NOTES ON COLOUR INHERITANCE IN JVIAIZE. 261 



Hayes, is clearly preferable. It is hoped to put it to the test 

 by growing a further generation. 



So far everything is in accordance with the hypothesis of 

 East and Hayes. The last entry of the table appears to 

 constitute an exception, since we have here an apparent ratio 

 of 3 non-purple to 1 purple arising from a purple parent grain. 

 In order to account for this phenomenon it is necessary to take 

 into consideration the differences which the purple grains show 

 among themselves. 



All the purple grains of the plants recorded in entries 10 and 

 15 showed a curious intermediate tinge not readily definable. 

 These require further study, and will be omitted 'from the 

 present discussion. 



Many of the grauis tabulated as deep purple possessed more 

 or less a reddish tinge as opposed to purple, and it is possible 

 that in some cases a ratio of 3 purple to 1 red might have been 

 made out. It was found, however, that a sharp distinction 

 between red and purple could not be made by eye with any 

 real constancy. Both classes, if classes they are, vary much 

 in tint. In the three cases, however, in which the presence of 

 pale purples is recorded, there was never any doubt about the 

 distinction between a dark and a pale purple. Some of the 

 pale purples are very pale indeed, and scarcely distinguishable 

 from white , and it seems clear that the classes pale purple and 

 non-pm-ple may intergrade to some extent. It seems necessary 

 to suppose that some of the very pale purple grains arise by 

 partial failm'e of the inhibiting factor, though present, since we 

 can only account for entry No. 21 by supposing that it is possible 

 for a pale purple grain to carry the inhibiting factor, which 

 is nevertheless fully effective in the majority of the offspring 

 obtamed . This possibiHty is also suggested by East and Hayes. 



The supposition that pale purple grades into non-purple 

 may also explain the deficiency of pale purples in entries 13 

 and 19, where a ratio of 3 dark purples to 1 pale purple is 

 apparently to be expected. 



Among uncovered cobs from the same field, which arose from 

 promiscuous pollination, were several which contained a large 

 majority of dark purple grains and were evidently derived 

 from plants which were homozygous purple in constitution. 



