[15]

[15v]

[17]

(3

(Hybridism of Plants has been considered experimented & discussed with far greater care than that in Animals: Kölreuter & Gærtner almost devoted their lives to this subject, the care, the conscientious accuracy, & the astonishing amount of labour exhibited by these naturalists have been admirable. Kolreuter found that nearly every nearly all species when crossed & resultant hybrids were sterile wh with another &; in the few exceptional cases, he assumed, & probably often rightly that these reputed species shd be called varieties (* give list opposed + to those retried by Gærtner; ascertain by Table). (E Suppl.) Gærtner tried some of these same & followed his ingenious & excellent  plan of counting seeds, & he asserts that those which he tried were not quite so fertile as pure species, & hence he considers them as species. G. now over & over again asserts that no 2 species or their hybrids are ever so fertile as pure parents. Thus give examples from G. (A) Suppl.) of number of seeds in Datura &c &c some better example for I shall have to give Datura elsewhere

[in margin:] (a) Back of Page State that it is now probably impossible to know certainly what some of these plants are what he states to be quite fertile as Hibiscus manihot & vitifolius

(D) (N.B. I must reread p. 208-214 the discussion about ascertaining the maximum number. (Done)

Now, to know the maximum average of seeds somewhat difficult, but passing over this, & admitting his standard, & an excellent idea it must be considered. (I have myself tried the counting in a few cases.) The act of artificial impregnation, (through the castration, though requiring the slitting & even removal of petals, seems to be very little injurious) the applying pollen only once or twice, instead of repeatedly as done by Nature, (possibly quantity of pollen wrong) the isolation of plant in chamber (not green-house) & its being

[in margin:] In the hybrid to eliminate ill-effects of artificial process very right to take maximum; but it wd have been better, I think to have taken an average of wild plants.—  So he did.

[17v]

see pencil note on Page by Hooker's authority

(a 3d Fortsetstung p. 114. Hibiscus manihot & vitifolius are said to be quite fertile, but the plants now so called seem to be, (as shown me by Hooker) quite distinct.

p 118 Sida Crista again doubtful case.

Acta Acad. 1782. Par II. p 255. Kolreuter says Malva sylvestris fl. alba & mauritiana flor carnes.  & increase cross gave offspring of incorruptæ fœcunditales. But Gærtner in table does not gives only 1a: but one of the hybrid itself viz M. mauritiana sylvestris is marked (K). The reciprocity (p 197) as not equal as 10:7 10:7

p. 282. species close; offspring offspring not variable p. 168.—

No one seems to certainly in Steudel to has considered these Malvæ as the same.—

[18]

(3

Supp. p. 3. M. S.)

Gærtner: p. 195. In Dianthus pure species normal seed vary from in maximum 80 to 120 (A) (in hybrid cross. 2— 54.— — to maximum.). p. 216. Decimal table of Dianthus— Table also of Verbascum —p 218 Lychnis dioica with Cucubalus Lychnis flos.— Silene.

Gærtner p. 414 [few words illeg] gives. the list from Kölreuter, which will save me copying it, if it deserves being copied.— There are only 10, of which 3 seem to have been tried over again by Gærtner.— viz Datura. stramonium; & tatula & D. lævis. & the two Malvæ.— p 578 another & rather different list,— Linum not in 2d list.— In 2d list 2 more forms of Malva Malva & Hyosciamus (see to Kölreuter ). The Hyosciamus proved by Gærtner to be really a var. (I have looked to Köl & find the first list is the one to quote. If to this first list (p. 414 of Gærtner), Hyosciamus be added, then Gærtner's statement must be given: but had better be omitted, all about Hyosciamus..). N.B . I think I must quote Kölreuter case (a) state how careful an observer: that he in some few instances counted seeds. Kolreuter's observations apply more especially to the fertility of the resultant hybrid & not so much to first cross. It is, I think, impossible (I speak after trying with Dr. Hookers aid) to ascertain what some of Köl. plants were. But I think Gærtner rather begs the case when he asserts that all are less fertile than pure species. There is nothing to show that he tried all: from Table he certainly did not try all: On other hand Kölreuter into the difficulty at once by pronouncing all these to be vars. because they are fertile; though probably some are vars. It is important & shows how difficult the test of fertility is as applied to species to see the two best observers who ever lived, thus differing.—

[18v]

(a). p 21 of my abstract of Kolreuter — perhaps give cases of Mirabilis vulg. & dichotom. as very fertile, but here Kolreuter introduces new test; to which I tr may add the other test of very different fertility with of the two with third species.

