48 THE EVOLUTION OF PLANTS 



difference, for in such cases the ovary has come 

 to be little more than an extra seed-coat. 



However, what is important to us to note 

 is that the ancestors of the Angiosperms, if 

 we can ever find them, are likely to have had 

 something approaching to a closed ovary. 



There are other characters connected with 

 the reproduction of Angiosperms quite as 

 important as any we have already men- 

 tioned I mean especially the changes which 

 go on inside the ovule; these processes are 

 such as to distinguish this class at once from 

 all other plants. But at present we will not 

 dwell on these points, because they do not 

 happen to help us in interpreting that part 

 of the fossil record which we are going on 

 to consider. We may recall, however, that in 

 Angiospenns a nutritive tissue, the endosperm, 

 is developed, after fertilisation, in the embryo- 

 sac, to supply food to the embryo. In a large 

 proportion of Angiospenns, both Dicotyledons 

 and Monocotyledons, this tissue is still present 

 when the seed is ripe, and is only used up during 

 germination; this is the case for example in 

 Buttercups, Spurges, Grasses, and Palms. In 

 another large group of families the endosperm 

 is all, or nearly all, consumed while the seed 

 is ripening, so that in the mature condition the 

 embryo practically fills the whole seed: as in 



