596 CALVIN B. BRIDGES 



either one were pure for not-pink, (as in the next case below) 

 no pinks would have appeared in the offspring. Had the num- 

 bers been large enough, exceptional males should have appeared 

 in the proportion of three red to one pink among themselves. 

 A second non-disjunctional white female (of the twelve) I 

 crossed to a wild male and obtained: 



white 9 rede? white cf redcT 

 15 159 146 13 



Here no pink appeared although the* cross was externally the 

 same as the previous one. The interpretation is that the male 

 in the first case differed from the wild male in that it was hetero- 

 zygous for pink. 



To effectually show that the non-disjunctional males were 

 heterozygous for pink I mated four (of the nineteen) to stock 

 pink females, and as I expected, half of the females and half 

 of the males were pink: 



red 9 pink 9 redcf pinkc? 

 70 69 63 62 



To show, once more, that it was the female only that was 

 responsible for the production of exceptions in Fi I simulated 

 the cross of non-disjunctional white pink female by non-disjunc- 

 tional male given above by testing a non-disjunctional brother 

 of the first by a white female of another stock. The result was 

 simply red females 105 and white males 74, with no exceptions. 



A third male (of the nineteen) I tested to an eosin female 

 with still no exceptions appearing in Fi among the 143 red females 

 and 146 eosin males. 



There still remained to be tested only the normally produced 

 females (343) and males (278). Two pairs from these gave the 

 ordinary Mendelian result: 



97 80 30 107 73 » 35 



