R. C. PUNNETT 89 



And if the figures are arranged for the three different pairs of factors 

 the result is what might have been anticipated from the nature of the 

 mating, viz. coupling between D and N,and repulsion between D and F 

 and between N and F respectively, 



Families 48 — 51. 



The numbers shew that the reduplication series in the first two 

 cases is approximately on a 3 : 1 basis while that for D and F fits most 

 closely a 1 : 7 : 7 : 1 series. In the last case however the df class appears 

 so rarely that the precise nature of the reduplication must for the 

 present remain undecided. It is not impossible that it may be on a 

 1 : 15 basis which is rather to be expected in view of the fact that the 

 coupling series for these two factors is on the 15 : 1 : 1 : 15 system. 



It may be noted that the 3 : 1 basis for the relation between N and 

 D from this mating is in accordance with the 1 : 3 basis for the relation 

 between these two factors when the nature of the mating is Nd x nD. 



(7) Nature of mating, DFn x dfN. 



The material for the study of this mating (Table XI) consists 

 so far of two F^ families, 13, 52 and 13, 53, and of four F^ families, 

 13, 113, 114, 117, 123, all derived from an original cross of the same 

 type. All six families appear to be of the same sort, and I have 

 assumed the nature of the mating in the F^ families is what is known 

 to be the case in the two F.^ families. 



In families derived from this type of mating I had anticipated 