[19]

(a)in the offspring of self impregnated hybrids, (p 439.). there are sometimes throwing backs to generally the mother pure parent forms, but these are not more fertile, generally less fertile than pure the hybrid parent (Q): here again we see that abrupt assumption of pure ancestral form = not accompanied by increased fertility. Gærtner seems to have observed, it gives detail of the case viz of the hybrid Dianthus Armeria deltoides, which generated yielded seed for ten generations having sown itself in the garden for the first 6 or 10 8

[20]

* With respect to Hybrids this has been shown in a [2 words illeg], & I infer this it in regard to Mongrels, only from the description by Kolreuter of the reciprocal crosses of several varieties of Mirabilis (2d Fortst. p. 126; Nov. Acta Petrop. 1795, p. 333, 1797. p. 393), of two Stocks (2d Fortst. p. 128) & of Digitalis thapsi & purpurea, (Journal de Physique 1782 p. 285). In some few cases, I see it as like a plant [illeg] as especially in the genus Digitalis, true hybrids from reciprocal crosses are not alike; & this seems to have been the case with some of the mongrel cabbages. X (p. 10, & 11. Bastarderzeugung) described by Wiegmann. Moreover X in crossing true species the likeness of the reciprocal crosses (p 223) specially holds good in wild plants & which have not been introduced &  brought into the state of varieties.

[21]

Gærtner doubts disputes (Bastarder. p. 414, p. 539.) even Dianthus glaucus & deltoides, being quite fertile together & therefore being varieties; though Kölreuter (3d Fortsez. found p. 94 found that tr seed of the wild D. glaucus did not come up true) & I see that they are considered as varieties by Hooker & even Babington in their British Floras.

[22]

Suppl. to p. m.s. p. 3. (D)

Gærtner (p 202). He finds it not difficult to get average (I may say that I have likewise tried) of seeds in a state of nature & as here state of nature impossible in the fullest & finest fruit of any species for they do not differ to any very great extent.— But of the hybrid-fructification, he takes maximum, in order to avoid, as much as possible ill effects of treatment & in order to get this trusts only to experiments made on several flowers & plants during successive years — Strictly it ought to have been ascertained whether the how many of the hybrid seed, would have germinated, not by doing this to seeds from slow germination & would be exposed to many risks; & it seems to be feasible to that assume view of the greater fertility of hybrids to assume that all the seeds, which look good, are equally good in the hybrid as in the pure species,— Nevertheless for ascertaining the one question of fertility it is much to be regretted that he did not take artificially se castrated self impregnated & secluded plants as the standard of measure— To a certain extent it seems to me to weaken all his conclusions on this one very important point, though as we shall immediately see it by no means accounts for the universality of his result viz in the lessened fertility of hybrids..— He fully admits that artificial impregnation is injurious, (see Beitrage p. 366) (aa)  & the only reason that he gives for taking his standard for seeds yielded in state of nature is (p. 212) that the difference in number between them & the same species artificially impregnated is never so great, as between latter & the hybrid union of nearest species. Then why in table is i(a) so often affixed to cases of own species by own pollen? —

[22v]

(aa): To give one example, taken by hazard, he states (Bastarderzeug p. 385) that Lychnis vespertina naturally yields 210-230 good seed, but artificially fertilised with its own pollen gave 192 seed.—

[23]

a) Supplt. p. 4 M.S. (A)

Lychnis & Back Genera

I have ascertained that Gærtner artificially self impregnated, with apparently due isolation & every precaution the same, as he took with Hybrids 20 species, & to this may be added crosses of 13 varieties, such ( (*counting vars, such as no one has yet ever for one minute doubted are really varieties) making altogether 33 cases; (*& I sh here state that I have omitted certain plants in which he failed to make any cross whatever, as he perhaps in such cases did not Know the proper time or place of impregnation. I have, also omitted Leguminosæ on account of difficulty & injury in act of castration: the effect of these omissions is to make the proportion of cases in which he got (K) rather greater than it would have been, had I taken every case given by him without correction.)

