CASTLE. — INBREEDING, CROSS-BREEDING, AND SELECTION. 757 



Further Details concerning Experiments of S. O. Mast. 



Crosses were made between a stock of flies (series A) inbred for over 

 thirty generations, and two other stocks (series M and N) inbred for 

 from one to nine generations. 



History of the M and N Series. 



The M and N series were descended from two pairs (31 and iV) cap- 

 tured on decaying apples in an orchard three miles north of Cambridge, 

 October 1, 1903. Their progeny were inbred (brother with sister) as 

 a control on the results of the A series obtained in previous years, and 

 also to obtain a stock of known productiveness to use in crosses with the 

 A series. 



In Tables VIII and IX are shown the line of descent and the size of 

 broods of flies, each produced by a pair (brother and sister) descended from 

 one of the original pairs M and N, which were captured on decaying apples 

 as already stated. The successive generations are designated M and N 

 respectively, followed by a numeral to indicate the number of generations 

 of inbreeding which had occurred in its production. 



No sterile individuals were found in the ^series, while in the M series 

 one sterile female was found in the seventh generation, another in the 

 tenth, and a sterile male in the fourteenth. A second male, probably 

 sterile, was encountered in generation 5. In the three cases of sterility 

 first mentioned, the suspected individuals were proved to be sterile by 

 subsequently mating them with individuals that were known to be fertile. 

 The second matings, like the original ones, proved unfruitful. 



In the M series, the average brood in the first inbred generation was 

 213, while the average for the series as a whole is 280, an apparent 

 increase of fertility under close breeding. In the N series, generation 1 

 gave an average of 231 young, while the series as a whole gave 278.5 

 young to a brood. But the first inbred generation may have been lower 

 than the average because of less perfect methods of treatment. If we 

 divide the series into first and second halves, we find that series M is 

 more productive in the second half than in the first, while the reverse 

 is true of N. The average brood is in : 



Generations 1-7. Generations 8-14. 



if series ... 263 296 



N series ... 317 240 



