64 \VV.MAX REED GREEN. 



iiiewhat in error because of the number of offspring which 

 die too early for identification; but the error is not considerable 

 since all of the available evidence points to the conclusion that 

 all of the different kinds of offspring are about equally viable. 

 Inspection of this table shows no marked changes in the relative 

 number of kinds of offspring from the first to the later broods in 

 any of the generations in either the sexual or the asexual lines. 

 This would corroborate the conclusion already reached from the 

 numerous other experiments, namely that sex is in no way corre- 

 lated with the age of the mother. It was of the greatest interest 

 to find that the offspring of the F 3 generation in the sexual 

 line were no more likely to be males and sexual females than were 

 the offspring of the 3 generation in the asexual line. It seems 

 that if such selection were to have any definite effect that it 

 would at least begin to show by the third generation. Yet it 

 will be observed that in the second generation 33.3 per cent. 

 of the offspring in the asexual line is male, and 32.7 per cent, is 

 male in the sexual line; while in the third generation the corre- 

 sponding percentages are 35 per cent, and 31.5 per cent, respec- 

 tively. Of the stem mothers the percentage of males in the 

 offspring is 23.5 per cent., the lowest of all. It seems to me 

 however that the differences in no case are great enough to be of 

 any significance in a species of animals which displays so much 

 variation in x> many respects as this one does. 



The ratio of the asexual to the sexual females is seen to be 

 27.6 per cent.: 23.1 per cent, in the FI generation; 19.1 per cent.: 

 15.3 per cent, in the 2 generation of the asexual line, 21.9 per 

 cent.: 11.3 per cent, in the F 2 generation of the sexual line; 20.7 

 per cent.: 17.6 per cent, in the 3 generation of the asexual line, 

 and 28.2 per cent.: 12 per cent, in the FS generation of the sexual 

 line. The last ratio is striking as showing such a low percentage 

 of sexual females in the sexual as compared to the asexual lines 

 in the same generation. 



It will be seen that if the percentages of the males in these 

 five categories are averaged we get somewhat over 31 per cent., 

 which is very much lower than that observed in experiment 2, 

 which was 42 per cent. Since in the latter case the olispring 

 considered numbered over 1,700, it may seem that we should be 



