36

Mémoires du Museum d' Hist. Nat. Tom. VII.

Mem. sur la Fam. des Cruciferes. M. de Decandolle.

p. 178. the varieties of Raphanus sativus those which have large roots are cultivated for this end, those with small roots bear many more seeds & are used for making oil — so with the vars. of Brassica asperifolia. (case of Balancement).

p. 177. the Anastatica a rose of Jerusalem & some other Crucifers of the deserts roll up & are blown by the wind till damped & shed seeds. —

p. 215. "Their characters (of the order) (mem. M. Gay's paper in Annales) are so fixed that they seem to to admit no passage from one to the other. There is no intermediate step between a radicle on the bord ou sur le dos des cotyledons" &c &c

37

p. 216. Crucifers most nat. fam. in Vegetable Kingdom. —

p. 220 shows that in the orders formed by position of the embryo, which seems the best & most natural arrangement, there is a repetition of the form of pod & other generic characters — a sort of representation in each order — analogy Macleay w^d call it — Can hardly be external conditions.— Mem in this most nat. family have all sprung no doubt from same parent, & therefor have tendeny to vary same way.

This w^d certainly lead to belief that a priori tendeny overrules chance selection. — Refers to M. Correa de Serra Memoires Carpologiques for similar examples. — See to this.

It is extracted in early part of these memoir or annales

38

p. 227. 227 The Cleomas of all the Capparides have evidently most affinity with Crucifers, but no particular genus of latter can be picked out. For if we look to the unilocular crucifers they are only so by abortion, & if they were placed next to Cleomas other affinities w^d be broken. — so with other particulars ( exactly Waterhouse.)

p.228. He originally arranged Crucifers in a circle, (which is only possible in very natural families), but not so practicable as linear arrangement.

39

p.226. (a) In Crucifers & all eminently nat. fams. (ie clearly descended from one stock) it is very difficult to point out order of relations of its genera with the related families. — He doubts real passages (like Waterhouse) & attributes it generally to our ignorance. — (In not nat. families, descended from more than one stock, true passages might be expected.) — Crucifers have undoubted relations with Papaveracees fumariacees & Capparidées — but where commence or end the series of genera? —

Shall I say that the crambe approaches the fumariees than any other genus, on account of its fruit uniloculaire, monospermirridescent— but this affinity is weakened by structure of cotyledons & the fruit in reality has two loges. — (how exactly like the case of rodent-like marsupial).

39 verso

(a) Hooker remarked to me, that natural groups of plants cannot hardly ever (thus monocot & Dicot. or Orchideae) be separated by any one character — This is different from Animals. — Shows plants more unbroken series, descendants from more parents — or parents less widely separated. —