196 A. FRANKLIN SHULL. 



is not the same as the distribution in the well-fed families of 

 Tables IV. and V., about whose nutritive conditions we know 

 only that which size of family tells us. The argument that the 

 numerous male-producing females of the starved line were 

 ■produced as a result of high nutrition of their parents, loses much 

 of its weight when it is shown that these male-producers were 

 not chiefly in the largest families. 



It is not to be asserted that nutrition has no effect upon male 

 production. Indeed, Whitney ('14) has presented new evidence 

 that qualitative differences of nutrition do affect male production. 

 It is not clear what relation Whitney's results have to the ques- 

 tion of periodicity of male production, whether changes of 

 nutrition can be made to destroy the rhythm, or wholly to 

 alter the interval, or merely to modify the extent of male produc- 

 tion. My own starvation experiments, referred to above, left 

 the intervals between the periods of male production unaltered, 

 but the waves of male production and the intervening periods 

 of female production were rendered less striking. I attributed 

 the effects shown in these experiments to the chemical nature 

 of the medium, and not to nutrition. Until experimental 

 evidence indicates the contrary to be true, it is safest to assume 

 that nutrition also, when it affects male production at all, does 

 not alter the interval between periods of male production, but 

 merely the extent of male production. 



To summarize: Three lines of Hydatina, bred through many 

 months, showed fairly regular periodicity in the production of 

 males. One line exhibited relatively abundant male production 

 every month; another every two months; while in the third the 

 interval varied from three to five months during the period of 

 observation. The fact that the interval between the periods of 

 many males is quite regular in some lines, and is not the same, 

 in all lines reared simultaneously, indicates that this periodicity 

 is due to an internal factor. Hundreds of families were examined 

 to determine whether the largest families, which were presumably 

 offspring of the best nourished parents, contained the greater 

 number of male producing females, as Mitchell assumes they do. 

 In well nourished lines there is some doubt whether there was 

 any correlation between size of family and number of male 



