No. 577] SHORTER ARTICLES AND DISCUSSION 



39 



Classes 25-29 30-34 35-39 40-44 45-49 



Frequencies 6 34 67 43 33 



Classes 50-54 55-59 60-64 65-69 



The wing-length manifestly varies so as to form a pretty good 

 probability curve ; what the ' ' gene ' ' is doing, I do not under- 

 take to say. 



It is, of course, conceivable that the variation here observed in 

 actual wing length might be due to variation in general body 

 size, larger flies having longer wings. To determine this point 

 measurements of tibia-length were made on the same flies, and 

 in the case of each individual the ratio was computed between 

 wing-length and tibia-length. These ratios are distributed as 

 follows : 



Ratios 70-.79 .80-.89 .90-.99 1.00-1.09 1.10-1.19 1.20-1.29 



Frequencies 2 7 26 49 47 23 



Ratios 1.30-1.39 1.40-1.49 1.50-1.59 1.60-1.69 1.70-1.79 



It is evident that there is no constant relation between wing- 

 length and tibia-length, and so between wing-length and general 

 size, with which tibia-length is closely correlated. Again we ob- 

 tain a good probability curve. Does the " gene " vary or are 

 we dealing also with additional modifying " genes "? "We are 

 confronted here with the same problem as in the case of the rats. 



But it is possible to assume that the considerable variation 

 shown by vestigial wings in Drosophila is purely somatic, " phe- 

 notypic," not due to genetic causes, and so would not show any 

 effects if subjected to selection. So it was thought in the case of 

 the plus and minus variations in the hooded pattern of rats, 

 before the experiment was made, hut experiment has shown, even 

 to Mr. Muller's satisfaction, that the variations are in part due 

 to genetic causes and that selection slowly and surely changes 

 the range of variability. Is it safe to assume the contrary for 

 Drosophila in the absence of all experiment ? 



Mr. Wenrich has also studied the wing-length of " extracted " 

 vestigial flies obtained in the second generation from a cross 

 between pure vestigials and normal flies, and he finds that the 

 variability is regularly increased as compared with that of the 

 uncrossed vestigial race. This again is parallel with what occurs 

 when hooded rats are crossed with wild or with Irish rats, and 

 indicates that similar causes are at work in the two cases. Such 



