348 CALVIN B. BRIDGES 



with many of them of an extra dark color. This extra dark color 

 likewise occurred in a fair proportion of the eosin sons. The 

 whole result of culture M67 seemed inexplicable unless it was 

 assumed that the white which here appeared was not the original 

 sex-linked white which had been running through the experiments , 

 up to this point, but was a new and complex color. 



The disappearance of the true white is easily accounted for when 

 one remembers that the culture from which this supposed white- 

 eosin female was drawn was extraordinarily complicated as to eye- 

 colors by the presence of eosin and ^dark' eosin, of white-eosin 

 compounds and of 'dark' white-eosin compounds, of the true 

 white and of the new complex white, and as to inheritance by the 

 occurrence in the same culture of two sex-linked and of two non- 

 sex-linked characters, by the multiple allelomorphism of white and 

 eosin and by the occurrence of exceptions by non-disjunction. 

 Because of the dark modification it had been impossible to dis- 

 tinguish between the normal eosin color and a white-eosin com- 

 pound raised to about this same color by the action of the dark 

 modifier, and the mother chosen had really been pure eosin and 

 not a dark white-eosin compound as supposed. 



The true nature of the new white was not fully established 

 until one of the white females from M67 was outcrossed to a bar 

 male with the result shown in table 7. Here the surprising fact 

 came to light that the white female when outcrossed gave no 

 white sons, and the 'exceptional' daughters likewise were not 

 white, but were eosin. These facts prove that the white mother 

 did not carry the sex-linked white, but was a modification of 

 eosin, and that the modifier was an autosomal recessive. These 

 conclusions were confirmed in the clearest fashion in the next 

 generation by the offspring of culture M92 whose parents were a 

 regular bar daughter and an 'exceptional' bar son both from M74 

 (table 7). 



The colorless flies reappeared only among those sons which 

 were already eosin, while the not-eosin flies, both males and 

 females, showed no effect whatever of the action of the modifjdng 

 gene. The Fo was a good example of a 12 : 3 : 1 ratio. The new 

 gene then differs from the two creams and dark only in this 



