STRUCTURE OF SEED. 213 



thus, Euphorbia Canariensis still presents remains of 

 them at the end of one, or even two years. 



We have not as yet any certain proof of cotyledons 

 furnished with stipules, if this do not happen in Trapa 

 natans, where the ascending twin filaments, which are 

 observed near the base of the stem as far as the origin 

 of the cotyledons, appear to be true stipules. 



The primordial leaves which we see sometimes fully 

 developed in the seed with the cotyledons, as in the 

 French-bean, or which are developed immediately after 

 them, are always of an analogous nature to the true leaves 

 of the plant ; but they often differ — 1st, in form ; thus, 

 those of the French-bean are simple and heart-shaped, 

 instead of being trifoliate with oval leaflets ; the differ- 

 ence, however, rarely proceeds as far as this ; 2dly, in 

 size, which is usually less ; 3dly, in position, which, in 

 Dicotyledons, is opposite, or nearly so, even in species 

 where the leaves become alternate as the plant advances 

 in age ; sometimes the change takes place suddenly, as 

 in the French-bean, which has the first two leaves oppo- 

 site, and all the others alternate ; at other times gra- 

 dually, so as to show that the alternate position is a 

 simple degeneration, owing to the mode of development. 

 The contrary takes place in Monocotyledons, which have 

 their leaves alternate. It results from this, that when 

 we see a plant with the lower leaves opposite, we may 

 be pretty certain that it belongs to Dicotyledons ; and if 

 we find one with them alternate, there is great pro- 

 bability that it is a Monocotyledon. 



