426 SCIENCE PROGRESS 



nucleus as giving to the effete gametes a new impetus to 

 development, rather than determining the nature of that de- 

 velopment. That this may be necessary is shown by the case 

 of cyclops mentioned above, in which the product of fusion of 

 a supernumerary sperm and polar body aborts after a few 

 divisions. 



Whatever the true history of this obscure process may turn 

 out to be, direct advantage would seem to accrue to the embryo 

 from the supply of a food substance of the same double origin 

 as itself. 1 This is brought out very well in the hybrid maize 

 races alluded to before, where the dissimilarities are obvious. 

 It is very significant to find that whatever the nature of the 

 endosperm produced by the cross, the embryo bears the same 

 endosperm character. 2 We cannot doubt, therefore, that the 

 power to benefit by a particular kind of endosperm is inevitably 

 correlated with its production. 



The unravelling of the history of the phylogenetic evolution 

 of the process of endosperm formation should prove one of the 

 most interesting developments in botany, and if accomplished 

 will go far to solve the problem of the origin of Angiosperms. 



1 See " Recent Work on the Results of Fertilisation in Angiosperms." Annals 

 of Botany, xiv. 1900. 



2 Bateson, " Reports to the Evolution Committee." 