Well of the 33 cases 33-17=16 only 17 gave (K) or normal fertility; ie nearly just above half.— So that we may assume that the injurious effect of art. impreg. will lessen in every other case the number of seeds yielded. But if we look to the hybrid crosses between pla distinct species, Gærtner is never wary of saying that he never got once the full number or what he w. designate as K": considering the table at end alone, it w. seem work on when Lychnis considered that he occasional once or twice got full number in crossing Lychnis diurna & vespertina & reverse , which however some Botanists as we shall presently see consider only varieties.—

(* In table the cross between Mathiola annua & glabra is marked (K), but in text the contrary is distinctly stated p. 102 p 197 &c) I have observed a few other contradictions in Table compared to Text, which makes me a little doubtful—

(over

[23v]

It must be observed that we are not speaking here of fertility of resultant hybrids. in which (* judging from the table, (without the hybrids Malva Mauritiana-sylvestris. Lychnis vespertina -diurna, & the hybrid from this & L. diurno-vespertina;— Verbascum-albo-pyramidatum, (I suspect a mistake) Lobelia-cardinali-fulgens, & reciprocal Hybrid, though this last & Lychnis & others contradicted in text p. 385] he also expressly states that in no one case, is a Hybrid so fertile as pure parent.

[24]

(B

Supplement. to p. 4

I have just said that Gærtner maintains p. 334 & p. 540 that no hybrid whatever has same fertility of pure parents, & this was Kolreuter conclusion, (with the exceptions named, now let us consider fertility of hybrid character & which from this very reason he Kolreuter considered as only varieties); & this remark is, applicable applied by them not only to the first generation of hybrids, self impregnated, (for of course ?? I do not here refer to crosses with either parent or other species), but to successive generations. In the successive generations of hybrids & successive generations (The most prominent fact as shown by remarks of Kolreuter & Gærtner in that the regard to fertility of hybridism is that it is becomes excessively variable. In the seedlings raised from the same first union of the two pure species, even out of the same capsule, & under the very same condition, the fertility is very variable (*Gartner p. 366 (Kolreuter 1 Fortzet. p. 14)): so again the same hybrid produced one year will sometimes (p. 391 p. 385) be far more fertile than a similar hybrid raised another year (perhaps quote case ( Aquilegia vulgaris canadensis thus the hybrid one year was so fertile as to sow itself, another year bore extremely few seeds —). (It is very singular that when an abnormal Exceptional type (p. 244) is raised amongst a set of hybrids, ie one departing from the usual external character of the the such hybrids in question it is usually almost always quite sterile, even though the normal hybrids may have considerable fertility. These "exceptional types" (p 245) approach more resemble one of the pure parents, which a priori it might have been thought would have added to their fertility. I may add that generally those hybrids which normally are not intermediate between the two pure parents, but take strongly after one forming the "decided type" of Gærtner

(p. 286—) are generally very sterile.—

When a Hybrids are crossed with pollen of either parent for successive generations, although the plant offspring gradually assume the character of to the pure parent & acquire fertility, (p.450) but yet perfect fertility specially p 350, 355 on under side is the element last (p. 459.) acquired. a plant (p. 460) (p. 449) could even come to perfect quite resemble its pure parent & yet be totally sterile!

[in margin:] Put in Back References to Laws of sterility

[24v]

I may say already shown that during "reduction" fertility does not increase with resemblance to pure parent (showing by way how independent of classificatory resemblance) (especially p. 350, 355 G on male side)

(C

Suppl. to p. 4.

(and during the absorption "reduction" the fertility (Gær p 461; Kolreuter 2 Fort. p. 98) is excessively variable) 

(All observers write unanimously that very many species cannot be crossed — under gradation)

Gærtner B. 10, 32, 384 Hybrids sensitive to ill treatment

[25]

p. (C

Supplement

Thus the hybrid names characterised by Gærtner (p 447) Nicotiana paniculato-rustica, (which means that the great-great-grandmother was pure N. paniculata, all nearer relations being pure N. rustica (a)) though differed only from pure N. rustica in producing less seed.

So Kölreuter in all respects made similar observations & a course pair of reciprocal crosses reciprocal between those two species in a reverse Experiment to the last of Gærtners (3d Fort. p. 47) found that the hybrid resulting was, as above, from four some crosses of the pure N. paniculata during 4 generations with N. rustica, was quite like N. paniculata in appearance but its pollen was not so good & specially in Autumn contained more empty & useless grains than posses the pure pollen.—  Again Kolreuter (N. A. 1797. p 373 (p. 25 my M.S.) that the hybrid grand child of from Mirabilis jalappa or longiflora Jalappa vulgaris & longiflora, but having 3/4 blood of M. longiflora (male) was more sterile than that the its hybrid mother; & this grandchild, without any other cross with its own pollen produced fewer seedlings, on great grandchildren of pure parents of which several were quite sterile,— vide Back of Page (III)

