20 Trans. Acad. Sci. of St. Louis. 



In this table we see only a slight difference in the 

 mean duration of life of the mated and nnmated males, 

 and in this case it is in favor of the nnmated individu- 

 als ; the same is true of the females. When we compare 

 the mean for all females with that of all males, we are 

 surprised to find a much greater difference, and strangely 

 enough, this is in favor of the males. The mean dura- 

 tion of life for this entire lot coming from fluffy cocoons 

 is 10.72 days, or nearly a day more than the normal 

 moths of the same year and place. (See Table 10.) 

 When we compare the figures for these two parts of the 

 population, we see that the mean length of life is greater 

 in the fluffy lot in practically every class. Why should 

 it be so unless the insects in the large, fluffy cocoons are 

 endowed with greater vigor or a greater supply of re- 

 serve energy which carries them just a little further in 

 life? May it be that they are thus well supplied with 

 reserve nutriment as a direct result of abundant food 

 supply of these individuals in the larval stage, and that 

 the spinning of the larger cocoons is likewise merely the 

 outcome of this accelerated growth? 



Our next lot of material is that known as the South 

 St. Louis Cecropias. Of the 236 insects which emerged 

 from these cocoons between May 7 and June 13, 142 fur- 

 nished longevity data. (See Table 12.) 



Here we find that the mean duration of life of the 

 mated males was greater than that of the unmated ones, 

 while just the reverse was true for the females, and here 

 too the mean duration of life of all males taken together 

 was greater than that of all females. The mean for the 

 whole population, 9.32 days, is the lowest of the Cecropia 

 lots taken up in this paper, but it is greater than some 

 of the 1911 lots. 



The New York Cecropia material consisted of 143 in- 

 sects which emerged between May 1 and June 4, 110 of 

 which were used for these tabulations. (See Table 13.) 



In this group we find precisely the same relation of 

 length of life to sex and mating as in the foregoing lot; 



