, though not a tree, declines in health & growth in proceeding northward.

4. Journal of Hort. Soc. on the local habitation of Plants vol I. p. 46.

5. Alph. De Candolle Géograph. Bot. p. 428, 453, 455.


the striking/48/ones given by Herbert> of the same species growing in very different situations in different countries,—as Herbert instances the Orchis monorchis & militaris in England on chalk Banks, & in reed‐beds on the edge of Lake of Brienz,/48v/the common milkwort (Polygala vulgaris) in England on dry upland pastures, in Zante on alluvial & very moist meadows./48/—Such cases are probably in main part though not exclusively due to other plants more vigorously occupying the sites in one or both countries which the species in question would most enjoy; for with so flexible a constitution there would be few sites on which such plants could not exist.

To show how one plant can influence others, & like‐wise many animals, I am tempted to give one very common case. In Staffordshire on the estate of a relation, where I had ample means of acquiring all particulars, there was an extensive barren heath, never touched by the hand of man; but on one side several hundred acres had been planted about 25 years before with larch & Scotchfir, nothing whatever having been done, except small holes having been dug, & the whole enclosed. The effect on the native vegetation was quite remarkable in the very great change in the proportional numbers of the plants found on the Heath; & in the presence of 12 species (not counting grasses & carices to which I did not attend) not growing/49/on the Heath; of these twelve, three had never been observed elsewhere in the neighbourhood by a relative who had attended pretty carefully to the botany of the district. The change in the insects must have been even greater; for six insectivorous birds were extremely common in the wood & were not to be seen on the Heath; where two or three other insectivorous species lived, but did not frequent the plantations. I was interested by one particular: young oaks were springing up of all ages by hundreds, in parts at the distance of a mile from any oak‐tree, here & there actually appearing as if they had been sown broad‐cast; but I was assured that this never had been the case; & the woodmen told me that there was not the least doubt how they came there; that they had repeatedly seen rooks dropping acorns in their flight across the woods: there was no rookery near, & the line of flight would take the birds across the heath where there were no oaks, so that this <curious> most efficient means of dispersal must have been wasted for centuries, until the decay of the leaves of the fir‐trees & the growth of other plants had made a bed on which the acorns soon after being dropped could germinate. I have given instances/50/to show what an effect the introduction of a single quadruped can indirectly produce on the vegetation


of a country; & here we see that the introduction of a tree, with no other change whatever, can produce as great an influence on other plants, birds & insects.

Make the ground quite bare, as on a railway cutting, & it may be almost said to be chance by what plants it will be at first covered, being dependent on the nature of the soil, the kinds of plants growing near, the means of diffusion & number of their seeds & the direction of the wind; but in a few years, notwithstanding that the number of the seeds of the first occupants will probably have been increased a million‐fold, the proportions will greatly change, & ultimately become the same as on adjoining old Banks. Many curious accounts have been published of the change of vegetation when a N. American forest has been burnt or cut‐down & then left to nature. This has been called rotation; & it seems pretty clear1 /51/that in our meadows & woods, when not suddenly destroyed that there is a real rotation, like that followed by farmers & probably dependent on the same causes, viz chiefly exhaustion of the various chemical elements in the soil required in different proportions by the different families of plants. The same principle probably comes into play in causing the beautiful diversity of plants in our meadows & woods: the good farmer every fifth or seventh year plants the same crop on the same field; but nature raises her crops altogether in exact proportion to what the soil can support, each kind slowly changing its place, with this great difference that she is not the determined enemy of any bird, insect or slug, & cares not what or how many plants overmaster the others. But when a forest is burnt down, whilst still in full vigour, & a very different vegetation, as is invariably the case, springs up, it seems doubtful whether this should be called rotation in the above sense; the change would rather appear to be due to what seeds are ready in the ground, or quickest brought there; on the rate of growth of the seedlings & their immunity from animal attacks. In these cases, the trees/52/ reassume in the course of ages the same beautiful variety in the same exact proportions as in the surrounding virgin forest: this has been noted in many parts of the world, as over the ancient American ruins in Central & North America.2 On how many & complex contingencies must this wondrous battle prolonged over centuries have been determined by which each species has recovered its rights!

1. Alph. De Candolle. Géograph. Bot. p. 448, 472.

2. An Enquiry into the Origin of the Antiquities of America by J. Delafield, [p. 55 seems most apt.]


It is indeed a wonderful conflict, on which I cannot cease marvelling. Causes appearing to us most trifling are potent. In the Staffordshire Heath formerly alluded to, a small portion had been broken up & attempted to be cultivated, for two or three years; but had utterly failed & was planted with fir trees at the same time with other parts of the heath; & 25 years afterwards, the undergrowth was so different that the lines of separation could be most easily traced. In walking over the most barren heath where four or five plants held absolute sway, I have often been surprised to see a line of turf along small pathways: is this owing to the heath being mechanically destroyed? or do/53/animals follow the paths & occasionally, though rarely drop a little manure? Manure may be directly injurious to the Heaths; but I have noticed in a neglected field of my own, that manuring caused a marked decrease in the hard‐heads (Centaurea nigra); yet this plant certainly likes manure, but the more vigorous growth of other plants must have checked its increase. In this same field I have observed in different summers, an obvious difference in the proportions of the several plants; showing how rapidly a slight change in season allowed one species to increase overanother. So again in old meadow land, which has been ploughed years ago, the same species may be observed in the slightly damper furrows & slightly dryer ridges, but in different proportions: in this (& other such cases) there can be no doubt that the plants growing both in the furrows & on the ridges, could for a time cover the field, if all the other plants were exterminated, but that having to struggle with other plants, the slightest difference in dampness, determined the proportional numbers in either case./

54/The old divine Jeremy Taylor says, "Tell me why this turf this year brings forth a daisy, & the next year a plantane."1 No one can answer. But let it not be called chance. The chemist may throw a dozen salts into solution & may hope to predict the result; the naturalist cannot do this with the living beings dispersed by ten thousand ingenious contrivances all round him; but when we see the virgin forest reassuming its beautiful variety apparently in the same exact proportions, over the ancient Indian ruins, we must see how little of what we call chance has to do with the final result. This struggle, this war of nature, becomes only in the least degree intelligible to us, by keeping steadily in mind that each plant would cover the ground for a period if left to its natural powers of increase; for no one will doubt if four‐fifths of our British plants were suddenly exterminated, the remaining fifth would soon decently clothe the land./54 v/One may wonder why

1 [Taylor, Jeremy: 'Of Modesty', in Holy Living.]


any one or half‐dozen of the most vigorous plants in England, annually producing thousands of seeds, growing in all sorts of ordinary stations, existing here in the middle of their range & therefore well capable of bearing somewhat more heat & cold, damp & dryness, why such plants do not monopolise the whole surface. But assuredly/54/every single plant, even the most vigorous & predominating in its nature, is habitually destroyed in multitudes at some period of its life from the seed upwards, either annually or at recurrent periods, by means, which we very/ 55/seldom can perceive; the only difference between the weak & strong being that proportionally to the number existing at any time the weak one has been destroyed during former generations, or has been prevented increasing, more than the one called strong.

In considering the facts now given, & many similar ones known to any naturalist, one caution is perhaps necessary. Although certainly the most different organisms very often act & react on each other in the most complicated way; yet from such cases exciting our surprise we may perhaps be led to attribute too much to this mutual action from remote parts in the polity of nature. That part of the complex term struggle for existence, which is more correctly expressed by dependency, generally relates to organic beings remote in the scale of nature; & individuals of the same species are hardly ever dependent on each other, excepting in their sexual, parental & social relationship. But we have seen how dependency graduates into a struggle for existence. On the other hand that part of the idea, more correctly expressed by the word struggle, applies in its fullest force between individuals of the same species. When we remember that individuals/55a/of the same species, whether animals or plant, live on nearly the same food & are exposed to the same dangers & difficulties, it is in itself probable that the struggle will be here most severe at some period of life. Probably it will be nearly equally severe between the individuals of two varieties, when they meet, & secondly between closely allied species or between organisms, however different in structure, if they have nearly related habits & encounter each other. /55a v/What can be more remote than a locust & a ruminant quadruped, yet they must often powerfully affect each other. In the cases of rare species, having few individuals thinly scattered, we may infer that the struggle, as far as organic beings are concerned, is chiefly with <other> distinct species <or conditions of existence>./55 a/And lastly the struggle will often be very severe with the external conditions of existence independently of the co‐inhabitants of the district.

We have some evidence how powerfully allied species affect each


other: every one has heard how the Norway Rat has exterminated the Black Rat under the most different climates & circumstances of all kinds from the Polar circle1 to within the Tropics, in the New & Old world: in New Zealand2 the Black Rat had previously almost expelled a previously introduced species: in Färoe3 "the decrease of the mouse has been in proportion to the increase of the Rat," so that the common mouse, which was the earlier inhabitant, has been almost exterminated. /55b/Even with varieties of our domestic animals it has been found by experience4 that other breeds of sheep cannot exist on the mountains of Cumberland with the Herdwick breed, "for they stand starving best." If one species of Swallow were to increase we might expect that other Swallows would suffer more than other Birds; & so it seems to be, for with the late curious increase in parts of the United States of the Hirundo fulva, the Barn swallow has decreased.5 When the red‐legged Partridge increases, the common Partridge decreases; so it has been observed with the Pheasant & black‐grouse. Again Fish with allied habits must chiefly affect fish; & thus the shad (Clupea sapidissima) has increased in the Hudson, in parts full twenty‐fold, owing to the erection of a dam, & the consequent decrease chiefly of another species of Clupea.6 In Russia the small Asiatic Cock‐roach (Blatta asiatica) has everywhere driven before it the great cock‐roach.7 /55c/ The [ ] Leech exterminates the [ ] when placed in the same pond. And to‐go to the other extreme of the scale how fatally does civilized man cause the extermination of savage men. I have said that the struggle is often severe between organic beings & their conditions of existence, independently of the co‐inhabitants: this chiefly holds good on the confines of life, as in the extreme arctic regions or on the borders of a desert like the Sahara. When animals & plants actually perish from cold or drought, there cannot be said to be any struggle between the individuals of the same species; but between the constitution of each & the destroying element. But more generally, the cold or drought for instance, kills by lessening the food, & then there may be most truly said to be a struggle between the individuals of the same species or of species with allied habits. To give one instance to show how during such periods one variety may indirectly

1 Lloyd Field Sports of N. Europe vol. 2 p. 321.

2 [Darwin left a blank space here for a reference he never supplied.

3 Landt Description of Faroe p. 213.

4 The Northmen in Cumberland by R. Ferguson p. 22. 1856.

5. Dr. Brewer in N. American Journal of Science, vol. 38 p. 392.

6. Mr. Adams in N. American Journal of Science Vol 20. p. 150.

7. Pallas Travels in Russia vol. I. p. 16.


master another: in La Plata, during/55d/the great drought, the cattle perish chiefly from famine & the Niata breed would be utterly exterminated, if not protected, for from the peculiar shape of the jaws they cannot feed on twigs of trees so well as the common cattle when all the dried up herbage has been consumed; 1 but if there were no bushes whatever in the country probably the Niata cattle would pass through the ordeal as well as the common breed; both with greatly reduced numbers.

Hence, I think, we may conclude, that as a general rule, the struggle for existence in its strictest yet never simple sense is most severe between the individuals of the same species, & next between the individuals of two distinct varieties, or species, or even classes if their habits are somewhat allied. In all cases, the struggle being ruled & modified by multiform relations./

55e/Facts apparently opposed to there being a severe struggle in all nature:—I will now give the few cases which alone have seemed to me to throw doubt on the struggle for existence. Perhaps the most striking is the existence of species, even locomotive species as mammals, confined, without any physical barrier & with no difference in conditions appreciable by us, to a very small locality, but there very abundant; for it might be argued that if there be such a power of increase, & as the species is abundant in the locality in question, showing that the conditions of its existence are there favourable, why does it not spread.—Many instances in all classes could be given of facts of this nature: Mr. Bentham has often insisted to me, how remarkable it is that certain plants should be found in a single spot, as the Pyrenees & no where else in the world; & should there be abundant; & therefore apparently not like a species on the point of extinction.2 Some local species have been known to exist in the same place/56/for one or two centuries.3 But by far the most remarkable case of this nature on record, is that of certain species & even varieties of land‐shells in Madeira & P. Santo, are positively stated by Mr. Wollaston4 to swarm on

1. [In the MS. the caret for the insertion of this final clause, added between the lines, is placed thus: 'feed ^on twigs.']

2 Bartram in his Travels (p. 466) speaks of "a singular and unaccountable circumstance" namely that he found a Franklinia (Gardenia) alatamaha growing plentyfully over two or three acres in E. Florida, but that he never met with elsewhere. Mr. Wollaston (Variation of Species p. 153) gives plenty of cases of common insects, though extremely local insects, in Madeira.

3 A1. De Candolle. Geograpn. Bot. p. 471.

4 On the Variation of Species p. 132. Helix Wollastoni is one of the most striking cases, & the varieties, as so considered by Mr. Wollaston, of H. polymorpha obey the same law.


certain hillocks on these islands, where they are also found fossil, & that they occur no where else either fossil or recent in the whole group, which has been thoroughly well investigated. The superficial calcareous beds in which these very local land‐shells occur, include a few extinct species, & I am informed by Sir C. Lyell that the island has undergone considerable change since their deposition: hence we must conclude that these land shells, each on its own site, has swarmed probably for several thousand years, & yet have just held their own place & have never spread!/

57/In cases like these latter in which each district has a representative species, filling as far as we can perceive the same place in the economy of nature, the difficulty is, perhaps, not quite so great as‐it at first appears; for let us take one of those common land shells,/57 v/which we positively know, from the extraordinary numbers occasionally appearing in favourable seasons, can rapidly increase, & is therefore habitually kept under by checks of some kind; & let us suppose it to inhabit two points [,] hillocks a few miles/57/apart, I should think that probably the inhabitants of those two hillocks were the lineal descendants of the first colonists, without‐having in many cases been at all intermingled; for although no doubt the checks would fall much heavier at some times on the inhabitants of the one hillock than on the other; yet if they were not wholly exterminated on the one, the rapid power of increase common to these & almost all the lower animals, together with their slow power of travelling, would allow the survivors of the hillock which had suffered most to breed up their numbers before they could be invaded by the inhabitants of [the] other hillock, though they would be to a certain extent by the inhabitants of the intermediate low land; but during another season the lowlands might be invaded by highlanders. The result would be different with slow breeding animals having rapid powers of travelling as with birds, or plants having seeds easily blown by the wind. Thus far I can admit, the weight of /57 bis/slow diffusive progress, to which Mr. Wollaston1 attributes so much importance. The result would, also, be very different if the land‐shell inhabiting one hillock was a variety having the smallest advantage over the individuals in the intermediate tract & on the other hill, for then it would surely spread; but in the Madeira case we may suppose that each species or variety long inhabiting its own hill is at the very least as well adapted to the conditions (I do not mean mere climatal conditions) there occurrent as to the conditions of the other hill.

Those cases in which a plant is absolutely confined to one

1 Variation of Species p. 125, 130, 153.


small area, & is there very abundant, without close representative species in other adjoining districts, seems to offer more difficulty. On a less striking scale, the same difficulty is often encountered, namely in plants being very abundant on one spot, but not found anywhere/58/else in the district or even Kingdom, & yet without any perceptible difference in the conditions: These, however, are exceptional though not very rare cases, the common rule apparently being that very local plants or animals1 are not numerous in individuals. But the fact which has struck me the most, is that given by Alph. De Candolle, that some few "social plants" are social2 to the extreme limits of their range, or are not thinly scattered as might be expected, & when consequently we must suppose that the conditions have begun to be unfavourable. If social plants could help each other like some social animals, from which the term social has been borrowed, there would be no difficulty, for then as far as they could range, they would range in company. But there seems to be no essential difference, only one in degree, between a social plant, & one numerous on any one site. Al. De Candolle has shown in his admirable discussion on this subject,3 that most social plants are thus inhabiting peculiar or unfavourable sites as salt‐marshes, heaths, arctic regions, beneath water &c, & where consequently as only few plants can grow there peculiarly adapted plants grow together in great numbers. Hence, also, in islands, inhabited/59/by only few species, they are very apt to be social; as they are wherever the conditions are very uniform. But the fact which has seemed to me to show that there is no essential difference between very common plants & social plants, is that some naturalised plants are social in their adopted country, —as is eminently the case with the cardoons & thistles on the plains of La Plata, & not, as far as I can make out, in their native home. Nevertheless it seems to me that many plants, both those commonly called social, & those abounding in numbers in some one spot & not elsewhere found in the neighbourhood or even in the whole world, may be said, in a somewhat strained sense, to help each other, so that if they did not live in numbers, they could not live all.—

It follows from the doctrine of the struggle for existence that every plant is checked in its increase in the seed, seedling, or mature state. For simplicity let us suppose in any plant that the main check falls on the seed, owing to its being devoured by some

1 A1. De Candolle. Geograph. Bot. p. 470.

2. Geograph. Bot. p. 462. M. De Candolle instances the Cistus & Lavenders &c on the plains in the south of France: some alpine plants: & forests of trees in the Arctic regions.

3. Geograph. Bot. p. 460.


bird or insect: the argument will be just the same if applied to the seedling & we/59a/suppose a great loss by slugs or other animals. We must bear in mind that in all probability that this will not be the sole check; a certain percentage of seed, for instance, perishing by not getting buried &c. Now from a thousand/60/ plants of the same kind growing together, there will be a far better chance of many seeds being preserved than from a dozen plants,—that is as long as the increase of the bird or insect which preys on the seed is checked by some other agency & is not determined by the seed of the plant in question: if with the increase in seed the numbers of its devourers increased in the same ratio, then it would make no difference in the proportion saved whether there were a thousand or a dozen plants; but if the devouring birds or insects could not thus increase, owing to the want of food in winter, or owing to being preyed on by other animals &c, & this would very often be the case, then there would obviously be more seed saved from the thousand plants than from the dozen.—We see this often practically illustrated; a farmer notices a peculiar ear of wheat, & plants the seed in his garden, but it is notorious that without he carefully protect his dozen wheat plants, he will hardly save a seed owing to sparrows: I have seen this occur & in the same year: I raised some hybrid Radishes & <with all sorts of protection> had the greatest difficulty/61/in saving a few seed out of thousands of pods from the attacks of another bird, the green‐finch.—Yet in a large plot of seed Radishes or in a field of wheat, plenty of seed can be secured. Beyond a doubt, there would be great difficulty in a small colony of radishes or wheat establishing itself in my garden, supposing that they could sow themselves. In animals we have seen the same thing occur in small colonies of foxes & antelopes naturally establishing themselves as described by Sir John Richardson, in N. America, though these instances occurred near the limits of their range.

Another & quite distinct cause may come into play in determining that a social plant could not exist beyond the limit in which the conditions were so highly favourable, that large numbers could grow together: in dioicous plants there must be at least two individuals near each other, & if the fertilisation of the plant be due to the wind, & not to insects, bearing in mind that they will be planted by chance, it seems almost necessary that there should be a good many together in order to be thoroughly fertilised & produce their full complement of seed.1 Now we have seen in

1 I have previously shown in our third chapter that many trees are dioicous & monoicous, & they are apt to be social.


the third chapter that there is good reason to believe that many/62/plants are what Sprengel called dichogamous; & when the fertilisation is not aided by the voluntary flight of insects, these could seed well only when growing in masses: I believe many Grasses are in this predicament, namely depending to a great extent on other individuals for their fertilisation; & are not visited by insects; & grasses are commonly social.

From these two considerations, more especially the first one, (& it is likely there are other considerations overlooked by me) I think we can to a certain extent see why a plant/62 v/may, or rather must, exist socially in numbers together, even near the confines of its range, if it can exist at all: we can, also, see why a plant or animal may exist in/62/large numbers in one spot & not spread; for when once established in numbers it might escape destruction by its enemies, but when thinly scattered in colonies, (owing to the severe struggle going on) all might easily perish. Hence this fact which seems at first paradoxical, & is so if we look chiefly to climatal or soil conditions as of predominating influence, ceases to be paradoxical when we look at all organic beings as periodically struggling for existence with their utmost energy against their enemies. Authors have often spoken of the occupation of the soil, as a powerful/63/element in distribution: in the strict sense of the word, if we remember that plants undergo a natural rotation & that seeds are disseminated in a multitude of ways, I think it can have very little influence: in the sense above given, namely that plants or animals when once established in numbers, by their very numbers escape destruction, I have no doubt this occupation is potent.

Another class of facts seemed at one time to me opposed to there being a severe struggle in nature; namely animals having recovered in a state of nature from severe injuries, as evidenced by the fossil Hyaena1 which had part of its upper jaw entirely worn away; or by the famous Mylodon described by Owen with a fractured skull. Mr. Couch caught a cod‐fish with no eyes, yet in good condition.2 /63 v/Mr. Blyth mentions two nearly blind Indian crows; but these very singularly were fed by other members of the flock.— Rengger describes rickety Jaguars with short legs as not very uncommon in Paraguay./68/Lame birds have been noticed for several years building in the same nest. Birds, more especially rooks, have not very rarely been observed with their upper & lower mandibles crossing & distorted; & this has been observed even in the case of a /64/Woodpecker (Picus erythro-

1 [Buckland] Phil. Transact. 1828. p. 85.

2 Transact. Linn. Soc. Vol. XIV p. 72.


cephalus) which one would have thought would have most severely suffered from such a malconformation.1 All these cases show only that the struggle for existence is periodical & not incessant, of which fact we have plenty of other evidence: in the first very severe winter the rooks with the crossed bills would no doubt be cleared off.—

<In some cases the term struggle is not very appropriate; for instance in the Misseltoe (Viscum); as it can hardly be said to struggle with any other beings, though evidently dependent on them: if it increased in an inordinate degree it would greatly injure the few trees on which it can grow: it would probably be actually exterminated if the Thrush genus which it helps to feed became extinct; & Kölreuter has shown that its. fertilisation is dependent on certain insects:2 probably deficient means of dispersal is a principal check in this case.>/

65/Finally I must allude to an opinion, which I have repeatedly seen advanced, but probably without deliberation;—namely that the numbers of any species depend on the number of its eggs or seed, & consequently not on a struggle for existence at some period of its life or its parents' lives./65 v/This belief has^probably arisen from the larger animals, which can seldom be supported in very great numbers in any country, producing few young; but most of them can protect their young; nor is this relation invariable, as we see in the Crocodile, & amongst Birds in the ostrich./65/The number of the eggs is no doubt one element in the result but by no means one of the most important. How many rare fish there are existing in very scanty numbers,' yet annually producing thousands of ova! Years ago‐I was struck with this in finding a large sea‐slug (Doris) at the Falkland Isld , very rare & yet on calculating the number of the eggs of one individual, I found six hundred thousand. The Condor lays only two eggs & yet in parts it is quite as common, (for I have seen between twenty & thirty take flight from one cliff) as the American Rhea, which lays between twenty & forty eggs & even more: but we need not go so far, the Kitty‐wren, (Sylvia troglodytes) lays on an average just twice as many eggs as the other British wrens or Sylviadae, yet we see no corresponding relation in numbers.3 /65a/The Picked

1 Mr. Blackwall Researches in Zoology has collected several cases, p. 173‐6: see also [Sheppard & Whitear] Transact. Linn. Soc. Vol. 15, Part I, p. 9.

2 [Koelreuter (Erste) Fortsetzung (1768), p. 72.]

3 [On the verso of the manuscript sheet, fol. 65, ending here Darwin wrote: 'Put a remark that fertility is most important in rapidly increasing but not in final results. This is crucial difference. In the ultimate number no doubt other elements are far from unimportant.']


Dog‐fish (Squalus acanthias) actually swarms on many coasts & yet is said to lay only six eggs; whereas the Cod‐fish sometimes lays above three million & a half.1 Again many Diptera increase at such a rate, that Linnaeus has stated that three flies of Musca vomitoria would devour a horse as quickly as a Lion:2 yet there are other flies, which produce only a single egg, or rather pupa, <at a birth & probably> in their whole life, and yet such flies/66/ are by no means rare, as all who have had their horses tormented by the horse‐fly, (Hippobosca) must well know. Amongst plants, I have looked through lists, in which a few of the most abundant plants of a country are marked, & have often noticed amongst them the bearers of the fewest seeds. But the most conclusive evidence of all may be derived from fossil tertiary shells; we have numerous cases of a shell formerly rare & now common in the same region, or the reverse case; & I presume no one will imagine that these shells laid a different number of ova at the two periods. There is an old Eastern fable that the locust lays ninety‐nine eggs, that if it laid the hundreth it would overrun the world; this fable is probably as false as it is old.

Upon the whole none of the facts, which seem at first to deny that all organic beings have at some period or during some generation to struggle for/67/life are of much weight; on the other hand the several remarks & illustrations given in the foregoing pages, imperfect as they are, appear to me conclusively to show that such struggle, often a very complex nature, does truly exist. I have found myself that much reflexion is necessary fully to realise this struggle & dependence of one being on another: our great ignorance of the complete biography of any one single plant or animal makes us slow to believe in the multiform & often extremely obscure checks to their increase. Look at any piece of wild ground, & notice that hundreds, often thousands of seeds annually produced by each plant & disseminated by a hundred ingenious contrivances; —think of the number of eggs produced by each insect, worm & snail,—each animal strives to live, each plant will live if it can,— & yet the average number cannot possibly long increase: go from spot to spot, till you reach the confines of life, & the same story is

1 Yarrell British Fishes vol. 2. p. 401; Fleming's Philosophy of Zoology vol. 2 p. 356.

2 [Darwin probably remembered this from Lyell's Principles of Geology, for in his copy of the 9th ed. (1853) an X is marked in the margin of p. 673, where this statement is attributed to Linnaeus (on the authority of Kirby and Spence —Introduction to Entomology, see 1815 ed. I, 250—who give no source reference) whereas the same assertion in Erasmus Darwin's Temple of Nature (London, 1803), Additional Notes IV, p. 17, is not marked in Darwin's copy. Linnaeus added this statement in the '12th edition' of the Systema Naturae (Stockholm, 1767) to his entry for Musca vomitoria (see Tom. I, pars II p. 990).]


predetermined. Everywhere, the rate of increase,/68/if unchecked, will be geometrical; whilst the means of subsistence on the long average will be constant; & we know in our slow‐breeding larger domestic animals, how large & rapid the result of this ratio has been in an unstocked country. We must regret that sentient beings should be exposed to so severe a struggle, but we should bear in mind that the survivors are the most vigorous & healthy, & can most enjoy life: the struggle seldom recurs with full severity during each generation: in many cases it is the eggs, or very young which perish: with the old there is no fear of the coming famine & no anticipation of death. Philosophical writers, such as Lyell, Hooker,1 Herbert &c. have most ably endeavoured to make others appreciate the struggle & equilibrium of life, as clearly as they‐ do themselves; & I should not have discussed this subject at length, had it not been in many ways of great importance for us; & had I not occasionally met with‐good observers of nature, who by such remarks,‐ as that the number of the individuals of a species was determined by the number of its eggs:—or that when an island partly subsides into the/69/ocean, it will become (as if not already) crowded in‐an extraordinary degree with living beings,—show as it seems to me, an entire ignorance of the real state of nature. Nature may be compared to a surface covered with ten‐thousand sharp wedges, many of the same shape & many of different shapes representing different species, all packed closely together & all driven in by incessant blows: the blows being far severer at one time than at another; sometimes a wedge of one form & sometimes another being struck; the one driven deeply in forcing out others; with the jar & shock often transmitted very far to other wedges in many lines of direction: beneath the surface we may suppose that there lies a hard layer, fluctuating in its level, & which may represent the minimum amount of food required by each living being, & which layer will be impenetrable by the sharpest wedge./

70/Corollary on the relation in Structure of organic beings. It follows almost necessarily from what we have seen of the struggle for existence, dependent on the habits of animals & plants, that the structure of each organic being stands in most intimate relation to that of other organisms. For habit generally goes with structure, not withstanding that in most great families, a few species having the same general structure can be picked out with habits in some degree aberrant. It is very important in order, as I believe, to

1 Hooker & Thomson Flora Indica: see the remarks in Introduction p. 41. [cf. notes 2, 3 p. 175, and 1 p. 188.]


understand many facts in geographical distribution, the steps towards extinction, & the principle of natural selection, fully to appreciate how intimately visible structure, by which we discriminate species from species & genus from genus, is related to the structure of other organic beings. Obviously every living being has s its constitution adapted to the climate of its home; but this seems to produce scarcely any visible difference in structure :/70 v/ thus in every kingdom we have a few species keeping identically the same structure under the most opposite climates—look at Poa from Equator to T. del Fuego, up to limit of snow in Cordillera./ 70/Thus species of such tropical genera as the Elephant & Rhinoceros, inhabited during the glacial epoch very cold countries, with no essential difference in organization; for their woolly covering however important for their habits cannot be /71 /looked at as an important difference in structure. It has often been noticed that many tropical families of plants send out one or two species, having of course the structures of their family, into the cool temperate regions; on the other hand, such northern genera as the Rose & willow have each a species inhabiting the hottest plains of India. I presume that many highly succulent & vascular plants are so far related to a hot climate that they could not exist where severe frost would burst their textures; but it would seem that much caution is required in drawing all such conclusions. For instance seeing the vast number of Heaths at the Cape of Good Hope, & hearing2 that every family of modest size, even leguminous & compositous plants, there have some & often many species with heath‐like foliage, it would appear a safe induction that heath‐like leaves were related to a dry & moderately hot climate; yet our heaths inhabit damp & cold mountains. We find animals & plants/ 72/inhabiting the most abnormal stations, as hot & sulphureous springs & deep caverns into which a ray of light never penetrates, & yet not displaying any great difference in structure from species of the same genera inhabiting ordinary stations.

Whether an animal or plant lives, breathes or moves on land, air or water certainly influences the structure in a most important manner; but even in these cases there is a secondary & perhaps equally important relation to the coinhabitants of the same element. Whether an animal feeds on vegetable or animal food, plainly influences structure, though here the relation is between organic beings, either alive or dead, & often of a special nature. Moreover if we run over in our mind the various structures of the commoner

1 Hooker, Himalayan Journal vol. 2. p. 255.

2. Drege & Meyer. Zwei Pflanzengeograph. Doc. Flora 1848 B. 2 p. 26.


animals, we shall see that the manner of obtaining their prey or food & of escaping danger from other living beings is almost equally influential on their structure.

As the relation of plants in structure to other organic beings is not so obvious as in animals, I will briefly run through the life of a plant in/73/the abstract, & which will serve as a summary for parts of this chapter. Beginning with the flower, which has its dangers from flower‐feeding beetles &c, I cannot doubt from the facts given in our third chapter, that the beauty of the corolla, the scent in night‐bloomers, the positions of the nectary & of the stamens & pistils to each other stand in many cases in direct relation to insects of special genera & classes. When the seed is matured, animals in‐multitudes prey on‐it; & it will escape destruction by its size, hardness, defences, chemical nature or mere number. Its dispersal in some cases depends partly on hooks or on agreeable pulp: even the down of a thistle is perhaps important to it, in as much as the ground is thickly covered by other plants & thickly sown every year: under this same relation to other plants, the period & rapidity of germination will be all important. So again the amount of nourishment surrounding‐ the embryo within the seed, we may believe is given to certain plants that in their earliest days they may succeed in struggling with other plants. The seedling has its special enemies as has the mature plant, which/74/sometimes defends itself from‐animals by prickles, more often by its chemical composition, & which often gains the day over other plants by rapid growth or mere height, at the same time protecting & shadowing other plants, & feeding them with its decayed leaves.

One set of plants will allow another set to live only on some bare chalk banks, though not perfectly suited to them; but the relation of different plants to each other growing on the same plot of ground must be equally important. Cut a piece of turf & look at the inextricable mass of roots, each growing rapidly in the line where it can find food: it is like a battle between voracious animals devouring the same prey. The power of each plant in an entangled mass to get its food apparently will depend on their different periods of activity & on the depth & manner of growth of their roots. Each plant requires certain inorganic bases & a certain amount of moisture; but this in many cases will depend as much on other co‐existing plants as on the nature of the soil; for even with regard to moisture one sees in hot summer how the grass though shaded is often dryed up under a tree. To give one example,—/75/the turnip can beat many weeds from over-


shadowing them> by its rapid growth, & so as farmers say cleans the ground; this rapid growth, I may add, apparently stands in relation to the enormous destruction which this plant suffers during its early state alone from the Haltica & saw‐fly. The turnip is said to contain but a small percentage of the salts of phosphorus, yet farmers find it adviseable to give it phosphate of lime, owing to its rapid growth, rather than to wheat which has ultimately to assimilate a larger percentage of phosphorus, but is a slower grower. So that amongst plants struggling together in a soil very poor in phosphorus, it is quite possible that one requiring much phosphorus might beat another requiring but little of this substance.

From these several considerations I think we may safely conclude that a plant or animal if naturalised in a new country, under exactly the same conditions of climate & soil as in its native country; but associated with a different set of organic beings, would in fact be generally placed under quite as new conditions as if the climate had been somewhat modified. Under an extremely different climate it would not/76/become naturalised. It would probably be quite unimportant to the naturalised organism, whether the greater number of its compatriots were to it new or old forms; those which stood in some relation to it would alone be important, & then in the highest degree; & these influential forms might be as different as possible in the scale of nature, but more commonly those having somewhat similar habits & therefore often systematically related would be the more important. We may put the case in another point of view; let us in imagination wish to alter the structure or constitution of any being so that its numbers might increase: on the confines of its range we should have to change its climatal constitution & in doing this we should not have, judging from analogy, much to alter its structure; even in the midst of its range, as we see the proportional numbers of the inhabitants of a country are changed according as the season is wet or dry &c, we might in some cases increase its numbers by a similar change: always having to do this without deteriorating in the slightest degree its multiform relations to the other in‐habitants of the same place. But these relations are so numerous, so complex & so important that we may believe that it would/77/ probably be easier to make some slight change in structure in respect to the other co‐inhabitants in order to allow its numbers to increase. How totally ignorant we are how this could be effected, we shall immediately perceive, if we ask ourselves what we should alter. In the case of single species of a Family or Order inhabiting a country, or in such cases as the Misseltoe, we can perceive that


the altered structure would have to stand in relation to beings systematically far removed: we may imagine a greater power of penetrating the bark of the apple, or the berries being rendered more attractive to birds might aid the misseltoe to increase in numbers but it is all the wildest conjecture. Very commonly the altered structure would have to be in relation to nearly allied forms & here the difficulty of imagining a favourable change of structure is even greater. Would mere increase in size & strength prevent the black rat from yielding in so many quarters of the world to the Norway rat: this is quite doubtful, at least the great size of the occidental Blatta has not saved it from its puny Oriental congeners. What change could we make in the Barnswallow of the United States to allow it to‐withstand the inroads of the allied Hirundo fulva? And so we may continue to puzzle ourselves in infinitely numerous cases./

78/I have discussed this subject at some length, for it seems to me most important under many points of view, that we should fully realise our ignorance, & never forget, that though the constitution of each being is necessarily related to the climate of its country, yet that not only in animals but in plants, much, probably far the greater part of the structural differences between species & species stands in the most direct yet generally unperceived relation to the other organic beings of the same country.

