264 NOTES ON rOLOUR INHERITANCE IN MAIZE. 



statistical proof of Mendel's law established up to the present 

 date. 



East and Hayes found in certain cases a ratio of 15 yellow 

 to 1 non-yellow among the progeny of a self -pollinated hetero- 

 zygotc between yellow and white, indicating the presence of 

 two separate dominant yellow-producing factors. No similar 

 ratio ever apjicared in my earUer experiments. In two 

 experiments I described the appearance of certain very pale 

 yellow graiiLs, scarcely distinguishable from white, amongst 

 the progeny of plants of very mixed ancestry. The discrimi- 

 nation of this type of grains was a matter of so much difficulty 

 that the precise ratio in which they occurred was not definitely 

 ascertainable, but they were stated to have made up about 

 10 jKT cent, of the total number of grains. 



There can be very Uttle doubt that the real ratio was 12 

 yellow : 3 white : 1 very pale yellow, the pale yellow being a 

 separate recessive character, and therefore quite distinct from 

 either of the yellows studied by East and Hayes. The non- 

 yellow grains of maize studied by me were never by any means 

 pure white in colour, and the very pale yellow grains might 

 almost as well have been called " dark white." In the state of 

 knowledge of Mcndeliaii phenomena existing in 15)04, it was 

 almost juslifiable to group together both classes of non -yellow 

 grains, i.e., " white ' and recessive yellow, and to record the 

 simple Mendclian ratio between yellow and non-yellow in the 

 exceptional cases in which the " very pale yellow " character 

 appeared. It must be admitted that my earlier account of 

 this i)henomen()n was not clearly expressed, but the account 

 of the deviation from a strict Mendelian ratio then intended 

 was to all intent^s and purposes the same as that now given. 



Amongst non-purple grains produced by the self-fertilized 

 plaiitH recorded in llu» present note, the ratio of yellow to 

 whit*.', in cases where both colours appeared, usually approxi- 

 mat^'d to 3 : 1. Entry No. 20 of the table constitutes an 

 exception, since among non-purple grains there are only 11 per 

 cent. «)f non-y«l|()\\s. [f this should prove to represent a 

 ratio of If) ; 1 similar to those described by East and Hayes, 

 it in the first example of the kind which I have witnessed 

 (exjK.*etation 67 i 1 per cent.). 



