THEORY OF REVERSION 231 



the sixteen-square Table on p. 226 be true, they must 

 be distributed as shown in the last two columns of 

 the foregoing Table. Nothing short of the carrying 

 out of an experiment of this kind on a large scale 

 will suffice to prove the truth of the Mendelian theory 

 of reversion in this case. 



Assuming that it will stand this test — and I have 

 reasons for believing that it will — we see that we 

 have here a consistent theory of reversion for the 

 first time. Reversion, according to this theory, is 

 due to the meeting in one zygote of the two factors 

 necessary for the production of the ancestral character. 

 These factors had, presumably, at some period 

 become separated and lodged the one in one and the 

 other in the other of the two strains which, when 

 they are mated, produce the reversionary hybrids. 

 In the case of flower colour in Pisum it is not 

 difficult to make a suggestion as to how this 

 occurred. The wild Pisum, of which I have grown 

 plants from seed kindly given me by Mrc Arthur 

 Sutton, has a purple flower. But this is not a hetero- 

 zygous purple but homozygous (i.e. of the formula 

 PPBB), because it breeds true to purpleness and 

 never produces pinks and whites. Where, then, did 

 the pink come from ? This question may be answered 

 by making only one assumption, which is amply 

 warranted by the frequency of analogous instances 

 of the same occurrence throughout the vegetable 

 kingdom. This assumption is that a white-flowered 

 variety of the pea arose by mutation from the purple- 

 flowered by the sudden and simultaneous loss of both 



