Life Cycle of Hydatina Senta 



Z^l 



TABLE V — continued 



about three generations in each period. The two series cannot 

 be compared generation for generation, for the thirtieth genera- 

 tion of the well-fed Hne occurs five days earher than the thirtieth 

 generation of the starved hne. 



All the major crests and depressions of the curve of well-fed 

 families correspond to similar crests and depressions of the curve 

 of starved families; but the curve of the starved line is more nearly 

 uniform, its highest and lowest points are well within the extremes 

 of the curve of well-fed families. This indicates that the same 

 agent is producing the major fluctuations in both lines, but that, 

 that agent operates to a greater degree upon the well-fed families 

 than upon those that were starved. I can find only one factor 

 that meets these requirements, namely, the quantity of the food- 

 culture employed. 



Experiment IV. The results of the preceding experiment were 

 controlled by three repetitions of it. In each pair of controls 

 the two lines were derived from sister individuals, one closely 

 following the other in the family. 



In A, the parents were derived from the sixteenth generation 

 of the starved line in Experiment III. The best food was used 

 in each. 