[page] 213




Darwin's Pocket Diary records two periods of work on his chapter on natural selection, namely March 1857 and the spring of 1858. The first draft, completed on March 31, 1857, was written on sheets of gray wove foolscap which are distinguishable from the bluish gray paper used for the later interpolations and revisions. The outline of this original form of the chapter appears in the original table of contents made before the later revisions:

19 Comparison of Mans & Natures Selection.

26 Extinction & Divergence plays part.

28 Crowd of Difficulties.

33 Causes favourable & unfay. to selection.

35 Isolation (38) in regard to Intercrossing

41 Illustration of isolation & intercrossing Madeira & Galapagos

43 Varieties keeping separate.—46 in Plants

49 Intermediate varieties rare    50 why not met with over Continents

54 Large number of individuals favourable for Selection

56 Summary on causes fay. & unfay.—57 Malay

59 Slowness of Selection

63 [pencil addition] Theory applied to Races of Man.

Aside from one comment on a separate slip of paper dated June, 1858 (fol. 53 A, printed here as note 1, p. 263), the dating of the additions and revisions of the manuscript is uncertain. The valuable clarification on folio 21, By Nature, I mean the laws ordained by God to govern the Universe', Darwin might have added above the line at almost any time, after he had completed the lines where it is inserted. Still probably most if not all the additions and revisions were made sometime in the period between April 14 and June 12, 1858 which according to his Pocket Diary, Darwin devoted to 'Divergence and correcting ch. 6.' Darwin had mentioned divergence in the original draft of the chapter but only briefly, whereas in the revised version he devoted over forty new pages to this topic. The following table gives the numbers of all the pages or scraps whose pale blue gray colour indicates they were written later than the original draft:

12A 37 v 49
13 v 38 51 (bottom half starting 'slowness of selection')
16v 38 v
17v a &B 39A to 62
26* 40 64 to 76.
26a to 26 nn 40A

The revision of the later part of the chapter added to its length so that the original folio 58 later became folio 77, and (after cancellation of a final sentence running on to the original folio 59 and the addition of a new con‐cluding sentence) now ends the chapter.



[completed March 31, 1857]

1/How will the struggle for existence, which we have discussed in the last chapter, act? Annually during thousands on thousands of generations, multitudes have been born more than can survive to maturity. The least possible weight will turn the balance which shall live & which die. Look at the young in the same litter or nest, something must determine which shall live & procreate its kind. If two beings were absolutely alike in all respects, during the whole course of their lives, it might be truly said to be chance, which of the two should come to maturity & procreate their kind. But such absolute identity can hardly be predicated of any living beings; & certainly, as has been seen in the fourth chapter, there is a considerable amount of variability in nature. A large proportion of the variation, which does occur, may be quite unimportant for the welfare of any particular organism, & such variation would not in the least be affected by the struggle for existence. On the other hand, any variation, however infinitely slight, if it did promote during any part/2/of life even in the slightest degree, the welfare of the being, such variation would tend to be preserved or selected. I do not say that it would be invariably selected, but that an individual so characterised would have a better chance of surviving.

If we reflect on the infinitely numerous & odd variations in all parts of the structure of those few animals & plants, on which man may be said to have experimentised by domestication, & again on the many, though slight variations which have been noticed in a state of nature, it would be most strange if in the course of thousands of generations, not one variation added to the welfare of some varying organic being; in thinking of this we should bear in mind how multifarious, singular, & complex the relations for each living being are in habits & structure to other organic beings & to climate, both for securing food & escaping many dangers, during the various stages of life. Again we should bear in mind that whole treatises have been written, showing what numerous, what trifling, what strange peculiarit[i]es are inherited, or tend/3/ to be inherited, that is appear in some of the offspring or reappear in their descendents. An individual, therefore, which from having some slight profitable variation, was preserved or naturally selected, would in many cases, tend to transmit the new, though slight modification to its offspring. Moreover the causes, which


from their extremely complex nature we are forced generally to call mere chance, which produced the first variation in question would under the same conditions often continue to act; & assuredly these causes would be eminently likely to act on individuals having some inherited tendency <however slight>, in this same direction: so that the cause of the variations & inheritances would act & react on each other, thus giving fresh & fresh opportunity for natural selection to seize on & preserve whatever modification of structure habit, or constitution, was in any degree useful./

3 v/On the other hand, any modification if in the slightest degree injurious would be rigidly destroyed. In the struggle for existence, during the long course of generations, individuals thus characterised, would have a very poor chance of surviving. Even if the injurious modification from the nature of the conditions, or from a strong principle of inheritance, appeared again & again, it would be rigidly rejected again & again./

3/I can hardly imagine any change in structure habits &c so slight that it might not be useful to an individual/4/of a species, & hence be selected. It seems at first to be simple chance which individual insect shall fall a prey to a bird; yet birds are guided by their eye‐sight; & we so often see leaf‐eating insects are green or those living on bark, mottled‐brown, we may believe that a slight change in the shade of colour, might in the long run cause such individuals better to escape destruction & leave offspring with the same inherited tint. Colour is thought an unimportant character by naturalists; but when we see as it has been fancifully1 said that "the ptarmigan is lichen in summer & snow in winter, that the red‐grouse is heather, & black‐grouse peaty earth"; & when we remember the main check to the increase of our game birds, is owing to birds & beasts of prey, I can see no reason to doubt that in birds varying in colour as does the red‐grouse, that the finest tints of colour might be selected owing to such individuals suffering less. Such selection would perhaps/5/the more readily be effected with birds & insects when they invaded a new district, or slightly changed their habits, which certainly occurs, as we see with insects attacking our exotic plants. I observe in many German & French pigeon‐books, that people are cautioned not to keep

1 See, also, some good remarks on the colour of those birds giving them a better chance of escape from birds of prey by Mr. C. St. John in his Tour in Sutherland‐shire, Vol. 2, p. 179. [As Prof. John L. Brooks has pointed out to me, Darwin's next quotation about the ptarmigan derives from Blyth's article, 'An attempt to classify the "Varieties" of Animals...' 'Mag. Nat. Hist. 8 (1835), 51. Note that Darwin's statement that 'the ptarmigan is lichen in summer' derives from Blyth's phrase, 'lichen rock', rather than from the 'mossy rock' of the original statement by Robert Mudie.]


white pigeons, as they suffer much the most from hawks. Nor let it be said that the occasional destruction of individuals of a particular colour could have no influence on the colour of the whole body; for it is well known how effective is the destruction of any lamb with a tinge of black in keeping the flock pure.

Again take a beast of prey, pressed for food owing to the destruction by a dearth of the animals on which it feeds; what a trifle will determine which shall survive; the least superiority in power of scent, a shade of colour so as to be less conspicuous, (I have noticed that a prowling white‐piebald cat is far easier seen by birds than a tabby), the power of springing an inch further may well determine its success, /6/when life depends on success: in such cases one meal lost may be the turning point; here it may be truly said that the last straw breaks the camel's back./6 v/And success will depend not only on the vigour of the moment, but often on the condition in which the animal has been able to keep itself during several previous months./6/Or again look at the surprisingly large annual destruction of shrews by cats either by mistake or for sport, as shown by the number found killed but not devoured on our gravel‐walks: supposing for the moment that this destruction is a main check to the increase of shrews, may we not believe that an individual born by chance with an inheritable stronger odour, & so a little more repugnant to the prowling beast of prey would have a better chance of escaping; & from this individual others still more offensive might be selected, till a shrew was formed with an odour as insufferable to man & beast, as that of some foreign allied animals.

A sudden or great variation most rarely, some will /7/say never, occurs in nature; but if it did, & were profitable it of course would be selected; but small modifications, let them appear ever so trifling, if in the least influential on the welfare of the being, I can see no reason after the most careful consideration to doubt would tend to be preserved or selected. They would, also, tend to be inherited; & slight modification might thus be added to slight modification in any given direction useful to the animal;—just as in our domestic animals & plants modifications useful to man have been added together & rendered permanent by artificial selection./

7a/Natural selection may act at any time of life; for variations appearing at one period tend, as we have seen, to reappear at the corresponding period: thus peculiarities in the caterpillar or coccoon of the silk‐moth are inherited; & any modification in a caterpillar or coccoon useful to it, might be naturally selected & made permanent /7a v/and so it might be through however many


stages of existence, or alternations of generations to use Steenstrup's expression,1 any animal may pass. Thus also, /7a/ the embryo might be modified by selection in relation to the mother's womb: in Yorkshire according to that excellent writer Marshall2 big‐buttocked calves were selected, until they were found to destroy many cows during calving, & thus a deviation of this kind, if left to nature, would be soon eliminated: on the other hand if this deviation were useful in any way to the embryo, or to the calf after birth, no doubt in the course of time the parental structure might become modified by selection to allow of such births; for facility in parturition is undoubtedly hereditary./

7a v/In the Tumbler pigeons, the beak has been rendered so short by long‐continued selection, that Mr. Eaton3 says he is "convinced that better head & beak birds have perished in the shell than ever were hatched, the reason is that this amazingly short‐faced bird cannot reach the shell with its beak, & perishes in the shell if the Fancier does not extricate it." But by longcontinued selection a shell thinner at the right end might be naturally obtained, for we know that the eggs [of the] common Hen often vary in thickness./

7a/So again any modification in either sex separately, whether useful to that sex alone, or in functional relation to the other sex, or to the flock or to the young, might be selected & become attached to that sex alone; for/[Folios 7 b to 7k are missing]

8/creation of each living thing endowed with a small limit of variability, or with the theory of a great amount of slow modification; & it will be the object of this work in the latter chapters to make this comparison. But for the present, in order to explain my principles, I must assume that there is no limit or no close limit to variation during the long course of ages.

From what we have seen in the first chapter the main cause of variability seems to lie in a change of the conditions of existence, perhaps aided by abundant food. That many countries have suffered great changes during the same geological period, that is within the period of existence of the majority of the same species, no geologist will dispute. Reflect for a moment on the vast changes of climate & of the level of the sea, during the glacial epoch,—a mere sub‐division of one geological period. Now let us/9/take the case of a country subjected to some climatal or other change;

1 See the wonderful facts given in Steenstrup's most interesting work, translated and published by the Ray Society.

2 [Rural Economy of Yorkshire, II, 183.]

3 A Treatise on the Almond Tumbler, p. 33.


the proportional numbers of its inhabitants will be altered & organic beings better adapted to the new climate will flow in from the surrounding countries, as they certainly did into Europe during the glacial epoch. But if the country were cut off by some impassable barrier, from the adjoining warmer or colder or dryer countries, as the case might be, or if one supposed country was an island, then new beings could not immigrate, & fewer of the old inhabitants would be exterminated for there would not be new beings to take their place; the majority would suffer & then would decrease in numbers; but some few, which were previously just able to reach so far south (supposing for the moment that the change was from warmer to colder) under the new conditions would be favoured & would increase in numbers. Bearing in mind how intimately each organism is related to other organisms, & even to the proportional individual numbers of each, for one organic being in large numbers may well be far more influential for good or evil to another, than if in small numbers, there can, I think, be no doubt/10/that in our imaginary country the selections of nearly every inhabitant would be seriously disturbed both by the change of climate, & more especially by the changed proportions of the other inhabitants & by destruction of some few./10 v/Each being would be placed under conditions, such as the world had never exactly seen before./ 10/Moreover the changed conditions <of existence> would tend to make some of the organic beings more variable than heretofore. Under such circumstances, it seems to me that it would be quite extraordinary, if in some few at least of the slightly varying organisms, no profitable variations better fitted for the new & complex combination of conditions occurred./10 v/A very slight modification would often suffice to give some advantage between the struggling inhabitants; for we have before seen, that the severest struggle, leading even to the extermination of one, often lies between closely allied & therefore very similar species of the same genus.—If any such profitable modification did/10/occur, I cannot doubt but that it would be slowly though steadily selected; & the variety thus selected would gain strength & increase in numbers. Under the above circumstances, which though imaginary must repeatedly have occurred in the world's history, the conditions would probably be most favourable for some rapid selection & consequent modification of forms; nevertheless I think we may conclude that there does not exist a land in which the process may not be going on slowly. Everywhere organic beings present individual differences,/11/& some few more marked variations. No country can be named in which all the inhabitants are perfectly


adapted to its conditions of existence: this may seem a rash assertion, but I think it can be fully justified. Each being in its native country no doubt is adapted to its conditions of existence as perfectly as the other coinhabitants, in proportion to the average number of the individuals of its kind; but not one country, still less not one island can be named which does not possess many organic beings naturalised thoroughly well as far as we can judge.1 /11 v/M. Alph. De Candolle has insisted strongly on this fact of the universality of naturalised plants & has drawn the foregoing inference from it. The number of naturalised plants in Europe & N. America is probably in great part due to great changes effected by agriculture; but I think Sir C. Lyell has shown that the action of man on other organic beings though more potent, does not differ essentially from that of any other animal when introduced naturally into a new country.2 In the case of many plants naturalised in the uncultivated parts of many islands, man has probably in no ways influenced the conditions. No one will assert that the existence of the cat, rat &c in New Zealand, of introduced monkeys in Cape de Verde Isld —of horses & cattle in La Plata &c &c is due to changes effected in the natural state of their countries through man's intervention./11/Now does this not show, that in the natural polity of each land there were places open, which could be filled by other beings more perfect, not by any ideal standard, but by actual proof, in relation to the previous inhabitants & to the climatal conditions of that land? Nor let it be said that individual differences are so slight that the most careful selection could make no sensible change by adding them up during a long course of ages; for man, even during mere scores of years, has certainly thus acted on differences so slight/12/as to be inappreciable except by an eye long educated. Therefore I conclude that there is no land, so well stocked with organic beings, or with conditions so unvarying, but that in the course of time, natural selection might modify some few of the inhabitants & adapt them better to their place in the great scheme of nature. I may here add that hereafter we shall show good reason for believing that it is not the oppressed & decreasing forms which will tend to be modified, but the triumphant, which are <increasing in numbers, extending their range, & coming into new relations,> already very numerous in individuals, widely diffused in their own country & inhabiting many countries, which are most variable & so will be most apt to be modified & so become under new forms still more triumphant./

1 [Pencilled addition:] so has Bunbury in Linn. Transact. [21 (1854) 188‐9.]

2 [Principles, 9th ed. (1853) p. 664.]


12 A/Illustrations of the Action of Natural Selection. In order to make it clear how I believe natural selection acts, I must beg permission to give one or two imaginary illustrations./12/Let us take the case of a wolf, which preys on various animals, securing some by craft, some by strength & some by fleetness; & let us suppose that the fleetest prey, a deer for instance, had from any change whatever increased in numbers, or other prey had decreased in numbers during that season of the year, when the wolf is hardest pressed for food; I can under such circumstances see no/<12>13/reason to doubt that the swiftest & slimmest wolves would in the long run be preserved & selected; always provided that they retained strength to master their prey at this period or some other period of the year when compelled to prey on other animals./13 v/I can see no more reason to doubt this, than that the Breeder can greatly improve the fleetness of his greyhounds by long‐continued & careful selection. /13/The same process would tend to modify the deer in order to escape the wolf slowly rendered fleeter; though it might happen that some other & incompatable modification might be more important to this animal, as getting food during some other season. Even without any change in the proportional numbers of the animals on which the wolf preyed, a single cub might be born with an innate tendency either of instinct or structure leading it to pursue certain prey; nor can this be thought very improbable seeing that of our cats, one naturally takes to catch rats & another mice, & according to the excellent observer Mr. St. John one to bring home winged game, another hares & rabbits, & another to hunt on marshy ground & almost nightly to catch woodcocks & snipes, how if any innate slight change of habit or structure benefitted our wolf, it would be more likely to survive & procreate many young, than the other wolves; & some of its young would/<13>14/probably inherit the same tendency, & thus a new variety might be formed, which would either supplant or coexist with the parent form. Or again with our wolves, those inhabiting a mountainous district might readily be led chiefly to hunt different prey from those on the lowlands; & from the continued selection of the best fitted individuals in the two sites two varieties might slowly be formed, which would, cross & blend where they met, but to this subject of intercrossing we shall soon have to return; I may add that according to Mr. Pierce there are two varieties of the wolf in the Catskill Mountains in the <United States>,2 one with a light grey‐

1 Wild Sports & Nat. History of the Highlands. 1846, p. 40.

2 ['A Memoir on the Catskill Mountains...', Amer. J. Sci., 6 (1823), see p. 93.]


hound like form which pursues deer, & the other more bulky with shorter legs & which more frequently attacks the shepherd's flocks.

If the individual numbers of a plant depended chiefly on the wide dispersion of its seed, so that some might fall on a proper site, any plant which had its seed furnished with pappus a little better adapted to be wafted; or with pulp more agreeable to Birds, would have a better chance of being dropped where it could germinate & reproduce its kind; & I can see no reason why nature should not thus select the most dispersable seed than that gardeners should be able to go on selecting varieties having more & more differences in seed, pod, or fruit./

15/Let us now take a more complex case: some plants excrete a sweet juice apparently for the elimination alone of something injurious from their sap, as in the case of the glands at the base of the stipules of some Leguminosae; & this juice is greedily sought by insects. Let us suppose the juice to be excreted at the inner bases of the petals, & insects in seeking the juice would be apt to get dusted with pollen, & carry it on to the stigmas of other flowers of the same kind, & so cross them: this, as we have every reason to believe, would make more vigorous seedlings which would have the best chance of surviving; & some of these seedlings would probably inherit the nectar‐excreting power; & those individual flowers which excreted most nectar would be most visited by insects, & oftener crossed, & so in the long run would gain the upper hand. In order to increase the amount of nectar, the nectaries & with them the petals might become modified, as well as the position of stamens & pistils in relation to the particular insect which visited the flower; some insects like ants being of not the slightest service to the plant; others as Bees being very useful in fascilitating intercrosses. We might have taken for our example, insects devouring pollen instead of nectar, & as pollen is formed for a definite object its destruction appears at first a simple loss to the plant; yet if a little was occasionally or habitually carried to another plant, owing to the visits of the pollen‐devouring/16/insects, & a cross thus effected, although nine‐tenths of the pollen were destroyed, it might still be a great gain to the plants; & those individuals which produced more & more pollen & had larger & larger anthers would be selected. Indeed this process of selection of larger & larger anthers might be carried on, merely that some of the pollen might escape destruction, without any indirect advantage being gained by the pollen being robbed, in the same manner as many plants probably produce thousands of seeds, in order that a few may escape destruction.


When our plant had by natural selection been rendered so attractive to insects, that unintentionally on their part they regularly carried pollen from flower to flower; & how effectually they do this, <the result of Kölreuter's artificial fertilisation of flowers, the same number being left to insects, clearly shows> I could easily show by many striking facts; then another process might commence. No naturalist doubts the advantage of what has been called the "physiological division of labour"; hence we may believe that it would be an advantage to a plant to produce only male organs in one flower or one whole plant, & only female organs in another. If then an individual plant tended to fail, in either sex in the different flowers of the same individual, or on all the flowers on different individuals; nor does this seem/17/very improbable,/17 v/as it can be shown that the two sexes in the same flower are sometimes rendered sterile in different degrees, when the plant is exposed to changed conditions of life, & as we see in/17/nature how many gradations there are between dioicous, monoicous & polygamous plants; then if this incipient division of labour profited the plant in the least degree, it might be increased by natural selection, until one plant had separated sexes.

Lastly let us turn to nectar‐feeding insects in our imaginary case: let us suppose that the plants of which we have been slowly increasing the nectar by continued selection was a common plant, & that certain insects depended in main part on its nectar for food. Now/17 v/I could give many facts, showing how eager Bees are to save time, & to visit flowers as rapidly as possible—for instance their habit of cutting holes at the bases of flowers, which they can enter with a little trouble—bearing this in mind,/17/I can see no reason to doubt that an accidental deviation in the size or form of the body, far too slight to be appreciated, or in the curvature or length of the proboscis &c might profit a moth, fly or Bee, so that an individual so characterised would more rapidly obtain food & so/18/have a better chance of living & leaving descendents with a tendency to a similar slight deviation of structure./18 v/For instance the tube of the corolla of the common red & tall incarnate clovers do not on a hasty glance appear very different in length; the Hive‐bees can easily suck the nectar out of the latter, but not out of the common red clover; so that whole fields of the plant offer precious nectar on which the welfare of the community depends, in vain to our Hive‐bees. On the other hand I have elsewhere experimentally shown that the fertility of clover depends in the closest manner on the visits of Bees, which by moving parts of the corolla push the pollen on to the stigmatic surfaces./18/Thus


I can understand how a flower & Bee might slowly become either contemporaneously or one after the other modified & adapted in the most perfect manner to each other.

I am well aware that the doctrine of natural selection exemplified in the above imaginary examples, is open to the same objections, which were at first launched out against Sir Charles Lyell's noble views on "the modern changes of the Earth, as illustrations of geology", but we now very seldom hear the action of the coast‐waves, for instance, called a trifling & insignificant cause, as applied to the excavation of gigantic valleys or to the formation of the longest lines of inland cliffs. In our imaginary examples, it may be observed that natural selection can act only by the preservation & addition of infinitesimally small inherited modifycations each profitable to the preserved being; but as modern geology has almost banished such views as the excavation of a great vally by a single diluvial wave, or cataclysms desolating the world, so will Natural Selection, if it be a true principle, banish the beliefs of/19/the continued creation of new organic forms, & of any subsequent, great & sudden modifications in their structures.

We must add to the effects of natural selection, the direct action, probably very small & almost certainly slow, of climatal conditions; —we must not forget, & I believe this to be of very wide application, that in the modification of one part, either during the same or during an earlier period of life, other parts will be altered according to the complex & unknown laws of the correlation of structures, for instance a selected modification of the larva would almost certainly influence mature forms;—we must allow something in the higher animals for the effect of habit & disuse, of which again the action must be always slow;—but over all their causes of change, I <fully believe> am convinced that Natural Selection is paramount.

Comparison of nature's selection with man's selection. From the facts given in the two first chapters, it cannot be doubted that man can do, & has done, much in the modification of animals & plants by the artificial selection of variations. But he labours under great disadvantages: he selects only by the eye & acts therefore on external characters alone: he cannot perceive slight constitutional differences; nor the course of every/20/nerve & vessel: he can by no means tell whether all parts & organs are correlated perfectly, but only so far that life & tolerable health are preserved. Far from allowing each being to struggle for life; he protects each to the utmost of his power, both during youth & times of dearth & from all enemies. Instead of selecting steadily


from generation to generation, he only occasionally selects; & his judgement is often bad or capricious: he & his successors never go on selecting for the same precise object for thousands of generations. Even when most carefully selecting he sometimes grudges to destroy an animal, imperfect in some respect, as it comes up to his standard in some other respect. Each being is not allowed to live its full term of life & procreate its kind, according to its own capacity to exist. He does not always allow the most vigorous males to be the fathers of their breed. He often begins his selection with some striking abnormal form, differing widely from anything observed in nature, & of no use to the being selected. From migrations, changes of agriculture &c, he often unintentionally changes the conditions to which his products are/21/exposed; or intentionally crosses them with individuals brought from another district or country, as was done in the darkest ages. He selects any peculiarity or quality which pleases or is useful to him, regardless whether it profits the being & whether it is the best possible adaptation to the conditions to which the being is exposed: nor does he regularly exercise the selected peculiarity: he selects a long‐backed dog, or long‐beaked birds & trains it to no particular course of life;—he selects a small dog or bird & feeds it highly;—a long limbed animal & exercises its fleetness only occasionally or not at all like the Italian greyhound. And lastly, to repeat, he can judge by external characters alone, & not from the perfect action & correlation of the whole organisation during the whole course of life.—

See how differently Nature acts! By nature, I mean the laws ordained by God to govern the Universe.' She cares not for mere external appearance; she may be said to scrutinise with a severe eye, every nerve, vessel & muscle; every habit, instinct, shade of constitution,—the whole machinery of the organisation. There will be here no/22/caprice, no favouring: the good will be preserved & the bad rigidly destroyed, for good & bad are all exposed during some period of growth or during some generation, to a severe struggle for life. Each being will live its full term & procreate its kind, according to its capacity to obtain food & escape danger. Nature will never select any modification without it gives some advantage to the selected being over its progenitors under the conditions to which it is exposed. Every selected change will be fully & regularly exercised. Nature will not commence with some half‐monstrous & useless form; but she will act by adding up deviations so slight as to be hardly or not at all appreciable by

1 [This sentence was added above the original line of text.]


the human eye. Natural conditions remain constant for enormous periods, or generally change very slowly, so will the consequent variability be slight, & the selection very slow. Nature is prodigal of time, & can act on thousands of thousands generations: she is prodigal of the forms of life,/23/if the right variation does not occur under changing conditions so as to be selected & profit any one being, that form will be utterly exterminated as myriads have been. No complications are too great for nature: a contingency happening once in a thousand generations may lead to the extermination of a variety: she can gradually select, either simultaneously or successively, slight changes adapting the selected variety to a score of other beings, most widely apart in the great scale of nature.

Can we wonder then, that nature's productions bear the stamp of a far higher perfection than man's product by artificial selection. With nature the most gradual, steady, unerring, deep‐sighted selection,—perfect adaption to the conditions of existence,—the direct action of such conditions—the long‐continued effects of habit & perfect training, all concur during thousands of generations. Here we meet with no hereditary useless monsters. All who have reared animals & plants believe that trueness is dependent on long‐continued & careful selection, & on exposure to the same conditions. How incomparably truer, then, must nature's varieties, called by us species/24/when strongly marked, be, when compared with the varieties reared by man. Now trueness or the absence of variability, is the most important characteristic mark of a species in contrast with a variety, second only to the sterility of hybrids, & not second to this in the eyes of some, as Gaertner & Herbert whose studies would naturally have led them to attribute the greatest importance to the laws of breeding. If we admit, as we must admit, that some few organic beings were originally created, which were endowed with a high power of generation, & with the capacity for some slight inheritable variability, then I can see no limit to the wondrous & harmonious results which in the course of time can be perfected through natural selection.

It may, perhaps, be here worth notice, that amongst barbarous nations, there will be little intentional/25/selection, & the animals in great degree will be left to struggle for life without aid under conditions nearly constant; & it has been remarked that in such cases the breeds approach much more closely in character to true species, than amonst civilised nations.

Seeing what man has done in a few thousand years, I have sometimes wondered that nature considering the perfection of her


means has not worked quicker, than geology teaches us to believe she has in the modification of organic beings. But from what has gone before, & from what will presently follow, we may see that there are most powerful retarding agencies always at work.

The forms produced by natural selection, if quite modified, will be called species, if only slightly different, will be called varieties; if no further variation occurs in the right direction by which the variety may be further profited, I can see no reason why a variety may not remain in that state during an enormous lapse of years/26/; & we have seen in the fourth chapter, that some varieties such as the land‐shells in the calcareous superficial beds of P. Santo certainly are of high antiquity.

But that a variety should remain constant during whole geological periods is excessively improbable; for we have seen in our 5th Chapter in how important a manner the structural differences of each organism is most intimately related to those of the other coinhabitants of the same district; & as all these are struggling for supremacy, & will hence constantly tend to be modified & become improved, if one variety be so fixed as not to vary at all in a fitting direction, & so become through natural selection adapted to those other changing organic forms to which it is related in the polity of nature, it will be exterminated./1

26*/Extinction.—The general subject of extinction will be discussed in a future chapter on palaeontology. But extinction must be here noticed, as bearing in a very important manner on the theory of Selection. As man in any country improves his breeds, he neglects the less improved & these gradually disappear. Hear Youatt2 on the cattle of northern Yorkshire: at the commencement of the 18th century the ancient black cattle were the only breed. To them succeeded the long‐horns, which by degrees spread over the whole northern & midland counties; but much valued as they were, they were after a time "swept away, as if some by some strange convulsion of nature". For they had to give way to the short‐horns, & these for the last century have maintained their ground; & no doubt will do so, until some better breed be formed, if better can be. So it has been with innumerable varieties of our cultivated plants; "old sorts being fairly beaten out by new & better ones."3 Thus it has been, & thus it will be, with man's

1 [See appendix for short cancelled passage. Fol. 27 is gone, replaced by fols. 26 and 26 a to 26 nn.]

2 Cattle. Library of Useful Knowledge, 1834, p. 248, 199.

3 [Anon.] Gardener's Chronicle, 1857., p. 235.


productions. In nature, the same species existing in two now separated areas, might become modified in one or both, & the resultant forms might continue, whilst/26 a/separated, to exist for any length of time. Such forms are often called by naturalists representative or geographical species, races or varieties: they are maiden knights who have not fought with each other the great battle for life or death. But, whenever from the union of the two areas, they meet, & come into competition, if one has the slightest advantage over the other, that other will decrease in numbers or be quite swept away. But as we see in a vast number, perhaps in a large majority of cases, that the varieties of the same species, & the species of the same genus, inhabit the same country, or divisions of it not separated by impenetrable barriers, generally the varieties as well as the species will have come into competition with each other & with their parents from an early period or even from the very commencement of their formation; and as a form can be selected by nature solely from having some advantage, at least in the spot where the selection is going on, over its parent form; the parent will be almost infallibly there exterminated by its own offspring.

Hence, we may, I think, safely conclude, that/26 b/natural selection (like man's selection) almost necessarily entails a nearly proportional amount of extinction;—one species whilst forming beating out another, & one even the finest variety, if having any kind of advantage over another, taking the place of & exterminating the less favoured & less modified variety. It is in each country, a race for life & death; & to win implies that others lose.

Principle of Divergence.—This principle, which for want of a better name, I have called that of Divergence, has, I believe played a most important part in Natural Selection. To seek light, as in all other cases, by looking to our domestic productions, we may see in those which have varied most from long domestication or cultivation, something closely analogous to our principle. Each new peculiarity either strikes man's eye as curious or may be useful to him; & he goes on slowly & often unconsciously selecting the most extreme forms. He has made the race‐horse as fleet & slim as possible & goes on trying to make it fleeter; the cart‐horse he makes as powerful as he can: he selects his Dorking‐fowls for/ 26c/weight & disregards plumage; the Bantam he tries to get as small as possible, with elegant plumage & erect carriage: a pigeon has been born with slightly smaller beak, another with slightly longer beak & wattle, another with a crop a little more inflated


than usual, another with a somewhat larger & expanded tail &c; his eye is struck & he goes on selecting each of these peculiarities, & he makes his several breeds of improved tumblers, carriers, pouters, fantails &c, all as different or divergent as possible from their original parent‐stock the rock‐pigeon; the intermediate, & in his eyes inferior birds, having been neglected in each generation & now become extinct. It is the same with his dress, each new fashion ever fluctuating is carried to an extreme & displaces the last; but living productions will not so readily bend to his in‐ordinate caprice./26c v/<Moreover, far more fancy‐pigeons will be kept, (I do not mean those kept as food) after they have become broken up into very distinct breeds, than when fewer & more simi‐lar birds existed; for each fancier likes to keep several kinds, or one fancier keeps one kind & another becomes famous for another breed.>

26c/Now in nature, I cannot doubt, that an analogous principle, not liable to caprice, is steadily at work, through a widely different agency; & that varieties of the same species, & species of the same genus, family or order are all, more or less, subjected to this influence. For in any country, a far greater number of individuals descended from the same parents can be supported, when greatly modified/ 26 d/in different ways, in habits constitution & structure, so as to fill as many places, as possible, in the polity of nature, than when not at all or only slightly modified.

We may go further than this, &, independently of the case of forms supposed to have descended from common parents, assert that a greater absolute amount of life can be supported in any country or on the globe; when life is developed under many & widely different forms, than when under a few & allied forms;—the fairest measure of the amount of life, being probably the amount of chemical composition & decomposition within a given period. Imagine the case of an island, peopled with only three or four plants of the same order all well adapted to their conditions of life, & by three or four insects of the same order; the surface of the island would no doubt be pretty well clothed with plants & there would be many individuals of these species & of the few well adapted insects; but assuredly there would be seasons of the year, peculiar & intermediate stations & depths of the soil, decaying organic matter &c, which would not be well searched for food, & the amount of life would be consequently less, than if our island/ 26e/had been stocked with hundreds of forms, belonging to the most diversified orders.

Practice shows the same result; farmers all over the world find that they can raise within the period of their leases most vegetable


matter by a rotation of crops; & they choose the most different plants for their rotation: the nurseryman often practices a sort of simultaneous rotation in his alternate rows of different vegetables. I presume that it will not be disputed that on a large farm, a greater weight of flesh, bones, and blood could be raised within a given time by keeping cattle, sheep, goats, horses, asses, pigs, rabbits & poultry, than if only cattle had been kept. In regard to plants this has been experimentally proved by Sinclair1 who found that land sown with only two species of grass, or one kind of grass with clover, bore on an average 470 plants to the square foot; but that when sown, with from 8 to 20 different species, it bore at the rate of about 1000 plants, "& the weight of produce in herbage & in hay was increased in proportion." It is important to observe that the same rule holds for different & not very distinct varieties of the same species when sown together; for M. L. Rousseau, a distinguished practical farmer, on sowing fifteen varieties of wheat/ 26 f/separately, & the same kinds mixed together found on actual measurement that the latter "yielded a much heavier crop than that obtained on far better land on which the unmixed wheats were grown for the purpose of the comparative trial."2

We see on a great scale, the same general law in the natural distribution of organic beings; if we look to an extremely small area supposing the conditions to be absolutely uniform & not very peculiar./26 f v/Where the conditions are peculiar & the station small as compared with the whole area of the country, as Alpine summits; Heaths salt‐marshes, or even common marshes, lakes & rivers, &c.—a great number of individual plants are often supported, belonging to very few species: so it is with Fresh‐water shells; so it is with the marine inhabitants of the arctic seas. But even in these cases, though the individuals appear to be very numerous compared with the species, yet even in these cases, the coinhabitants belong to very different types; for instance Dr. Hooker has marked for us all the plants in Britain, which he thinks may be called truly aquatic: they are, [ ] in number, & they belong to [ ] genera and to [ ] orders.— With respect to the number of individuals to the species, we shall have to return to this subject in our chapter on geographical distribution, & I will here only say that I believe it mainly, but not wholly, depends, on the manufacturing, if I may so express myself, being

1 The author of Hortus Gramineus Woburncnsis, in Loudon's Gardener's Mag. Vol. I.,1826 p 113.

2 Gardener's Chronicle & Agricult. Gazette, 1856, p. 859. See, also, p. 858. and 1857, p. 179 [Samuel Taylor 'The Thick and Thin Sowing Discussion' pp. 178‐9.]


small in size (& sometimes in duration); that is that the number of individuals is small in comparison with the numbers of individuals of the commoner species which inhabit ordinary stations: for we have seen in our 4th Ch. that it is species which most abound in individuals which oftenest present varieties, or incipient species./ 26 f/Supposing the conditions to be absolutely uniform & not very peculiar or unfavourable for life, we seldom find it occupied by any two or three closely allied & best adapted forms, but by a considerable number of extremely diversified forms. To give an example, I allowed the plants on a plot of my lawn three feet by four square which was quite uniform & had been treated for years uniformly, to run up to flower; I found the species 20 in number, & as these belonged [to] 18 genera & these to 8 orders & they were clearly much diversified./26f v b/The most remarkable ex‐ception to this rule, under conditions not apparently very peculiar, is one given by Mr. C. A. Johnsl who says that he covered with his hat, (I presume broad‐brimmed) near to Lands End six species of Trifolium, a Lotus & Anthyllis; & had the brim been a little wider it would have covered another Lotus & Genista; which would have made ten species of Leguminosae, belonging to only four genera! The wretched soil of Heaths, though covered thickly with one or two species of Erica, supports very little life, as judged by their extremely slow growth, & yet, selecting the very worst spots, I have very rarely been able to find a space two yards square, without one or two other plants, belonging to quite different orders, not to mention a good crop of Cryptogams.

