190 Experimental Zoology 



degree the abhorrence of defilement by contact with other castes 

 or with certain animals. 



George Darwin, who examined the question, as far as the sta- 

 tistics were known, found no clear evidence of any evil effects 

 caused by close intermarriage, or at most the evil was very small 

 in amount, and out of all proportion- to the prejudice against 

 consanguineous unions. One might easily offer a number of 

 possible explanations of the basis for this feeling, but such specu- 

 lations would have little scientific value in the present state of 

 our knowledge. 



Experiments with the Pomace Fly 



The most extensive series of experiments on inbreeding that 

 have been made in recent years are those of Castle and his pupils. 

 They have used the fruit or pomace fly, Drosophila ampelophila, 

 which can be easily kept in confinement, and will breed through- 

 out the year in a warm place. The whole life-cycle may 

 be completed in eleven or twelve days (egg three, larva three, 

 pupa three days, and two days before the imago begins to lay 

 eggs). 



Brothers and sisters were paired in each generation. A pair 

 of pupae of the same parentage was put into each jar. If the 

 flies proved to be of different sexes, they were left together; if of 

 the same sex, they were rearranged. The larvae left the fruit 

 on which they had fed and pupated on the top or sides of the 

 glass. They were collected and counted, and the productive- 

 ness of the pair calculated on this basis. Sometimes no eggs 

 were produced, which was at times due to the sterility of the 

 male, at other times of the female, as shown by pairing them with 

 other individuals. 



The longest, or A-series, was carried through 59 generations 

 extending over three years. We can appreciate the extent of the 

 series if we compare it with the time required to produce the same 

 number of generations of mankind. If we allow three genera- 

 tions to a century, it would take 20 centuries to give 60 genera- 

 tions. In other words, a corresponding experiment on man would 



