IN THE TROPICS : III. 



L21 



F3 



Grains of each form and colour from the three series describ- 

 ed above were sown, and pollination was effected by growing 

 plants of G. W. S. among them. Owing to the time of flowering 

 of the latter variety having been miscalculated, only a few 

 plants set any grains, and such as did so produced only a few 

 grains per plant. In the cases described enough grains re- 

 mained to indicate the nature of the plant in question. 



We have already seen reason to suppose that heterozygote 

 grains may be white as well as blue. We may now hope, by 

 the examination of F 3 . derived from white F^ grains, to 

 distinguish which of these white grains are pure homo zygotes 

 and which heterozygotes ; since the latter may be expected to 

 produce a certain number of black grains on again pollinating 

 by white, whilst pure white grains so treated will produce 

 only white. The following table sums up shortly the results 

 obtained in this generation : — 



Table 16. 



Number of F., Plants, which produced Grains 



(F 3 ) of the characters stated below, 

 when pollinated by White. 



... .... . Black and All 



All White. white _ Black _ 



F.j Grains 

 belonged 

 to Series. 



1 



•1 



( folour of 

 F 3 Grains 



I White 

 ) I Mack 



\ White 

 } Black 



White 



Black 



Total. 



Hi 



21 



19 



15 



22 



8 

 17 



8 

 54 



31 



21 



29 



IT 



27 



Besides the information with regard to the offspring of white 

 Fg grains, the table shows that black grains always produced 

 some black offspring. This indicates that the two germinal 

 nuclei of any pollen grain always bear the same allelomorph, 

 and so do the oosphere and combined polar nuclei of any 

 female flower. Consequently segregation of the characters 

 under consideration takes place prior to the separation of tho 

 two germinal nuclei in the pollen tube. 



