92 THE ORIGIN OF MONOCOTYLEDONS [CHAP. 



that the proportion of families would be greater 

 which are more or less aquatic. Such proves 

 to be the case, when percentages are taken ; for 

 such amount to 33 per cent., while among dicoty- 

 ledons they are only 4 per cent. Such a marked 

 difference, one must suppose, points to some con- 

 nection between an aquatic habit and structure ; 

 for since dicotyledons are more than four times 

 as numerous as monocotyledons, the chances in 

 favour of a preponderance of the former occurring 

 in rivers, lakes, ponds, etc., is much the greater. 

 Thus, e.g.., Mr Guppy found about twenty-eight 

 seeds and fragments of plants capable of grow- 

 ing, of dicotyledons in the Thames and Lea ; but 

 only about fourteen of monocotyledons. 1 I will 

 now consider some points of similarity between 

 these two classes of angiosperms. 



Commencing with the embryo, the first 

 observation is that the development from the 

 egg-cell to the pro-embryonic condition is the 

 same in both classes ; i.e. up to the period and 

 even beyond it, of the quadrature of the terminal 

 cell. The visible differences begin when the 

 globular pro-embryo becomes unsymmetrical in 

 monocotyledons. With regard to the origin of 

 the single cotyledon in the latter M. Ph. van 

 Tieghem observes that, of the two cells formed 

 by the first longitudinal partition of the embryo 

 cell at the end of the suspensor, if they are 



1 " The River ^Thames as an Agent in Plant Dispersal," Journ. 

 Lin Soc. Bot., vol. xxix. p. 333. 



