276 Inheritance of Glume Length and of Colour m Oats 



particoloureds of exactly the same nature are thrown, the selfed whites 

 gave one particoloured ear out of every four in the next generation ^ 

 Now this last is a feature which marked certain non-purples in the 

 second cross. One is inclined to ask whether there is really any difference 

 between the particoloureds of the two crosses. Surely in the circum- 

 stances the cP hypothesis cannot alone account for the particolouring 

 in the first maize cross any more than, as will shortly be questioned, the 

 action of an inhibitor can alone be responsible in the second. 



Here, however, a useful interpolation may be made to note that 

 nearly all the above segregations of the first cross, with the exception 

 of the 9:3:4 ratio, are also to be found among the results set out in 

 this paper. It is even possible that the ratio 9 flushed : 3 streaked : 

 4 non-coloured is present too; for if the two groups in Table IV 

 throwing all three colour types be added together, then comparison 

 with expectation is as follows : — 



Observed. 163i^ : Z1S : blN. 

 Expected. 141i^ : 47>Sf : mN. 



Bearing in mind the small number of plants, one can hardly say the 

 discrepancy is very much worse than in East and Hayes' 



Observed. 638 purple : 210 particoloured : 306 non-coloured. 

 Expected. 649 „ : 216 „ : 289 „ ^ 



To return to the second maize cross. Here East and Hayes hold 

 that the cP seeds are whites and that the particoloureds develop when 

 the colour inhibitor I undoubtedly operating in this cross is present 

 heterozygously along with P and C. The fact that in this case no 

 particoloureds breed true favours the conception. But if this be so, 

 why do quite as many of these light purples (particoloureds) yield ratios 

 of 3 particoloured : 1 non-purple and 1 particoloured : 3 non-purple as 

 give 1 : 1 (possibly 9 : 7 or 7 : 9) ? [Note that the first of these is also 

 thrown by streakeds in the wheat cross, but not the others, which come 

 from non-coloureds.] Also, unless the non-purples lacked P and C, 

 they must by the theory be homozygous for I and therefore incapable 

 of throwing particoloureds. But this is not so. For in two cases the 

 following proportions were found ^: 



' East and Hayes, p. 78. 



^ East and Hayes, p. 72. Table 21 A. Starcliy seeds only, i e. the most favourable 

 Eelection. 



» East and Hayes, p. 100. Table 25 E. 



i 