To show the degree of diversity in our British plants on a small plot, I may mention, that I selected a field, in Kent, of 13_ acres, which had been thrown out of cultivation for 15 years, & had been thinly planted with small trees most of which had failed: the field all consisted of heavy very bad clay, but one side sloped & was drier: there was no water or marsh: 142 phanerogamic plants were here collected by a friend during the course of a year; these belonged to 108 genera, & to 32 orders out of the 86 orders into which the plants of Britain have been classed. Another friend collected for me all the plants on about 40 uncultivated, very poor, acres of Ashdown Common in Sussex; these were 106 in number, & belonged to 82 genera & 34 orders; the greater pro‐portional number of orders in this case being chiefly owing to the presence of water & marsh plants on the Common: the vegetation was, however, considerably different in other respects, no less than nine of the 34 orders, not being found on the field of thirteen

1 Phytologist, Vol. 2, p. 908.


acres in Kent.—/26f/To give another example of a small area having singularly uniform conditions of life; namely one of the low & quite flat, coral‐islets having a wretched soil, composed exclusively of coral‐debris, but with a fine climate; for instance Keeling Atoll, on which I collected nearly every phanerogamic plant, & these consisted of 20 species1 belonging to 19 genera & to no less than 16 different orders!/

26g/The extreme poverty of the floras of all such islets may be partly due to their isolation & the seeds arriving from lands having different Floras, but chiefly to the poverty & peculiarity of the soil; for coral‐islets, when lying close to large volcanic groups, have an almost equally poor & closely similar flora: the extreme diversity of the plants, the twenty in the case of Keeling islands, belonging to sixteen orders, can, I think, only be accounted for by the fact that of all the plants of which the seeds have been borne across the sea in the later periods of the natural colonisation of the island, those alone, which differed greatly from the earlier occupants, were able to come into competition with them & so lay hold of the ground & survive.

As with plants so with insects. I may premise that entomologists divide the Coleoptera into 13 grand sections, & then into families, sub‐families &c. Mr. Wollaston2 carefully collected during several visits all the Beetles on the Dezerta Grande, a desert volcanic islet about four miles long, & in widest part only three‐quarters broad, lying close to Madeira; & he found 57 species, belonging to 47 genera; & these to all 13 grand sections, except two, which being aquatic forms, could not exist on this waterless islet. Again on the Salvages, an extremely small volcanic isld . between Madeira, & the Canaries, six beetles were collected, & these/26 h/belonged to six genera, to six Families, & to three of the grand Sections!3 As a general rule, I think we may conclude, that the smaller the area, even though the conditions be remarkably uniform, the more widely diversified will its inhabitants be: for to this very diversity, the power of supporting the greatest possible number of living beings, all of which are struggling to live, will be due.

There is another way of looking at this subject; namely to

1. Described by the Rev. Prof. Henslow in Annals of Nat. Hist., 2. Ser., Vol. I, 1838, p. 337.

2. Insecta Maderensia. 1854.

3 In the volcanic Galapagos Islands in the Pacific, I carefully collected all the Coleoptera during several weeks; but omitting two probably naturalised species, I got only 24 species, which have been described by Mr. Waterhouse in Annals & Mag. of Nat. History Vol. 16. 1845, p. 19.—The 24 species belong to 18 genera, to 17 families & to 10 out of 13 grand sections. So here again we see the same rule as in other cases in the text. ['Lundy Island' added in pencil.]


consider the productions naturalised through man's agency in several countries; & see what relation they bear to each other & to the aboriginal productions of the country, i.e. Are they closely allied to, that is do they generally belong to the same genera with, the aboriginal inhabitants of the country? Do many species of the same genus become naturalised? If we looked only to the inorganic conditions of a country, we might have expected that species, belonging to genera already inhabiting it, & supposed on the common view to have [been] adapted by creation for such country, would have formed the main body of the colonists: or/ 26 i/the many species of certain favoured genera would have been the successful intruders. On the other hand, the principle of diversity being favourable to the support of the greatest number of living beings would lead to the expectation, that land already well stocked by the hand of nature would support such new forms alone, as differed much from each other & from the aborigines. Alph. De Candolle1 has fully discussed the subject of naturalisation: He shows that 64 plants have become naturalised in Europe (excluding species from neighbouring regions) during the last three centuries and a half; & these 64 species belong to 46 genera & 24 orders; of the genera, 21/46 are new to Europe.2 Again in N. America, 184 species have become naturalised & these belong to 120 genera & to 38 orders; of the genera, 56/120 are new to N. America.3 A list of the naturalised plants in Australia & on many islands would give similar, but much more striking results. The number of new genera naturalised in Europe & N. America, reciprocally from each other, is the more remarkable when we consider how much allied the two floras are; & that a very large proportion of the/26 k/naturalised plants inhabit land, cultivated nearly in the same manner, which would favour the introduction of allied forms & many forms of the same groups. Hence, I think, we may conclude that naturalised productions are generally of a diversified nature; & as Alph. De Candolle has remarked native

1 Geographie Botanique, p. 745, 759, 803.

2 In some respects small areas, not including in the sub‐regions many indigenous representative species, are best for comparing the native <indigenous> & naturalised productions. De Candolle gives a list (p. 645 et seq.) (in large type) of 83 plants, which he considers as certainly naturalised in Great Britain: these belong to no less than 71 genera: & of these 31/71 are new to Britain. The indigenous genera include on an average about 2.8 indigenous species: the naturalised only 1.1.

3 Dr Asa Gray seems to consider many more plants as naturalised, than does De Candolle for in his Manual of the Botany of the Northern United States (2nd Edit.) he gives a list of 260 naturalised plants, belonging to 162 genera, of which no less than 100 are new to America. The naturalised genera include on an average 1.6 species: the indigenous include 2.6.—


floras gain by naturalisation, proportionally to their own numbers, far more in genera than in species.

If we turn to animals, we find, though our data are very scanty, the same general fact: no where in the world have more mammals become well naturalised than in S. America (cattle, horses, pigs, dogs, cats, rats & mice); & yet how extremely unlike is the native mammalian Fauna of S. America to that of the Old World!

The whole subject of naturalisation seems to me extremely interesting under this point of view, & would deserve to be treated at much greater length. It confirms the view that in natural colonisation, for instance in that of a coral‐islet, diverse forms very different from the few previous occupants, would have the best chance of succeeding. It shows us, & by no other means can we form a conjecture on this head,/261/what are the gaps or still open places in the polity of nature in any country: we see that these gaps are wide apart, & that they can be best filled up by organic beings, of which a large proportion are very unlike the aboriginal inhabitants of the country. Consequently we might perhaps from this alone infer, that natural selection by the preservation of the most diversified varieties & species, would in the long run tend, if immigration were prevented, to make the inhabitants, more & more diversified; though such modified forms would for immense periods plainly retain from heritage the stamp of their common parentage.

The view that the greatest number of organic beings (or more strictly the greatest amount of life) can be supported on any area, by the greatest amount of their diversification is, perhaps, most plainly seen by taking an imaginary case. This doctrine is in fact that of "the division of labour", so admirably propounded by Milne Edwards,1 who argues that a stomach will digest better, if it does not, as in many of the lowest animals, serve at the same time as a respiratory organ; that a stomach will get more nutriment out of vegetable or animal matter, if adapted to digest either separately instead of both. It is obvious that more descendants from a carnivorous animal could be supported in any/26m/country: if some were adapted, by long continued modification through natural selection, to hunt small prey, & others large prey living either on plains or in forests, in burrows, or on trees or in the water. So with the descendants of a vegetable feeder more could be supported, if some were adapted to feed on tender grass &

1 [Milne‐Edwards, Introduction à la Zoologie générale.. .Paris, 1851 see p. 35, pp. 55‐7, and art. 'Organisation' in Dict. class. hist. nat., vol. 12, Paris 1827, pp. 332‐44.]


others on leaves of trees or on aquatic plants & others on bark, roots, hard seeds or fruit.—

Perhaps I have already argued this point superfluously; but I consider it as of the utmost importance fully to recognise that the amount of life in any country, & still more that the number of modified descendants from a common parent, will in chief part depend on the amount of diversification which they have undergone, so as best to fill as many & as widely different places as possible in the great scheme of nature. Now let it be borne in mind that all the individuals of the same variety, and all the individuals of all the species of the same genus, family &c, are perpetually struggling to become more numerous by their high geometrical powers of increase. Under ordinary circumstances each species will in the briefest period have arrived at its fluctuating numerical maximum. Nor can it pass this point, without/26 n/some other inhabitants of the same country suffer diminutions; or without all the descendants of one species becoming similarly modified in some respect so that they better fill the place of their parentspecies; or without (& this would be the most effectual) several varieties & then several species are thus formed by modification, so as to occupy various new places, the more different the better, in the natural economy of one country. Although all the inhabitants of the country will be tending to increase in numbers by the preservation through natural selection of diverse modifications; but few will succeed; for variation must arise in the right direction & there must be an unfilled or less well‐filled place in the polity of nature: the process, moreover, in all cases, as we shall presently see, must be slow in an extreme degree.

Let us take an imaginary case of the Ornithorhynchus; & suppose this strange animal to have an advantage over some of the other inhabitants of Australia, so as to increase in numbers & to vary: it could, we may feel pretty sure, increase to any very great extent, only by its descendants becoming modified, so that some could live on dry land, some could feed exclusively on vegetable matter in various stations, & some could prey on various animals, insects fish or quadrupeds. In fact its descendants would have to become/ 26 o/diversified, somewhat like the other Australian marsupials, which, as Mr. Waterhouse has remarked, typify in their several sub‐families, our true carnivores, insectivores, ruminants & rodents. Moreover it can, I think, hardly be doubled, that these very marsupials would, profit by a still further division of physiological labour; that is by their structure becoming as perfectly carnivorous, ruminant & rodent as are our old‐world forms; for


it may well be doubted (not here considering the probable intellectual infirmity of the marsupialia in comparison with the other or placentate mammals) whether many marsupial vegetable feeders could long exist in free competition with true ruminants, & perhaps still less the carnivorous marsupials with true feline animals. And who can pretend to say that the mammals of the old world are diversified & have their organs adapted to different physiological labours to the extreme, which would be best for them under the conditions to which they are exposed? Had we known the existing mammals of S. America alone, we should no doubt have thought them perfect & diversified in structure & habit to the exact right degree; but the vast herds of feral cattle, horses, pigs & dogs,/ 26p/at least show that other animals, & some of them as the horse & solid‐horned ruminants, very different from the endemic S. American mammals, could beat & take the place of the native occupants.

In Chapter Iv we have seen on evidence, which seems to me in a fair degree satisfactory, that on an average the species in the larger genera in any country oftenest present varieties in some degree permanent, and likewise a greater average number of such varieties, than do the species of the smaller genera. It is not that all the species of the larger genera vary, but only some, & chiefly those which are wide‐rangers, much diffused & numerous in individuals. In the same chapter we also saw that/26 p v/the species in the larger genera are thought by highly competent judges to be more closely related together, being clustered in little sub‐groups round other species, than are the species in the smaller genera; & this closer affinity & grouping of the species in the larger genera, & the fact that there is no unfailing test by which to distinguish species & varieties,/26 p/all to a certain extent confirm the view that varieties, when in some degree permanent, do not essentially differ from species, more especially from such species, as are closely allied together. Hence I look at varieties as incipient species./

26 q/I have lately remarked that the formation of new varieties & species through natural selection almost necessarily implies (as with our domestic productions) much extinction of the less altered, & therefore less favoured, descendants from the same original parent‐stock, whose places they occupy in the struggle for life. Hence, though the larger genera may be now varying most, & must, according to our theory, have varied largely, so as to have become modified into many specific forms, yet such large genera must have suffered a large amount of extinction, & very many intermediate & less modified forms have been wholly swept away.

[page] 236

// fol. 26s b begin//

[page] 237

// fol. 26 s b cond //


Nevertheless, I think we may infer that in any given country, on the whole, there will have been rather less extinction, proportionally to the whole amount of extinction within any given period, amongst the larger than amongst the smaller genera. For the species which vary most & thus give rise to new species, are chiefly the very common & much diffused species, & therefore the most favoured forms, which would naturally be the least liable to extinction; & such common & much diffused species tend to belong to the larger genera. Indeed it seems to me that the simple fact of a number of allied species, beyond the average number of allied species, inhabiting any country; shows that there/26r/is something in common in such groups of species, or genera, which is favourable to them, & consequently that they would suffer proportionally less from extinction than the smaller genera. Therefore, from the species of larger genera tending to vary most & so to give rise to more species, & from their being somewhat less liable to extinction, I believe that the genera now large in any area, are now generally tending to become still larger. But what will be the end of this? for we do not find in nature genera of indefinite size, with in‐numerable species. Here in one way comes in the importance of our so‐called principle of divergence: as in the long run, more descendants from a common parent will survive, the more widely they become diversified in habits, constitution & structure so as to fill as many places as possible in the polity of nature, the extreme varieties & the extreme species will have a better chance of surviving or escaping extinction, than the intermediate & less modified varieties or species. But if in a large genus we destroy all the intermediate species, the remaining forms will constitute sub‐genera or distinct genera, according to the almost arbitrary value put on these terms,—according to the number of intermediate forms which have been destroyed,—and/26 s/according to the degree of difference between the extreme species of the original genus. Nevertheless the modified descendants from the common parent‐stock, though no longer forming what is called the same genus, may still go on becoming more & more numerous, & more & more diversified.

The complex action of these several principles, namely, natural selection, divergence & extinction, may be best, yet very imperfectly, illustrated by the following Diagram, printed on a folded sheet for convenience of reference.*/26s v/This diagram will show the manner, in which I believe species descend from each other & therefore shall be explained in detail: it will, also, clearly show

* Fold‐out version can be found at the end of the book.


several points of doubt & difficulty; /26s/Let A to M represent the species of a genus, numerically large compared with the other genera of the same class in the same country, & arranged as naturally as can be done, so that A & M are the two most distinct forms in all respects. The unequal distances of the letters may represent the ordinary way in which the species, even when as in this imaginary case all are closely related together, yet stand unequally related in little sub‐groups. This genus may have one, two or even more varying species. Any of the species may vary; but it will generally be those species which are most numerous in individuals & most diffused; & this shows that such species have already some advantages over the other inhabitants of the country. From our principle of divergence, the extreme varieties of any of the species, & more especially of those species which are now extreme in some characters, will have the best chance,/26t/after a vast lapse of time, of surviving; for they will tend to occupy new places in the economy of our imaginary country. I do not mean that any of these points are of invariable occurrence, but that in the long run such cases will prevail. The extreme species A and M will differ in very many respects; but for convenience sake we may look to any one character, & suppose A the most moisture‐loving & M the least moisture‐loving species.

We will first take the simplest case. Let M inhabit a continuous area, not separated by barriers, & let it be a very common & widely diffused & varying plant. From the fact of M. being very common & widely diffused, it clearly has some advantages in comparison with most of the other inhabitants of the same country; but, we will suppose, that it might become still more common, if retaining the advantages which it already has, it could endure still more drought. It is a varying species; & let z1 ‐m1 represent numerous, extremely slight variations of many kinds, produced at intervals, of which m1 alone is a more drought‐enduring variety. As m1 tends to inherit all the advantages of its parent M, with the additional advantage of enduring somewhat more drought, it will have an advantage over it, & will probably first be a thriving local variety, which will spread & become extremely common & ultimately, supplant its own parent. We may now repeat the process, & let the variety m1 vary in a similar manner; perhaps/ 26u/many thousands of generations may pass before m1 will produce another variety m2 , still more drought‐enduring & yet inheriting the common advantages of m1 & M; but if this should ever occur, the same results, as before, will follow: & ultimately, by repeating the process, there may be produced m10 , which may


either be, according to the amount of difference thus acquired, a very strongly marked variety, or a sub‐species, or good species, enduring far more drought than M & probably with correlated differences of structure. In each stage of descent, there will be a tendency in the new forms to supplant its parent, though probably, as we shall see, very slowly, & so ultimately cause its extinction. But if M had originally inhabited a country separated by barriers into distinct districts, in one or more of which the varieties M1‐10 had never originated or had never been able to enter, M and m1 & ultimately m 10 might be living contemporaneously, but separated:/26u v/or, again, if m 1‐10 had been produced, capable of enduring more drought, but not at the same time enduring an equal amount of moisture with the parent M, both parent & modified offspring might coexist: the parent (with perhaps a more restricted range) in the dryer stations, & m1‐10 in the very dryest stations./

26u/It should always, be borne in mind that there is a wide distinction between mere variations & the formation of permanent varieties. Variation is due to the action of external or internal causes on the generative systems, causing the child to be in some respects unlike its parent; & the differences thus produced may be advantageous or disadvantageous/26 v/to the child. The formation of a permanent variety, implies not only that the modifications are inherited, but that they are not disadvantageous, generally that they are in some degree advantageous to the variety, otherwise it could not compete with its parent when inhabiting the same area. The formation of a permanent variety must <can> be effected by natural selection; or it may be the result, generally in unimportant respects, of the direct action of peculiar external conditions on all the individuals & their off‐spring exposed to such conditions. We shall best perceive the importance of the difference by glancing at our domestic breeds: in our truest breeds, innumerable slight differences are continually occurring & can be detected by measurement, but only those differences which improve the breed in the often fanciful eyes of the Fancier are rendered permanent by the animals so characterised being carefully preserved, matched & largely bred from; all other slight differences being lost, by the animals not being largely bred from, & from indiscriminate crossing. If, however, the process of selection were continued for a long time by two Fanciers, under very different conditions of climate or food, some subordinate differences would probably arise between the two lots, owing to the direct action of such conditions. Now in our diagram, the letters z1 ‐m1 , z2 ‐m2 &


c represent all sorts of successive slight variations, of which m1‐10 , the most drought‐enduring varieties alone have been naturallyselected & been rendered permanent.

This natural selection has been possible, owing to there having been/26 w/a place in the economy of our imaginary country, which the descendants of M, from inheriting all or some of the advantages over the other inhabitants which made M a very common species, could seize on, when rendered more drought‐enduring.

With respect to the process by which each new & improved variety supplants its parent, this must often have gone on in two slightly different manners, differing, however, only in degree. In those animals which are highly locomotive & of which two individuals unite for each birth, there can only seldom have arisen as we shall hereafter see, within the same continuous area, especially if of not very large size, distinct varieties, for they would become blended by such free crossing. In such cases, modifications must be effected quite insensibly by the natural selection of mere individual differences; nearly in the same way as many of our domestic breeds throughout whole districts have been insensibly changed from their ancient state. So that in our diagram the letters m1‐10 may represent in the case of the higher animals, not recognizable varieties, but mere ideal steps in a real, yet insensibly gradual, change of structure. In organic beings which do not cross freely/26 x/& which are more stationary, & which are capable of propagating at a great rate, a variety might easily be formed in one spot (more especially if in some slight degree isolated) & might not spread & supplant its parent‐stock, until it had become developed by the continued natural selection of similar extremely slight or individual differences into a distinct & plainly recognizable variety./26x v/I am inclined to think from the frequency of local varieties, though the subject must remain very doubtful, that this latter process has been a very common one, for a variety would often be unable to supplant its parent, until it had become considerably modified so as to have a decided advantage over it. For instance in the imaginary case of the varieties m1‐10 which are supposed to inherit all the characters of M, with the addition of enduring more drought; these varieties would inhabit stations, where M could not exist, but in the less dry stations m1‐10 would have very little power of supplanting their parent M; nevertheless during unusually dry seasons m1‐10 would have a great advantage over M.& would spread; but in damper seasons M, would not have a corresponding advantage over m1‐10 for these latter varieties are supposed to inherit all the characters of their parent. So there


would be a tendency in m1‐10 to supplant M, but at an excessively slow rate. It would be easy to show that the same thing might occur in the case of many other new characters thus acquired; but the subject is far too doubtful & speculative to be worth pursuing. I will only add that/26 x/with the most freely crossing & locomotive animals, when inhabiting an area, separated by barriers only to be passed after geographical changes or through some most rare accident, a similar process must often have occurred; for in such cases, distinct & plainly marked varieties might have been insensibly formed in the different districts by the selection of mere individual differences; & when these districts became united, so that the varieties could mingle & come into competition, the best variety would supplant the other varieties or the parentstock.

To return to our diagram. I do not suppose the process generally to have been so simple as represented under M, where a simple variety m1‐10 in each stage of descent has been naturally selected. We have seen that not only more species, especially the very common species, in the larger genera in any country present varieties in some degree permanent, but that each such species on an average tends to present a greater number of varieties, than do the species, especially the rarer species, in the smaller & less flourishing genera. As varieties from a species tend to inherit the advantages which/26 y/made the parent common, these varieties will ultimately tend to be common & to vary; moreover they descend from a variable stock, & are still exposed to the conditions which made their parents vary, hence for this cause they will be liable to vary. Consequently there will be a tendency in the original varying species, after a vast number of generations to produce an almost infinite number of varieties; but our principle of divergence explains how the most diversified varieties will generally have decided advantages over the less diversified & intermediate varieties, causing their extinction & thus reducing the number of varieties living at any one time. These remarks are illustrated in our diagram under A., which species, after many generations represented by dots, is supposed to have varied largely, & to have produced these varieties a1 , d1 l1 in some degree permanent; of these, again after many generations & much variation; the two extreme varieties a1 and l1 , are supposed to have produced other varieties in some degree permanent; of which the extreme varieties have again reproduced others, represented finally by a10 & l10 . In the diagram I have been able to represent only one other branch proceeding from f5 , & giving rise to a third variety h10


which being the extreme form in its own branch has the best chance of surviving/26 z/& seizing on some place in the natural economy of the country inhabited by the genus.

By continuing the process represented in the diagram, the forms marked a10 , h10 , l10 , may be made different in any degree, till they would be universally be [sic] ranked as good species; & the number of such new forms would continually tend to increase. These new species will generally have supplanted, perhaps by a very slow process their several parents in each stage of descent & their original common parent A,—that is if formed in one continuous area, or as soon as they came into competition with each other if formed in different areas. The original species A. was supposed to be the most moisture loving plant; & if for simplicity sake we imagine a1 more moisture loving & l1 less moisture loving, but inheriting some of the advantages which made A in the great & complex battle for life a very common species; & the offspring of these varieties to be continually selected on the same principle, a10 will have been rendered so moisture loving as to have become semi‐aquatic, & l10 far less moisture loving than A; & in the third branch, h10 , about as moisture loving as A, for it has descended from f5 which was more moisture‐loving than A, and subsequently has become less so. Not that I at all suppose the diversity is ever thus confined to one point; for as a1‐10 becomes moisture‐loving & as l10 becomes less moisture‐loving both would under the extremely complex conditions to which all organic beings are exposed, come to be exposed to new dangers & /26aa/would have to gain some other advantages over other organic beings with which they would have to compete. So that in love of moisture & in many other respects, a1‐10 , h1‐10 , l1‐10 would come to differ or diverge more & more from each other & their original parent‐stock.

A little reflexion will show the extreme importance of this principle of divergence for our theory. I believe all the species of the same genus have descended from a common parent; & we may call the average amount of difference between the species, x; but if we look at the contemporaneous varieties of any one species, the amount of difference between them is comparatively extremely slight & may be called a. How thus can the slight difference a be augmented into the greater difference x; which must on our theory be continually occurring in nature, if varieties are converted into good species? The process feebly illustrated in our diagram, I believe, explains this; namely the continued natural selection or preservation of those varieties, which diverge most in all sorts of respects from their parent‐type, (but still


largely inheriting those advantages which made their parents generally dominant & common species) so as to fill as/26 bb/many, as new, & as widely different places in the economy of nature, as possible.

A glance at Diagram 2. will perhaps render this plainer. The varieties a1‐10 , l1‐10 may be here again for simplicity be looked at as more & less moisture loving plants; & everything is the same as in diagram I (the third branch h6‐10 cannot be introduced) except that it is left to mere chance in each stage of descent, whether the more or less moisture loving varieties are perserved; & the result is, as graphically shown, that a10 & l10 differ in this respect; & so in other respects, hardly more than did the first varieties (a1 l1 ) which were produced.

In regard to the difference between varieties & species, I may add that varieties differ from each other & their parents, chiefly in what naturalists call unimportant respects, as size, colour proportions &c; but species differ from each other in these same respects, only generally in a greater degree, & in addition in what naturalists consider more important respects. But we have seen in Ch. IV, that varieties do occasionally, though rarely, very slightly in such important respects; and in so far as differences in important physiological characters generally stand in direct relation to different/26 cc/habits of life, modifications however slight in such characters would be very apt to be picked out by natural selection & so augmented, thus to fit the modified descendants from the same parent to fill as many & as widely different places in nature as possible. We shall, also, see in a future chapter that a large part of the differences in structure between species may be accounted for by the mysterious laws of correlation; by which, I mean, that when one part is modified, (or the whole animal at one age, as with the larvae of insects) other parts necessarily become altered through the correlated laws of growth. That there is no obvious & unmistakeable difference between the differential characters of species & varieties, is plainly shown by the number of debateable forms in the best known countries, which are ranked by one good naturalist as true species, & by another as mere varieties.

Our principle of divergence has another very important bearing. In the diagram, A. has given rise to three new species, & M to one. The other species of the genus, B to L, are supposed to have/ 26dd/transmitted unaltered descendents. Hence, even supposing that A & M have been supplanted as I believe will usually have been the case, by their modified & improved descendants, the genus will have become not only more divergent in character


(a10 more aquatic than A; & m10 more drought‐enduring than M.) but numerically larger. The original species A to M were supposed to be closely allied, but yet to exhibit traces, as is so general, of being divided into sub‐groups. The sub‐groups, after the formation of the new species, will be slightly altered & increased in number; for a10 & h10 will be closely related together from common descent from f5 , & closely yet less closely with l10 from descent from their common ancester A.; and they will all differ as much, generally more from B, than did A. So again m10 having constantly diverged from the characters of M will now stand more distant from L, than M originally stood. This is represented in the Diagram III. And from the reasons already given, I believe there will be a constant tendency in the modified descendants of A & of M to go on thus producing more & more new specific forms & thus more & more modified or divergent.

What will be the limit to this process in nature? Though many genera are large, they do not include an indefinite number of species. I believe that there is no limit/26ee/to the number of species tending to be formed from the most favoured forms in any country (or those which have any [sic] the greatest advantages over the coinhabitants), except the number of species which the country is capable of supporting; but such modified descendants, or new species, after a long period will have to be ranked not in the same genera, but in distinct genera, families or orders. For if we suppose the process illustrated in diagram I. to have long continued & the modified descendants of A to have become extremely much multiplied and diversified in many ways, they will tend to take the places of & thus exterminate the species B.C.D. &c, which originally were nearest related to A. & were not then such common & flourishing species. So if M had left several modified & divergent descendants, it would have been with L, K. &c./26ee v/It may be here worth observing, that although the new species in taking the place of the old (their great uncle) may have acquired through natural selection, some of their characters; this kind of resemblence would be called by naturalists that of analogy, & the real affinity of the new species would be with their real parents: thus 10 might come to simulate some of the character of B, from occupying its place in nature yet the real affinity of 10 would be with A.—/26ee/Continue this process, & all, or nearly all the original species (A to M) will become extinct. In Diagram IV. this is represented, E & F alone now having descendents, whether or not modified. And the final result will be, that we shall have two large groups of modified descendants,/26ff/coming


from the two species, generally the extreme species, (A & M) of the original genus, and differing as much as natural selection could make them from each other & from their two parents, which at the first start differed much: assuredly these two new groups of new species would be ranked in different genera, which would be very distinct, if all the original intermediate species from B to L. had been exterminated, but somewhat less distinct if some of these species (as represented in Diagram iv.) had left descendants, whether or not modified.

Now for a moment let us go back many stages in descent: on our theory the original twelve species A to M are supposed to have descended & diverged from some one species, which may be called Z, of a former genus. But now, according to the result given in the last paragraph, Z will have become the ancestor of two or three very distinct groups of new species; & such groups, naturalists call genera. By continuing the same process, namely the natural selection of generally the most divergent forms, with the extinction of those which have been less modified & are inter‐mediate, Z may become the ancestor of two very distinct groups of/26gg/genera; & such groups of genera, naturalists call Families or even Orders. But to this subject, we shall have to return in our Chapter on Classification.

I have previously remarked that there seems to be no limit to the number of modified descendants, likely to proceed from the most favoured form in any country,—the most favoured always tending to diverge in structure & take the place of & exterminate the less favoured & intermediate forms,— except the total number of species, which the country is capable of supporting. But it may be objected that as natural selection, extinction & divergence must have been going on since the dawn of Life, why have we not an infinite number of species, almost as many species, as individuals? We shall presently see that natural selection can act only with extreme slowness. Nor do we by any means know that the maximum number of species, which any country would be best fitted to support, has anywhere been as yet produced: the fact that there is no country which does not support several, often many, organic beings naturalised by man, without, as far as we know,/26hh/having caused the extinction of an equal number of the indigenous productions, renders it probable that such countries were capable of supporting a greater number of specific forms than nature had supplied them with. Even the Cape of Good Hope, which is apparently the richest district in the world in different kinds of plants has received, as I am informed by Prof. Haney


from [ ] to [ ] naturalised species. Many geologists, indeed believe that the number of species in the world has gone on increasing from the earliest geological days; but I am sorry to own that the evidence on this head seems to me quite insufficient./

26 hh v/It might indeed be argued from the enormous list of shells, found in the eocene Paris basin, & even in the ancient Silurian system of Bohemia, as so admirably worked out by .Barrande, that at these periods & in these places, a greater number of species existed than anywhere at the present day. But it may be doubted how far such comparisons are in any instance trustworthy; for we have reason to suppose that the duration of each sub‐division of each geological formation is so enormous, that it is not fair to compare all the species found in one such sub‐division with all existing within an area at the present day. Barrande's "colonies " show, according to Sir C. Lyells explanation of them,1 what changes of climate or currents must have taken place within certain definite periods: the Glacial epoch within what may be called the present period, should teach us caution, for far lesser changes than the glacial epoch, not easily to be detected in ancient geological formations, might alternately bring in & displace, & apparently mingle many organic beings, which never really co‐inhabited the same area./

26hh/But if the time has not yet arrived, may it not at some epoch come, when there will be almost as many specific forms as individuals? I think we can clearly see that this would never be the case. Firstly, there would be no apparent benefit in a greater amount of modification than would adapt organic beings to different places in the polity of nature; for although the structure of each organism stands in the most direct & important relation to many other organic beings, and as these latter/26ii/increase in number & diversity of organisation, the conditions of the one will tend to become more & more complex, & its descendants might well profit by a further division of labour; yet all organisms are fundamentally related to the inorganic conditions of the world, which do not tend to become infinitely more varied. Secondly as the amount of life & number of individual beings, whether or not much diversified, also primarily depends on such inorganic con‐ditions; if there exist in any country, a vast number of species (although a greater amount of life could be supported) the average number of individuals of each species must be somewhat less than if there were not so many species; & any species, represented by but few individuals, during the fluctuation in number to which

1 Supplement to Manual of Geology, 2nd Edit. 1857. p. 34.


all species must be subject from fluctuations in seasons, number of enemies &c, would be extremely liable to total extinction. Moreover, whenever the number of individuals of any species becomes very small, the ill‐effects, as I believe, of close inter breeding would come into play. Lastly we have seen in our Chap. IV & shall presently again see, that the amount of variations, & consequently of variation in a right or beneficial direction for natural selection to seize on & preserve, will bear some relation within any given period, to the number of individuals living & liable to variation during such period: consequently when the descendants from any one species have become modified/26 kk/into very many species, without all become numerous in individuals, which [we] see hardly ever to be the case with all the species of the same genus or family, there will be a check amongst the less common species to their further modification: the lesser number of the individuals serving as a regulator or fly‐wheel to the increasing rate of further modification, or the production of new specific forms.

Subject to these restraining influences, I can see no limit to the number of modified descendants, which might proceed from the most favoured forms, whatever they may be, now living in the world. If we return to look to the future, as far into the remotest futurity as the Silurian system lies in the remote past, our theory would lead to the conclusion that all organic beings which will live at that far distant period, will be descendants from a very few of our contemporaries; perhaps from not so few, owing to the increasing complexity of the organic world, as our existing organisms have descended from; for our animals have descended, according to our theory, from four or five ancestral types & our plants from apparently still/26ll/fewer; & if we rashly dare trust to mere analogy, all our plants & animals from some one form, into which life was first breathed.

Taking a more modest glance into futurity, we may predict that the dominant genera, now abounding with common & widely difused species, will tend to be still more dominant for at least some considerable lapse of time, & will give rise to new groups of species, always diverging in character, & seizing on the places occupied by the less favoured forms, whether or not their near blood relations, supplanting them & causing their extermination. The great & flourishing genera both of plants & animals, which now play so important a part in nature, thus viewed become doubly interesting, for they include the ancestors of future conquering races. In the great scheme of nature, to that which has much, much will be given.


Finally, then, in regard to our principle of Divergence, which regulates the natural Selection of variations, & causes the Extinction of intermediate & less favoured forms, I believe it to be all /26 mm/ important as explaining why the average difference between two species of the same genus, the parents of which by our theory once existed as mere varieties, is greater than the average difference between two such varieties. It bears on, & I think explains, the classification or natural affinities during all times of all organic beings, which seeming to diverge from common stems are yet grouped like families within the same tribes, tribes within the same nations, & nations within the same sections of the human race. We shall, also, hereafter see that these views bear on palaeontology & explain why extinct forms either fall within existing groups, or, as is so frequently the case, are in some slight degree intermediate between them.

