D.— ZOOLOGY 107 



saturnid moths (female heterogamety) which are always quoted as sup- 

 porting the contention that in this group males commonly outlive the 

 females, but, in point of fact, an analysis of their figures by MacArthur 

 and Baillie (1932) shows that the contrary holds true for five out of six 

 of the species. Graf (1917), working with the potato tuber moth, found 

 that the males died first in 221 out of 275 pairs in the experiment, and 

 concluded that the length of life of males is less than that of females. 

 In the gypsy moth (female heterogamety) Goldschmidt (191 7) records 

 that the males are so precocious and so short-lived that they are often gone 

 before the later females are ripe. Very full information concerning the 

 mortality of the sexes of the codling moth (female heterogamety) is con- 

 tained in certain publications of the United States Department of Agri- 

 culture. The studies were made on a very large scale, repeated over many 

 years and at stations widely separated, covering all sections of the conti- 

 nental United States. The most useful data from this source include the 

 records of longevity of imagoes from observations on large numbers of 

 mated moths reared in their natural sex ratios in large out- door oviposition 

 cages. From the figures given it is evident that at all stations and in all 

 generations (excepting only the first broods) the males are the shorter-lived. 

 The weighted average duration of life for females is 10-23 days, that for 

 males 9*40 days. The difference between this is over twenty-five times 

 its probable error and is highly significant in view of the large number 

 on which it is based. MacArthur and Baillie (1932) use these figures very 

 effectually to maintain their conclusion that Lepidoptera afford no excep- 

 tion to the rule that the larger destruction and shorter life of males, 

 irrespective of whether these be homo- or hetero-gametic, cause them to 

 become relatively fewer at advanced ages. 



A multitude of other references could be quoted, but were this done 

 nothing very essential to this discussion would have been added. Every- 

 where one finds support of a kind for the contention that the male, 

 irrespective of his actual sex chromosome constitution, is the shorter-lived. 



Though much remains unclear, demanding further and more critical 

 examination, enough has been gleaned from this necessarily incomplete 

 survey to show beyond all doubt that there is nothing unique about the 

 human sex ratio and that the factors which operate to distort it are exactly 

 those which produce the same results in other forms. Sex-linked lethals 

 and other recessive disadvantageous X-borne genes are undoubtedly 

 concerned in the production of a sex incidence of mortality that bears 

 more heavily upon the heterogametic sex. But it is impossible, in the 

 face of all the evidence that has been presented, to conclude that all the 

 inequality in mortality and longevity between the sexes is due to the action 

 of such genes. Wildtype stocks of Drosophila shown to be free from 

 expressed lethals have exhibited the usual sex differences in longevity. 

 There are autosomal genes, as Gonzalez (1923) has shown, which com- 

 pletely reverse the usual mortality of the sexes and allow males to outlive 

 the females by a very significant number of days. So that mortality 

 and longevity are controlled through the whole genetic and environmental 

 complex ; they are not different and separate characters, but expressions 

 in time of the organisation of the body. 



