p. G. Bailey 225 



The above table shows that Trow's special hypothesis fits the 

 figures better than the general hypothesis ^ if the assumption be made 

 that the NF series is the secondary series, and that the DF relation- 

 ship is on the 1 : 15 basis. The agreement may be purely accidental 

 but it is interesting to note that Punnett fqund certain strains in which 

 the fundamental N F series was itself on a 1 : 1 basis. 



(8) Nature of mating DFn x df N. 



The values obtained in this case are l=n = l, and m= 15, i.e. the 

 apparent relations are the real relations. The N D and N F fundamental 

 repulsion series are reduced to a 1 : 1 : 1 : 1 basis. 



The only additional data^ bearing upon this question are those 

 furnished by Gregory (2) in his description of the results obtained 

 from a cross of the nature MSG msg x msg msg. Trow (4), p. 315, 

 discusses these results from the point of view of the special hypo- 

 thesis, and he shows that the observed numbers agree fairly well with 

 those obtained by calculation on the assumption that the SG series is 

 the secondary one. On the other hand it can be shown that, if in 

 reality the general hypothesis applies to this case, the numbers fit in 

 very badly with any simple primary series. It should be noted that 

 the observed MS series closely approximates to the fundamental MS 

 series, and that no fundamental SG series has yet been described. 

 Consequently it is by no means clear that the case, which Gregory 

 has described, is really comparable with those described by Punnett. 



Conclusion. 



The general hypothesis adopted above, although it admits the 

 possibility of a difference between the fundamental and the primary 

 series due to the interaction of the reduplication series one upon 

 another, does not postulate a differential interaction. The special 

 hypothesis on the other hand does postulate a differential interaction. 



1 If the complex series calculated on the general hypothesis be taken, the agreement 

 between the numbers found and the numbers obtained by calculation would be even better 

 than on Trow's simple hypothesis. 



2 Morgan and Cattell (5, 6) have described certain crosses with Drosophila which 

 involve three factors. The results, however, are complicated by the phenomena of sex 

 limitation, and by differential death rates. Moreover it is not clear in each case whether 

 the given relationships are to be looked upon as fundamental or primary. Nevertheless 

 it is interesting to note that in the most satisfactory case, namely that involving black 

 body colour B, red eye colour R, and long wings L, the secondary relationship for Land B 

 i.e. 1-9 : I calculated upon Trow's special hypothesis closely approximates to the relation- 

 ship found by experiment. 