The relation of all past & present beings may be loosely compared with the growth of a few gigantic trees; that is if we suppose that from each of the innumerable twigs, innumerable buds are trying to sprout forth, & that the other buds, twigs & branches have the best chance of growing from getting more light. The buds & twigs may represent existing species, & all beneath their living extremities may represent extinct forms. We know that the twigs proceed from lesser branches, these from larger & these from main limbs, from the trunk, & that the several branches & limbs are of very unequal/26nn/sizes; & this grouping of the branches may represent the natural classification of organic beings. In our living trees we can trace in the gnarled & leafless branches the connecting links; but so imperfect are our palaeontological records, that we can only here & there find a form which may be called a forked branch, with its two arms directed towards two now distinct groups of organisms. As we know that the gnarled branches were at successive periods tender twigs crowded with buds, so we may believe that every organic class, whether or not now having lineal descendants on the earth, swarmed at each stage of descent under diversified forms of life. Many a smaller & larger branch, & even some main limbs have utterly perished, from being over topped by the ever diverging budding twigs; so it has been with whole groups of organic beings. Here & there a branch is still alive, carrying only a few twigs & buds; & these will represent the organic groups having few species & fewer genera, which are now on the road to extinction. As buds give rise by growth to fresh buds, & these, if vigorous, branch out & give rise to many a diverging branch still branching out, & causing the death of many a feebler twig &


branch on all sides & beneath, so by generations I believe it has been with the great Tree of Life, which fills the crust of the earth with fragments of its dead & broken branches, & covers with its ever living, ever diverging & marvellous ramifications, the face of the earth./

27, 28/Long ere this, a crowd of difficulties will have arisen in the reader's mind, overwhelming my theory of natural selection, more especially when applied to organs or beings widely different in the same great classes. Some of these difficulties are indeed great enough almost to crush my belief; but many, I think, are only apparent. Is it possible to believe that the eye with its admirable correction for spherical & chromatic aberration, & with its power of adapting the focus to the distance, could have been formed from the simplest conceivable eye, by natural selection? Is it possible for the instinct of a bee, which produces a cell constructed on the highest geometrical principles, to be thus perfected? I confess that my mind recoils from such an admission; yet, reflecting on the known gradations in so wonderful an organ as the eye amongst existing animals,—a mere/29/small fraction of those which have lived,—I can see no logical impossibility; & as far as probability is concerned, a safe conclusion can be drawn, as it seems to me, only from the general phenomena of organic beings, as indicative whether each being has been simply created or has been produced by the common laws of generation with superadded modification. But these questions, & likewise the general subject of instinct shall be discussed in separate chapters.

What shall we say of small & apparently trifling organs, yet most useful to the animal possessing them, as the eye‐lash, or a tail serving as a fly‐brush; could these have been produced by natural selection, which is in fact selection for life & death? But I have already shown how cautious we should be in deciding what trifle may turn the nicely‐suspended balance of life in the great struggle for existence. Again how could a swimming animal be turned into a crawler, or a walking animal into a flyer: how/30/ could they live in an intermediate state? Undoubtedly nothing can be effected through natural selection except by the addition of infinitesimally small changes; & if it could be shown that in cases like the foregoing, transitional states were impossible, the theory would be overthrown. This being so, it may be further asked, do we not meet in certain members of a class organs, which, as far as we can see, are absolutely new creations, & which cannot be some other part or organ modified by natural selection in accordance with the laws of morphology? We shall see that such cases are surprisingly few & hard to find.


Again it has often been urged that if species were subject to change all nature would be in confusion & the limits of no species distinct; but this argument depends on the assumption that the change is rapid & that many species are simultaneously undergoing change. If species were as distinctly defined, as some authors pretend, systematic/31/natural history would be a far less difficult subject, that those authors will find if they will take up for description almost any group, especially a varying group of species; but to this subject I shall immediately recur. So again it has been said, if species were subject to change, we should find plain evidence of such change in our collections of fossil remains; but the force of this objection, in main part, lies in the supposition that the records of geology are as ancient as the first commencement of life, & that they are far more perfect than some of our most experienced geologists have shown good reason for believing that they are in truth. I will here only ask those who make this objection, can they believe that at some future geological epoch, fossil remains will tell that which we do not now know, namely what were the exact steps by which the various British breeds of sheep & oxen have descended/32/from some one or two parent stocks. It should be remembered we do not mean forms intermediate between horse and tapir, but between both of them & some unknown common parent.

Lastly why do two species when crossed, either yield few or no offspring, & these more or less sterile, & why do those varieties which we may safely conclude are descended from a single species yield abundantly fertile offspring? To this important subject I will devote a chapter. And all the foregoing great difficulties, & some curious special cases shall be stated in detail, as fairly as I can, & be discussed. That some difficulties remain to be mastered will not be thought surprising by those who will make allowance for our ignorance oh what is daily passing round us in the living world; & our incomparably greater ignorance of the many past worlds which have teemed with life.—/

33/Causes favourable & unfavourable to Natural Selection.—Having given a pretty full outline of my theory, it will be necessary to discuss as well as we can, though very imperfectly, the circumstances, favourable or the contrary to natural selection. We have seen that variability is the foundation. The variation, whatever its cause be, must be inherited or tend to be inherited to be of any use. Certainly this tendency is very strong & applies to the most trifling changes; but it often fails; & the offspring instead


of taking after their parents resemble their grandfathers or more remote ancestors. We see this repeatedly perhaps oftenest, at least most plainly, where strongly marked varieties are crossed; but in all cases it must tend to retard natural selection.

Again the variation must be in the right direction to profit the individual, otherwise it will not be selected. I do not here refer to the direct effects of climatal conditions, for these must be quite unimportant, in relation to the numberless exquisite co‐adaptations of each organic being to other inhabitants of the area./

34/I am inclined to believe that in the polymorphous or protean groups of species, as they have been called, mentioned in our Ch. IV which we meet with in every great class, we see more fluctuating variability,—perhaps the very tendency to vary being inherited,—the variation being of no use in any one direction to the being in question, & therefore with no one character steadily selected, augmented & rendered nearly constant.

The expression of variation in a right direction implies that there is a place in the polity of nature, which could be better filled by one of the inhabitants, after it has undergone some modification: the existence, therefore, of an unoccupied or not perfectly occupied place is an all important element in the action of natural selection. I do not doubt, as previously remarked from the number of naturalised productions, that everywhere such open places ready to be filled exist; but it is obvious that such places or gaps will be more frequent, & it may be said wider, in districts favourable for life, but yet not thickly stocked with various forms. Districts subjected to some physical change & cut off from free immigration will be thus circumstanced;—for instance part of a continent separated by a desert or mountain‐barrier, into which after climatal changes,/35/the other inhabitants of the continent cannot freely enter; or better still a volcanic island, rising from the ocean at first with few or no inhabitants, but receiving an occasional stray colonist. Now both Mr. Wollaston & Alph. de Candolle1 have strongly insisted that isolated areas are the chief scenes of what they consider, like most naturalists, as the actual creation of new species & likewise of varieties. It is not, I may add isolation in the abstract which seems to affect organic beings; for the very same spot may easily be isolated for one set of beings & not to another: thus Madeira is not isolated for birds for annually birds are blown there from the mainland, & there is only one endemic or peculiar bird & that not a very distinct species: from what we know of the habits of land‐molluscs this

1 Variation of Species, p. 70. Bot. Geograph. [See II, 1092‐4., 1125.]


island must be closely isolated for them, & a large majority of the species are endemic; whereas there is not a single endemic sea‐mollusc, & these, little as we know of their means of dispersal, can hardly be so completely isolated as the land‐molluscs: again coleoptera are seldom strong flyers, & therefore would be here more isolated/36/than the other orders of insects, & Mr. Wollaston tells me that he believes that there are far more endemic species of Coleoptera than in the other orders. We have seen in the last chapter that birds, for instance, in the struggle for existence would be apt to come more into competition with other birds, than with other animals; & so land‐molluscs with land‐molluscs, & beetles with beetles: consequently a few beetles or land‐molluscs (whether we suppose them the remnants of an ancient population before the island was severed from the mainland, or as I think far more probable, occasional stray colonists) placed by themselves in this island would find themselves in a far more disturbed condition & with more places opened to them in their own scheme of nature, than would those other animals, which found themselves associated with all or nearly all their old compatriots with whom they had long struggled in their native land.

Isolation by itself will apparently do nothing; we can find on mountain summits, & in the lowlands innumerable instances of plants & insects with not another individual of the same species within a distance of many miles, & which we have no reason to doubt have long remained there, & yet are absolutely identical with the same species/37/from elsewhere. Isolation under a somewhat different climate introduces another element of change; but the fact which must strike every naturalist is that isolation under the same climate seems to have been eminently favourable to the production of new forms. The climatal conditions of Madeira could probably be paralled on the shores of Europe, as closely as the habits of most species require, judging from their ranges on the mainland; yet, as Mr. Wollaston has shown, those islets swarm with peculiar endemic Coleoptera & Land Molluscs. We see the effects, of isolation under the same climate in the numerous endemic species, both with whole groups & in the separate islets of the Galapagos & Sandwich & Canary archipelagoes, & in the West Indies, as far as some of their productions are concerned./ 37 v/In our chapter on Geographical Distribution, I shall enter on some details showing how extremely rich isolated islands are in endemic species in relation to their areas, as compared with an equal area on the most favoured mainlands. In the case of some of the above archipelagos it is known, & in/37/other cases it is


highly probable, that the inhabitants, excluding those peculiar to the archipelago, are differently grouped to what they are in the mainland, & differently on the separate islets, so that a colonist would be exposed in each to a somewhat different set of competitors. But to this subject, also, we shall have to return in our chapter on Geographical Distribution./38/From the foregoing considerations I conclude that the association of an organic being in any country with a different set of those beings, with which it comes into the most direct competition or dependence, as eminently favourable for natural selection for acting on whatever variations may occur, & so seizing on & filling up new places in the economy of that country. I look at this as so important as to be second only to variability, the basis on which the power of selection rests. Now an organic being could be particularly liable to become associated with new competitors, either when first by chance entering an isolated region into which few of its compatriots had entered; or when living there, after climatal or other changes had destroyed many of the inhabitants, & the isolation of the spot had checked free immigration of new & better adapted inhabitants. In this way, I think, isolation must be eminently favourable for the production of new specific forms. It must not, however, be supposed that isolation is at all necessary for the production of new forms;1 when a species spreads widely it will almost universally become associated with new competitors & there will often be some advantage gained by the selection of some modifications/38a/in its structure. I do not doubt that over the world far more species have been produced in continuous than in isolated areas. But I believe that in relation to the area far more species have been manufactured in, for instance, isolated islands than in continuous mainland.

The rate at which selection can act, depending on the chance production, as we must call it, of slight favourable variations; it might well happen, that of two forms undergoing modification, the one might beat out the other, if brought soon into competition; whereas if time had been allowed the other might have gained through selection some advantage, by which it could have held its own, when thrown into competition with the first. In this way, I can believe isolation may have played an important part in allowing two varieties from the same species to be considerably modified, before they are enabled to struggle with each other for existence./

1 See on this subject some excellent remarks by Dr Hooker in his Review of A. De Candolles Geographie Botanique in a note in Hooker's Journal of Botany vol. VIII p. 83. [p. 153]


38a A/Isolation, moreover, comes into play in lessening the amount of inter crossing, but here we are launched on a sea of doubt. That the majority of animals have their sexes separated or when united require the concourse of two individuals for the production of young is certain; & I think it has been shown in the third Chapter that occasional crosses will take place both/39/ with plants & animals far oftener than would at first be anticipated: but facts do not allow us to say that such occasional crossing is of universal occurrence. In those few cases, moreover, in which intermediate forms have been observed between two strongly marked varieties or reputed species, unfortunately we hardly ever know whether they are due to crossing, or to the intermedial action of external conditions & of the powers of natural selection. But as two individuals of most animals & some plants habitually unite for reproduction; this crossing will obviously retard, perhaps obliterate, the process of selection by dragging back the offspring of a selected variety towards its parental type. Let us suppose a stray gravid female or a pair of any animals to reach a small isolated island; if their offspring instantly varied & the old died, there would be no crossing, but such an improbable supposition may be quite disregarded; but if after several generations when the island was pretty well stocked some of their offspring slightly varied in any favourable direction; these would be selected or preserved, & though they would in all such cases be apt to cross with the parent‐form;/39A/yet the offspring from such crosses would have a stronger inherited tendency to vary in nearly the same favourable manner, as did the first variety; & natural selection would by preserving such individuals continually augment the tendency; until all the individuals might become insensibly modified in the same favourable manner. Just in the same way as a large herd of cattle may be modified by crossing even with a single bull of an improved shape & by the continued selection of the crossed offspring most like the Bull; & this would be much facilitated if the conditions of the country had <any> the weakest tendency occasionally to produce animals of the desired character./ 40/I am inclined to believe, that wherever very many individuals of a freely crossing & highly locomotive animal existed, the retardation of any selected modification from crossing would be so strong, that it could hardly be overcome, without indeed the tendency to vary in some particular direction was extremely strong. Hence I infer that some degree of isolation would generally be almost indispensable. This isolation may result from the nature of the area; or from the varieties as soon as produced, keeping to


a certain extent separate; & we shall immediately see that some partial separation of varieties, can & does take place in nature. That isolation from locality is important with highly locomotive, freely crossing animals, I infer from the fact, that with birds & mammals, the varieties & close & very doubtful species, (not here considering mere monstrosities, such as albinoes &c) generally inhabit distinct areas.

On the other hand, with organic beings, such as most plants, which do not cross for each birth or which are not highly locomotive so as to cross with individuals over a wide area, or which when favoured can increase at a great rate, I can well believe that a small body of any selected variety might be/40a/more quickly formed & hold their own against the ill effects of crossing, without being completely isolated. Though in such cases, isolation, at least partial isolation at first, would be favourable to their natural selection.

I have just taken the case of the selection of a variety of a freely‐crossing animal, on an isolated island; if we suppose the same process to be going on, in some favourable spot, but open all round to the inroads of the parent or unaltered form, there would be crossing with the parent form, not only at first in the actual birthplace of the variety, but all round its confines, where there might be neither the same tendency to vary nor the same place in the polity of nature open & ready to be filled up by the selected forms; in such cases, the retardation from crossing would be extreme./

41/In all these cases of crossing, we should remember the facts given in the third chapter, which convinced me that the offspring from two varieties have a greater amount of vigour & fertility, which would give them an inherent advantage, however slight, over the parent forms; tending thus to obliterate the variety, but on the other hand leaving descendants with some inherited tendency on which the same original cause of variation, we may believe, would be very likely to react. I am tempted to give an illustration of the effects which I should attribute to isolation in regard to crossing: in Madeira there are 20 land‐birds1 which breed in the island, & of which only one does not inhabit Europe or Africa; but besides these twenty, 26 stray species from the continent have been observed. Of these stragglers about 17, as I am informed by Mr. Harcourt, appear every two or three years, & some of

1 Excluding Grallatores; see Mr. E. Vernon Harcourt's excellent paper on the ornithology of Madeira in Annals & Mag. of Nat. History June 1855.I am infinitely obliged to Mr. Harcourt for having given me much valuable information on this subject.


them almost annually, & occasionally in little flocks, which has been noticed in the case of the starling, rook &c.—This being/ 42/the case, it seems most improbable that individuals of those species which breed on the island, should not likewise be occasionally blown there from the continent, although of course it is almost impossible to prove this. Therefore I should infer that the Birds of Madeira have not undergone modification, in the first place because the small island is well stocked with the same species, which have long struggled together in other & not very dissimilar lands, & secondly because, any slight tendency to change, which I believe would occur as the conditions cannot be identically the same, is checked by an occasional cross with quite unaltered forms having no tendency to vary;—the crossed offspring having greater vigour & hence a better chance of surviving.

What a contrast is presented by the Galapagos Islands, situated in a most tranquil climate, without any storms to blow birds from the mainland, which is nearly twice as far off; in this considerably larger group we have 26 land‐birds, of which 25 are endemic or peculiar to the archipelago! Of these 26 species, 8 belong to one endemic genus Geospiza, & five others belong to three sub‐genera closely allied to Geospiza; there are three closely allied mocking‐thrushes, & two tyrant‐flycatchers; so that I imagine that there were only/43/l4, perhaps only 11 original stray colonists, which arrived at different periods, & which had to fill the places in the economy of nature, occupied by 20 birds in the very much smaller island of Madeira; hence I suppose that nearly all the birds had to be modified, I may say improved by selection in order to fill as perfectly as possible their new places; some as Geospiza, probably the earliest colonists, having undergone far more change than the other species; Geospiza now presenting a marvellous range of difference in their beaks, from that of a gross‐beak to a wren;1 one sub‐genus of Geospiza mocking a starling, another a parrot in the form of their beaks. In this archipelago, moreover, there could be little retardation, or none, from crossing with unaltered forms from the continent.

I have remarked that in animals of which two individuals unite at each act of reproduction some degree of separation must be if not actually necessary, yet most advantageous. This may arise from a selected individual with its descendants, as soon as formed even into an extremely slightly different variety, tending to haunt a somewhat different station, breeding/44/at a somewhat different season, & from like varieties preferring to pair with each other.

1 Zoology of the Voyage of the Beagle, Pl. 36 to 44.


The following facts show that this is possible. After matching for experiment the most distinct breeds of Pigeons, the birds, though paired for life, seemed to me to show plainly a liking each for its own kind, so that I was led to ask Mr. Wicking, who has kept a larger stock of various breeds together than any man probably in Britain, whether he thought the different breeds, supposing that there were plenty of males & females of the same kind together, would prefer to match together; & he without having any theory unhesitatingly answered that he was convinced that they would:/ 44 v/it has, moreover, often been remarked that the Dovecot pigeon, the ancestor of all the breeds seems to have an actual aversion to the several fancy breeds.1 /44/It has been asserted2 that sheep of different breeds turned out together tend to separate, one sort taking to the more upland another to the lowland pastures: in the Shetland Islands3 two breeds of sheep have long kept distinct, the one haunting the mountain summits, the other the lower lands. In the Falkland Islands, Capt. Sulivan assures me/ 45/that the herds of white & brown cattle tend to keep separate, though neither are quite pure: the white haunt the mountains, & contrary to what might have been expected, they breed about a month earlier than the brown. In the New Forest4 the herds of brown & pale‐coloured deer have long kept separate, without intermingling. We have seen in the Catskill Mountains5 two varieties of the wolf hunting different prey. In N. America, Sir John Richard‐son6 says that "there are two well‐marked & permanent varieties of the Caribou deer that inhabit the fur‐countries; one of them confined to the woody & more southern districts, & the other retiring to the woods only in the winter & passing the summer on the Barren Grounds ":so that these annual migrations are different; the woodland variety retiring more inland in September, the other more southward. So in Tasmania, Mr. Gould informs me that there are two very slightly different varieties of [ ] one of which migrates & the other does not. Many instances could be given of

1 The Dovecote by the Rev. E. S. Dixon, p. 155.

2 [A group of four notes are missing at this place in the manuscript. Darwin published a revised version of the text on fols. 44‐5 in The Variation of Animals and Plants under Domestication, 1st ed. (London, 1868), II, pp. 102‐3. On the MS., he pencilled a 'U' over the portions used. It seems likely that he transferred this sheet of notes to his later MS. The citations are supplied from the published text, ch. 16, notes 6 & 7: 'For the Norfolk sheep, see Marshall's "Rural Economy of Norfolk", vol. ii. p. 136.']

3 ['See Rev. L. Landt's "Description of Faroe", p. 66.']

4 [Darwin attributed statement to Bennett: 'White's "Nat. Hist., of Selbourne", edited by Bennett, p. 39.']

5 [Note missing.]

6 Fauna Boreali‐Americana, p. 239, 250.


Birds of the same species inhabiting the same country, some of which migrate and some do not & which can be distinguished by very slight differences. In all such cases there would be some tendency for varieties having such different habits to keep distinct./

46/We have seen in the fourth chapter how the Common Ravens in Faroe drive away the pied Ravens, though sometimes pairing with them: the hooded & common crow haunt different districts which must check their crossing, for when they meet they often cross; but here we have to do with forms considered as species by most ornithologists. So again in India reputed species of Coracias, as I am informed by Mr. Blyth, intermix & blend on the confines of their range. So do, to give one instance in insects, the Carabus purpurascens of Western Germany & the eastern C. violaceus; at least where they meet there is a reputed third species C. exasperatus, which presents varieties undistinguishable from the two foregoing species.1

In the case of plants, as there is reason to suppose that in the majority of cases or at least in many cases only an occasional cross occurs, there will be less retardation in natural selection from this cause; more especially as any favoured variety might rapidly increase, & hold its own, on exactly the same principle, that seedraisers cultivate large plots of the same variety in order to get pure seed & lessen the ill‐effect of an accidental cross. A variety might, also, easily affect a slightly different station & seed/47/at a different period on a hill‐top for instance as is known often to be the case. Indeed there are innumerable instances of varieties of plants occupying particular sites or whole districts in the midst of the range of the species: thus the Centaurea nigricans, which Prof. Henslow, as we have seen has proved by culture to be only a variety of C. Nigra, occupies Hampshire to the exclusion of the common forms. The primrose & cowslip are sometimes found mingled though generally affecting slightly different stations. Although there can be little doubt that crossed varieties of plants will have an advantage from their inherent vigour; yet we shall see in our Chapter on Hybridism that there are some few curious but well ascertained facts showing that between certain varieties the pollen of one far from having a prepotent fertilising power on the other variety, is less influential. This leads me to remark, that although facts are greatly wanted to support the hypothesis, that sterility may supervene between varieties slowly formed by natural selection, I think I shall be able to show in the same chapter that

1 Erichson's Report in Ray Soc. Reports 1841‐1842. p. 161.


this is not in itself very improbable. At least I shall be able clearly to show that the difficulty in crossing species & the sterility of their off‐spring, by no means follows laws, as if simply/48/ordained to keep species distinct. On the hypothesis that sterility at last supervenes between varieties formed by nature, & called by us species, there will obviously be not the least difficulty, where this has happened in keeping such varieties for ever distinct: But on this hypothesis it may be very important that two varieties during the early formation until converted into species should be isolated or kept apart.

If in opposition to the general facts, given in the third chapter, there do exist organisms, of which two individuals never, or only at intervals of thousands of generations, unite or cross, then these cannot be kept uniform by intercrossing & selection cannot be thus retarded. In such cases the formation of new varieties & species will be stopped only from the absence of a new place in the polity of nature, from the want of variability, the variations not being inherited, the offspring taking after its grandfather or more remote ancestor, instead of its parent./

49/The number of the individuals of any species must form one important element in the formation of new species through natural selection. Several considerations incline me to lay considerable stress on this. We have seen in Ch. Iv, on evidence which seems to me satisfactory, that it is actually the common species abounding with individuals which oftener present varieties; & I there gave the obvious reason, that when many individuals existed there would be a better chance within a given period of variations arising, which might in some way prove beneficial to a selected variety. Just in the same way, as an/49A/agriculturist with a large stock of animals to work on, will have a better chance of gaining a prize for the standard of perfection than will one having only a few animals to select from: so again it is nurserymen, who raise large crops of our different flowers, who generally succeed in getting new & prettier varieties. As in each country all the variable forms are striving through selection to get the upper hand, there is not unlimited time for any one; & if/50<55>/any particular form be not modified it will run a good chance of being left behind in the race & being thus exterminated.

On the other hand a large number of individuals will apparently be injurious by favouring intercrossing with the selected forms. But we have not facts enough to guide our conjectures on these complex points: it may be that varieties, even amongst organisms which do not freely cross, generally arise on a small spot, partially


isolated, in the midst of the range of the parent‐species; & that they remain there till so much modified, as to spread largely by overcoming the parent form; sometimes crossing with it on its confines with selection continually acting on the crossed forms./

50 v/The supplanting of a parent‐species by a variety, which inherits all the characters & advantages of the parent, with some superadded advantages, will generally be an extremely slow process, as already explained in a former part of this chapter; for instance a variety more capable of enduring drought or resisting some insect will have an advantage over its parent only in the dryer spot, or chiefly where the enemy abound, yet during fluctuations of seasons, when very dry or when the hostile insect is unusually abundant, it will everywhere have an advantage, & tend to spread & supplant its parent./

50<55>/As perfectly isolated spots, such as islands, are often small, selection will be here retarded by the fewness of the individuals; but at the same time the competition will be less severe & there will be less danger of the extermination of a new variety from their being fewer forms to give rise to other new & victorious varieties or species. The greater number of open places in the polity of nature in islands, especially if stocked by chance colonists only at long intervals, could probably more than counteract the evil from the fewness of individual numbers;/51 <56>/Certainly, oceanic islands abound out of all proportion to their area, with endemic forms, in comparison with continents; but for reasons hereafter to be given, I suspect that the formation of species through nat[ural] sel[ection] has been slower. Considering the whole world, from the fewness of the completely isolated spots, & from the difficulty of the subsequent diffusion of new forms therein produced, such isolated spots, will probably not have played a very important part in the manufacturing of species.

Slowness of Selection.—From the various considerations now advanced, we can see that the formation of new species must be an extremely slow process. New places in the polity of nature for the occupation of the modified descendants of any species can be formed in most cases only at an extremely slow rate. Such new places will be due to physical changes, which will act either directly on the habits or requirements of the inhabitants, or, in a more important manner, indirectly, by causing the extermination or change in proportional numbers of some of the species; by immigration, also, not only will new places be opened to the immigrating species, but the economy of many of the old inhabitants may be


thus most seriously affected. All such changes will generally occur either very slowly or at long intervals. Secondly we require for the formation of new species, variability, & repeated variation of/52/ the most diversified nature, in order that changes of structure may occur in the right direction. Variability will depend on the conditions, more especially on changing conditions, to which the organic being is exposed; & the amount of variation will in part depend on the number of varying individuals. Selection acts only by the addition of infinitely small & numerous variations in some given & advantageous direction; & the process will be stopped by want of inheritance in any such characters & retarded by intercrossing.

I can well believe that many will exclaim, that these causes are amply sufficient wholly to stop all modification through natural selection: I do not believe so; but the result must be judged of by the general phenomena of nature. That changes will usually be extremely slow, I fully admit; & I am convinced that a fair view of the geological history of the world accords perfectly with an extreme degree of slowness in any modification of its inhabitants./

53/On the absence of intermediate forms or links between species of the same genus.—One of the most obvious difficulty on our theory, is if two or more species have descended from a common parent, & have been so slowly modified by numerous small changes, why do we not see all around us, or find embedded as fossils in the earth, innumerable varieties or the finest links closely connecting in an unbroken chain such species? This subject must be discussed here at some length, & likewise in our chapter on palaeontology. That such links must, on our theory, have existed, or do now exist, I fully admit. With respect to the nature of the links it is difficult always to keep clear of one source of deception, namely the expectation of finding direct links between any two species which we are considering: an example from our domestic breeds of pigeons will make what I mean clear; if we take a carrier & Fantail pigeon & consider their origin, we have not the least reason to expect graduated links between them, namely birds with longer beaks slightly covered with wattle & at same time with tail slightly expanded; but what we should find, if we had records of every bird kept by fanciers, during the last few thousand years, would be varieties intermediate in character between carriers & the rock‐pigeons, & between fan‐tails & rock‐pigeons:/53 v/The rockpigeon, being in its general characters intermedial between these two breeds, though not having a long‐beak covered with


any wattle, or having its tail at all expanded.1 /53/So again, still more strong, if we look to two species/54/remote in character; for instance the Horse & Tapir; from not having any idea, what on our theory, was their common ancestor, it is hardly possible to avoid the conclusion that numerous forms directly intermedial between these two must have existed: whereas it might well happen that the common ancestor was fully as unlike in many of its characters a horse or a tapir, as these two animals are from each other, yet being in its general organisation intermediate between them, though, perhaps much more nearly resembling one of these two genera, than the other.

From what we have already seen in this chapter, it seems probable that each variety, whether arising insensibly from the slow modification of the whole parent‐stock, or when formed in a separate area, or on some one spot within the same area with its parent, & subsequently spreading, will tend in the long run to supplant & exterminate its parent‐stock; for its formation is due to some new advantage gained under the conditions to which it is exposed, & it will generally largely inherit the advantages of its parent. This process will be continually repeated. In all these cases we could obtain a chain of intermediate gradations, only by discovering fossil remains of extinct forms; for of those living at one time & within one area we should see only the parent‐stock and one or two varieties, which if destined to become triumphant will increase in numbers & range & so ultimately supplant the parent; the parent, I may add,/55/being ranked as the variety, as soon as its range became less than that of the conquering variety. In the cases of insensible modification we should not at any one time see within the same area, a variety recognizably different from the parent, only mere individual differences.

Why in those classes, of which fossil remains are capable of being preserved & have been abundantly discovered, we do not find innumerable links connecting recent with extinct species, will be most conveniently discussed in our chapter on palaeontology. I think several fairly good reasons can be assigned. I will here only add that the whole force of the difficulty rests on the as‐sumption that our geological records are not only nearly continuous in time, but during each period nearly continuous in space; for otherwise varieties, which seem at first to be so frequently local

1 [Fol. 53 A] June 1858. I doubt whether I have got intermediate links yet clear. An animal rarely ranges over whole continents from climate—if it ranges to some extent then it will get into new conditions, but they will change rather abruptly, & only few cases—so we ought not to expect infinite gradation at same time only over moderate area, over which climate will let it range.


could only rarely be preserved. We should, also, remember that the definition of the term species is arbitrary; if an extinct form be found to a certain extent intermediate in character between two existing species, as is of such frequent occurrence; this may be fairly viewed on my theory as one of the intermedial links; the extinct form may have been the actual ancestor of our two species, or/56/more probably it may be an early & less modified descendant of the common ancestor, either in the direct line of descent of one of the two species or in a collateral & extinct line; 1 but all naturalists would rank our in some degree intermediate fossil as a distinct species, without they likewise discovered every intermediate grade between it & one of the living species; but that this should be asserted obviously requires the collection of very many specimens, which generally must have been embedded at slightly different periods & over a considerable area: supposing moreover this to have been effected, as occasionally has been the case, nothing more is thought about it; it is only the case of two forms at first ranked by our palaeontologist as two species & subsequently proved by a second palaeontologist to be merely varieties. Con‐chologists now doubt whether certain sea‐shells, living on the shores of N. America & Europe should be ranked as species or varieties; when the present day has become a miocene or eocene epoch is it probable that the palaeontologists of that far future epoch will find fossilised intermediate links between these now living & doubtful forms. He who does not expect this, has no right, as far as I can see, to expect now to find all the fossil links between a recent & closely allied fossil shell./

57/Looking now to the present time alone, if we travel for instance southward over a continent; we find at the point whence we start many species very common, but as we travel southward some of them become, more or less abruptly, rarer & rarer, till they disappear; but as they disappear, other closely allied or representative species, apparently filling nearly the same place in the economy of nature, take their place, at first being rare, & then becoming more or less abruptly, common. The two species, both comparatively rare, often commingle in neutral territory which

1 What I mean may, perhaps, be best understood by turning to the diagram printed at p. [236‐7] Let a10 & l10 be two now living forms with all their ancestors extinct. If A should chance to be discovered it will be strictly intermediate, though it might in many of its characters far more resemble a10 than l10 ; if a2 were found, it would be a lineal ancestor of a10 , & it would be largely intermedial between a10 & l10 , for it had diverged but little from the parent‐type A. So it would be with f3 or i3 , for they are early & collateral descendants from A, which have become extinct & transmitted no descendants.


is narrow. Every naturalist must have been struck with very‐many such cases amongst the birds & mammals of large continents: it may be observed with plants in ascending mountains, & with shells, as discovered by the dredge, in the descending depths of the sea.1 Why in such neutral or border territories without any barriers dividing them into sub‐regions and under apparently quite intermedial conditions do we not commonly find intermediate & graduated forms, connecting the two species, which are supposed by our theory to have originally descended from a common parent. That we most rarely find such forms is most certain; the two species, even selecting the most locomotive & freely crossing animals, on comparison, will be found in every single respect as distinct, as if specimens had been taken from the metropolis of each species. This/58/for a long time, formerly appeared to me a most serious difficulty; but the difficulty is largely due, as I believe, to common yet erroneous views on several points in nature.

In the first place we should be very cautious in concluding that because a continent is now continuous, it has remained in this state during the whole period of existing species. How many extensive areas have been greatly elevated within the period of existing shells; & what wonderful changes of level are shown by erratic boulders now scattered over the low‐lands & mountain‐summits, & which have been borne on ice‐rafts over the sea. What an enormous amount of recent depression of level may be inferred from the structure of living coral‐reefs. Even when we have no direct evidence, the form of the land sometimes leads to the conclusion, as in the case of the southern extremity of Africa, which is so extraordinarily rich in species, that it formed at no very remote epoch, a large archipelago of islands. It is probable that very many single volcanic islands have within the recent period existed as a group of islets; like those forming the little Madeira group which are inhabited by many distinct species & distinct varieties. Even when there has been no change of level, desert tracts may formerly have intervened, where the land is now continuously fertile. If we look at some of the larger volcanic/ 59/islands, or read Mr Webb & Berthelot's account of Teneriffe, we shall see that some of the valleys are almost as perfectly separated for some organic beings from each other by lofty spurs as if divided by arms of the sea.

In such isolated fragments of land, groups of the same species

1 See Prof. E. Forbes numerical observations on this head in his Report. Brit. Association on the Aegean Sea 18 [43] p. [174 & passim.]


might become differently modified, for they would be associated, especially after any changes in climate &c, with different sets & different proportional numbers of competing associates; & in such cases there could be formed no graduated intermedial links by crossing; nor would the more important conditions of life, in relation to other organic beings, graduate insensibly away between one of the isolated fragments of land and another. After reelevation, if the new forms had been sufficiently modified not to cross & blend with each other, each would spread as far as it could, & would mingle in the intermediate territories with other forms proceeding from different birth‐places. On each of the once isolated spots & there alone, we ought to find if our geological records were perfect, intermedial links between the new forms and the states under which they formerly existed.

Nor should we forget the facts, already given, of varieties of the most freely crossing animals, sometimes keeping apart, or breeding at different seasons &c, which would greatly lessen or prevent the formation/60/of intermediate links by crossing, though it would not often lessen the function of such links in relation to the inter‐mediate state of the conditions of life. Unfortunately in those cases, in which intermediate varieties have been found, connecting two races or two closely allied species, we hardly ever know, whether they have originated from crossing or from the direct & graduated action of climate, or from natural selection having fitted those intermediate forms for intermediate conditions of life. And our ignorance on this head greatly adds to the perplexity of this whole subject.

Although I believe the former broken & isolated state of parts of now continuous areas, & in a lesser degree the voluntary separation of the varieties of the higher animals, have played a very important part in the formation of species since become commingled, or just meeting in a border territory; yet I do not doubt that many species have been formed at different points of an absolutely continuous area, of which the physical conditions graduate from one point to another in the most insensible manner. But here lies a source of deception; we are so much struck with the evident manner in which the heat or moisture graduates away, in going from one latitude to another, that we can hardly avoid/ 61/overlooking the more important relations of organic beings to each other. We have every reason to believe, from what we see in gardens & manageries, that almost all organisms can withstand more heat, cold, moisture or dryness, than they are exposed to within their natural range; the definite limit to the range of most


species, under gradually increasing unfavourable conditions, being the presence of other competing forms better adapted to such conditions. So that in going for instance southward, the decreasing numbers & final disappearance of any species, is not by any means wholly due to the extremely gradual change of climate, but to the sudden presence of other competing forms, or the sudden absence of others, on which our species may chiefly depend for food; & the relation of the prey or fo6d will again depend on other organic beings; all nature being bound together in an inextricable net‐work of relations./61 v/A change in climate is very obvious, but the struggle for existence, depending on many contingencies & chiefly on other organic beings often far removed in the scale of nature is extremely obscure; & it is most difficult to keep this steadily in mind. Hence we have no reason to expect that in going southward that any one species ought to be insensibly modified in relation to the slowly changing climate, but chiefly in relation to each, new set of those organic beings, with which it comes into the most direct competition or stands in some relation; & the zone with really intermediate conditions, will depend in chief part on the range of other organic beings. As we see that the range of most organisms is in some degree defined, the species becoming, generally within a rather narrow space, rare & then quite disappearing, the zone with really intermediate conditions for any two species will generally be narrow, & therefore cannot support any vast number of varieties intermediate between such two species.

