430 T. H. MORGAN AND C. B. BRIDGES 



it is the presence of the color producer in simplex in the male, 

 and in duplex in the female, that is responsible for bicolorisms 

 in the eosin why then are not other sex linked mutations such 

 as vermilion, also bicolored, for the same relation obtains there? 



To account for the fact that the daughters were eosin and not 

 red in this cross of eosin female by white male, we assumed that 

 all white flies backed' not only the color producer but a color 

 determiner also, i.e., white was assumed to be a double recessive. 

 We further assumed that the linkage between these two factors 

 was well nigh absolute for the following reason: If a white male 

 is mated to a red female there should appear in F2, unless the 

 linkage is complete, four types of males — red, white, eosin and a 

 new type which should be the single recessive, or simple white. 

 We have records of approximately 150,000 flies from such a 

 cross and its reciprocal and yet only two classes of males ap- 

 peared — red and white. On the assumption that white repre- 

 sents a double recessive, we were therefore justified in concluding 

 that the linkage was complete. 



An alternative interpretation for this set of phenomena and 

 the one we now believe to be the better one, is the treatment of 

 red, white, and eosin as a set of triple allelomorphs. We assume 

 that in a certain locus in the sex chromosome a determiner 

 mutated to produce white. In another fly a separate mutation 

 in the same locus produced eosin. Each is a simple sex-linked 

 recessive with respect to the normal determiner carried by the 

 red fly. White and eosin are mutually exclusive with respect to 

 their common locus in any one sex chromosome. For example, 

 in the female where two sex chromosomes occur, either chromo- 

 some can carry in the cormnon locus eosin or white or the 

 dominant form of these determiners, i.e., red. Since in the male 

 there is only one sex chromosome, he exhibits directly his com- 

 position as red, or white, or eosin. 



We now think of the white as a simple primary mutation and 

 of eosin as another primary mutation which is, however, char- 

 acterized by bicolorism. When an eosin female is mated to a 

 white male two normally recessive determiners enter the same 



