302 



A. FRANKLIN SHULL 



decrease of cytoplasm^ or to both, the absolute sizes of all nuclei 

 included in this series, totaling about 1900, were collected. This 

 comparison is possible, since all drawings were made to the same 

 scale. The sizes given are square inches of area in the drawings. 

 To reduce the probable errors involved, the nineteen generations 

 were divided into an early, middle, and late period of six, seven, 

 and six generations, respectively, and the nuclear sizes averaged 

 for the three groups separately. The results are as follows : 



Size of nuclei of yolk gland in first six generations 0.443 ± 0.0052 



Size of nuclei in seventh to thirteenth generations 0.469 ± 0.0065 



Size of nuclei in last six generations 0.532 ± 0.0054 



TABLE 2 

 Relative nuclear volume of the yolk gland of Hydatina senta for each of nineteen 

 generations, including two ivell-marked periods of male-production. 

 Compare with table 1 



The increase in absolute nuclear volume is nearly proportional 

 to the corresponding increase of relative nuclear size shown in 

 table 2, indicating that the relative increase is due to enlargement 

 of the nuclei, while the volume of the cytoplasm remained nearly 

 stationary through the nineteen generations. 



It is difficult to see any relation between nuclear volume and 

 the number of male-producers (upper curve). Although the 

 curve of nuclear volume has a slight appearance of bimodality, the 

 second hump begins rather too early to be held responsible for the 

 second 'wave' of male-producers. In this connection it should be 

 remembered that those external conditions known to affect the 



