CONCLUSION 303 



ments, swim to the necks of the archegonia, through a 

 cavity, filled with sap, which is formed in the upper part 

 of the nucellus. These plants thus present a beautiful 

 transition between the Cryptogamic and Phanerogamic 

 methods of fertilisation. The male cells are conveyed 

 for a certain distance by the growth of the pollen-tube 

 containing them, but they have to complete the journey 

 to the ovum by means of their own movements. 



This remarkable discovery confirms, in the most 

 striking way, the theoretical conclusions at which 

 Hofmeister arrived forty years ago. 



From Gymnosperms to Angiosperrns is another great 

 step, and here the gulf has not yet been bridged. In 

 Angiosperrns the female prothallus has almost dis- 

 appeared, and even the archegonia are no longer recognis- 

 able. The embryo-sac (the equivalent of the megaspore) 

 proceeds, after only a few preliminary divisions, to the 

 formation of the ovum, and the development of the endo- 

 sperm is postponed till after fertilisation, obviously an 

 expedient arrangement, for it is not formed at all unless 

 it is wanted. The processes in the pollen-grain are also 

 simplified. The great characteristic of Angiosperrns is 

 the high development of the flower and fruit. Not only 

 does the megaspore remain enclosed in the sporangium 

 or ovule, but the latter is itself enclosed in the ovary, so 

 that fertilisation has to take place through the mediation 

 of the stigma and style. The remarkable development 

 of the floral leaves, characteristic of most Angiosperrns, 

 is connected with the occurrence of pollination by insects, 

 for which so many Angiosperrns are specially adapted. 

 ' At present we are not in a position to determine 

 either the relation of Angiosperrns to Gymnosperms, 

 nor that of Monocotyledons to Dicotyledons. The latter 



