JOHN S. DEXTER. 



Their supposed gametic constitution and the results to be ex- 

 pected in the FI generation are shown in Table VIII. 



TABLE VIII. 



The Pi generation consists of members of the Fa generation of Experiment II. 

 (see Table III.). 



Pi. YRXYWX = YRQ NWXYW = NWtf 



Gametes YRX, YWX NWX, YW 



Fi. (i) YRX NWX = NR 9 (3) YRX YW = YR<? 



(2) YWX NWX = NW9 (4) YWXYW = YW<? 



The table shows that there should be produced in equal 

 numbers NR females, NW females, YR males, and YW males. 

 The actual hatch consisted of 



NR females, 116 Females, 82 



NW females, 116 YW males, 100 



NW males, 2 



I am quite unable to account for the occurrence of the two 

 NW males, except through mutation. Rather than offer this 

 rather wild guess, I shall not attempt to explain it at all. It 

 will be observed, however, that it could not possibly have been 

 that the YR females were not virgin when the experiment was 

 begun, for no known mating will produce an NW male from such 

 a female. Also, it would be an unbelievable coincidence that 

 should produce on the same day the fifth after the beginning 

 of the hatch two such males on account of contamination of the 

 food. 



As in experiment III., I mated the virgin NR females to the 

 YW males, so that Table VI. accounts for the results in the F 2 

 generation in the same way that it did for experiment III. The 

 discussion, also, of the probabilities for the F% generation in that 

 experiment applies equally well to this experiment. 



I started only one bottle of this kind (No. 47). From it were 

 produced: 



NR females, 2; YR females, 81; NW females, 89; YW females, i; 

 NR males, i; Females, 82; N W males, 90; YW males, o. 



A comparison of these figures with those of bottle No. 44 

 of experiment III. will show a very close agreement of the one 

 to the other. 



