318 THE AMEBIC AN NATURALIST [Vol.LYI 



Here ag-ain there are a number of small arrays and 

 gaps towards the right-hand side of the table, due as be- 

 fore to the method by which the material was got. 



The regression of duration of life upon density is 

 shown graphically in Fig. 2. 































































A 





























A 



V 









/ 



























































\ 



















































































It is apparent from inspection here as before that the 

 regression is not clearly linear, but rather indicates an 

 optimum density in the region of 35 to 45 flies per bottle, 

 with a diminished expectation of life at both lower and 

 higher densities. The constants are 

 r = - .132 + .014, 

 v= .283 +.013, 

 C = .0629 + .0066. 



The criterion of linearity is nearly 10 times its prob- 

 able error, and w^e may therefore conclude for the Sepia 

 stock, as for the wild stocks, that statistically the regres- 

 sion of duration of life upon density of population is 

 sdgnificantly skew. 



The data for the short-lived Quintuple stock are given 

 in Table III. 



Owing to the fact that the Quintuple stock is character- 

 ized by low fertility, as well as short duration of life, 



