1 88 A. FRANKLIN SHULL. 



(p- 253). Whitney (1914), on the other hand, has no hesitancy 

 in ascribing the periodicity of male production to external 

 factors; thus, in mentioning earlier work of his own, "he was of 

 the opinion that whatever the potent factor was that sometimes 

 caused only females to be produced, and at other times caused 

 nearly all males to be produced, it must be an external factor." 

 On the basis of his recent work he attributes this periodicity to 

 alternation of the active and quiescent states of the protozoan 

 food of the rotifers. 



In my own work on Hydatina during several years, certain 

 lines have been bred so long, and so many families completely 

 reared, that further light may be thrown upon the rhythm of 

 male production. In the following pages evidence is first pre- 

 sented, bearing on the regularity of the periodicity of male 

 production, and the probable independence of this periodicity of 

 the environment. Later the supposed evidence that male pro- 

 duction is correlated with nutrition is discussed. 



Regularity of the Periodicity of Male Production. 

 Conditions Necessary to the Demonstration of Periodicity. 



In parthenogenetic lines that produce many males, rhythm is 

 not as easily demonstrated as in lines producing few males. If 

 the proportion of males in a line producing many males be repre- 

 sented by a curve, there are so many irregularities in it, so many 

 minor humps even in the periods of few males, that the larger 

 humps are less striking. If the lines produce few males, on the 

 other hand, the whole curve may be so lowered that the periods 

 of depression are below the base line; that is, there are no males 

 at all in these periods. In such cases the rhythm may be quite 

 striking. 



Further difficulty in detecting rhythm is introduced by rearing 

 only one family in each generation. There are great individual 

 differences between families taken at the same period, so that 

 the family chosen may or may not be an average of all families 

 that might be reared at the same time. Furthermore, the method 

 of publishing the results, merely giving the total for the one 

 family in each generation, often makes the rhythm appear less 

 definite; because it often happens that in three or four successive 