It comes to this that if the majority of the living forms in any country, as every one can see with care, are defined in their character and do not insensibly blend together, then the relations in range & in all other respects of any one form undergoing modification will tend to be defined: if organic beings had been in a wholly preponderant degree related to climate alone, then the range & specific modification of any form undergoing modification would have been related in an indubitably [?] <clear> manner to the insensibly changing climate./

62/Whether we ought, on our theory, to find many cases of two species closely connected by intermediate links in the narrow zone, which is really intermediate in all its relations to the two bordering species, must depend on whether at the same period many species are undergoing modification & on whether intermediate varieties, when once formed are likely to endure for long periods. Every fact in geology seems to show that species change very slowly & therefore I conclude that but few species are under‐


going modification at any one period; but as the process by our theory is excessively slow, some such cases ought to occur in every large area. I believe that they do, & in our Ch. iv several cases have been given of varieties connecting two forms, which have been considered by several naturalists as good species. The cases on record are probably few compared with those which exist in nature; for varieties or sub‐species or species (for there is no rule to follow in knowing what to call such forms) seem to be scanty in individual numbers, & hence would be observed, generally, only in countries which have been well worked.

The truth of intermediate varieties being individually rare is of importance to us. Mr./63/Wollaston1 has stated his opinion that this is the case, & he informs me that it is founded upon his observations on insects & land‐molluscs; and from his immense experience in collecting, few naturalists have a better right to express an opinion. I applied to Mr. H. C. Watson & to Dr. Asa Gray for their opinions on this head; as from their critical knowledge of the floras of Great Britain & the United States, everyone would place great confidence in their judgment. Both these botanists concur in this opinion, & Mr. Watson has given me a list of twelve nearly intermediate varieties found in Britain which are rarer than the forms, which they connect. But both these naturalists have insisted strongly on various sources of doubt in forming any decided judgement on this head./

64/Therefore, as it seems to me, we ought to expect to find only some few cases of intermediate varieties, inhabiting a narrow zone between the areas inhabited by any two species which they closely link together. But it may be asked, if varieties intermediate in character between two bordering species are ever once formed in such narrow intermediate zones, why do they not endure for as long a time as the species which they connect? & if they did so endure, cases of linking varieties could hardly fail to have become in the course of time with species after species undergoing modification far commoner in nature than they seem to be. I think some sufficient reasons can be assigned why they should not last for very long periods. As they inhabit a narrow zone (for we have seen zones with really intermediate conditions must generally be narrow) they can hardly be, & do not seem to be, numerous in individuals, so that they would be in some degree liable to extinction from great fluctuations in seasons, or any extraordinary increase of enemies. They are, also, bordered on each hand by forms adapted to the somewhat different physical conditions, to greater heat or

1 Variation of Species p. 5.


cold, moisture or dryness &c, to the coinhabitants of the bordering regions, so that if during [a] few successive seasons the temperature became higher or lower &c, they would/65/be liable to invasion on either hand; & if they had not great powers of endurance or of migration, or if any slight obstacles intervened to migration, they would be liable to be wholly extirpated. Moreover in the case of any two species having moderately wide ranges & commingling, as is so often the case, in a narrow border territory, if we suppose this border territory to have been once peopled by a chain of intermediate links connecting the two bordering species, we can see that these latter from having wider ranges would be more abundant in individuals, than the intermediate forms in the narrow intermediate zone; and on the principle already explained of a large number of individuals greatly favouring the production of favourable variations, one or the other of the two bordering species would have a better chance of being modified or improved so as to seize on the place of the intermediate links, & perhaps even to invade the territory of the other bordering species.

Finally, then, I suppose, that a large number of closely allied or representative species, now inhabiting open & continuous areas, were originally formed in parts formerly isolated; or that the varieties became in fact isolated from haunting different stations, disliking each other, breeding at different times &c, so as not to cross./66/That amongst those organisms, of which two individuals rarely (or never) unite for reproduction, that varieties have arisen on some one spot & from having some advantage over their parents either during occasional times or at all times has spread (perhaps sometimes crossing on their confines) & have supplanted their parent‐forms; & this would be most readily effected in small & isolated districts. That amongst organisms of all kinds, I suppose, that many species have been formed on different points of open & continuous areas, of which the physical conditions change insensibly, & that in such cases linking varieties have been formed, but that these would not tend to be infinitely numerous & spread over a wider space, for they would by no means be related solely to the insensibly changing climate, but in an equally or more important manner to the somewhat definite ranges of certain other organic beings. Such linking varieties (whether produced by the action of natural selection or of external influence in an inter‐mediate degree, or by crossing) seem, as might have been inferred from their theoretically restricted range, not to be abundant in individuals; & hence, I believe, would be apt to be exterminated by fluctuations of seasons, extraordinary increase of enemies &c,


and by the inroads of the bordering species, which/67/they link together. And lastly, I believe that these bordering species would have a better chance, owing to their greater individual numbers, of being modified & improved, so as to seize on the places of the intermediate & linking varieties. I am well aware, that if I wished to treat my subject as a mere advocate, it would have been better to have slurred over all these complex actions & contingencies, which apparently must affect the formation of new species, & of the relative importance of which I cannot judge; but my object is to point out all difficulties, as plainly, as lies in my power./

68/Summary of Chapter.—During the severe struggle for existence, to which all organic beings, owing to their high rate of increase, are exposed, during some period of their lives or during some shortly succeeding generations, Natural Selection acts by the simple preservation of those individuals which are best adapted to the complex contingencies to which all are related. Natural Selection can seize on plainly marked variations or on the slightest modifications, on mere individual differences even though inappreciable by the human eye, if in any way whatever advantageous to the individual, from its egg state to as late a period as the powers of generation last & can transmit any new character. As pecularities are often, probably generally, inherited at corresponding ages, it can modify the egg or seed, the larva, or pupa, without causing any change in the adult form except such as necessarily follows from correlation of growth. As peculiarities are often inherited by the corresponding sex, it can modify each sex in relation to the other; and the individuals of the male sex may be modified by sexual selection, enabling them to struggle for supremacy with other males, like natural selection modifies both sexes that they may struggle for supremacy/69/with other & distinct organisms. Sexual selection will also aid natural selection in giving most offspring to the most vigorous males, under whatever conditions they live. Natural Selection will scrutinize every habit, instinct, constitutional difference, every organ external & internal, will preserve the good, & rigidly reject the bad. It may pause in its work for thousands of generations, but whenever a right & fitting variation occurs, without error & without caprice natural selection will seize on it. From the several reasons already assigned, the process in all or nearly all cases will be excessively slow.

The greater the variability the better the chance of favourable variations. Individual differences seem to be of almost universal occurrence; a larger amount of variability apparently depends


mainly on changed conditions of life. The chance of favourable variations occurring will, also, stand in some close relation to the number of the individuals of the varying species. External conditions will, also, act directly on the individuals differently exposed & so modify them to a certain/70/limited extent: as will, also, use & disuse; but to these subjects we shall have to recur in a future chapter.

Intercrossing will prevent or retard the process of natural selection; but here we are involved in much doubt. Those animals, which move much about & unite for each/70 v/birth will thus be kept truest to their parental type; or if undergoing change will be modified in an insensible manner, without any recognizable variety being formed at any one period. It may, however, be otherwise in those cases, in which varieties of the most freely crossing animals,/70/from their very first commencement, haunt some distinct station or breed at different periods &c. Those organisms which rarely cross, & which are capable of increasing at a quick rate, may be formed on some one spot, & thence spread with little retardation from intercrossing.

The direction in which natural selection will act & its very power to effect any thing will mainly depend on there being places in the natural economy of any country not filled up, or not filled up as perfectly as possible. And this will depend on the number, nature, & relations of the other inhabitants of the region, in a far more important manner than on its physical conditions. Look/71/ at the woodpecker or the Bee or almost any other animal or on plants (though here the relations to other organisms, as we have seen in our last Chapter, are less plain, though not less certain) & see how clearly their structure is related to other organic beings: a woodpecker or bee may inhabit the hottest or coldest, the dampest or driest regions, yet how essentially similar is its whole organization. Hence I infer that the association of an organism with a new set of beings, or with different proportional numbers of the old inhabitants, as perhaps the most important of all elements of structural change. If a carnivorous or herbivorous animal is to be modified, it will almost certainly be modified in relation to its prey or food, or in relation to the enemies it has to escape from. Change of climate will act indirectly in a far more important manner than directly, namely in exterminating some of the old inhabitants or in favouring the increase of others. The immigration of a few new forms, or even of a single one, may well cause an entire revolution in the relations/72/of a multitude of the old occupants. If a certain number of forms are modified


through natural selection, this alone will almost certainly lead to the modification of some of the other inhabitants. Every where we see organic action & reaction. All nature is bound together by an inextricable web of relations; if some forms become changed & make progress, those which are not modified or may be said to lag behind, will sooner or later perish —

When a district is isolated, so that after any change in its physical conditions, new beings cannot freely immigrate, or enter only by a rare accident, the relations between its inhabitants will assuredly in time become greatly disturbed. Hence I infer that isolation would be eminently favourable to the production through natural selection of new specific forms. Isolation will also to a certain extent lessen the retarding influence of intercrossing. It will facilitate the supplanting of the parent type by its modified offspring, & lastly it will give time for a variety to be sufficiently changed so as/73/not to blend with, and to hold its own against, other varieties formed elsewhere, with which it may hereafter be thrown into competition.

As each new variety is formed through natural selection, solely from having some advantage over its parent, each new variety will tend to supplant & exterminate its predecessor. In regard to the intermediate links by which each new species must once have been closely connected with its parent, we could expect generally to find such only amongst fossil remains. In those cases however in which a species, ranging over a continuous area, is at the present day in the act of breaking up into two or more distinct species, we ought to find intermediate links in that narrow border territory which is really intermediate in all its organic & inorganic conditions; but we have no reason to expect to find many such cases, & we do find some. The intermediate links in such border territories, from reasons already assigned, would be liable to early extermination.

As a general rule we have seen that widely diffused species, abounding with in‐/74/dividuals, & belonging to large flourishing genera, are those which vary most. Of the varieties descended from any one species, the most divergent, or those which differ most from each other & their parents in all respects, will in the long run prevail, for they will be enabled to fill more & more widely different places in the polity of nature. It follows from this that the amount of difference which at first may have been very small between any two varieties from the same species, in each successive set of new varieties descended from the first two, will steadily tend to augment as the most divergent or different will


generally be preserved. From reasons already given, namely from the number of different places in the polity of any country not being indefinitely large, and from the individual numbers of each species necessarily being small where very many species exist, which will render such poor species liable to accidental extinction, and will check further modification,—the number of species inhabiting any country will not increase indefinitely; and as the most divergent are those which are the most likely to succeed, the intermediate forms, whether called varieties/75/or species of the same genus or of distinct genera, will tend to disappear.

The groups already large being those which vary most, & the principle of divergence always favouring the most extreme forms, & consequently leading to the extinction of the intermediate and less extreme, will taken together give rise to that broken yet connected series of living & extinct organisms, whose affinities we attempt to represent in our natural classifications. For all organic beings during all time seem to have been related to each other like twigs diverging from the same branch, branches from the same limb, and limbs from the same main trunk representing the common ancestor of a whole class of organisms, with many an intermediate branch and limb now lost.

Finally, then, in regard to the several contingencies favorable to natural selection, I am inclined to rank changed relations or associations between the inhabitants of a country from opening up new places in its polity, as the most important element of success. The amount of variability, which is largely contingent on the/76/number of individuals, as of secondary importance; though perhaps time being given for each new variety to be perfected before being thrown into competition with other varieties, may be almost equally important. A diminished amount of inter‐crossing is probably the least important element. But the subject is far too much involved in doubt for us to be enabled to weigh <to strike any balance between> these several contingencies./

76 A/Thus far I think we may with safety conclude that a large tract of land, stocked with nearly similar species, if by subsidence converted into a group of islands, like those of the great Malay archipelago, would in the course of time be eminently favourable for the production of new forms; & such archipelagoes are known to be extraordinarily rich in species. In the course of time, after our supposed subsidence, we might expect the destruction of some species through climatal changes and the occasional introduction of stray colonists; oscillations of level when downwards would cause more destruction, when upwards would extend the area, &


make new stations;—/77<58>/and these combined causes would act powerfully on the relations of the inhabitants to each other & would thus open new places in the polity of nature for natural selection to fill. Such changing conditions would also add to the variability of many of the organisms. In such large islands, there would be plenty of individuals to act on; intercrossing, at least on the confines would be prevented; & time would be allowed for the varieties in all the islands to be strongly marked & perfected so as to have a better chance of escaping annihilation, when thrown into competition with other & more favoured varieties, formed elsewhere. Those organisms which were originally common to the whole region, before the first great subsidence, might become converted into new .forms, whether called varieties or species, in each separate island, or in some of them remain unaltered, according to the nature of the organic forms with which they had to struggle in each island after it had undergone physical changes. If we now suppose our archipelago, through renewed elevation, to be reconverted into continuous land; then of the several forms produced from the same parent‐species in each former island, some would probably remain on the spot to which they had been adapted, some would spread, & if only slightly different might become blended by crossing with other varieties, or they would exterminate them, or if sufficiently distinct might live commingled with them. But in the case both of varieties & species, the most divergent, or those which had become most modified so as to fill the most diverse new places, would have the best chance of surviving.

[page] 275





The writing dates for chapter VII are not clear from Darwin's Pocket Diary, which lumps it together with the following chapter. Having finished chapter VI at the end of March, 1857, Darwin presumably commenced writing chapter vii at the beginning of April; for, to judge from the identical four illustrative facts appearing both in a paragraph in Darwin's letter of April 8, 1857, to Hooker1 and on folio 8 of chapter VII, it seems likely that both passages were written at nearly the same date. Probably by June 5 Darwin's writing had almost reached folio 105, for on that date he mentioned in a letter to Hooker2 that: 'I have been so much interested this morning in comparing all my notes on the variation of the several species of the genus Equus and the results of their crossing.' This topic, the next to the last discussed in the chapter, forms the subject of folios 105 to 113, and Darwin presumably started to write it up soon after he had reviewed his relevant notes.

Darwin had probably completed chapter VII by July 5,1857, when he mailed to T. H. Huxley a fair copy of folios 41 to 44 together with the following letter:

Down, Bromley, Kent

July 5 [1857]

My dear Huxley

Will you be so kind as to read the two enclosed pages as you said you would, and consider the little point therein referred to. I have not thought it worth troubling you with how far and in which way the case concerns my work, the point being how far there is any truth in MM Brullé and Barneoud. My plan of work is just to compare partial generalisations of various authors and see how far they corroborate each other. Especially I want your opinion how far you think I am right in bringing in Milne Edwards view of classification. I was long ago much struck with the principle referred to: but I could then see no rational explanation why affinities should go with the more or less early branching off from a common embryonic form. But if MM Brullé and Barneoud are right, it seems to me we get some light on Milne Edwards views of classification; and this particularly interests me. I wish I could anyhow test M. Brullé's doctrine; as in Vertebrates the head

1 ML, no. 56.

2 L & L, II, 101‐2; NY, I, 459.


consists of greatly altered Vertebrae, according to this rule, in an early part of the embryonic development of a Vertebrate animal, the head ought to have arrived more nearly to its perfect state, than a dorsal or cervical vertebra to its perfect state. How is this? I have been reading Goodsir, but have found no light on my particular point. The paper impresses me with a high idea of his judgment and knowledge, though, of course, I can form no independent judgment of the truth of his doctrines. But by Jove it would require a wonderful amount of evidence to make one believe that the head of an elephant or tapir had more vertebrae in it, than the head of a Horse or Ox. Many thanks for your last Lecture. How curious the development of Mysis!

yours very sincerely


Do you know whether the embryology of a Bat has ever been worked out?1

14 Waverley Place, July 7, 1857

My dear Darwin—

I have been looking into Brullé's paper, and all the evidence I can find for his generalization (adduced by himself) is contained in the extract which I inclose—Let us dispose of this first—Paragraph No. 1. is true but does not necessarily either support or weaken his view, which rests on paragraph No. 2.— Now this paragraph is a mass of errors—You will find in my account of the development of Mysis that the antennae appear before the gnathites are any of them discoverable—& Rathke states the same thing with regard to Astacus—and I believe it to be true of Crustacea in general.

The second statement, that the legs do not appear until the buccal appendages have taken on their adult form is equally opposed to my own observations & to <all> those of all who have worked in this field.

It would have been very wonderful to me to find Brullé resting such a generalization on such a basis, even had his two affirmations as to matter of fact, been correct. But as they are both wrong—one can only stand on one's head in the spirit—

Next as to the converse proposition marked 3). It is equally untrue—Mouths antennules backwards The appendages in Mysis

1 CD. MSS. box 145, pp. 152. W. R. Dawson, ed., The Huxley Papers a descriptive catalogue...(London, 1946), p. 26, lists this as Darwin letter no. 22 and gives the date July 5, 1857.


& in Astacus appear in regular order from before backwards wholly without respect to their future simplicity or complexity—and, what is still worse for M. Brullé, the ophthalmic peduncles, which as you know well are the most rudimentary & simple of all the appendages in the adult make their appearance at the most very little later than the mandibles & increase in size at first out of all proportion to the other appendages

M. Brullé bases his whole generalization upon what he supposes to occur in the Crustacea—whereas the development of both Astacus & Mysis—affords the most striking refutation of his views Tant pis pour Brullé'!

And now having brûler'd Brullé (couldn't help the pun) I must say that I can find no support for his generalization elsewhere—There are two organs in the Vertebrata where developmental history is especially well qualified to test it—the Heart & the Nervous system—both presenting the greatest possible amount of variation in their degree of perfection in different members of the vertebrate series—The heart of a Fish is very simple as compared with that of a Mammal & a like relation obtains between the brains of the two—[Darwin's comment: 'Good'.] If Brullé's doctrine were correct therefore the Heart & Brain of the Fish should appear at a later period relatively to the other organs than those of the Mammal—I do not know that there is the least evidence of any‐thing of the kind—On the contrary the history of development in the Fish & in the Mammal shews that in both the relative time of appearance of these organs is the same or at any rate the difference if such exist is so insignificant as to have escaped notice—

With regard to Milne Edwards views—I do not think they at all involve or bear out Brullé's. Milne Edwards says nothing, [CD.: '?See to this'.] as far as I am aware about the relative time of appearance of more or less complex organs—I should not understand Milne Edwards doctrine as you put it, in the paragraph I have marked: he seems to me to say that, not the most highly complex, but the most characteristic organs are the first developed —Thus the chorda dorsalis of vertebrates—a structure characteristic of the group but which is & remains excessively simple, is one of the earliest developed—The animal body is built up like a House —where the Judicious builder begins with putting together the simple rafters—According to Brulle's notion of Nature's operation he would begin with the cornices, cupboards, & grand piano.

It is quite true that "the more widely two animals differ from one another the earlier does their embryonic resemblance cease"


but you must remember that the differentiation which takes place is the result not so much of the development of new parts as of the modification of parts already existing and common to both of the divergent types.—

I should be quite inclined to believe that a more complex part requires a longer time for its development than a simple one; but it does not at all follow that it should appear relatively earlier than the simple part. The Brain, I doubt not, requires a longer time for its development than the spinal cord. Nevertheless they both appear together as a continuous whole, the Brain continuing to change after the spinal cord has attained its perfect form. The period at which an organ appears therefore, seems to me not to furnish the least indication as to the time which is required for that organ to become perfect

You see my verdict would be that Brullé's doctrine is quite unsupported—nay is contradicted by development—so far as animals are concerned—& I suspect a Botanist would give you the same opinion with regard to plants—

Ever yours faithfully


[Passage copied by Huxley from Brullé, Ann. sci. nat. ser. 3 zool. 2 (1844), 282‐3:]

1) En suivant, comme on l'a fait dans ces derniers temps les phases du developpement des Crustacés, on voit que les pièces de la bouche et des antennes se manifestent avant les pattes; celles ci ne se montrent que par suite des developpements ultérieurs —2) De leur côté, les antennes sont encore fort peu développées que les pièces de la bouche le sont déjà plus; enfin c'est lorsque les appendices buccaux ont revêtu la forme qu'ils doivent conserver que les pattes commencent à paraître. II en résulte donc cette conséquence remarkable [sic], que les appendices se montrent d'autant plus tôt que leur structure doittêtre plus complexe. On trouve, en outre, dans ces développements divers une nouvelle preuve de l'analogie des appendices. Ainsi les pattes n'ont pas de trans‐ formation à subir elles ne se montrent que quand les autres appendices ont déjà revêtu la forme de mâchoires ou d'antennes. 3) Done dans un animal articulé les appendices se montrent d'autant plus tard qu'ils ont moins de transformations à subir: c'est le comple‐ ment de la loi précédente." On peut par conséquent juger du degré d'importance et de complication d'un appendice par l'époque même a laquelle it commence à se manifester


[As immediate reactions to Huxley's letter, Darwin jotted down in pencil the following:]

VII, 41 A/There is only one point in which I cannot follow you. — Supposing Barneouds I do not say Brullés remark were true & universal, i e that the petal which has to undergo the greatest am't of development or modification begins to change the soonest from the simple & common embryonic form of the petals; then I cannot but think it wd throw light on Milne Edwards proposition that the wider apart amore different> the classes of animals, the sooner do they diverge from the common embryonic plan.—which common embryonic plan, may be compared to the similar petals in the early bud.—the several petals in one flower being compared to the distinct, but similar embryos of the different classes.—I see in my abstract that M. Edwards speaks of the most perfect & important organs being first developed & I shd have thought that the char[acteristic] organs wd be developed.

These comments Darwin developed in his reply of July 9, 1857, in which he thanked Huxley and mentioned his decision that he would 'not allude to this subject, which I rather grieve about, as I wished it to be true; but, alas! a scientific man ought to have no wishes, no affections—a mere heart of stone.'1

Also as if to reject this discussion on Brullé, Darwin took his pencil, altered the numbering of folio 45 to read '40 to 45', and he half cancelled the corresponding entry in his table of contents.



1/We have seen in our first & fourth chapters that changed conditions of existence, especially if accompanied with excess of food, seems to be a main cause of variation. But it must be owned that we are profoundly ignorant in regard to the first cause of variation. We do not know, whether the change in the conditions must be in some degree abrupt to cause much variation as d think> may, perhaps, be inferred from such changes alone affecting the fertility of organisms; or whether a much slighter change over a longer period would not be equally effective. We can assign no sort of reason why one organism varies greatly under domestication, & why another varies hardly at all: why in a state of Nature, most, but not all the species of certain whole groups are excessively variable; & we do not even know whether this latter sort of protean variation is the same as ordinary variation. Ignorant as we thus

1 ML, no. 52.


are in regard to the primary cause of variation, yet when varieties do appear, we/2/can sometimes, in a very dim & doubtful manner point out some of the laws governing the changes in structure, as was attempted in the first chapter. Here I shall further treat on this subject; & compare domestic varieties with those naturally produced, & both together with the forms called by Naturalists species.

If it can be shown, even partially, that species differ from each other in a similar manner & apparently according to similar laws, as do varieties, it strengthens our view, that species are only strongly marked varieties with the intermediate gradations lost. The old cosmogonists believed that fossil shells, resembling but not identical with living shells, had been created within the solid rock; & they asked why God should not have thus formed them? The paleontologist would probably now reply, that we see in the fossil & living shell plain evidence of similar structure, & therefore he would affirm that their origin & formation must have/3/been alike. So I believe that the similarity of the laws in the formation of varieties, & in the so‐called creation of species, indicates that varieties & species have had a like origin; & not that the one has been due to the nature of surrounding causes, & the other to the direct interposition of the Hand of God.—

The laws which obscurely seem to govern variation, & which were briefly alluded to in our first Chapter, together with some others not then mentioned, may be grouped under the following heads. (1) The immediate action of the <external> conditions of life. (2) The effects of habit & disuse (3) The correlation of growth, namely the manner in which the modification of one part affects another part, either through quite unknown relations, or by such relations as that called by Geoffroy St. Hilaire balancement or compensation, by which the large development of one part is supposed to cause the reduction of another; or by such as the early arrest of development in a part,/4/—the period, at which any modification supervenes, any early change of structure affecting parts subsequently developed;—multiple parts strongly tending to vary in number;—homologous parts varying in a like manner & tending to cohere &c.— (4) Parts developed in any species in an extraordinary manner <& rudimentary parts> tending to vary. (5) Distinct species presenting analogous variations; & a variety of one species, resembling in character another species: reversions to ancestral forms. (6) The distinctive characters of varieties more variable than specific characters; specific characters more variable than generic: secondary sexual characters variable./4a v/Lastly,


varieties occuring most frequently amongst those species, which are most closely allied that is those which fall into the larger genera—<also amongst the more common species, (or those which are the most vigorous in any region & are consequently most abundant in individual members?) also amongst those which have widest ranges.>It, also, seems that the species in the larger genera, are apt not only to be the most variable but to have the widest ranges & to be the most abundant in individuals. From the facts to be given under the last head we gain, if the view that varieties & species do not essentially differ be true, a slight but deeply interesting prophetic glance into the far future of the organic world; we can dimly see whither the forms of life are tending; where about in the great scale of Nature new species will arrive, & where old forms will tend to disappear./

4 bis/The immediate or direct action of external conditions. When we find that certain individuals of a species placed under peculiar conditions, are all or nearly all affected in some particular manner, especially if all are soon affected, & more especially if the modi‐fication does not seem of any use to such individuals, so that probably it is not the result of selection, then I should be inclined to attribute the effects to the direct action of the conditions of existence. But it is most difficult to eliminate <the power of> selection:/4 bis'/thus we have reason to believe that climate produces some immediate & direct effect on the woolly covering of animals; but when advantage of this is taken by man & a long‐wooled animal is produced by artificial selection, it would be wrong to attribute such wool to the immediate action of climate; & so it would be in the case of natural/4 bis/selection.

From the facts given in the first chapter, I think we may in some case attribute greater size, early maturity, & the nature of the hairy covering &c to the immediate action of food & climate./

5/The time of flowering in plants, & of breeding in animals no doubt is affected by climate; & a more curious difference has been observed in a Lizard, namely that it is oviparous in dry Northern Chile & viviparous in humid Southern Chile.1 But in such cases we can seldom, perhaps never, separate the various elements of change; we cannot tell whether it be cold or damp or lessened or different food which has produced any given result. The wretchedly dwarfed & often distorted state of the shells in the Baltic may be safely attributed to the brackish waters; for the shells grow more perfect as they approach the open sea. Few Naturalists, however,

1 M. Gay, Annal. des Science Nat. Zoolog 2d series. Tom V. p. 224.


would rank such shells, or the stunted plants on a lofty mountain, as varieties: But I can hardly see where to draw a line of separation: I presume that it is assumed that these dwarfed states are not hereditary; & this would be a valid distinction; but we have previously seen how difficult it is even to conjecture what is inherited in a state of nature./

5 bis/In some cases of shells having an immense range, as that of the common Buccinum undatum from the North Cape to Senegal,1 which presents a perfect series of intermediate grades between the extreme northern & southern forms; I presume that the modification may be attributed to temperature: but in cases, where we have a strongly marked variety, at the northern & southern ends of the range, with a narrow zone inhabited by an intermediate form, of which I have observed marked examples with cirripedes, it would be rash to attribute the difference to climate, for natural selection probably has come into play & according to my views is in the act of making two species. In regard to colour, Forbes2 says "it is easy for the practised con‐chologist to distinguish specimens of the most painted shells, gathered on the southern coasts of England, from those taken on other parts of our shores:" So it is in a marked degree with the tints of certain shells, specified by him, which range from the shallow laminarian zone into great depths./

6/In the case of insects, if we read the accounts given by Oswald Heer3 & Wollaston on the changes which the same species undergo in ascending mountains, & in approaching the pole, generally but by no means always becoming darker‐coloured we can hardly avoid attributing the change to climate. So again, Mr. Wollaston4 clearly shows that residence near the sea‐coast tends to make insects lurid, & affects them in various ways. In regard to Birds, it will suffice to quote Mr. Gould,5 whom no one will accuse of running varieties together, & he says that birds of the same species are brighter coloured in the interior of continents than near the coast, which he attributes to the greater clearness of the atmosphere far from the sea.6 /6A/It is well known that in animals with fur, the skins are much more valuable, the further North they are

1 [Deshayes,] Annales des Sciences. Nat. 2d series Zoolog. Tom v. p. 291.

2 Report Brit. Assoc. on British Marine Zoology 1850 p. 254.

3 Quoted by Bronn. Ges[ch]ichte der Natur B. 2. s. 96.—and Mr. Wollaston on the Variation of Species p 39‐41.

4 Ib. p 57‐64.

5 Zoolog. Soc. Meeting May 8. 1855.

6 From the character of the species, not varieties, inhabiting very dry districts, as the Galapagos Archipelego,—the deserts of Peru & Northern Patagonia, it would appear as if dampness was an element in the bright colouring of birds & insects.—


collected.1 In plants several cases are on record of the same individual or all its seedlings changing in a few generations, without the aid of selection, the tint of its flowers when brought from its native home into our gardens.2 /6 A v/Cold seems to lessen the intensity of the colours of flowers, as is asserted to be the case with some on high mountains, & as has been observed by the Dutch cultivators with their Hyacinths.3 /6 A/Moquin Tandon gives some instances of plants acquiring by variation more fleshy leaves, when growing near the sea.4 It has often been asserted that the same plant is more woolly when growing on mountains than on lowlands, & Moquin Tandon5 asserts that this change occurred with several species from Pyrenees when placed in the Botanic Garden at Toulouse: but Dr Hooker informs me that the Anthyllis vulneraria is glabrous in the Alps & woolly on hot dry banks:/7/ moreover Dr Hooker after tabulating some Alpine floras does not find that in truly alpine species the proportion of woolly plants to be large. He is inclined to believe that dryness has a stronger tendency to produce hairs on plants.6

Most of these variations are apparently of no service to the organisms thus characterised, & therefore not having been affected by selection, may be wholly attributed to the immediate action of the conditions of existence. Small & unimportant as are the modifications, it deserves notice, that they almost invariably tend in the same direction with the characteristic differences of the species peculiar to the districts under comparison. Thus, how incomparably more beautifully coloured are the sea‐shells of the Tropics compared with those of the cooler temperate regions. It is, also, well known that shells confined to great depths are almost colourless. Alpine species of Coleoptera are generally dark‐coloured; & Mr. Wollaston expressly states as every collector must have noticed that beetles confined to the sea‐coast are generally "lurid‐testaceous or pale brassy"./7 bis/Species of plants living near the sea frequently have fleshy leaves; those of dry & hot countries woolly leaves; those in tropical regions brilliantly coloured flowers. Arctic quadrupeds are thickly clothed with fur. The species of birds, which are confined to the interior of continents, according

1 Bell's British Quadrupeds on the Ermine Stoat: see Bronn's Ges[ch]ichte der Natur. B. 2. s. 87.

2 Dr. Hooker on the Climate & Vegetation of the Sikkim Himalaya p 49. in regard to a Rhododendron; & see, also, Mr. H. C. Watson's account of the Azorean Myosotis.

3 Moquin‐Tandon Eléments de Teratologie p. 42.

4 Elements de Teratologie Vegetale p. 73.

5 Id. p. 62.

6 Moquin Tandon gives, also, several facts in corroboration of the same view. Ch. 2.—p. 65.


to Mr. Gould, are more beautifully coloured than those which inhabit the coasts & adjoining islands. In all these cases, the species, which according to our views are only strongly marked varieties, are naturally affected in the same manner, but in a stronger degree, as the forms admitted by naturalists to be mere varieties.1

In some cases the action of external causes, which I have called immediate, from its influencing apparently without selection, all the individuals exposed to it, /8/seems indirect in its influence; by which, I mean, that very different conditions will produce the same result. Thus Dr. Harvey, the highest possible authority on sea‐weeds, says2 that the Fucus vesiculosus at the Canary Islands, where the heat is too great for it, appears under a nearly similar form, as in the Baltic where it is injured by the brackish water & mud; & he adds that no one "would be prepared for the fact that the heat of the tropical sea would exercise the same trans‐forming power on a particular plant as the mud & fresh‐water of a colder climate." In other cases, also, it would appear that an organism presents a nearly similar range of variation under what‐ever condition it is exposed: thus to give a very trifling instance, the common Polygala has blue, white & purple flowers in the cold humid island of Faröe in 62° n.3 in England & southern Europe. The Juncus bufonius which ranges from the arctic regions to the equator "in every region seems to present the same variations in its size & branching.."4 These cases, which I believe to be not common, though Dr. Hooker thinks a good many could be collected,/ 9/lead us back to the perplexing facts of polymorphous species & genera, discussed in the fourth Chapter; they show us how ignorant we are on the subject of variation, & how prepotent an influence, the organisation of the species has on the causes, whatever they may be, of variation.

Upon the whole, I think, we must attribute some effect to the immediate action of external conditions; but I am inclined to think it is very little. Innumerable instances could be given of organisms of all kinds exposed to an immense range of climatal & other conditions,5 & yet not varying in the least, & although, as

1 [Here Darwin added in pencil:] 'If Buckman did not use selection, here allude to his facts as strongest evidence of direct action of food & cultivation.'

2 Sea‐side Book 1849 p. 66. [Cf. ML no. 56: Darwin letter of April, 8, [1857] and L & L II, 90‐1 Darwin letter of April 12 on theme of this paragraph.]

3 Landt, Description of Feroe p. 180 see Herbert in Hort. Journal vol I p 48 (?) or ask Bentham Lat. of Zante.

4 D. Don on Indian Junci in Linn. Transactions vol. 18. p. 324.

5 Perhaps one of the most striking cases, is that given by Göppert (Wiegmann's Archiv fur Naturgesch. 1837 p. 210) of unaltered species growing on hot soil


Mr. Wollaston has remarked, we ought by no means to infer because these causes have no influence on one species, they will have none on another; yet I think we may to a certain extent be guided by <the frequency of such cases of non‐variation.> As I consider those forms which are ranked by most Naturalists as independently created species, as only strongly/l0/marked varieties, the high degree of generality of the fact, that the tropical & temperate, & temperate & arctic zones, are inhabited by species, often closely allied, <of the same genera,> as strongly confirmatory of the view, that climatal conditions have no great influence on organisation; but to those, who look at species as independently created, these latter facts will have no weight.

