490 



Mr. R P. Gregory. 



The intermediate Aaaa would be expected to show the dominant character in 

 less degree than the normal heterozygote ; it is to this class that the 

 intermediates already described are to be assigned. 



The intermediate characters do not, however, provide the most favourable 

 opportunity for putting the hypothesis with which we are dealing to a critical 

 statistical test, because the range of variation among the intermediates is- 

 sufficient in some families to make classification by inspection a matter 

 of difficulty. This difficulty will no doubt decrease as one becomes more 

 familiar with the new forms, but for the present all that can be said is that 

 the results of the experiments are in general accord with the present 

 hypothesis. 



A more critical test is, however, provided by some experiments relating to 

 the characters of thrum-eye and pin-eye, and red stigma and green stigma. In 

 these cases no intermediates have as yet occurred and it may be assumed 

 that one " dose " of the factor is sufficient to bring about the development of 

 the dominant character. The results of these experiments are set out in 

 Tables I and II. 



Table II. — Crosses of Green Stigma and Eed Stigma. 







self. 





Pj x recessive. 





F 1 plant. 













G-reen. 



Red. 



Form of cross. 



Green. 



Eed. 



72 3 /13 

 72 5 /13 

 72 6 /13 



13 

 17 

 1 





 

 



72 3 /13 as ? 

 72 5 /13 as ? 

 72 6 /13 as ? 



5 

 28 

 14 



2 

 6 

 4 



68713 



44 



2 



68713 as ? 



Ditto 



Ditto 

 68713 as <J 



15 

 1 



3 

 48 



1 

 

 

 17 



1 



In these results two kinds of heterozygous Fi plants are clearly shown to- 

 exist. Thus, in the crosses of thrum x pin (see Table I), the Fi 68713 gave 

 41 thrum 1 pin when selfed, and 65 thrum 15 pin when crossed with the 

 recessive. These numbers may, I think be regarded as representing 

 respectively the ratios 15 : 1 and 3 : 1, and the Fi-plant may therefore be 

 identified as AAaa. The thrum parents from which the other Fi's were 

 obtained each gave a small number of recessives in the Fi-families from 

 crosses with recessive plants. It is, therefore, not surprising to find that 

 most of the Fi thrum plants derived from their crosses have given F 2 's- 

 approximating to the lower ratios 3 : 1 and 1:1; that is to say, the majority 

 of the Fi's are of the constitution Aaaa. 