If we now turn to In hybrids crosses fertilised with their own pollen, Gærtner repeatedly states that those generally even the more fruitful hybrids always never increase but often decrease in fertility lessen in successive generations (p. 418 to 421) so that even somewhat the offspring cannot even be fertilised by the pollen of One of the pure parent

(As in the first cross between the pure species, the degree of fertility is very variable. Fertility is indeed (p 554) the most variable of all characters. He states expressly (p. 420) that in no case has he known more fertility in second generation of hybrids with own pollen than in first with own pollen— Again

[25v]

or to use the language of Breeders there wd be 1/16 blood of N. paniculata

[26]

Supp to Supp to p. 4 C.

Kolreuter describes another case (a) (ZZZ) (Nova Acta 1795. p. 324 p. 376 ) in which a hybrid with 3/4 of Mirabilis vulgaris ( [illeg] ) & M. — 1/4 of M. longiflora, produced seven seedlings which were sufficiently fertile; & three 3 of these 7 produced with their own pollen fifteen plants, which yielded few seed only in autumn: nevertheless, one of these yielded nine seedlings which produced the seeds of which were. mostly good for nothing. These latter seedlings were the gr-gr gr grandchildren of the two pure parents. Although in this & the last case we have 3/4 not a simple hybrid, but a hybrid with 3/4 of its blood of one pure parent, yet we see when allowed to breed fertilised with own pollen for some successive generations how sterile it Keeps.. Nova (all Nova Acta 1795. p. 324. & 1797 p 375,)

[27]

(D

Suppl. to p. 4

p. 553 G. Bastard & p 421 generation, also each generation, however; it yielded less seed & at the 10 its fertility was quite lost. (this Dianthus armeria deltoides hybrid was further remarkable not varying as is so general with hybrids during successive descents.)

fr—. retaining its hybrid form quite constant —

[28]

(Supplement to p 4. D.)

Kölreuter gives very similar facts to these & I will abstract the 2 most striking experiments made on the Marvel of Peru. Two hybrid plants from a cross the union of Mirabilis Jalapa f & M. longiflora m ga spontaneously produced without artificial fertilisation 16 plants, ie grandchildren of the pure parent species, most of which were very little sterile, but one produced nine great-grand-children of the pure species. Four of these nine were very sterile little fertile & together pro yielded only ten plants, which were excessively infertile sterile; & one only yielded three three great-great-great gr. gr. gr. grandchildren of pure ancestor.

The fertility of the 3 not stated, but they do not seem to have yielded any offspring. (Nova Acta 1795. p. 332 to 1797. p. 373, p 381, p 382, 392, 403.—)

Another case is more interesting. Kölreuter found (Nova Acta 1793. p. 394) that when the hybrids from Mirabilis Jalapa (f) was fertilised with pollen of & M. dichotoma, (m.) it yielded (or from the reciprocal union) were little or not at all less fertile than pure parents; & he shd. even have doubted whether they were true hybrids had it not been for their "portentous stature",— & curious fo character in Hybrids, presumably hereafter to which has been alluded to.— discussed in an earlier chapter One of these hybrids, M. Jalapâ-dichotoma, spontaneously produced 28 seedlings, of which 14 were more fertile than itself,— some of the 14 were exceeding in fertility even the pure

[29]

(E

Suppl. to p. 4.

grandfather mother M. Jalapa: but the other 14 were considerably less fertile than the hybrid parent, some producing "very few (per pauca) see good seeds." (Nova 1795 p. 316.). Subsequently he describes in more detail (Nova Acta 1797 p. 383-89) some of the 28 grandchildren of of that two fertile species from one of these, children most fertile (feracissimus) produced four plants great grand children of the pure parent were reared: the, 34 plants, sprung from 8 of the most fertile (& therefore very fertile) "feraccissimus" fertilatalis compianæ of the 28 fertile grandchildren of these same  2 pure species; of the 34, only 13 ever pr one produced an abundance of seed, & nine were excessively sterile, the remainder in intermediate degrees, only once three great, great grandchildren were raised. Here we see an evident decrease in fertility in the grandchildren of the first very fertile Hybrid.—

