390 C. W. MITCHELL AND J. H. POWERS 



of series from 1 1 to I 5. If present, such tendency was weak and 

 did not show itself in any phenomena capable of definite record, 

 unless it be that the earlier mass cultures were a little more prone 

 to undergo the upward transition than were those of the later 

 series of /. It is our judgment also that it required more delicate 

 manipulation of feeding technique to induce the transition at the 

 close of our experiments than it did at the beginning. But too 

 much weight should not be placed upon this judgment; its value 

 is only suggestive. 



More or less paralleling the last saccate series, / 5, two new races 

 of humped stock were again reared, derived however in a some- 

 what different way, and calculated to test a new point. By 

 April 19 the series K had reached the seventy-seventh genera- 

 tion, having produced, through a period of five months, an all 

 but unbroken succession of humped generations. Suddenly in 

 the seventy-seventh generation a single unusually small family 

 occurred of but five individuals of which two were saccates, one 

 transitional, and two of the prevailing type. We desired to test 

 the hereditary qualities, sexual and asexual, of these accidental 

 reversions, coming as they did in the midst of the strong humped 

 series, K. Would they, after all, show the potential of their series 

 or would it turn out that they had dropped permanently to the 

 potential of 7? One of the saccates was accordingly isolated as 

 the progenitor first of a mass culture. Even this mass culture 

 answered our main query by an almost immediate saltation to 

 the humped form. And in less than a week there followed a 

 second saltation to the campanulate and cannibalistic type. It 

 was from some of these huge cannibals that males and young 

 females were secured which produced resting eggs. Subjecting 

 these to conditions favorable to hatching, two young were secured 

 to become the parents of the series which, having the same point 

 of origin, we group together and term K 2. They were bred for 

 but ten and eleven generations respectively before being dis- 

 carded. Their histories are given in table 6. A glance at this 

 table shows that these series, like those derived from the humped 

 line J, again rose to the humped condition in the earliest genera- 

 tions, that is, in the second and third, respectively. The rotifers of 



