184 Reproduction, in Simocephalus vetiilus 



increasing frequency in later generations, but also in later bronds of 

 earlier generations. Now all the four broods which gave males in lay 

 experiment were first broods. Yet broods of a much higher order were 

 tested and furnished parthenogenetic females only. 



Also according to Papanicolau, a potent factor in producing sexual 

 forms is not only the order of the brood, but the oixler of the brood to 

 which the parent belonged. The greater part, but far from all, of each 

 generation in my experiment was descended from first broods in the pre- 

 ceding generation. This however was also the case in Papanicolau's main 

 experiment (Tafel I), in which the line ended altogether in sexual and 

 degenerate forms in the twentieth generation. A direct test of the 

 influence of the order of the brood was also made. It will be seen that 

 in generation 22 two sixth broods were tested for sex. Now the parents 

 of these sixth broods were members of the tenth brood of a specimen 

 of generation 21 which was itself a member of the eleventh brood of an 

 individual of generation 20, itself descended from the tenth brood of 

 generation 19, the parent of which was a member of the ninth brood of 

 generation 18. This is a far severer test than any which Papanicolau 

 describes, and yet all were parthenogenetic females. (Breeding from 

 late broods like this makes each generation fall far behind, in point of 

 time, the generations bred chiefly fi'om first broods. The sixth broods 

 of generation 22 were contemporary with the first broods of generation 

 36 of the main part of the line.) 



The third of our list of Papanicolau's conclusions has been no more 

 lay experience in this experiment than were his other.s. He found 

 degenerate broods extremely common, increasing in frequency pari 

 passu with the sexual forms. I have had very few of these degenerates, 

 and they certainly showed no tendency to increase as the line grew 

 older. 



As mentioned above, the gTeat majority of the individuals of this 

 line were bred under the conditions which I have for convenience 

 termed " normal " as distinct from the intentionally abnormal conditions 

 under which a part of the line was bred. The " normal " and abnormal 

 environments resulted in similar sexual conditions however. The 

 abnormal environments included feeding with a protophyte culture 

 which produced certain specific abnormal characters (described in my 

 paper of 1913), cultivation in chemical solutions, and at high and low 

 temperatures. In the case of the high temperature, part of the line 

 was subjected to the abnormal condition for four or five successive 

 generations. Naturally abnormalities and a higher death rate were 