 It is impossible to study the many details given (I here give only one or two only Considering these facts given by Kölreuter & Gærtner, & the general statement of the latter that the successive generations of simple hybrids, & as we have seen of less simple hybrids with more of one species than of the other in its constitution, all becomes less & less fœcund, it seems, difficult to avoid the full belief that this is a general law. But there is one statement, repeatedly made by Gærtner, which makes me doubt suspect that if the the more fertile numerous hybrids of the same Kind could be left, in numbers in numbers growing freely together themselves this increase of sterility would be diminished lessened

[30]

(f

Suppl. to p. 4)

or obli eliminated. For Gærtner often states (p. 418, 421, 554) that hybrids, even little fertile Kinds, if artificially impregnated with their own pollen of their own hybrid Kind for some generations, decidedly improve in fertility. Now artificial impregnation according to Gærtner's method implies castration of the anthers at an early period, isolation in a chamber of the plant thus rendered female, & dusting with pollen necessarily of another flower & probably of another individual of the same Kind. But these several processes *, as we have seen, a few pages back are in so few degrees injurious that with pure species, half the plants thus treated with their own pollen, are rendered less fertile only of 33 pure species thus artificially impregnated only 17 gave the full number of seeds; so that it makes it very surprising that the fertility of the hybrids shd thus by this very means have thus been injured.— (But we shall presently have to consider some very curious & well ascertained facts, showing that several exotics & even some few indigenous species, though having good pollen, as demonstrated by its actions of on other species, will not but fertilise their own pis carpels, though these can be fertilised by the pollen of other species; demonstrating that the male & female organs are mutually sterile towards themselves each other, but not to certain other species).— I believe that this is explicable only by the advantage of a cross which has elsewhere been discussed, this even overbalancing the injury consequent on the process of artificial fertilisation. Now

[30v]

* Beiträge zur Kenntniss der Befruchtung. C. Gærtner p. 332. p. 365.—

[31]

(g

Suppl. to p. 4)

when the hybrids as above stated are artificially fertilised with own pollen, they receive the pollen of another flower or individual & would thus I cannot doubt be greatly beni receive the benefit of a cross. But — on the other hand when hybrids generally be few in number of the same Kind are left to impregnate themselves, on account of the extreme difficulty of preventing crosses with the pure parent or with other species or other hybrids, it is in almost all cases necessary to isolate them at the time of flowering, the so that they would not could only be fertilised by the own pollen of the individual flower & not receiving the benefit of a cross, their own strong tendency to sterility would be increased (a) If left it were possible to leave them freely growing in numbers naturally in the open air, insects & other agencies, would in this, as in other cases, ensure occasional, if not pretty frequent crosses, So that I doubt whether the by Gærtner's method the sterility of successive generations has not been increased. As Kolreuter experimentised on several species of Mirabilis & Nicotiana, I shd suppose infer that the same precautions were taken as by Gærtner to prevent crosses, with other species, & consequently with other individuals of the same hybrid, & the same injurious results would ensue.— If this explanation be not correct, I cannot conceive what other one can be offered.

[31v]

(a) It must, however, be confessed that the strongest case of Gærtners, viz that of Dianthus armeria-deltoides which grew for 6-8 generations in the garden seems opposed to the view here given, for probably there were several plants.

[32]

(h

Suppl. to p. 4

If we Let us now consider the facts given by the third greatest experimentiser Hybridiser who ever lived the Hon. W. Herbert, late Dean of Manchester; who experimentised for during even a longer period than Gærtner, but who never kept quite so systematic records or published in such detail. The general results at which he arrived agree very closely with those of Kolreuter & Gærtner, but with one great difference, that he attributes much more fertility of in many cases to the first cross union of two species & likewise to the resultant hybrid.

He says (Amaryllidacæe p. 354) "it is certainly not correct as a general law, though some have stated it, that the number of seeds in one pericarp in some other in hybrid than in case of natural impregnation; it is true in some cases & the reverse occurs in others."

This great difference may be partly due to mere chance in the nature of the plant on his having accidentally experimentised on on more favourable plants; & possibly to Herbert having worked much partly perhaps some cases he may have overlooked, in slight degree of sterility; but chiefly, I am inclined to think,, to Herbert's renowned famous skill as an Horticulturalist & to his having abunt ample appliances as green-houses & hot-houses, instead of a mere chamber as apparently was the case with Gærtner: for it is certain —(Gærtner, Bastard. p. 10, 32, 384.) that Hybrids the fertility of Hybrids is more influenced by proper treatment & by external conditions than by are pure species.— (To give Herberts principal facts.