Acclimatisation.—Though climatal conditions may have no great influence on organisation or visible structure, yet it is notorious that the great majority of organic beings are adapted, within moderately narrow limits, to the climate <of the regions> which they inhabit. When, therefore, a Naturalist meets an animal with a very wide range, for instance the Puma in the reeking hot forests of Central America, on the dry deserts of Patagonia, in the damp cold woods of Tierra del Fuego & up to the limits of eternal snow on the Cordillera, he is much surprised; for he is accustomed to meet for instance, one species confined to the Tropics, another to the temperate & another to the cold regions; his surprise is, also, increased, from falsely attributing (as I believe) far too much weight/11/to the relations between climate & visible structure; climatal conditions are manifest; but the more important conditions determining each creature's power of getting food & escaping dangers are obscure in the highest degree. Nor must we overrate the degree of adaptation in the constitution of each living being to the climate of its own restricted home: when a new plant is introduced from a foreign land, until actual trial we cannot closely tell what range of climate it will endure. Even plants confined to certain islands, & which have never ranged, as far as we know beyond the narrow confines of their home, are found to endure very different climates: look at the Snowberry tree (Chiococca

above burning coal; & other similar cases given by Humboldt in regard to certain grasses on the edges of hot‐springs. Many plants have enormous ranges (see Hooker Introduct. New Zealand Flora p. x) & remain unaltered; some range from the base of the Himalaya & other mountains up to an immense height. A land‐shell, the Nanina vesicula ranges from the hot plains of India up to 10,000 feet (Huttons Chronology of Creation p. 202) on the Himalaya, where a Toad has an immense range (Hooker Himalayan Journals vol. 2. p. 96)—For wide range of insects see Mr. Wollaston's excellent discussion, p. 29‐31 in his Variation of Species.—It would be easy to accumulate innumerable examples.—


racemosa) how difficult to eradicate from our shrubberies, who would have ever supposed that it had been naturally confined to the West Indian islands? <Those who think each species created, as we now see it, will. Must we say that such island plants were created for the prospective chance of the island becoming joined to the mainland & then the plants in question spreading?—>

Nevertheless there can be no question that very many, probably most organic beings are pretty closely adapted to their own & no other climate; & if the species/12/of the same genus are descendants from one common parent, many of them must in the course of ages have become accustomed to very different climates. Is this possible? I think the following facts, though few from the nature of the case, show that plants at least do become in some degree acclimatised. Dr Hooker states1 that he has found a great difference in the hardiness of individuals of several Himalayan plants, depending upon the height at which the seeds were gathered: he instances seedling Pines, which taken at the height of 12,000 feet, were hardy in England, whilst those from 10,000 feet were tender; & so there is a great difference with the Rhododendron arboreum according to the height at which the seeds have been collected. Mr. Thwaites, the curator of the Botanic Garden at Ceylon, whose accuracy is well known, writes to me, that he finds "that individuals of the same species are acclimatised to different elevations,—being more & more impatient of cultivation at any station, according as they have been transported to it, from stations of greater & greater altitude." Again Mr. H. C. Watson has cultivated a variety of a British Lysimachia brought home from the Azores, & found it was decidedly tender.2 /

13/I think, also, that there can be little doubt that the varieties & sub‐varieties of our domestic animals & plants become in a slight, though very slight degree, acclimatised each to its home: I infer this from the caution incessantly given in works ancient & modem of agriculture, <in all countries,> even in the old Chinese Encyclo‐pedias, not rashly to change the breed of any animal or race of plant from one to another district, more especially in wild moun‐tainous districts./13 v/The horses from Algiers stood the climate of the Crimea better than those of Europe: Merino‐sheep from the Cape of Good Hope "are far better adapted" to India than the same breed from England:3 the cactus introduced into India from Canton, Manilla, Mauritius & the Kew Botanic gardens were un‐distinguishable to the eye, but the Cochineal insect perceives a

1 Flora of New Zealand Introduct. p. xi.

2 Phytologist. vol 2. p. 976.

3 Royle Productive Resources of India, p. 153.


great difference, for it will flourish only on the Indian plants, supposed to have been formerly imported by the Portuguese.1 / 13/Different dogs have extremely different capacities for standing heat, but then their probable origin from distinct species renders this case of no value. No one, I presume doubts that the Negro & Laplander have very different constitutions in regard to climate.

Again we have some instances, but here also from the nature of the case but few, of animals naturally extending their range, though we do not know how far the individuals actually become acclimatised to their new homes: thus Audubon gives several instances of Birds, which undoubtedly/14/have extended their range much further northward during late years in the United States.2 Thus, also, there can be little doubt that owing to the introduction of cattle, a vulture (Cathartes atratus) in S. America now ranges many hundred miles further south than it originally did three centuries ago.3 The innumerable instances of plants, not cultivated by man, & of some few animals ([)insects for instance (]) not domesticated, which have been naturalised through his agency in many countries under different climates show clearly that organic beings can adapt themselves, whether or not becoming acclimatised, to new conditions. Look at the common mouse & rat which have run wild on the hottest & dryest volcanic & coral islets under the equator, & in Faroe in the north & at the Falkland Islands in the south; it is opposed to all probability that these species had aboriginally nearly so wide a climatal range. The Fallow‐deer is feral in Barbuda in the West Indies, & can live on the shores of the Baltic; but it is superfluous to give other instances.—

These facts lead me to believe, that many organic beings by slowly extending their range, can become acclimatised. Whether the acclimatisation is/15/effected by mere habit, or by the natural selection of individuals born with a constitution, fitted either to greater heat or cold, it is impossible to say: probably both actions concur. The spreading of any organism, in those cases in which there is no physical barrier, will depend, mainly, on the nature of the other inhabitants, that is whether there be any place which it can seize in the polity of nature. If there be such place animals & plants will, sometimes extend their range, even although the climatal conditions are in some considerable degree unfavourable

1 Id.p.59

2 [See Appendix for a group of Darwin's reading notes etc. attached here.]

3 Zoology of the Voyage of the Beagle [Part III, Birds] p. 7. The Rio Negro is about 500 miles south of Monte Video, where according to tradition they did not formerly exist, having come there from still further north.—


to them, as we see with the Elephant reduced in size in India north of Lat [ ]; & with the Capercailzie,1 in Northern Scandinavia; & with the dwarfed trees in the northern parts of Scotland & the United States. But the spreading will, also, depend upon how closely the organism has become rigidly acclimatised to the conditions of its native home. Nearly all our domestic animals & some plants have great climatal flexibility of organisation, as we see in their cultivation & in their becoming feral under such different climates; & in their generally retaining perfect fertility under sudden & great changes of climate. Although in many cases we do not know/16/what were the parent forms & what their natural ranges, or how many aboriginally distinct species are now blended together in our domestic races; yet if we look at the whole body of our domestic productions or even if for instance we run through the shorter catalogue of our domesticated Birds—there can be no doubt that they live under a much greater diversity of climate than do an equal number of organisms taken at haphazard in a state of nature. The arguments given towards the close of our second Chapter have convinced me that our domestic productions were not aboriginally selected from having this constitutional flexibility, though doubtless they are far more useful from possessing it; half‐civilised man could neither know, nor would he care, whether the animal which he was taming or the plant which he was cultivating was thus constituted; he would not care for this more than did the Laplander when he domesticated the Rein‐deer, or the inhabitants of the hot deserts of the East when he domesticated the Camel. Hence then, I conclude, from the very general, though as we have just seen, not/17/universal constitutional flexibility of our domestic productions, either that organisms in a state of nature possess this same quality far more generally than we should expect from their natural ranges, or that the simple act of domestication gives this constitutional capacity for bearing climatal changes in a high‐degree. It may be doubted, whether if the wild parent‐form or multiple parent‐forms of the Horse, the goat the Fowl &c the maize, tobacco, rice, wheat &c were suddenly carried from their wild native state into the various climates under which the domestic races now flourish, they would be prolific & healthy. If this doubt be correct & an organic being subjected to domestication or change of some kind, has its constitutional adaptation to special climate so far broken down, that

1 L. Lloyd Field Sports of the N. of Europe Vol I p. 284. in Lapland this bird seldom weighs more than 9 or 10 pounds, whereas in the southern parts of Sweden it not seldom exceeds 17 pounds in weight.—


it acquires a general degree of flexibility, then we can perhaps understand a statement insisted on by M. Alph. De Candolle, which long appeared to me very strange;—namely that with the progress of knowledge, plants in a state of nature are found to divide themselves into two opposed categories, "les unes locales et ordinairement tres locales, les autres tres repandues."1 <For according to this notion, as soon as a plant begins to spread, it would be in predicament of a domesticated production & would gain flexibility of organization & might spread very far.— >/

18/Finally then I conclude that most animals & plants are capable of spreading beyond their present confines, when no physical barrier is opposed to their progress; the main & general check being the presence of other & better adapted organic beings; a second check being their native acclimatisation but that this may be overcome by habit & natural selection; & that when overcome, the being tends to gain a general degree of flexibility of organisation, allowing it to spread very widely, as far as climate is concerned; its means of obtaining food & escaping danger being then the sole but powerful checks to extension. On this view, such facts as the former existence of a rhinoceros & elephant adapted to a glacial climate—the wide extension of man himself,—of his domestic productions & of those accidentally transported by him —are not exceptions to a general law: it is only that these animals have lost their special acclimatisation & have regained their normal constitutional flexibility./

19/Effects of use & disuse on structure.—That constant action will increase the size of a part & that this increase becomes hereditary, I think can hardly be doubted from the facts given in the first chapter for instance the size of the mammae in our cows & goats when habitually milked, the more muscular stomach of owls & gulls fed on vegetable matter; & the great weight of the bones of the legs of the domestic duck &c. On the other hand from disuse parts decrease in size, as we see in the wings of the duck & of the Cochin China fowl. (?) Nor is this at all surprising because as we have seen parts become visibly more developed, or atrophied from accidents & operations, during the life of an individual.

1 Geographic Botanique p. 484

(a) [On verso of this folio, Darwin pencilled the following remarks:' Col. Sykes. Fowl from India, native home, bred readily in this country—screw loose—we must say that act of domestication by itself in a being never transported to other country gives flexibility to endure climate!

(a) A screw loose—this fact of when adapted & enabled to beat two sets of organisms is enabled to beat many more sets, must be far more important element —yet above must come into play.']


In a state of nature, the same variety cannot be observed during very many generations; the conditions of existence when they change change most slowly; <& if a sensible modification did occur in any form, that form would naturally be considered as a distinct species,> hence we cannot recognise the effects of use & disuse in varieties in a state of nature. But if we look at species, as only strongly/20/marked varieties, we frequently meet with structure analogous to that resulting from disuse under domesti‐cation. Thus the great logger‐headed Duck1 of Tierra del Fuego, which so much surprised the old voyagers, & which I have often watched, cannot use its wings more than a fat Aylesbury duck, & is under any extremity incapable of flight. Feeding, as it chiefly does in the great beds of floating kelp, it does not require wings to escape from danger, to which it would hardly be more exposed, than the ocean‐haunting Penguins. The islands of Mauritius, Bourbon, Rodriguez, of North, South & Middle New Zealand, & of Philip all have had birds, incapable of flight; & when we remember that no beast of prey inhabited these islands, & that ground‐feeding birds usually take flight only to escape dangers, I should attribute, their almost wingless state to disuse./

21/In New Zealand, the birds incapable of flight, belong, as we know from Prof. Owens wonderful discoveries, to 3 or 4 very different orders; & therefore I should infer that at least so many birds had colonised these islands ages ago, & had since given birth to the score of birds in this state now inhabiting these islands.2 But as several of these belong to the ostrich family it may be supposed that one at least of the original colonists, arrived, we know not how, at these islands in an already almost wingless state. But in regard to the other almost wingless birds of New Zealand & of the other specified islands, it seems to me probable that they arrived by flight & that their wings since became almost atrophied from disuse in their new & protected homes. In ostriches which inhabit continents & great islands, as we see that they/21A/ can escape danger by their fleetness, & in close quarters by their dangerous kicks, quite as well as any small quadruped, disuse together with the increasing weight of their bodies may well have rendered them incapable of flight. The fact of so many birds with imperfect wings inhabiting oceanic islands, naturally leads us to/

1 Micropterus brachypterus Eyton.. Zoology of Voyage of Beagle. [Part III] p. 136.

2 Nov 21/57 conversation with Owen I think 3 types Rallidae—Aptornis either distinct or a Parrot—& Dinordinae [sic], which includes Apteryx [?]—If there could be winged Dinordinae—these might have come by flight—If Dinordinae close to Rallidae or other winged Birds then perhaps always wingless—Though I should think even Struthioidae were once winged.—


22/Mr. Wollaston's1 remarkable discovery of the frequently apterous condition of the Beetles at Madeira; for no less than 200 species out of the 550 coleopterous inhabitants of this island, have their wings in various stages of reduction & are incapable of flight; & this undoubtedly is a wonderfully large proportion./22 v/The more wonderful, as winged Beetles would during the whole existence of Madeira as an island have had a better chance of getting there than aboriginally wingless species; just on the same principle that many European birds have by their wings reached Madeira; & that the only mammals existing there are the winged Bats. We see clearly the tendency in the beetles of Madeira to be wingless in the fact mentioned by Mr. Wollaston, that 17 genera here have wingless species, which genera usually have winged species in other parts of the world. Moreover of the /22/29 endemic genera, that is genera strictly wholly confined to the island, no less than 23 have all their species incapable of flight! Still more remarkable is Mr. Wollaston's conviction, & no one can be a more capable judge, that some few of the very same species, common to Europe & Madeira, are wingless on this island & winged on the continent; & he gives full details in regard to three of them. Here, then, I may add we have another case of varieties in a particular locality marking the species, which are exposed to the same conditions; or as I should look at the case we here have permanent & strongly marked varieties, called species, very naturally possessing the same character with the less‐strongly marked forms, called by naturalists/ 23/varieties.

In regard to the origin of the apterous condition of the Madeiran coleoptera; as Mr. Wollaston repeatedly remarks, that the Beetles on the more exposed rocks lie concealed during the almost incessant winds, & immediately appear in numbers, when the winds lull & the sun shines, something may, perhaps, be attributed to the mere disuse of their wings just as with the males of the silk‐moth. But I am inclined here to lay far more stress on the principle of selection with its antagonist action of destruction. Beetles from not being powerful flyers are very liable to be blown out to sea, as I have <repeatedly> witnessed, & this would naturally happen far oftener on a small island than on a continent; therefore on an island active individuals with a strong tendency to use their wings would be oftener destroyed, & sluggish individuals with their wings reduced in size, however little the difference might be, would in the course of ages be oftener preserved, & would leave offspring with the same inherited tendency; & this process ultimately,

1 Insecta Maderensia p. xii. and Variation of Species, p. 82 to 87.


through continued selection, might render the beetles quite safe from being blown to sea, by rendering their wings rudimentary. As the danger would be obviously greater, in the smaller & more exposed islets, I‐ have ascertained through Mr. Wollaston's kindness,/24/that on the Dezertas, a mountainous rock near Madeira, four miles long & about three‐quarters in breadth, there are 54 Beetles; & that of these, 26 are winged & 28 wingless, which is a proportion one‐fourth larger, than the Dezertas ought to have had in accordance with the proportions of the winged & wingless coleoptera in the whole archipelago./24v/In working out the proportions, the insects believed by Mr. Wollaston to have been introduced by the agency of man have been left out on both sides.—On the Dezertas, however, the number was only three. If I had contrasted the Beetles on the larger island of Madeira itself, with those on the Dezertas alone, the proportions would probably have been greater than that given in the text./24/From the Salvages, a little rock, between Madeira & the Canary islands, six Beetles are known to Mr. Wollaston, & four of these are apterous: at Kerguelen island, Dr. Hooker found only one beetle & one moth, & both were apterous.

Any beetle which from not being a ground‐feeder or which absolutely required wings for any purpose, would on the principle above explained run great risk of utter extinction; without indeed its conditions of life were so highly favourable that it could bear great occasional loss from being blown to sea. Now one of the most remarkable features in the entomology of Madeira, strongly insisted on by Mr. Wollaston1 is the entire absence or extreme rarity of certain whole Families & /25/great genera of Coleoptera, which abound in species on the mainland of Europe under a similar climate. Thus to take the Families alone of Cicindelidae there is not one species; of the following great groups only one in each up to the present day, has been discovered, namely Buprestidae, Elateridae, thalerophagous Lamellicoms, Telephoridae, Oedemeridae, Silphidae & Pselaphidae. No one but an entomological collector will fully appreciate this most remarkable fact. In considering this list it occurred to me that these very Families (the remark does not apply to all the genera) were exactly those which from their habits of life do actually use their wings far more than other Coleoptera: accordingly I enquired from Mr. Wollaston whether this was not the case, & he has gone through the whole list &, with the exception of the Pselaphidae, says that undoubtedly it is so. Therefore I think we may with some safety conclude that

1 Insecta Maderensia p. x.


a vast majority of those Beetles, the habits of which did not allow them to subsist without wings & therefore did not allow them to become apterous through selection & disuse have been exterminated: & this conclusion supports the former one on the origin of the apterous species./

25 A/On the other hand, in those classes of insects which are not ground‐feeders & are rapid & powerful flyers, this very power might save them from utter destruction, by allowing them to battle against the wind. Such insects might even have their wings enlarged by natural selection; and Mr. /25/Wollaston1 says he is by no means certain that this is not actually the case with the Lepidoptera & some flower‐feeding beetles, which if they are to live at all, must have wings & use them. Hence I can see no difficulty in two directly opposite processes going at on the same time with different members of the same great class; some having their wings reduced by selection & disuse, others having them increased, —just as Pigeon Fanciers during the few last centuries have decreased & increased the length of beak of the tumbler & carrier pigeons, both derived from the same stock. The turning point will have been when an insect first arrived on the island, whether, according to the nature of its food, its individual numbers were increased by its flying less & so running less chance of being blown to sea; or flying better so as to conquer the winds./

26/Such, I believe, to be the explanation of the conditions of the wings of the insects on Madeira; but it must be plainly confessed, that doubt is thrown on it, from the fact, discussed by Mr. West‐wood2 that in many parts of the world, there are insects belonging to various orders, of which individuals may be taken either winged or perfectly apterous; of this fact the common Bed‐bug is a well known instance. It has been thought that the wings are developed during hot seasons, but the evidence seems to me hardly sufficient. The facts anyhow plainly show that there is something in regard to the wings of insects, which we do not in the least understand.

Loss of tarsi.—

We will now turn to another somewhat analogous case: Kirby has remarked3 that in certain Scarabaeidae, (dung‐feeding beetles) the anterior tarsi of the males are generally broken off: he examined seventeen specimens in his own collection " & not a single one had

1 Variation of Species p. 87.

2 Modern Classification of Insects. Vol [II, pp. 473, 158, 431.] Also Mr. Wollaston Variation of Species p. 43‐45.—

3 Introduction to Entomology vol. 3. p. 337. [Actually p. 338.]


a relic of the anterior tarsi; "/26 v/& in Onitis apelles they are so rarely present, that the tarsi in this beetle have been supposed by some authors not to exist./26/ I remember formerly, when largely collecting in this Family, having made the same observation; & Mr. F. Smith of the British museum tells me that he, also, has observed it. This frequent, & almost habitual loss of a portion of the/27/front limbs of the males is not common to all the genera, having the same general habits, for it is not observed in Copris or Onthophagus./27 v/I do not suppose that the tarsi are lost by the males fighting: at least in Lethrus, in which the males are known to fight furiously, the tarsi were quite perfect.—/27/If mutilation were inheritable, as many authors believe,—if cutting off a dog's or cat's tail tended to make them produce tailess offspring,—then we might have expected some result from this almost habitual loss of the tarsi; but I cannot believe in mutilation being inherited. Nevertheless so constant a loss clearly shows that the anterior tarsi are of not much service to the insect & therefore probably are not much used; & disuse, I do not doubt causes atrophy & is inherited. Now in the genus Onitis above referred to & likewise in Phanaeus,1 members of the Scarabaeidae, the tarsi are "very slender & minute", & may be said to be quite rudimentary; indeed in the Brit. Mus. I could not find any specimen of Phanaeus with tarsi, & in another genus, Ateuchus, (which includes the great sacred beetle of the Old Aegyptians) it is well known that the tarsi of the front legs are absolutely deficient/27 v/It would be easy to bring forward cases of the atrophy or entire disappearance of parts apparently from disuse; but as these occur in all the individuals of the species, & as I cannot illustrate them by analogous losses merely in individuals or varieties, I have not given them in the text. Many parasitic Crustaceans have their limbs atrophied when attached for life to fishes. In another totally distinct Kingdom, there is a striking case in as much as it occurs in nearly full‐grown individuals in the Pholas lamellata; this shell has been described as a distinct species, but has been shown by Mr. W. Clark2 to be the half grown animal of Pholadidea papyracea, which after it has domed its shell, does not any longer require its foot for boring, & consequently the whole large muscular foot is "depauperated &

1 M. Brullé (in Annal. des Science. Nat. 2 series Zoolog. Tom 8. p. 284) asserts that in Phanaeus the males are deprived of tarsi, whereas the females almost always have them. He adds that in Onitis, the females of some of the species have tarsi, whilst in other species neither the males or females have them. I do not know whether M. Brullé was aware of the frequent accidental loss of the tarsi in several other coprophagous genera.—

2 Annals & Mag. of Nat. History vol 5. 1850. p. 12: see also Dr. Fleming's British Animals [p. 451.]


finally obliterated."/27/Hence I am inclined to attribute the very small size or loss of the tarsi in these beetles, wholly to disuse./

28/Blindness.—I have one more class of facts of the same nature to bring forward. It is well known that moles & some allied genera owing to their subterranean habits have either very small yet perfect eyes,/28 v/as in European mole, in which the eyelids are hidden under thick fur, & are one‐third of the size of the head of a middle‐sized pin;/28/or, eyes excessively minute, & fairly covered over by the hairy skin, so that if they have any vision at all, it must be confined to the dimmest perception of mere light.—The burrowing Aspalax, (a Rodent & therefore belonging to another order of animals) is in the same predicament; its eye being excessively minute & covered not only by skin, but by a tendinous expansion. Now in S. America there is a very common rodent, the Tucu‐tucu (Ctenomys Braziliensis), more subterranean in its habits even than the mole: I heard of a Spaniard who had often caught them, & without my making any remark, he stated that "invariably very many are found blind": he procured me some specimens, which I kept alive, & one of them was evidently stone‐blind; I preserved it in spirits & Mr. Reid dissected the eye, & found that the blindness had apparently been caused by inflammation of the nictitating membrane. As blindness tends to/ 29/cause atrophy & as diseases of the eye are believed to be strongly hereditary (especially with horses), I can see no difficulty in believing that the eyes of the Tucutuco might be reduced by disuse & disease to the state of those of the Aspalax: yet as inflammation of the eyes must be injurious to any animal, & as the Aspalax can live in its blind state, it may well have been that the absolute closing of its eyes was effected by the continued selection of smaller & smaller eyes & more closely shut eyelids.—

It is well known that in the deep caves of Styria there are many blind insects, & Crustacea arachnidae & a reptile the Proteus: in the caves of Kentucky there are, also, blind insects crustaceans, fish, & a Rat./29 v/The various stages of abortion of the eyes in these Kentucky animals is very curious: some have no trace of an eye, some have a rudiment, & the Crustacean has the footstalk for the eye without the organ,—it has the stand for the telescope without the instrument./29/Now as the existence of useless eyes could hardly be injurious to these animals, I should attribute their blindness to simple disuse. Although it is trenching on a distinct subject, I may remark, that many of the cave animals


of/30/Europe & No. America, though exposed to closely similar conditions of existence, are except in their blindness very little allied.1 According to my views, these animals were not created in their respective caves, but American animals must have got into the Kentucky caves, & European animals into those of Styria, slowly penetrating, century after century into the profounder abysses, & gradually have become blind by disuse: they would, also, become modified in any other way, through selection gradually fitting them for their new & dark homes. Now in regard to the Kentucky caves, Prof. Dana informs me that the Crustacean is2 /

31/In the discussion on the Madeiran insects, I remarked that it was quite possible that natural selection might at the same time be enlarging or reducing the wings of different insects of the same class. In the caves of Kentucky I think we have evidence of something analogous in regard to the eyes of the animals; the contest, however, being here between selection enlarging & disuse alone reducing these organs. The blind cave Rat, instead of having rudimentary or no eyes, has eyes of an immense size; & Prof. Silliman Jun. who kept this animal alive, thought that after a period & when accustomed to the light, it acquired some slight degree of vision. Now if we may suppose that this animal did not habitually live in the utterly dark parts of the caverns, we may suppose according to our principles, that the individuals with infinitesimally larger eyes & a more sensitive optic nerve had been continually selected, until some American rat from the outside world, had been converted into this strange inhabitant of darkness, with its large [?] eyes, blue fur & long moustaches.3 /

32/In the depths of the ocean, & in deep & dark wells some Crustaceans as Calocaris & Niphargus are blind.4 Now though I am not aware that any Fish inhabiting very deep water is normally blind, yet it seems to bear on the above facts, that the Gadus lota5 at the depth of 100 fathoms has its air‐bladder frequently atrophied, often accompanied by total blindness. On the other hand, it has been "remarked that fishes which habitually descend

1 Trans. Entomolog. Soc [ ].

2 [Here Darwin broke off in mid‐sentence. On the lower half of the sheet he pencilled the following memoranda: 'Fish & Rat.—

In the caves of Styria I have failed in finding out the affinities of the insects, but one or two are even thought to be only varieties of European insects.—Proteus has American & European species Look in Dict Class. for range of each genus & write to Dana to ask']

3 [Darwin left a pencilled question mark at the end of this sentence of which the last half dozen words were scrawled in pencil.]

4 E. Forbes. Report Brit. Assoc. 1850 p. 254.

5 Prof. Jurine in Mem. de la Soc. d'Hist. Nat de Geneve Tom 3. p. 149.


to great depths in the ocean have large eyes".1 And one most remarkable fact is on record, (which is worth giving, though of a most perplexing nature.) M. Eudes‐Deslongchamps gives with great detail two cases2 of eels taken from wells about 100 feet in depth, which had their eyes of immense size, so that their upper jaw in consequence projected over the lower. But here comes the remarkable fact the first specimen was shown to Agassiz, & he/ 33/ thought it was specifically identical with the common Eel. One of the wells was within the precincts of a prison; & it seems impossible to conjecture how the eel got in; & it seems, moreover, quite incredible that such an alteration could have supervened during one generation: it is, also, most improbable that there should be a race of subterranean eels, for, I believe it is well established that the eel invariably breeds in the sea. Surrounded with difficulty as this case is, we apparently have in the large eyes of these eels, & in the blind Gadus from the deep parts of the lake Leman, a parallel case to the opposite condition of the eyes of the Kentucky cave‐fish, crustaceans &c contrasted with the large eyes of the cave‐Rat./

34/Correlation of growth.—In the first chapter I briefly alluded to several laws, appearing to govern variability. These laws are most imperfectly known; & I will here recapitulate them, adding a few remarks, more especially in regard to a comparison of the structure of those forms, recognised as varieties, & those which are generally supposed to have been formed by distinct acts of creation. Physiologists admit a principle, called "nisus formaticus", which repairs, often in a wonderful manner, accidental injuries; & I think we may infer, that if any part were greatly increased or altered in form by continued selection, this "nisus" would give corresponding size to the vessels & nerves &c, without the direct aid of selection though this might always come into play.

I alluded in the first chapter to the mechanical action, attributed by Vrolik, to the shape of the bones of the pelvis of the mother on the head of the human embryo in different races./34 bis/In various groups of Birds, the form of the kidneys differs remarkably, & M. St. Ange3 attributes these differences to the varied shape of the pelvis, which would seem to have acted mechanically on them; & the form of the pelvis probably stands in direct relation

1 Sir John Richardson Encyclop. Brit. [8th ed.] art. Fish. ['Ichthyology'] p 219.

2 Mem de la Soc.Linn. de Normandie vol. 5 1835. p. 47, and vol 7. 1842. [p. xxix.]

3 Annales des Science. Nat. 1 Ser. Tom 19 p. 327.


to the different powers of locomotion. So again in Snakes, Schlegel1 has remarked that the varied positions of the heart & of the lungs, the riband‐like liver with the gall‐bladder removed from it, the anomalous position of the kidneys & organs of generation, all stand in direct relation to the shape of the body, formed for crawling, & to the manner of swallowing: how much of these remarkable modifications ought to be attributed to direct selection acting on slight variations in these important organs, & how much to the indirect, & almost mechanical action of changes in the form of the body & of the mouth, it would be very difficult to say.

In our first chapter I showed that Isidore Geoffroy St. Hilaires law of the multiple parts whether physiologically important or unimportant varying much, in number, holds good both in regard to varieties/35/& to species; I presume that this stands in relation to a greater or less amount of plastic matter, out of which the multiple organs have to be developed, having been accumulated at an early embryonic age.

Homotype <Homologous> parts tend to vary in a similar manner, owing, it may be supposed, to their similarity at an early embryonic period; or one part tends in its variation to imitate another part of <the same homotype> nature. Thus the great anatomist Meckel, has insisted, as stated by Isidore Geoffroy Saint‐Hilaire2 "que les muscles du bras, de I'avant‐bras et de la main ne s'ecartent presque jamais de leur type normal par le nombre, et la disposition de leurs parties, sans tomber dans les conditions qu'offrent dans l'etat regulier, les muscles de la cuisse, de la jambe & du pied; et reciproquement."

Homologous parts both in animals & plants seem to have a strong abnormal tendency to cohere or unite;3 & the variations thus caused, can often be so closely paralleled by normal structures, that it is difficult to believe that the parallel is accidental./

35a/Moreover it would appear that multiple parts are especially apt to be variable in form as well as in number. M. Isidore Geoffroy4 insists on this; & M. Moquin‐Tandon5 observes that "les organes répetés le plus de fois sont aussi ceux dont le developpement est le plus variable." As this "vegetative repetition", to use Prof. Owen's expression, is a sign of a low or little specialised organisation, the foregoing remarks on the variability of multiple parts seems

1 Essay on Serpents, Engl. Translat. p. 26

2 Hist. Gen. des Anomalies Tom. I p. 635.

3 Isi. Geoffroy. Hist. Gen. des Anomalies. Tom I. p. 541, 545.—For plants see M. Moquin‐Tandon Elements de Tératologie Vegetale. 1841. p. 248, 267.

4 Hist. des Anomalies Tom I. p. 60, 638 650. Tom. 3. p. 456.

5 Teratologie Végétale p. 124.


to fall under an observation often made by naturalists that the lower animals are more variable than the higher. And with plants, Dr. Hooker remarks1 that "variations in the floral organs are apparently more likely to occur the less the individual parts deviate from the normal type, the leaf; as if the more complete adaptation to a special function rendered them less liable to casual variation." Or as the/35B/case may be put, as long as an organ had to act in many ways, its exact form would probably not signify; just as a knife for cutting all sorts of things, may be almost of any shape, but a cutting tool for some particular object had best be of some particular shape; so with an organ as it began to be specialised through natural selection for some particular end, its particular structure would become more & more important; & this same natural selection would tend to keep the form constant by the rejection of accidental deviations, excepting indeed such few as tended to improve the organ; <& these it would seize on;> whereas until the exact shape or structure of the organ became important for its function natural selection would hardly come into play in checking any slight fluctuations in its form./

36/There can be no doubt, that many parts of the organisation of every living thing are correlated together, so that if one part changes, another part will tend to change, by a bond which we can sometimes see dimly but often not at all. Some instances were given in the first chapter of variations thus related; for instance hair & teeth believed by most physiologists to be of an homologous nature in the so‐called Turkish or naked dogs: now if we take a general survey of the mammiferous Kingdom, the two orders which are most anomalous in their teeth, namely the Edentata & Cetacea, are certainly most anomalous in their dermal covering; as we see in whales, contrasted with seals, & in the armour of the armadillo Mylodon &c & Ant‐eater. I presume the remarkable fact of the seedling Cuscuta not having cotyledons, though germinating in the soil, stands in direct correlation with the mature plant being parasitic on the elaborated sap of other plants & so not requiring leaves.

As I have said the bond of correlation is often quite hidden from us; remember the blueness of the eyes & deafness in cats,—the nakedness of young pigeons & their colour—/37/constitutional differences & complexion &c. So in the gravest & in unimportant monstrosities Is. Geoffroy1 remarks "que certaines anomalies coexistent rarement entre elles, d'autres fréquemment, d'autres

1 Flora Indica Introduct. p. 29. & Ranunculaceae p.2.

2 Hist. des Anomalies &c Tom 3. p. 402


enfin presque constamment, malgré la difference tres grande de leur nature, et quoiqu'elles puissent paraitre completement independantes les unes des autres". In looking at organic beings in their normal state one incessantly sees throughout whole groups of animals & plants, having quite different habits, two parts of their organization having no apparent connection, yet almost identical throughout all the species: but it is most difficult in such cases to know whether there is any correlation in the parts. The mere fact of the community of structure in the two parts throughout many allied forms is no proof whatever, according to our theory, of any correlation of growth, for it may be wholly due to community of descent. And in the ancient parent of the allied forms, the two parts may have acquired their present structure & apparent con‐nexion, from having been independently modified for separate purposes through natural selection. <Just as the Fancier is now making by artificial selection the beak of his tumbler‐pigeons very short, & the feet very small, without, perhaps, there being any correlation whatever in the growth of these parts.> /37 v/But it would be rash even in this case positively to assert that there was no correlation; for it is well known that acephalous monsters are especially liable to have imperfect feet./

38/On the other hand, when in a group of species, the same part or organ differs in each, such differences are very generally, perhaps universally, accompanied by at least slight differences in the surrounding parts. Thus Prof. Owen/38 v/remarks1 that "he knows of no analogy in the whole mammalian series that would justify a belief' that the lower jaws should not be different in two genera, characterized by a difference in the number of their teeth./38/Such differences in the connected parts, when slight & apparently unimportant in function, may in all probability be attributed to correlation of growth.

As we can hardly suppose that internal & structural differences in the fruit on the same individual plant can be of use to the species, we must attribute the differences in the pericarps,—in their shape, their appendages, & even in the ovary itself with its accessory parts—2 of the central & marginal florets of many com‐positae, to some correlation of growth. Possibly it may be a case of compensation, yet there does not seem to be any direct/38 b/ relation between the state of the fruit & the presence or absence

1 Proc. Geolog. Soc. 1842. p. 692.

2 H. Cassini in Annal. des Sciences Nat. 1 series. Tom. 17. p. 387.—C. C. Sprengel in his Das Entdeckte gives figures Tab xx of the achenium in central & ray florets of Picris (Helminthia) & Tussilago; Thrincia offers another instance.


of the ray‐like corolla in the outer florets. Possibly the differences may be related to the mutual pressure of the flowers: at least the Decandolles1 are inclined, in the case of certain states of Dianthus polymorphus, to account for the abortion of the anthers & the greater length of the style "to the lateral compression of the flowers in the cymes." But it seems extremely doubtful whether this explanation can be applicable to the differences in the internal structure of the seed, which has been observed in the inner & outer flowers in some Umbelliferae: thus in Hasselquistia2 the seeds of the ray‐flowers are orthospermous & those of the disc coelospermous; & analogous differences have been observed in the Coriander; it is, I may add, to show how important these differences of this kind are that Decandolle has founded on them the classification of the order./38 b v/It is by no means the Umbelliferae with the densest heads, which have the corolla most frequently developed in the external flowers; & in the carrot it is the central flower which is developed in an unusual manner. Perhaps, this whole class of facts are in some way related to nutrient flowing more freely to the central or exterior florets, & may be connected with causes which tend to produce peloria in the line of the axis. But in some instances I suspect, that C. C. Sprengels view that the exterior florets are developed & one bit of calyx in Mussaenda to make flower conspicuous to insects.3 /

38b/Certain Leguminosae bear on the same plant flowers of two different kind, & with the/38 c/flowers, as I am informed by Mr. Bentham, the pod sometimes differs. Ad. de Jussieu has described4 two different kinds of flowers borne by certain species of Malpighiaceae; one flower of the ordinary sort, the other without a corolla or with a mere rudiment of it, two ovaries without a style &c; so that Jussieu remarks in these degraded flowers "the greater number of the characters proper to the species, to the genus, to the family, to the class disappears, which thus laugh at our classifications". Nothing is known of the use or meaning of the two kinds of flowers borne by these & other plants, but I presume that the internal & structural differences in the imperfect flowers, which, however, seed as well & often better than the perfect flowers, can be of no service to the plant, & must be due to some correlation of growth.5

To give an instance of a correlation, which I should attribute

1 Mém. Soc Phys. de Genève. Tom. 9 p 78.

2 Tausch in Annal. des Sciences Nat. 2 series Bot. Tom IV. p. 41.

3 [See appendix for later memoranda on a folio (watermarked 1858) also numbered 38.]

