THE VASCULAR CRYPTOGAMS 75 



ance with changes in the method of fertilisation. The 

 fact that the pollen-grains are set free and germinate, 

 clearly indicates that they mark the beginning of a new 

 generation, while the production of active sperrnatozoids in 

 some Gymnosperms links them directly with the Cryp- 

 togams. The female prothallus, though not less bulky than 

 that of Selaginella, remains for ever shut up in the megas- 

 pore, which itself never gets free from the sporangium ; 

 so on the female side all trace of an independent existence 

 of the prothallus has been lost, and, except for minute 

 developmental research, we should never have suspected 

 the presence of an oophyte at all. 



In the Angiosperms matters are still worse for the 

 sexual generation. On the male side, indeed, there is 

 no great change, except that the homologue of the 

 antheridium is harder to recognise, but in the embryo-sac 

 the prothallus is scarcely to be traced, and its very 

 origin has been pushed out of its proper place, so that 

 most of it (the endosperm) has come to be an after- 

 product of fertilisation. In fact, the sexual generation 

 in Angiosperms has become so thoroughly incorporated 

 with the asexual, that it seems almost an affectation 

 here to talk of alternating generations at all, and 

 certainly the existence of such an alternation would 

 never have been discovered except by the comparison 

 with Cryptogams. The clue afforded by the life-history 

 of the Ferns has thus enabled botanists to follow 

 accurately the true course of development in the higher 

 plants, which otherwise we should never have under- 

 stood. 



The regular alternation of sexual and asexual indi- 

 viduals is often modified in special cases among Ferns. 

 The modification may either result in a lengthening or a 



