14 Trans. Acad. Sci. of St. Louis. 



June 17, all of which were used in this investigation. 

 (See Table 8.) 



The table shows that they were extremely reticent 

 about mating in confinement. All our efforts to offer the 

 best opportunities brought successful results in the case 

 of only four females, and even these rejected the males 

 of their own species and mated with Cynthias. This 

 table also shows that the unmated females were much 

 longer lived than the males and also than the mated fe- 

 males, just as they were in the 1911 material, but as a 

 whole the duration of life was much greater in the 1913 

 Fall River material than in the 1911 Creve Coeur lot. 

 Comparing in greater detail these two lots from differ- 

 ent localities and years, we find the mean for the males 

 to be greater in the former lot by iy 2 days, and for the 

 females by an excess of 2 days. Only consider the mean- 

 ing of 2 days in the life of an insect, the duration of 

 whose life is from 2 to 12 days. 



It is indeed perplexing when we find that the Poly- 

 phemus from Massachusetts showed a mean much lower 

 than did those from St. Louis, and that the Prometheas 

 from there had a mean duration much greater than those 

 from this region. 



SAMIA CALIFORNICA. 



Out of a large number of cocoons from Sonoma Co., 

 California, only 14 insects emerged. Of these, 12 could 

 be used for longevity work. (See Table 9.) 



The table shows that they lived from 7 to 12 days, the 

 mean for the entire population being 8.75 days. It is 

 much regretted that the number available for this study 

 is insufficient for the much desired comparisons, although 

 it is interesting to note that these few give evidence of 

 the same tendencies as the other species thus far consid- 

 ered in that the female sex and the unmated condition 

 both tended toward longer life. 



