34 



no effect is produced, and in such cases the fecundation of the embryo- 

 sac nucleus probal^ly did not take place. 



It is not improbable that in some cases the second sperm nucleus 

 enters the embryo sac, but fails to unite with the two polar nuclei. In 

 such cases it may be able to form a spindle and divide separately, the 

 unfecundated embryo sac nucleus formed by the union of the two polar 

 nuclei also dividing separately. If this occurs, there would then be 

 formed, in the protoplasm of the embryo sac, nuclei of two distinct 

 characters, one group from the division of the embryo sac nucleus and 

 the other from the division of the sperm nucleus. 



The nuclei which form the endosperm reproduce free in the cyto- 

 plasm of the embryo sac, the cell walls which delimit them not form- 

 ing for some time, probably not until nearly the ultimate number of 

 nuclei have been formed. Before a very large number is formed they 

 move out from the center of the embryo sac and usually come to lie 

 near the wall, where the nuclear reproduction continues. In the 

 delimitation of the nuclei later b}' the formation of walls about them 

 the process of wall formation begins at the periphery of the embryo 

 sac and gradually extends inward. This process of endosperm forma- 

 tion, which is common in seed plants and is known to occur in wheat, 

 is doubtless the process occurring in corn. If. then, the union of the 

 second sperm cell with the embryo sac nucleus fails to take place and 

 they both divide separately, when the nuclei formed by their divi- 

 sion move out to the surface of the embryo sac the nuclei of male and 

 female origin would probably become more or less interspersed. In 

 their further division we would then have groups of cells of male 

 origin here and there, interspersed between those of female origin. It 

 is probable that when the nuclei become dispersed around the outer 

 portion of the embryo sac they largely retain their individual location 

 and sphere of influence, so that the nuclei resulting from the division 

 of one nucleus after it has assumed a location at the surface remain 

 near together and, ultimately, when delimited by walls, form a 

 group of ceUs derived from the same nucleus. If this is the case, 

 wherever a nucleus derived from the division of the second sperm 

 nucleus is located, after migration to the periphery of the embryo sac, 

 there would be formed ultimately an island of tissue which, being 

 derived entirely from the male parent so far as the nucleus is con- 

 cerned, would probably exhibit almost entirely the characters of the 

 male parent. This hypothesis would account for the occurrence of 

 such variegated kernels as were produced in experiment la (PI. I, 

 figs. 9-12). 



When the migration of the nuclei from the center of the embryo 

 sac to the periphery occurs, if the nuclei derived from the male nucleus 

 have remained grouped together, as would probably occur, while those 

 derived from the division of the embrvo sac nucleus have remained 