4 Archives du Mus. d'Hist. Nat. Tom. 3. p 82.

5 [Here Darwin later pencilled: 'Nectary & petiole [?] of column in Pelargonium'.]


wholly to natural selection, & not to the laws of growth;—winged seeds are never found1 /39/in an indehiscent fruit; or, as I should put the case, seeds could become winged through natural selection only in fruit which opened, so that the seeds which were blown furthest got an advantage over those less fitted to be acted on by the wind, & thus gradually became winged; & this could never happen through natural selection in a fruit which did not open.

Those who have studied monstrosities believe,2 that any affection of a part developed during the early life of the embryo tends to modify other parts of the organization subsequently developed. This seems so natural that it can hardly be doubted; & hence the later formed structures as they are necessarily subjected to the influence of all previous abnormal changes, are the most liable to monstrosities & variations. On the same principle monstrosities of axis of the plants almost always affect the appended structures.3 We may infer from these considerations that the same cause tending to produce a monstrosity or variation would produce different results according to the period at which it acted on the embryo. Perhaps we may to a certain extent understand those sudden & great variations,/40/called by horticulturists 'sports', whether in the bud or seed, by supposing that a modification takes place at a very early age of development & greatly disturbs the whole organisation. I think there can be no doubt that in those animals, which live an independent & active life in their larval condition, any great modification at this period would sensibly alter the structure of the mature animal; & as many insects, when mature, live for a very short time, & never even feeding, have nothing to do but procreate their kind, much of the difference between species & species, may well in many cases be almost wholly due to correlations with their larval condition: on the other hand modifications in the mature state will almost necessarily have been preceded by modification at an earlier age. It must not, however, be supposed that a great amount of change, caused by the continued addition through natural selection of small changes, of any one organ, or at any one period, necessarily causes a correspondingly great change in all other parts of the organisation; or at all other periods of life; for I think the facts given in the first chapter on the changes due to selection under domestication, show that such is not the case./

1 Alph. De Candolle in Annales des Scienc. Nat. 2 series. Bot. Tom XI. p. 281.

2 Isidore Geoffroy St. Hilaire Histoire des Anomalies. Tom 3. p. 392 Andral was strongly of the same opinion.—

3 Moquin‐Tandon Elements de Teratologie Vegetale. p. 113.


41/M. Brullé1 in a memoir on the embryonic transformations of the Articulata2 insists "qu'un appendice se montre d'autant plus tot, qu'il doit acquerir un development plus complet". In another part he strongly reurges the truth of this proposition, & asserts that the converse holds good. It would almost appear according to this view as if more time were required for the growth of a part which has to undergo greater embryonic modi‐fications, & that consequently its development had to commence earlier. M. Barneoud3 has shown something analogous in plants having irregular flowers; for he finds in an Aconite, in Orchidaceae, Labiatae & Scrophulariaceae, that at a very early age the petals are equal & similar; "mais bientôt on remarque entre dies, une difference de grandeur d'autant plus forte et plus précoce/42/que la fleur est plus irreguliere a l'etat adulte". <So that in these cases, the parts which have to undergo most modification from their archetype, grow quicker than the less modified parts.>

Prof Milne Edwards4 makes a different but somewhat analogous comparison: he does not compare parts in the same individual developed from similar & homologous elements, but the same functional system in quite different groups of animals; & he seems to think that according as the organs in question are most developed in any class, the earlier they appear in the embryo in that class: thus he contrasts the circulatory system in the Vertebrata, in which it is so highly perfected, with the same in Annelids. Indeed the main basis of all affinities, so strongly insisted on by Milne Edwards in this paper & elsewhere,5 seems to hang on the same principle,—namely that the more widely/43/two animals differ from each other, the earlier does their embryonic resemblance cease; thus a fish on the one hand, & mammals together with birds on the other hand branch off from the common embryonic form at a very early period, whereas mammals & birds being more closely related to each other than to fish, diverge from each other at a later period. This seems to accord with M. Brullé's principle

1 [Darwin had a fair copy made of the text running from the top of fol. 41 to the first paragraph ending on fol. 44. A few slight additions or changes he made on the fair copy are incorporated in the text given here. The fair copy was sent to Huxley, whose answering letter and Darwin's additional comments are in the introduction to this chapter. Presumably because of Huxley's criticism of Brullé, Darwin later wrote at the top of fol. 41: 'Do not copy this Heading or pages' and, changed the number of fol. 45 to read: '40 to 45' evidently intending to omit fols. 41 to 44.]

2 Annales des Sci. Nat. 3 Series Zoolog. Tom. 2. p 273, 282‐283.

3 Annal. des Scien. Nat. 3 series Bot. Tom. 6. p. 270, p. 287.

4 Annales des Sci. nat. 3 Series Zoolog. Tom. 3. p 176.

5 Annal des Science. Nat. 3 Series Zoolog. Tom. I p. 65.


that the more each part is changed from the common archetype the earlier it is developed; for as a fish differs in nearly all its organization from a mammal, more than a bird differs from the mammal, the fish as a whole would have to be differentiated at an earlier period than a bird. So with Mr. Barneoud's case, if we look at an irregular flower at a period between its earliest condition & maturity, the more irregular & modified petals from having grown at a quicker rate may be said to have been earlier developed. I presume that actual time is not referred to in any of these cases; only relative time one organ being compared with another; for, as is well known, the heart of the chick arrives at the same stage/ 44/of development with that of a mammal in a far shorter actual period of time.1

If the foregoing principle be really true & of wide application, it is of importance for us; for then we might conclude that when any part or organ is greatly altered through natural selection it will tend either actually first to appear at an earlier embryonic age or to grow at a quicker rate relatively to the other organs than it did before it had undergone modification.: consequently, as we have seen in the case of monstrosities this early formation will tend to act on the other & subsequently developed parts of the system. This same principle would, also, probably play an important part in the following so‐called law of balancement or compensation of growth./

40 to 45/Compensation or Balancement: Geoffroy St. Hilaire & Goethe brought forward about the same period this law, which has been admitted by some naturalists & utterly rejected by others: it seems to me that there are the gravest difficulties in proving its truth, & yet I must think that it holds good to a large extent. Goethe puts the case under a clear point of view, when he says2 "the budget of Nature is fixed; but she is free to dispose of particular sums by any appropriation that may please her. In order to spend on one side, she is forced to economise on the other side." That this sort of compensation holds good with the modifications which our domestic productions have suffered, I can hardly doubt after the facts given in our first chapter; for instance in plants rendered sterile & seedless by their artificial treatment the nutriment goes to the enlargement of the fruit./

46/In monstrosities this law seems, also, to hold: Isidore Geoffroy

1 [See appendix for long note removed from this MS. and placed in port folio C 40 f.]

2 Pictet on the writings of Goethe, translation in Annals & Mag. of Nat Hist. vol. 2 p 318 1839.


St. Hilaire gives the following example as the best out of hundreds, "dans lequel l'antagonisme de développement m'a semblé" aussi evident que possible. II existait en effet du côté" gauche un rein & une capsule surrenale de grandeur ordinaire, et du côté1 droit, un rein extremement petit et une capsule tres‐volumineuse."1 M. Moquin‐Tandon gives several cases of this same law in monstrosities in the vegetable kingdom.2

But the question which here more immediately concerns us, is whether we can discern this law in the structure of species in their normal condition. The case of the ribs being so numerous & the limbs absent in serpents has been advanced as one of compensation of growth; & it may be so,/47/but as, according to the principles of this work, a part may be diminished by disuse, & another neighbouring part augmented by use or still more effectually by continued natural selection (for instance the greatly lengthened palpi & antennae in the eyeless cave animals), I do not see how such results are to be distinguished from compensation of growth. Nevertheless so many cases of apparent compensation of growth can be advanced, that I conclude there must be some truth in the law. For, as Mr Waterhouse has remarked to me, it would appear that when any part is greatly increased, adjoining parts or organs do not retain their usual or typical size, but are actually diminished. The large size of the canine teeth & the smallness of the premolars in the Carnivora may be given as an instance. The great size of the thorax & the small size of the abdomen in the Brachyurous Crustaceans & the exactly converse case in the Macroura, have been advanced as cases of compensation: Adouin, for whose opinion/48/one must entertain the highest respect, insists most strongly3 on the mutual relation in development of the three divisions of the thorax, in the several great orders of insects.

The following great Botanists seem to believe in the law of compensation, not merely in monstrosities, but in plants in their normal state; De Candolle, the elder, Richard Moquin‐Tandon & Auguste de Saint‐Hilaire. This latter Botanist, (no relation of the founder of the law) gives as instances of compensation, the expansion of the petiole, & the abortion of the limb in many leaves;—the great development of the bracteae when flowers are not developed as in the crown of the pineapple;—& a crowd of instances in which the doubling of the organs in one whorl seems to cause the abortion

1 Hist. des Anomalies Tom I. p. 276. This case is quoted from M. Martin St. Ange.

2 Elements de Teratologie Vegetale. p. 155‐160.

3 Annales des Sciences Nat. Tom I(1 series) p. 111. & 416.


of the organs in the succeeding whorls.1 /49/Moquin‐Tandon, besides cases quoted from Decandolle & some monstrosities2 brings forward as a case of balancement, the elongated peduncles & bright‐colours of the rudimentary flowers in the Feather Hyacinth; & again the development of the corolla & the abortion of the stamens & pistils in the outer flowers of the Snow‐ball‐tree (Viburnum opulus); & something of the same kind would appear to hold good in the outer florets of many Compositae & some Umbelliferae. Ach. Richard3 believes that the great size of the bulbillas in certain Crinums causes the pericarp in these species to be almost rudimentary./

50/If one could feel thoroughly convinced of the truth of this law of compensation, it would be important. On our view of species having arisen like varieties we could understand its action. (& we need not call in fresh creations to play the part of the laws of simple growth.) The order of the development of parts would probably be seen to be a very important element in change; if compensation be as powerful a law as many have thought, for the first developed parts would be apt to rob & so cause the deterioration of subsequently formed parts. It would, also, I believe, throw some light on rudimentary organs & parts. I have sometimes been inclined to think that the supposed law of compensation might be put under a simpler form; namely that nature, like a careful manufacturer, always tries to be economical in her materials; & if any part or organ can be spared, whether or not any adjoining part be in consequence largely developed, it is spared, & matter so/51/saved. Animals belonging to very different classes, when parasitic within other animals & thus protected, offer instances of this truth: I am thinking of two Cirripeds namely Proteolepas & the male of Ibla which live within the sacks of other cirripedes, & in both of these & in no other member of the class the entire capitulum or carapace is absent & thus saved. In many such cases, I doubt whether it can be truly said that any other part or organ has been, either as cause or effect, developed in excess; but the less nutriment required, owing to some parts of the body under changed circumstances being through natural selection less & less developed, might be of service to any creature in the severe struggle for life to which all are exposed: just as on the same pasture a greater number of animals in a moderately thin state, could be kept alive, than of animals with a thick layer of fat./

1 Lecons de Botanique 1841. p. 145, 199, 619. Again in Annal. des Sciences Nat. Bot. 2 series Tom I. p. 333. he advances this law in relation to stamens & petals in certain genera.

2 Elements de Teratologie p 157.

3 Annal. des Sci. Nat. Tom 2. 1 series, p. 15.


52/A part normally developed in any species in an extraordinary degree or manner, in comparison with the same part in allied species, tends to be highly variable.—

Several years ago, Mr. Waterhouse1 published a/52 A/remark to nearly this effect; Professor Owen, also, seems to have come independently to a similar conclusion. I was formerly much struck with Mr. Waterhouse's remark, for I could see no reason why in a species, if looked at as an independent creation, a part developed in any highly peculiar manner or to an extraordinary size should tend to be eminently variable: on the other hand if a species be only a strongly marked variety, the cause of this variability, we shall see, is not of very difficult explanation./

53/I must here premise that our apparent law, which we are here going to discuss relates only to parts differing greatly from the same parts in species if not actually congenerous at least pretty closely allied: nor do I suppose that the rule is of universal application. To give an imaginary example, the wing of a Bat is a part developed in a highly remarkable manner in comparison with the front‐legs of other mammals, but our law would not here apply: it would apply only to some one Bat having wings developed in an extraordinary degree, or manner, compared with other closely allied Bats,. When several species within the same genus differ remarkably one from the other in some part or organ, which is uniform throughout the rest of the same Family, then according to our law, the part or organ in question should tend to be variable in the species of the genus. Our supposed law is applicable to any character, although attached exclusively to either male or female sex, if the character be very remarkable in comparison with the same part in the corresponding sex of the/ 54/allied species. Moreover as all secondary sexual characters, whether or not developed in any especial manner, may be con‐sidered as in some degree a departure from the typical structure of the group to which the species in question belongs,2 for instance the male Turkey, Fowl & Pheasant all depart a little more from the typical structure of the Gallinaceae than do the females; so does the female common glow‐worm depart far more from the typical structure of the Lampyridae than does the male,—hence

1 A. Nat. Hist. of the Mammalia. 1848. vol. 2. p. 452, note 1, "As a general rule where any species is characterized by a maximum of development of certain parts, those parts are more subject to variation in the different individuals of the species than are parts which approach more nearly to the normal conditions."

2 W. [actually John] Hunter's Animal Oeconomy Edited by R. Owen p. 47.—Westwood in Entomolog Discourse has made primarily [?] same remark. [Possibly in Addresses to the Entomological Society, 1851‐53, London 1851‐53.]


it seems to be conformable to our law, that all secondary sexual characters should be more variable, as I believe they are, than the characters common to the two sexes.

Before giving a list of the more striking facts, which I have accidently met with, I must remark that the cases implying extraordinary development cannot be very frequent; & secondly that it is very difficult to collect facts of this kind: I have experienced this myself, & have seen it in others, namely that it is scarcely/ 55/possible, on being asked, to call to mind relations of a complicated kind without going deliberately through every species in a group with which one must be thoroughly familiar. Having been struck with Mr. Waterhouses remark before I undertook the classification of the Cirripedia I attended to it & was astonished at its wide application; so that I generally found some most striking & remarkable character in a species of far less use for classification than I had anticipated owing to its surprising variability.1 Moreover from Cirripedes being hermaphrodite, the cases are the more valuable, as clearly showing that the law holds good without any relation to sexual distinctions. As Birds are generally remarkably constant in their structure, I have also particularly attended to those few cases in which, in comparison to closely allied birds, some part presents a very unusual character, & we shall immediately see how apt these characters are to be universally variable. These cases of Birds, together with my own experience with cirripedes, have/56/mainly convinced me that there is much truth in our supposed law.—/56 v/On the other hand I have been led to doubt its truth from not having noticed any analogous remarks in Botanical works, & I believe in the present state of Natural History Botanical generalisations are more to be trusted than those deduced from Zoology. I applied to Dr. Hooker on this subject, who after careful consideration, informs me that though some facts seem to countenance the rule, yet quite as many or more are opposed to it. In plants one large class of cases, namely secondary sexual characters are not present. Moreover, as Dr. Hooker has remarked to me, in all plants there is so much variability, that it becomes very difficult to form a judgment on the degrees of variability: in a Bird having a beak of unusual structure we are at once struck at any variation, as the beak in other birds very seldom varies; but with a plant, how difficult to judge whether an abnormal leaf or petal varies more than leaves or petals of ordinary forms!

In parts developed to a great size, a source of deception should be

1 Monograph of the Balanidae. Ray Soc. 1854. p. 155.


here noticed; namely that the variation, even if not really greater than in other species in which the same organ is of the usual size, would be far more conspicuous. But this source of doubt does not apply to parts developed not to a great size, but in an unusual manner./

56/Naturalists have repeatedly remarked that every part of the living frame can be shown to be variable in some or another species: hence, as a mere coincidence, I should have expected that some few instances would have occurred of parts developed in any remarkable manner, being likewise variable in the same species. But it must be remembered that instances of parts developed in great excess or very differently from the same part in allied species are not numerous; & secondly that the cases of variability in organs which are usually constant in form (of which fact we have several instances in the following list) are decidedly rare; therefore the improbability is very great of variability, itself rare, being a mere chance concomitant, of unusual development, also rare in the same part or organ in the same species. Hence, I think, we may infer that there is some direct relation between the vari‐ability/57/& the unusual, though normal, development of the same part. I may here add that many Naturalists believe that variability is related to the slight functional importance of the part: (I do not myself believe in this doctrine;) it is therefore worth notice that when a part or organ is developed in a remarkable manner in a particular species, the most obvious inference is that it is of at least as much, probably of more, importance to the species in question, than the same part or organ where less developed in the allied species; & yet, as we shall immediately see, it is nevertheless generally highly variable./

58/The Hystrix cristata has a skull readily distinguished by "the enormous size of the nasal bones," but these bones, & "the highly arched upper surface of the cranium "are subject to con‐siderable variation".1

The male Narwhal has, perhaps, the most anomalous teeth of any mammal, & here we have variability in the length, of the tusk & sometimes the second incisor is developed into a short tusk.2 /

1 Waterhouse. Nat. Hist. of Mammalia vol. 2. p. 452. [Darwin here jotted in pencil the following:] Zoolog. Soc. beginning of 1857 Owen Length of arm of ourang‐outang longest arm & very variable in length—Owen has some other cases about teeth I feel sure <Falconer tusks of Elephants but the sexual Elements certainly variable but not confined to a species or genus> To give a few instances of male sexual characters variable—Tusks of Elephants <Narwhal> Mane of Lion if Persian be same species—Deers Horns [1 or 2 words illegible.]

2 Scoresby. Arctic Regions Vol. I. p. 490.—Kane's Arctic Exploration vol 1 p. 455. I have looked in Penny Encyclop. & Dict. Class & can find out no other particular of variability. [Note continued at foot of p. 310.]


59/The Wax‐wing, Bombycilla garrula, is very remarkable from the wing‐feathers being tipped by scarlet horny points which differ a little in the male & female: Macgillivray adds "the prin‐ cipal variations have reference to the wax‐like appendages to the secondary quills".1

The Chimney Swallow, Hirundo rustica, differs from many of its congeners, by its forked tail, which is much shorter in the female: Macgillivray says that it exhibits little variation, except in the tinge of the red on its breast "& in the lateral tail‐feathers being more or less elongated."2

The Oyster‐Catcher, Haemantopus ostralegus, certainly has a re‐markable beak, & Macgillivray says " considerable differences occur in the size of the bird, & especially in the length & shape of the bill."3

The Cross‐Bill, Loxia Europaea, has a most singular bill, as its name implies: several ornithologists have been struck by its great variability: Macgillivray says " the variations which I have observed in adult birds are not remarkable, excepting in regard to size, & especially in that of the Bill, which varies considerably/60/in length, curvature & the degree of elongation of lower mandible." He then gives various measurements showing how remarkably great the variations are in this important & generally constant part of the Bird's structure:4 /60 v/The upper mandible, moreover, sometimes crosses from the right & sometimes from the left; & this variation is the more remarkable, as certain muscles are unequally developed on the two sides, in accordance to the side to which the upper mandible crosses over:/5

60/The long‐legged Plover or Himantopus forms a small genus with closely allied species, quite remarkable from their extraordinary length of the legs compared with their nearest allies. Mr. Gosse6 has carefully attended to the measurements of the legs in H. nigri‐collis, & he finds no two birds with exactly the same length of leg, there being as much as half an inch in length difference between the extreme specimens. This bird is likewise remarkable by its

[Reading note sheet, marked '12' in ochre:]

Gleanings from the Menagerie of Knowsley Hall 1850.

J. E. Gray p. 55. Though Horns in Deer (a sexual character) are of service for separating the species into groups. & though they have largely been used for specific distinction, yet Dr. G. finds, ['] yet it has been found that animals of same herd or even family & sometimes even the same specimen under different circumstances, in succeeding years have produced Horns so unlike in size & form, that they might have been considered as belonging to different species [*] See good case in Sir J. Richardson Fauna, p 241 [part I. Re variation in horns of Barren Ground Caribou, Cerous tarandus var. aretica.]

1 British Birds vol. 3. p. 535.

2 British Birds vol. 3. p. 560.

3 British Birds vol. 4. p. 155.

4 British Birds. Vol. I. p. 423.—

5 Yarnell in Zoological Journal. vol. 4. p. 459.

6 Birds of Jamaica p 388.


bill being slightly upturned; but Mr. Gosse finds this character well pronounced in only one out of 16 or 18 specimens.—/

60 bis/In Trochilus polytmus the curvature of the beak seems in some degree a sexual character, being according to Mr. Gosse1 plainest in the female; but the curvature "varies in the individuals, <& I possess several females whose beaks are more curved than in T. Mango.">

One of the species of Chamelion [sic] (C. bifurcus) is most extraordinary from its nose being divided & produced into two horn‐like protuberances; but H. Schlegel2 says that "the nasal prominences are subject to variation."

In the genus Cygnus, the trachea in some species follows the usual course, in others it makes the most remarkable convolutions, entering the breast‐bone, & these convolutions differ greatly in some of the species; in the Whooping Swan "the diameter of the trachea & the extent to which it enters the crest of the sternum varies"; in Bewicks Swan, also, the trachea is not constant, the horizontal loop being sometimes absent, & in some specimens it does not differ from that of the Whooper.3

61/Cirripedes.—In Conchoderma the valves are very abnormal in shape & astonishingly variable but then they are in some degree rudimentary. One species of Concoderma [sic] differs from all other Cirripedes in having curious ear‐like appendages to the capitulum & these, also, are very variable. Alepas cornuta differs from the other species of the genus in having horn‐like projections on the capitulum, & these are variable in shape & position. Balanus laevis differs from all other cirripedes in having the basis filled up with a cancellated structure; the extent to which this is effected is very variable & very often there is no trace whatever of this remarkable structure. In Chthamalus antennatus the third pair of cirri (legs) is very remarkable in having one of the rami wonderfully elongated & apparently developed to act as an antenna; but this elongation of the one ramus & the number of its segments, are marvellously variable; & the arrangement of the spines, which are of functional importance & generally constant, was equally or even more variable, being arranged/62/on two distinct plans. Acasta sulcata is unique in having the pedicel of the fourth cirrus developed into most beautiful, curved, prehensile teeth; but of this remarkable structure there was not a trace in some specimens

1 Birds of Jamaica p. 98.

2 Essay on Serpents Engl. Translat. 1843. p. 216.

3 E. Blyth in Calcutta Review 1857. p. 155. and Macgillivray British Birds vol. 4. p 651, 665.


from the same district which after the most careful examination I am fully convinced belong certainly to the same species: moreover similarly anomalous teeth on the lower segments of the cirri were also highly variable. These teeth are not mere spines but actual modifications of the margin of the limb: their presence caused also the abortion of the usual moveable spines. As I look at a strongly marked variety as not essentially differing from a species, I may advance as an illustration of our law, a variety of Balanus balanoides which I at first described as a distinct species: in this the segments of the posterior cirri had ten pairs of main spines—a number quite unparalleled in any other cirripede whatever; but on examining many specimens I found the number varying from seven to ten pairs!/

63/Lastly I may advance the case of the opercular valves in Pyrgoma & in the too closely allied genus Creusia: the opercula valves, I may premise, are of the highest functional importance, & stand in direct relation to the most important muscles in the animal's body. These valves present very slight differences in most of the genera of sessile cirripedes; but in Pyrgoma they differ in the most striking manner in the different species: I had not sufficient specimens in most of the species to ascertain whether they varied much as they ought to do according to our law; but in/63 v/Pyrgoma cancellatum, the ridge giving attachment to the great & important adductor scutorum muscle is developed in the most wonderful & abnormal manner, & it is variable: in P. dentatum our/63/law is fulfilled in a more marked manner: the scutum in this species has a special ledge greatly developed;—it has the articular ridge developed into a unique tooth‐like projection;—the whole outline of the tergum is most unusual & on the inner side there is a unique tooth; now all these extraordinary conformations varied/64/in so wonderful a manner, that it is no exaggeration to say that the varieties differed far more from each other in these important parts of the structure than do the other genera of sessile cirripedes in the same parts. Creusia spinulosa might be added to this list; but the variation in the opercular valves was so great & so hopelessly perplexing that after weeks of labour I had to give up in despair the determination of what to rank as species & what as varieties.1 /

65/Insects.—We will now turn to insects, & give some illustrations

1 See my monographs on the Lepadidae & Balanidae (p 155) published by the Ray Soc. Under the heads of the Genera & Species, above specified, full details are given.—


from several of the great orders. One of the most striking cases has been given to me by Mr. Wollaston, namely that of a beetle, the Eurygnathus Laterillei [actually Lateeillei]1 the female of which presents "the extraordinary anomaly" of its head being immensely more developed than that of the male; & Mr. Wollaston believes that the case is unparalleled in the whole vast order of Coleoptera: now this, though serving as a well‐marked specific character, is so excessively inconstant that "scarcely two females have their heads of exactly the same size "; in some there being only a tendency in this direction, in about two‐thirds of the specimens, the head being "literally immense"./65 v/The females of some species of Dyti[s]cus, which normally have their elytra deeply furrowed in a very remarkable manner, are sometimes quite without these furrows; yet such females have been caught in connection with the males.2 /65/Mr. Wollaston believes that the Harpalus vividus is the only species in this great genus, which has its elytra connate, "but this character, anomalous as it is, is far from uniform". In the whole genus Scarites the mandibles are in both sexes remarkably developed, compared with other Carabideous genera, & Mr. Wollaston informs that "in size they are imminently variable."—In the/66/Stag‐Beetle, & indeed generally in the Lucanidae, the mandibles in the males are enormously developed & are eminently variable not only in size but in the form of the terminal teeth; yet the mandibles, as Mr. Stephens3 has well remarked in ordinary cases "are dwelt upon almost with mathematical nicety." The astonishing variability of so important an organ as the mandibles & of some other organs in this & in many of the following cases, is rendered very striking if the same part or organ be compared in a set of females of the very same species, where they will be found to be almost absolutely identical in form.

Mr. White showed me a series of specimens of a magnificent Chalcosoma from the Philippines in the British Museum, in which the females were absolutely similar, but the males exhibited the/ 67/most surprising series of varieties in the curious horns on the thorax & head; these horns being five or six times as large in some specimens than in others & with great diversity in the teeth: so it is in Megasoma & many Dynastidae. So again in the males of many Scarabaeidae & of some Cetoniidae. To turn to another quite distinct group of Coleoptera; the males of some Staphylinidae are homed, & the horns are very variable, as in Bledius.

1 Insecta Maderensia. p. 20.—

2 Westwood, Modern Classification of Insects. vol. 1 p 104.

3 Illustrations of British Entomology Mandibulalata. vol 3. 1830 p 367.


The whole snout is much elongated in the male Attelabus & in some Curculionidae, & is in them very variable as I am informed by Mr. Waterhouse. In the male of the Truffle beetle (Leiodes) the thighs are much incrassated &, here again, as I am informed by Mr. Waterhouse they are very variable.—In the males of Choleva, the trochanters of the hind legs are liable to great variation.1 In the carrion feeding Necrodes littoralis, the males have incrassated & dentated femora, increased [sic. Stephens says "incurved"] tibiae/68/dilated tarsi, sculptured thorax, costated elytra, & everyone of these points is highly variable!2

To turn to another Order, the Homoptera: the Umbonia spinosa was pointed out to me by Mr. White as having most singular spinose projections of the thorax, in both sexes, & these are highly variable. Again in Fulgora or the Lantern‐Fly & in the Fulgoridae the forehead is most singularly dilated into a muzzle, sometimes even equalling the whole rest of the body in length! this strange projection differs greatly in the different species,3 is not confined to either sex, & is very variable in several species, as I saw in specimens shown me by Mr. White.

Lastly to take one other great order, the Hymenoptera, in which I am indebted to the highest authority Mr. F. Smith, for the following striking illustrations of highly abnormal characters in several species, being, as heretofore, highly variable. Both sexes of the Chrysis ignita4 are highly peculiar from the apex of the abdomen being armed with four teeth;/69/but these are so variable in length as well as in position, as to assume nine distinct types of form, & are occasionally nearly or quite absent! The male of the Andrena longipes, in some examples, but not in all, has an enormously large head in comparison with that of the female: in another species, Andrena fulva, large males have a long acute tooth at the base of the mandibles, but in smaller specimens this is reduced to a mere tubercle; & this form was consequently described by Kirby as a distinct species. In a male Saw‐fly, the Tenthredo femorata, a series of specimens "exhibits a wonderful difference in the development of the posterior femora." On the other hand the females of the two following Bees have a peculiarity very remarkable & confined to two or three species in their respective genera: namely in the Osmia fulva, two stout horns on the front of the head, & these vary greatly in length & shape,

1 Andrew Murray, Monograph of the Genus Catops 1856 p 14‐.

2 Stephens Illustration of British Entomology vol 3. p. 367

3 Westwood Modem Classification of Insects vol 2. p 428.

4 Entomologists Annual 1857 [p 31.].


being somewhat bifurcated when large & wedge‐shaped when small; & secondly in the Nomada lineola, two teeth on the labrum, & these vary so much in length/70/that the varieties have been described by Kirby as distinct species.1

In considering the foregoing facts, & others might have been added, we see that they fall under three heads, namely of some striking peculiarity being eminently variable, when attached to both sexes, or when attached exclusively to the male sex, which is the commonest case, or exclusively to the female sex. The cases seem to me too numerous & striking to be accounted for by the mere chance coincidence of variability & unusually great development; more especially when we bear in mind how remarkably constant in character many of the very same organs or parts are when not developed in any extraordinary manner in the other species of the same groups. Our laws seem to hold equally/70A/ good, & when all the species of a group differ somewhat from each other in some part, as with the opercular valves in Pyrgoma; as when a single species differs somewhat in some part from its congeners.—indeed the cases do not essentially differ from each other. As genera are mere conventional groups, I should have expected that when a set of genera were closely allied but yet differed from each other in some one organ to a marvellous degree that this organ or part would have been variable in the species of such genera. This, I think does happen sometimes, but certainly very far from always. Thus amongst the Homopterous Insects, we have numerous closely allied genera differing from each other in certain parts in the most extravagant & grotesque manner conceivable,—with ball‐spines, bladders, lanterns such as a child might draw out from his fancy—& yet, as I saw with Mr. White in the British Museum, these astonishing peculiarities did not vary much in the species of the several genera./

71/Mr. Waterhouse believes that the extreme diversity in the development of the mandibles & horns in the Lucanidae & Dynastidae is related to the manner in which the insect in its larval state has been nourished; & it deserves notice that in some species, for instance in the Angosoma centaurus there is, as I have been informed by Mr. White, much less variability than in the allied species. In the Lucanidae & allied families, the existence of males presenting a wide range of varieties in their secondary male characters, from extreme development to a close approach to the

1 [Later note on separate slip of paper:] F. Brauer advances in favour of the law as given in the Origin (Verhand. Zoolog. Bot. Gesell. in Wien. 1867 Dec. 4) some highly peculiar characters in the wings of the dimorphic females of the Neuropterous genus Neurothemis.


female condition, is so very general, that a collector is not satisfied, as I am informed by Mr. Waterhouse, until he possesses a complete series of this kind for each species; & this fact perhaps does indicate that there is here something quite unknown & different from the other cases of variation & abnormal development, One is at first strongly tempted to explain all these cases of variability in the secondary male characters by the hypothesis of a great diversity in the virile force of the males; on the same principle that the horns of deer are affected by emasculation, by the amount of food,/72/or by unnatural conditions as confinement on shipboard;/72 v/and I think this explanation may be true to a large extent. We must, however, be cautious in inferring loss of virile powers from loss of the secondary male characters; to give one instance; Sebright Bantam has not sickle‐feathers in the tail, yet a writer in Poultry Chronicle, shows that one thus deficient, was the father of an unusual [?] number of chickens.1 /72/But an analogous hypothesis, would be rather bold when applied to the several cases of variation in remarkable developments, characteristic of the female sex: in the Eurygnathus we should have to suppose that about one‐third of the females, namely those with small heads nearly like those of the males, were in some degree sterile. Moreover this view is clearly inapplicable to abnormal characters in no way. connected with sexual function, & common to the two sexes as in many birds & as in the hermaphrodite cirripedes, which have afforded us so many instances of parts unusually developed being highly variable.

But now let us turn to what we know in regard to domestic varieties: we have seen in our two first chapters that fancy breeds, —those which the fanciers are now improving (by selection) to their utmost,—are much more difficult to breed true or vary more in the admired & selected points, than breeds which have long inhabited any district without/73/particular care having been paid to them; I refer of course to pure breeds alone & not to fluctuating mongrel breeds which would necessarily be variable from crossing. For example compare the head & beak of the common & improved Tumbler, of the common & improved Carrier, with these same parts in any old [?] breed, as the Fan‐tail, in which these points have not been much attended to, & observe what an astonishing range of variation the head & beak present in the two former cases. The cause seems obvious, namely that in each of the later generations, individuals with certain admired points most strongly

1 [Additional pencilled comments:] Hewitt says they are generally deficient in virile force Hen‐tailed Game Cocks show no loss of virile powers.


developed have been selected, so that the particular characters in question, though the difference in each generation may have been so slight as to have been scarcely appreciable, have not been fixed by strict inheritance during a long course of centuries. Moreover we have seen in our first chapters that new characters, or those in course of improvement through Selection often become, from quite unknown causes, attached in a greater or lesser degree to one sex, far/74/most generally the male sex,—take for an instance the wattle in the improved carrier Pigeon; furthermore it would appear that those characters which become under domestication attached to one sex, are eminently variable.

