EYE COLOR IN DROSOPHILA MELANOGASTER 375 



of the side between the wild-type and (not-eosin) cream II 

 corners is zero, so that the parallelogram becomes reduced to a 

 triangle (diagram 2, b). In the case of pinkish and cream III 

 there is an intermediate condition in which the side between 

 eosin and the double type is disproportionately far greater than 

 the distance between the wild-type and the simple modifier 

 (diagram 2, c). There is another type of disproportionate modi- 

 fication exemplified by eosin sepia which might be called reversed. 



a General modifier b, Specific modifier c, Disproportionate modifier 



Wild-type Wild-type, cream || Wild-type 



V-Cream 



Eosin / / "^ Eosin / 1 / Eosin 



Eosin pink 



i'Eosin cream II ^ 'Eosin cream 



d, Reversed modifier e, Non-modifier 



Wild-type ^^^®P'^ Wild-type 



Ruby \\Pink 

 Eosin ^ / Ruby pink \\ 



Eosin sepia \\_ 



Diagram 2. A graphical representation of some types of modification 



Sepia is considerably darker than the wild-type, becoming a 

 deep blue-black in old flies. One might expect that sepia would 

 cause a proportional darkening of eosin so that eosin sepia would 

 be as much darker than eosin as sepia is darker than the wild- 

 type. This is not the condition that actually obtains. The 

 eosin sepia double form is lighter than eosin, about as represented 

 in diagram 2, d. There are several other curious types of dis- 

 proportionate modification. Thus, for example, ruby (sex- 

 linked) may be described as a 'non-modifier' of pink. The effects 

 of these two genes are in the same direction and of like amount, 



