64 E. CARLETON MACDOWELL 



To test the stability of the extra race further, three mass cul- 

 tures were started from the 4th and 6th inbred generations and 

 have been carried along from bottle to bottle with no further 

 selection. Occasional counts have been made of these cultures 

 and it has been found that the extra bristles have been retained. 

 Table 3 shows the per cent normals in mass counts. It is evi- 

 dent that there are more normals in some cases in these mass 

 cultures than in the inbred lines, but these normals are more 

 than likely potential extras whose greater frequency is due to 

 the fact that in mass cultures there are larger numbers of flies 

 which are smaller on account of the crowding. It will be dem- 

 onstrated later that there is a relation between the size of the 

 fly and the number of extra bristles. (See the discussion on 

 the influence of environment). 



INHERITANCE 



1. THE FACT OF INHERITANCE 



That a race of flies constantly bearing extra bristles has been 

 established, indicates in itself that the extra bristles are inherited. 

 To show that this is not pseudo-inheritance, due to certain con- 

 stant environmental conditions, a stock of wild flies caught in 

 New York has been carried along under the same conditions as 

 the mass cultures of the extra stock, and for a few generations 

 this wild stock was bred in separate pairs as were the inbred 

 selected lines. A small percentage of extras was found in this 

 wild stock (tables 1 and 4) but this did not increase in time, 

 nor did the progeny of separate pairs show any higher pro- 

 portions of extras. Conclusive evidence that extra bristles are 

 inherited and not primarily due to environment is afforded by 

 crosses, which show also the mode of their inheritance. 



2. THE KIND OF INHERITANCE 



The wild stock New York 1912 was found to have fewer 

 extras than any other wild stock examined and so from this stock 

 normal flies were selected for several generations to attempt to 

 reduce the percentage of extras. A preliminary set of four crosses 



