W. E. Agar 191 



to the fact that the cycle is too near one end or the (jther fur the opjX)site 

 condition to be evoked. 



Now it is quite possible to account for these observations without 

 invoking an internal cycle. Firstly, it is extraordinarily difficult to keep 

 all the conditions of the environment constant, and especially the great 

 difficulty of keeping food cultures constant for long periods imposes upon 

 the experimenter, who assumes that physiological changes in his animals 

 were not correlated with changes in the environment, the obligation of 

 stating very fully what precautions and tests he took to ensure that the 

 conditions really were constant. 



Secondly, even where the conditions are constant throughout the 

 whole experiment, an increased tendency to sexuality in, say, the 

 twentieth generation as compared with the tenth may be due to 

 the fact that the line has been living for a longer period in an 

 unsuitable environment — perhaps, for example, one deficient in some 

 essential constituent or " vitamine." As both Papanicolau and 

 Woltereck point out, not only the number of generations but also 

 the length of time during which a line has been subjected to the 

 experimental conditions is of importance in determining sexuality, for 

 sexual forms appeared in late broods of early generations and early 

 broods of late generations. In my experience with S. vetulus an 

 individual of any generation produces its fourth brood about the same 

 time as the members of its first broods produce their first brood of off- 

 spring. That is, the fourth broods of the nth generation are contemporary 

 with the first broods of the n + lth generation. If this held true for 

 Papanicolau's strain of S. vetulus, it is easy to calculate from his Tafel I 

 that sexual and degenerate forms appearing in the later broods of the 

 earlier generations actually appeared earlier, in point of time, than those 

 of later generations. 



I have described experiments (1913) where the effects of a peculiar 

 environment produced on a given generation of S. vetulus were still 

 detectable in their great-grandchildren, and Woltereck has produced 

 evidence to show that an environment acting on an individual may 

 determine the sexuality of its grandchildren. It is plain, therefore, 

 both that environmental effects may persist for some generations after 

 they were produced — in other words that it may act in cumulative 

 fashion — and also that the length of time and not only the number of 

 generations during which the line has been in the conditions of the 

 experiment, is of importance in determining sexuality. Taking this in 

 conjunction with the fact that under some environmental conditions no 



13—2 