Now if we look at species as only strongly marked & very permanent varieties, & consequently at all the species of a small group, as the descendants from some one form,—like the fancy pigeons from the Dovecot,—then those parts in which all the species agree will have been inherited by them for an enormous period, & ought to be thoroughily fixed in the breed: in such cases it will make no difference whether or not the part is developed like a Bat's wing, in an extraordinary manner: on the other hand any part developed in an extraordinary degree or manner compared with the same part in the closely allied species, according to our theory, will have undergone an immensely long course of modification through natural selection within a comparatively recent/ 75/period; for as natural selection acts only by the addition of successive extremely small changes, & as the part in question is developed in an extraordinary degree or manner, the process of addition must have required a very long time to have produced the given result; & all this must have taken place since the several species branched off from the common parent stock & therefore long subsequently to any considerable change in the other parts of their organization. Consequently, in accordance with the analogy of our improved domestic breeds, we might have expected that such parts or organs would be the least strictly inherited, with a strong tendency to reversion to the aboriginal parent form. Moreover we might have expected from the same analogy, that some of the comparatively late & extraordinary developments would have become attached to either sex, generally to the male sex, without as far as we can see profiting either sex; & furthermore we might have expected that such secondary sexual characters would have been highly/76/variable,—all facts which seem to hold good in nature. Nor we must forget that Sexual Selection, by which the variations in the secondary characters confined to the males alone, & useful to them in their struggle for the females are


added up & accumulated is less rigid than ordinary selection; the less successful males generally leaving some offspring; so that those secondary sexual characters which are of use to the male, would be less rigidly scrutinised & sifted than the characters on which the life or death of the individual male & female depended. On the other hand, if we look to the generally accepted doctrine of each species having been produced by an act of creation, I can see no explanation of the several facts given in the present section, showing that secondary sexual characters, especially if developed in an extreme degree, & generally that all parts developed in any very extraordinary manner, are apt to be highly variable./

77/A part so little developed, as to be called rudimentary, tends to be highly variable.—

The subject of Rudimentary organs will be treated of in a separate chapter; I refer here to this one point of variability, as standing in relation to our last proposition of parts developed in an extraordinaryly great degree being variable. The cause, how ever, I believe to be different: organs become rudimentary through disuse (aided, perhaps, by the principle of compensation & often by natural selection) & through the effect of disuse becoming hereditary at a period of life corresponding with that of the disuse. Disuse shows of course that the part in question is not useful to the Species, & therefore natural selection cannot come into play to keep fixed a part when become useless & rudimentary, namely by destroying all injurious departures from one fixed type. The continued existence of a rudimentary organ depends wholly on the strong principle of inheritance, as we shall, hereafter,/78/attempt more fully to explain. On the other hand, a part developed in an extraordinarily high degree is as I suppose variable, from not having become strictly inheritable,—from natural selection not having had time sufficient to overcome the tendency to reversion & to regulate its own work of adding up very many small successive modifications.—/1

79/Monstrosities: arrests of development.—As monstrosities can not be clearly distinguished from variations, I must say a few words on some of the conclusions arrived at by those who have studied the subject. Geoffroy St. Hilaire & his son Isidore2 repeatedly

1 [Pencilled comment later cancelled:]<The diverse branching of horns in Lucanidae & the nine types of abdominal points in Chrysis, shows not all reversion—there is fluctuation as well as reversion.>

2 Principes de Phil. Zoolog. 1830 p. 215. And Histoire des Anomalies 1836 Tom. 3. p. 437 et passim.


insist on the law that monstrosities in one animal resemble normal structures in another. So in the vegetable Kingdom M. MoquinTandon says, "Entre une fleur monstrueuse et une fleur normale it n'y a souvent d'autre difference que l'état accidentel de la premiere et l'état habituel de la seconde. La monstrosité est done en general, l'application insolite a un individu ou a un appareil, de la structure normale d'un autre individu."1 As the resemblance between a monstrosity & a normal structure is generally not very close, & as the comparison is often made with forms remote in the scale of nature, & as when all within the same great class/80/are included, a vast field for comparison is opened, I cannot avoid the suspicion that some of the resemblances given are simply accidental. But I imagine no one would account for all the resemblances on the doctrine of chance. To give two or three of the best instances from Mr. Isidore Geoffroy;—in the pig,—which has the snout much developed & which is allied, but, as Owen has shown, not so closely as we formerly thought to the Tapir & Elephant, a monstrous trunk is developed oftener than in any other animal: the frequent monstrosity of three, four or even a greater number of breasts in woman seems to stand in relation to the fact of most mammals having more than two mammae: Carps are very subject to a curious monstrosity causing their heads to appear as if truncated, & an almost exactly similar but normal structure is met with in the species of Mormyrus, a genus of fish belonging to the same Order with the carp.2 Notwithstanding such facts, &/81/many others could be given from the animal & vegetable Kingdoms, I cannot believe that in a state of nature new species arise from changes of structure in old species so great & sudden as to deserve to be called monstrosities. Had this been so, we should have had monstrosities closely resembling other species of the same genus or family; as it is comparisons are instituted with distant members of the same great order or even class, appearing as if picked out almost by chance. Nor can I believe that structures could arise from any sudden & great change of structure (excepting possibly in rarest instances) so beautifully adapted as we know them to be, to the extraordinarily complex conditions of existence against which every species has to struggle. Every part of the machinery of life seems to have been slowly & cautiously modelled to guard against the innumerable contingencies to which it has to be exposed.—

1 Elements de Teratologie vegetale p. 116 p. 342. The same view is taken by M. Auguste St. Hilaire in his Morphologie Vegetale p. 818.

2 Histoire des Anomalies Tom I. p 285. Tom 3. p. 353. p. 436.


As all vertebrate animals, for instance, pass/82/through nearly similar embryonic changes, we can see that arrests in the development of any part,—a doctrine on which M. Isidore Geoffroy lays much stress—will account for a certain degree of resemblance of many monstrosities to the normal structure of other animals, even when very remote in the same great class. A very frequent monstrosity in plants having irregular flowers, such as Snap‐dragons, is their becoming regular; & as such flowers are known to be regular in their early bud state, I presume that this monstrosity would be admitted to be an arrest of development; as an instance of how all monstrosities are governed by laws, it may be added that the flowers nearest the axis are much the most apt to become regular;1 thus I have seen a Laburnum tree with the flowers at the end of each raceme open & not having the proper papilionaceous structure.

Other monstrosities appear caused not exactly by arrest, but by abnormal development; thus in the case of a monstrous number of mammae or digits, it may be surmised that in the embryo of all vertebrate/83/animals there is a tendency at some very early age to produce several mammae or digits, & that this tendency from quite unknown causes occasionally becomes fully developed in animals normally having only one or some small number. There are other monstrosities connected with the doubling of parts, the union of distinct embryos &c, to which we need not here allude. And there are other monstrosities, apparently not to be explained by arrests of or increments of development, which are common to various animals & plants in the same great classes, & which I presume can be understood only on the supposition of similar abnormal conditions acting on organic structures having much in common,—so created according to the common belief, but according to our views due to inheritance from a common/84/though sometimes immensely remote stock. I will only further remark that according to these same views, a part or organ may in one creature become normally reduced in size or quite atrophied from disuse during successive generations, in another it may suddenly become so in a monstrosity by an arrest of development;—again in one creature a part may by long‐continued natural selection become greatly increased in size or number, in a monstrosity it may suddenly be thus increased by abnormal development; but the possibility of this diverse origin of similar parts, through normal & through monstrous formation, evidently rests on the common embryonic structure of the two forms; & how organisms remote in the same

1 Moquin‐Tandon Teratologie Vegetal p. 189.


great classes come to have a similar embryonic structure will be treated of in a future Chapter.—

M. Isidore Geoffroy Saint‐Hilaire makes one generalisation which concerns us & well deserves notice;—1 /85/namely, that the more an organ normally differs in different species of the same group, the more subject it is to individual anomalies: thus taking the case of monstrous deplacements of organs, he affirms, that "Les organes qui se déplacent le plus fréquemment sont aussi ceux qui présentent des déviations plus considérables de la position normale"

We will now proceed to some remotely analogous considerations in regard to varieties & species./

86/Distinct species present analogous variations; & a variation of one species often resembles the normal structures of an allied species: or more commonly resumes the general character of the group to which it belongs.

In the first Chapter I gave a few instances of variations produced under domestication, resembling in character distinct species; & as some of these might be called inherited monstrosities such case can hardly be distinguished from those alluded to in our last section. Our present section relates more especially to varieties produced under nature or in organisms not much affected by domestication. But I think the bearings of our present discussion will be best shown by first giving an illustration from trifling variations in that group of domestic varieties, which I know best, namely pigeons. In all the main breeds there are analogous subvarieties, similar in colours,—in having feathered legs, & turncrowned heads; in several of the breeds & in subvarieties of others, the lesser' wing‐coverts are chequered with white & the primaries white. None of these points have any direct relation to the aboriginal parent breed, the Rock‐Pigeon; yet, I think it/87/ cannot be doubted that these analogous varieties are due to the several breeds having inherited a like organization from a common source; this organization having been acted on by similar organic & inorganic causes of change: just as we know that children of the same family often show a remarkable parallelism in symptoms when suffering from disease.2

Some of these variations as feathered feet, chequered wing‐coverts &c are fixed in & characteristic of certain breeds & subbreeds; therefore when such character appears for the first time

1 Histoire des Anomalies Tom. I p. 281, 418, 650 &c.

2 Sir H. Holland [ ].


in a breed, the sub‐breed thus characterised presents an analogy to other breeds properly so characterised. On the other hand when a character of the above kind is lost; or to give another instance when a blue Pouter which ought to have all its primaries white is "sword‐flighted" that is has some of the first primaries coloured,—or when a Turbit which should have a white tail throws a dark tail1 (of which Mr. Tegetmeier has had an instance) these/88/are not new variations, but the partial reversions to the parent breed but not to the parent species. Of reversions to the aboriginal species I have given an excellent instance in my discussion on Pigeons, in the fact that all breeds occasionally throw blue birds, & that these always have the two black bars on the wing, generally a white rump & a white external web to the exterior caudal feathers, —all characteristics of the aboriginal Rock Pigeons. It deserves, as we shall presently see, especial notice that these just specified characters are frequently brought out by crossing two Pigeons neither of which are blue, or probably have had a blue bird in their race for several <many> generations: why the disturbance caused by a cross should have this effect we are perfectly ignorant. In respect to all cases of reversions to ancestral characters, I may revert to the only hypothesis which appears to me tenable; namely that in such cases the child does not in truth resemble its ancestor a hundred or thousand generations back more than its immediate father, but that in each generation there has/89/been a tendency to produce the character in question, & that this tendency at last for causes of which we are profoundly ignorant overmasters the causes which have for so long rendered it latent. This does not seem to me more surprising than that the merest rudiment or vestige of an organ should be inherited for numberless generations. Those who explain an abnormal & monstrous number of mammae in a woman from the fact of the number of mammae in vertebrate animals being generally greater than two, will admit that a tendency, as well as an actual rudiment, may be inherited for any length of time. Under this point of view reversion to an ancestral form is only an arrest of development,—or the appearance in the mature state of a character which ought to have been passed through in an earlier stage.

Supposing that we had reason to believe that all the breeds of pigeons had descended from one stock, but did not in the least

1 I may remark that in crossing various breeds I have clearly noticed that colour sticks to the caudal feathers than to any other part, & secondly to the few first primaries: I have repeatedly noticed in crossing black & white birds of very different breeds, the few first primaries are black, succeeded by white feathers.


know what its characters were, or the ancient character of any of the breeds we should be quite perplexed to conjecture, when an individual was born with a turn‐crown whether this was a case of reversion of a character formerly attached to the breed, or a new variation analogous to what had/90/at some former period appeared & become fixed in some other breed. In the case, however, of the blue birds, as so many characters appear together without, as far as we can see, any necessary correlation, /90 v/& as these characters arise from a crossing of distinct breeds,—a cause wholly unlike what must aboriginally give the blue colour—/90/we might have pretty safely inferred that the black wing bars, white rump &c were due to reversion. But whether, or not, we could tell which characters were due to reversions (either to the aboriginal species or to some subsequent but ancient breed) & which to new variations analogous to those already existing in other breeds or sub‐breeds, we should without hesitation put all down to a community of organisation from common descent./90 v/Those who believe, as I do, that our Fowls are all descended from the Gallus Bankiva have an analogous case in so many breeds, as was remarked to me by Mr. Tegetmeier, having sub‐breeds with their feathers edged or laced & other sub‐breeds with their feathers transversely barred or pencilled. <This latter character may be derived from the hen of the G. Bankiva (though transferred to the Cocks of some of our breeds) & may be ranked as a case of reversion;> is doubtful whether either class of colour‐marking can be attributed to reversion but both the lacing, & pencilling are variations analogous in one sub breed to another, & likewise to some other quite distinct species of Gallinaceae./

90/Now let us turn to nature; we have frequent instances of distinct species & strongly marked natural races, presenting analogous variations. Thus many Foxes, as C.lagopus, fulvus & vulpes present crucigerous varieties;1 the American &/91/European Bears both sometime have young with a white collar.2 So in the British brambles, our best authority3 says "nearly allied species are apt to sport in parallel varieties ............ so that the species being ascertained, the same designation & very nearly the same description will characterise the variety in each case": these nearly allied species are themselves looked upon by many Botanists as strongly inherited races.4 An excellent observer in Sweden, Anders‐

1 Sir J. Richardson Fauna Boreali‐Americana p. 84, 93.

2 Id. p. 15.

3 Dr. Bell Salter in Henfreys Bot. Gazette vol 2. p. 114.

4 See the account by Mr. Ed. Lees on Brambles coming true from seed, in Phytologist vol. 3. p. 54.


son1 describes a set of varieties (which have been described as species) of Carex ampullacea & vesicaria, which "present a perfect analogy of every form in one species to those of the other." As I look at species as only strongly marked varieties, I may adduce one other, but distinct case, namely the remarkable correspondence, as insisted on by Prof. Fries2 between/92/particular series of the American species, of Hieracium, with those of Europe. Now all such cases, of parallelism of variation would be ordinarily accounted for by the species having been created with a nearly similar organization: following the analogy of our domestic productions we should attribute it to community of descent.

If there had been a permanently crucigerous species of Fox (as some believe there is), then the crucigerous variety of another species would have been a case of variation analogous to a distinct species; so would it have been if our supposed crucigerous species had produced a variety without the cross. As we do not know the ancestors of organisms in a state of nature, whether ranked as varieties or species, we can very seldom tell whether their varieties, when resembling in character other species of the genus, are variations for the first time in the breed, or reversions to a state through which the species in question had formerly passed. But in some very few cases we can form from indirect evidence a conjecture on this head./92 v/Thus the British Stoat (Putorius ermineus) may be called a variety in as much as it does not regularly turn white in winter: it has inhabited this country since the Glacial period,3 & during that period, analogy from other countries can leave no doubt that it was always white in winter.4 /

93/I will now give in small type such cases as I have collected illustrative of one or a few species varying in a manner closely analogous to other species of the same group.—I may recall to mind my former remark on the difficulty of collecting such cases excepting by an author himself carefully going through the group with which he is most familiar. I think that the following cases are too numerous & precise to be accounted for by mere chance, more especially as the comparison is always made with allied

1 Henfrey Bot. Gazette vol 2. p. 251.

2 Bot. Gazette vol 2. p. 185.

3 Owen, British Fossil Mammals p. 116.

4 Blyth remark in Loudon [Mag. Nat. Hist.] vol. 8. p 50—makes me doubt the case—see about Irish Hare & then give up case [;] if Hare turns allude to it—Bell seems to think common in North—Will not do to quote, without Variable Hare offers a good case [Separate note slip:] Stoat Case Loudon Magazine Vol. 7. 504‐591 [vol.] 5. 294—718 ['Zoophilus' [vol.] 8. 51 [Blyth]. The Stoat turning white in Cornwall strongest case. [The first three references are to short communications of paragraph length.]


species, generally closely allied species. <has several times occurred to me in reading an account of a set of species of the same genus which have differed in some remarkable character, that I have truly predicted that I should find this same character described as variable in the individuals of some of the species> the list of cases of parts greatly or abnormally developed being highly variable, in so far as the variations bring back the species to the common type of the genus they might have been here introduced. All the cases, alluded to in Chapter 4 of varieties intermediate between two species, whether the one or both vary, in fact come under this head. And of variations of what systematists consider/94/trifling characters, as colour, size, proportion being analogous to other species of the same genus, innumerable instances could be given,—indeed a large part of the difficulty in identifying species seems due to varieties approaching in character other species—but in such cases it seems hardly possible to distinguish mere general variability, from variations having some direct relation to the structure of other species of the group. But as far as they go they confirm what I must consider a law, namely that variations whether in some degree permanent, or occurring only in single specimens of one species often assume the character of another species of the same group./

95/Prof. Vaucher shows (in Mém. Soc. Phy de Genève Tom. I. p. 300) that the modes of gemmation are constant in each genus with some few exceptions; for instance, the species of Syringa bred in two ways; & in the common Lilac the two modes of gemmation are sometimes seen even in the same bush.—

Decandolle states (in Mem. Soc. Phys. de Geneve Tom. 3. Part II. p. 67) that in the Lythraceae some of the very natural genera have some species with petals & some without; & that in Peplis portula the individuals indifferently have petals or none—.

In the Primulaceae, according to M. Duby (Mem Soc. Phys. de Geneve. Tom. x. p 406) Samolus is the only genus in which the ovary is adherent to the calyx; and in another genus in this family, namely Cyclamen, some individuals of one of the species have the ovary partially adherent./

96/Oxalis buplevrifolia has simple & lineal leaves unlike all the other species of the genus, but Aug. St. Hilaire found some individuals with this enlarged petiole surmounted by the three usual, though here small leaves. (Morphologie Vegetale p 143)

The alternation & opposition of the leaves is respectively constant throughout many great Families, but in the Salicaria & Polygleae the species, have either alternate or opposite leaves; & both often are found even on the same individual. (Aug. St Hilaire Morphologie Vegetale p. 183)

The torsion in the aestivation of the corolla was thought to be uniform in the Gentianaceae, but in Gentiana Moorcroftiana & Caucasica it is different from the rest of the family & in individuals of the latter species, it is found


to vary in the individuals (Decandolle in Annal. des Science Nat. 3 Series Botany Tom I p. 259)/

96 bis/In the Malpighiaceae, A. de Jussieu (Archives du Museum d'hist. Nat. Tom 3. p 86) says the leaves are always opposed, with the single exception of Acridocarpus; & even in this genus one may sometimes remark a tendency, especially in the lower leaves, & even a complete return to opposition

Decandolle has seen a variety of Geranium pratense (Mém Soc. Phys. de Genève Tom 1. p 443) with the two upper petals white & the three lower blue; "on retrouve ici dans les Geraniées à fleurs régulières, cette tendence a la disparité des petales si remarkables [sic] dans plusieurs Pelargoniens [sic]." Geranium & Pelargonium, as every one knows, are very closely related genera.— /

97/Aug. St. Hilaire (Mem. du Mus. d'Hist. Nat. Tom XI. p 49) says that in the different species of Helianthemum one may observe "toutes les nuances possibles entre le plancenta [sic] purement parietal et des loges parfaitement distinctes", & that "dans l'H. mutabile une lame, plus ou moins large, s'etend entre le pericarpe et le placenta"; so that in this species the degree of division of the ovary into lodges seems to be variable.

In the Compositae the presence of a ray to the outer florets is generally a constant character; but in Senecio, for instance, some of the species have a ray & some not; & of those species which have not, as the S. vulgaris, varieties are found having a small ray; on the other hand in the species ordinarily having a ray, as S. Jacobaea, varieties are sometimes found without a ray./

98/W. Herbert (Amaryllidaceae p 363) says that by crossing Calceolaria arachnoeides [sic in Herbert] which is purple with C. corymbosa, which has small purple specks on the yellow corolla, flowers were produced to the surprise of cultivators broadly blotched with dark & even blackish purple "but the subsequent discovery of a Chilian biennial species, which I shall call C. discolor, blotched with a reddish purple in a manner somewhat similar, shewed that such an arrangement of the colour was a natural variation of the genus, which the cultivator might therefore have expected, if all the natural species thereof had been previously known."

Nerine curvifolia fertilised by a hybrid curvifolia‐pulchella, produced seedlings, of which one produced a young crimson leaf, "such a remarkable seminal variation brings curvifolia in closer affinity with N. marginata, which is distinguished by a red margin to the leaf". (Herbert Amaryllidaceae p. 412)/

99/Gaertner says (Bastarderzeugung p. 50) that Lychnis diurna when growing in dry places sometimes has sharp teeth on the sides of the petal, as is the case with Lychnis flos cuculi. I may add/99 v/that I have seen a seedling Spanish Pink D. Hispanicus with its petals so deeply cut & the point so much elongated, as to call to mind the petals of Dianthus superbus.—/

99/Azalea with five stamens & Rhododendron with ten stamens, of which half are of inferior power are closely allied genera as everyone may see; & W. Herbert says (Journal of Horticultural Soc. vol 2, p. 86) he has raised seedlings of A. Pontica & Indica frequently producing 7, 8, or 9 stamens. Azalea nudiflora in N. America presents "numberless varieties some of them exhibiting ten or more stamens". Asa Gray Manual, p. 257 2nd Edit.

The northern species of Gladiolus present, according to Herbert, (Journal of Hort. Soc. vol 2, p 90) "a strange diversity of seed"; there being a winged


or foliaceous margin in some of the species, which totally disappears in other species; & in G. communis some varieties have it curtailed, & some almost obsolete.

Of the Oak genus some species are evergreen & some diciduous [sic]; & the varieties of Quercus cerris are so variable in this respect, that Loudon (Arboretum et Fruticetum vol 3. p. 1846) says its varieties "may be arranged as deciduous, sub‐evergreen & evergreen."—So it is with the genus Berberis. (Hooker Flora Indica p. 218)/

100/Moquin‐Tandon (Teratologie Vegetale p 138) says he has found a plant of Solanum dulcamara in which all the upper flowers had two or three stamens "beaucoup plus longues et plus grosses que les autres", & in S. tridynamum & Amazonium three stamens are habitually much more developed than the others.

Dr. Hooker. (Journal of the Linnean Soc. vol. 2. p 5 Bot.) believes that the Lobeliaceae ought to be included as merely a tribe of the Campanulaceae. For in the Lobeliaceae, "even the irregular corolla affords no good mark, for some states of the Wahlenbergia saxicola (one of Campanulaceae) have an oblique corolla, & unequal inclined anthers, of which two have the connective produced into an appendix, thus imitating a prevalent feature of the Lobeliaceae." The coincidence of these several imitative characters deserves attention.—1 /

101/The American wolf is generally esteemed a distinct species from the European: Sir J. Richardson says (Fauna Boreali‐Americana. Quadrupeds p. 76) "the black mark above the wrist which characterises the European wolf is visible in some American wolves but not in all".

The Didelphys Azarae has a broad black stripe on the forehead; & the D. crancrivora [actually cancrivora] has an indistinct dusky line; the D. Virginiana has occasionally "a small dusky stripe on the forehead (G. R. Waterhouse,—Marsupialia in Naturalist Library 1841. p. 84)

The genus Timalia (allied to the Thrushes) according to Swainson (Fauna Boreali‐Americana. Birds p 31) stands in a group in which the bill is either notched or entire,—a character generally of high importance; & in Timalia pileata some individuals "have the bill perfectly entire, some slightly, & others distinctly notched; all apparently being old birds, full plumaged, & not differing in the slightest degree in other respects"./

102/Yarrell has stated that the Little Ringed Plover (Charadrius minor) can always be distinguished from Ch. hiaticula by a dusky spot on the inner web of the outer tail‐feather; this feather being in C. hiaticula wholly white, but Mr. Garrett & Thompson have shown that this spot does occur in some specimens. (Nat. Hist. of Ireland: Birds: vol. 2. p. 103

The position of the Spleen differs much in various serpents, "so as sometimes to occur at a distance from the pancreas & isolated at the posterior surface of the stomach"; and H. Schlegel (Essay on Serpents. Engl. Translat. 1843 p. 55) says he has observed individual variations in this respect.—/

103/Mr. Wollaston remarks (Variation of Species p. 62) that "it is almost diagnostic of the genus Gymnaetron that its representatives should be thus (ie with blood‐red dashes on the elytra) ornamented typically, or else that those species which are normally black should, when they vary, keep in view, as it were, this principle for their wanderers to subscribe to".

1 [Note slip:] Hooker's Misc. 3/109. Mr W Wilson says the Andromeda polifolia like Vaccinium vitis idea sometimes has the stamens attached to the corolla. Lindley puts these genera in remote nat. Fam.!!


Mr. Waterhouse informs me that the Pachyrhynchus orbifer one of the splendid Curculionidae of the Philippine Archipelago, which is the most variable of the genus, in its variations typifies the regular markings of the other species. So again in varieties of Cicindela campestris the golden marks became united as in C. sylvicola; & on the other hand in varieties of this latter species the marks become disunited as in C. campestris.

The classification of the Fossorial Hymenoptera was mainly founded by Jurine on the neuration of the wings, & this has been adopted by all subsequent writers; but in Typhia [i.e. Tiphia] & more especially in Pompilus, there is a considerable difference in this respect between the species, & in some of these species there are even individual variations (Shuckard on Fossorial Hymenoptera 1837, p. 48, 40, 43.)/

104/I will now give some examples from cirripedes. Acasta fenestrata, & in a lesser degree A. purpurata present a very remarkable character (Darwin Balanidae p. 305) in the shell being perforated by six clefts or holes in the lines of suture; & we have a similar character in some varieties of A. sulcata. In the different species of Pyrgoma, the opercular valves on each side are sometimes quite separate & sometimes so perfectly calcified together that even the line of juncture cannot be distinguished; & in P. milleporae the degree of union (p. 368 idem) is very variable in different individuals. In Bal. improvisus (p. 251 id.) certain varieties have their terga closely imitating the form of the same valve in the allied B. eburneus: so again a very remarkable variety of the common Balanus balanoides is the form of the tergum & in the parietes being tubular makes a close approach (p. 275 ib) to the very distinct B. cariosus. Certain varieties (p. 453 ib) of Chthamalus stellatus & cirratus have the anterior ramus of the third pair of cirri elongated & antenniform, prefiguring, as it were, the remarkable structure of this same cirrus in Ch. antennatus./

104 v/<Chthamalus Hembeli presents a unique character in the walls of the shell, when old, growing inwards & replacing the basis (Darwin Balanidae p.450)>/

105/Lastly I will give in rather more detail the case which has interested me most, & which combines several considerations. The common Donkey <sometimes is destitute, even when not an albino, of the characteristic transverse stripe on the shoulders>//1 ... [a double shoulder‐stripe]/105A/is said to have been seen in the Koulan of Pallas, now generally admitted not to be the parent of the domestic ass.2 The Hemionus is well known to be characterized by not having the cross shoulder stripe, but a trace of this stripe is asserted to appear occasionally.3

106/The Quagga, though strongly banded in the front part of

1 [The information in the cancelled passage is repeated in Variation, I, 63, where Darwin adds the note: 'One case is given by Martin, "The Horse", p. 205.' The manuscript folio is sheared off at this point. The broken continuity of the text is restored from the corresponding passage in the first edition of the Origin, ch. v, p. 163. See also Variation, II, p. 43, on Hemionus.]

2 Dict. Class. d'Hist. Nat. Tom 3. p. 563. [Desmoulins, art. 'Cheval'.]

3 Horses, Naturalist Library by Col. H. Smith p. 318. Also E. Blyth in the India Sporting Review 1856. p. 320.


the body is without stripes on the legs; but one individual which Lord Derby1 kept alive had a few distinct zebra‐like transverse bars on the hocks.

Again in the Horse, dun or mouse‐coloured or eel‐back ponys & horses invariably have (I believe) a dark stripe down the spine, as in the Hemionous, sometimes a transverse shoulder stripe, as in the Donkey, & sometimes dark zebra‐like bars on the legs as I have myself seen. I have heard of cream coloured horses with the dorsal stripe in India & others with the transverse shoulder stripe in S. Wales & in other parts of the world. A friend has likewise informed me that he had a brownish horse with the spinal & shoulder stripe. I have been informed by two other friends that they have seen Roans with the spinal stripe. Chesnut horses, also, of very different breeds not rarely have a dark & well defined stripe down the back. Col. Hamilton Smith, who has given numerous most curious facts on this subject,2 believes that the Dun Ponies have originated in a distinct, wild race or species: they are found in Iceland, commonly (as informed by a friend) in Norway/107/ Spain, near the Indus, & in the great islands of the Malay Archipelago, everywhere characterized by the longitudinal stripe, occasional shoulder stripe & bars on the legs. It is a very ancient race, & existed (together with cream‐coloured horses <ch we have seen also have the dorsal & shoulder stripe> the times of Alexander & are either truly wild or feral in the East, & were so at no very remote period in parks in Prussia. It is admitted by Col. Smith that duns appear occasionally in herds of variously coloured horses, but he would account for all such cases by a cross at some time from his dun‐stock; I suspect, considering the wide range & antiquity of this colour & its occurrence in wild breeds, that it might be argued with much probability that this was the aboriginal colour of the <aboriginal parents> all our domestic horses. However this may be, the shoulder stripe & bars on the leg are now only an occasional appearance. /107 v/It seems to me a bold hypothesis to attribute the spinal stripe in roan, creamcolour & chesnut horses to a cross at some time with a Dun./107/ In regard to the chesnut colour said to be strongly inheritable Col. Smith, who admits so freely various wild stocks, doubts about there having been one of this colour as it is characteristic of every breed; & Hofacker3 shows that chesnuts/108/are bred from both

1 Gleanings from the Menageries of Knowsley Hall 1850. p. 71 a splendid work by Dr. J. E. Gray.

2 Horses, <p. xi Preface> p. 100, 156 to 163, <275> 280, <288>.

3 Ueber die Eigenschaften &c 1828. s. 12. [Darwin later noted in pencil:] I must see to this, per[haps?] I translated [?] wrong??


parents of different colour. The stripe is only occasionally present; it has been seen in common chesnut horses, in the heaviest dray horses. I have seen it in a remarkably small pony from India. Hence I believe that the chesnut colour & probably the Roan itself are variations, & the dorsal stripe an occasional concomitant of these colours.

Here then in the horse, Donkey, & other equine animals we have several cases under domestication & in a state of nature, of variations analogous in one variety to another variety & to allied species. Remembering in how remarkable manner in pigeons the blue colour & allied tints with black wing bars &c were brought out by crossing the most distinct breeds, let us see what is the result of crossing the various species of the Horse genus. But first, let me remark that it would appear that the Dun Ponys & chesnut Horses with these asinine marks often appear from the crossing of two breeds of the Horse: this certainly is the case with the socalled Kutch or Kahteawar breed1 "which are generally greys or light duns & almost invariably have the zebra marks on the legs with list down the back"; & these are bred from a Kutch mare & an Arab sire; & it is asserted2 /109/that Arabs are never duns. Now for crosses between species: Rollin [sic]3 asserts that the common mule between ass & Horse are particularly liable to the zebra marks on the legs. <Burchell's zebra (E. Burchellii) is not striped on the legs, but hybrids between it & common ass in two instances were plainly barred on the legs4 >. In Lord Morton's famous case5 of the hybrid from a male Quagga & a chesnut mare (not thorough bred), & in the two subsequent colts from the same mare & a black Arabian, the bars across the legs were "more strongly defined & darker than those on the legs of the quagga, which are very slightly marked": indeed it can hardly be/110/said that the Quagga has ordinarily any bars on the leg. Lastly & this seems even a more curious case than the last in regard to our present subject, the Hemionus differs from the Ass in having the spinal stripe but not the cross shoulder stripe & with the legs without any trace of bars but a hybrid figured in that splendid

1 Col. H. Smith. Horses p. xi on the authority of Major Gwatkin, Stud‐Master.

2 Col. H. Smith p. 211.

3 (Blank space for reference. See Roulin, Acad. sci., Paris Mém. divers savans 6 (1835), 338.] Mr. Martin in his History of the Horse p. 212 gives a figure of a Spanish mule with the strongest zebra marks on whole length of legs: especially front legs; I have seen a fine cream‐coloured mule with all four legs strongly barred.

4 Martin Horses p. 223. See also the splendid drawings in Dr. .J. E. Gray Gleanings from the Menagerie of Knowsley Hall 1850.

5 Philosophical Transactions. 1821. p. 20.


work, the Knowsley Menagerie, has all four legs with transverse bars; there are even some zebra‐like stripes near the eyes, & on the shoulder there are three short transverse stripes. This last character reminds one of the variety of the common Ass & Koulan with a double shoulder stripe. Dr. J. E. Gray further informs me that he has seen a second hybrid quite like the one figured. Here we see most plainly what an extraordinarily strong tendency there is for the bars to come out in crosses between those species of Horse, which have naturally plain legs.

I will only further remark that in Hybrids from the zebra & ass or Hemionus in which as the one parent has striped legs, stripes on the legs might be expected, it is clear that the stripes are more plainly developed on the legs than elsewhere as may be seen in two figures in the Knowsley/111 /Menagerie, & still more plainly in a Hybrid figured by Mr. Geoffroy & F. Cuvier1 in which there are hardly any stripes on the legs. In one of these hybrids between Ass & Zebra there is a double cross shoulder stripe./ 111 v/In two hybrids from the common ass with plain legs & Burchells Zebra, the legs are barred quite as plainly, perhaps rather more plainly, than in Burchells Zebra.2 /111/Again in the offspring from a Bay mare & a hybrid Ass‐Zebra, the bars on the legs are to be seen, & I was assured at the Zoological gardens were extremely conspicuous when the animal was young. It may be noticed in connection with dun Ponys that in several of these hybrids, dun or slate‐like tints prevail.

What shall we say on these facts? Those who believe in the independent creation of species—& if there does exist such a thing as a species distinct from a permanent variety, undoubtedly these equine animals offer perfect examples—will say that they have been created with an organization so much in common, that under certain unnatural conditions & crosses, characters appear which mock those in animals created in other & remote countries: they/ 112/will have to admit that the bars on the legs of the zebra were so created & more strongly inheritable than the bars on the body; but that the similar bars occasionally appearing on the ass or on the several above hybrids are due to variation. It seems to me far more satisfactory to follow the striking analogy of domestic pigeons & attribute all the cases to one common cause, viz community of descent. Let it be remembered that the races of domestic Pigeons differ more from each other in external appearance than do the several equine species; & that in all the races when from

1 Hist. Nat. des Mammiferes 1820. Tom I.

2 Gleanings from the Knowsley Menagerie; the skin there figured I have seen in the British Museum: see also Martin on the Horse p. 223.


simple variation or crossing a blue tint appears (comparable to the dun in Horses) almost invariably the black wing‐bars appear (comparable to those on the legs of the horse, ass &c) often accompanied by other characters, as white rump &c (comparable to shoulder stripe &c). But although these colours & markings appear in the several breeds of Pigeons the form of head & body &c do not alter; & so it is with the equine animals when they become occasionally striped & barred.1 From the facts previously given, it is possible that the bars & stripes on the several equine animals might be analogous variations from the/113/several species having inherited a common organization, but the concurrence of several characters & more especially the characters being brought out by crossing—a cause wholly unlike that which produces the bars on the aboriginal parts seems to me clearly to indicate reversion to ancestral character;2 —this ancestral character being latent in the young of each generation & occasionally brought out when the organization is disturbed by a cross or other cause: hence, probably, it is that the stripes on the legs of the common Donkey are said to be plainest in early youth, as they were in the complex cross of Ass, Zebra & Horse. If is to my mind very interesting thus to get a glimpse into the far past, millions of generations ago, & see a dun‐coloured animal, wi