190 THE entomologist's RECOltt). 



I would point out, iu passing, that the male genitalia of the hybrids 

 differ from tliose of the males of both parental species, and the female 

 organs tend to^vards gynandromorphism. On the other hand, the 

 abortive sexual organs of the pupte show no tendency whatever towards 

 a confusion of the sexes, nor do they differ in any respect from those of 

 S. acdlatus or S. pojndi iu this stage. 



This is, I suijpose, only what might have been expected, as the 

 imaginal characters are subjected to the influences of both natural 

 selection and iutra-selection, while the rudimentary pupal organs escape 

 all these influences, except the slow reducing action of intra-selectiou iu 

 the embryonic stage. 



A comparison of my uotes on the larvae with those of Mr. House and 

 Mr. Edwin Birchall, shows that there are considerable discrepancies 

 between them. The hybrid larva; reared by these gentlemen resembled 

 "S. pupuli more nearly than those of *S'. ocellatm, although Mr. House 

 records that after the 3rd moult his larvge began to lose their *S'. populi 

 traits, and finally were more like -S'. oceUatm. There is, of course, 

 every probability that different broods may vary considerably in their 

 tendency to follow either parent species. With one exception 1 see no 

 need either to add to or alter the remarks I made on this subject iu 

 the paper published in the Knt. Ilecunl, vol. vi., pp. 180-181. I there put 

 forward as a possible cause of the non-hatching of ova resulting from 

 the cross between a 3 S. pnjmli and a J />'. ocellatus, the theory that 

 the small size of the eggs of S. ocellatm would ^jrevent, or prematurely 

 stop, the development of the embryonic larva ; without altogether dis- 

 carding this theory, I will venture to put forward as an alternative 

 that the hatching of eggs from one cross and non-hatching of those from 

 the other may be accounted for as follows : The female Ijciug the more 

 conservative sex, has a larger proportion of ancestral determinants in 

 its germ plasm than the male ; so that when the more progressive 

 male of the less specialised species (N. nedlatm) pau's with the con- 

 servative female of the more highly developed species (.S', populi), the 

 cross is between individuals that are more nearly related as regards 

 the constituents of their germ plasm than would be the case were 

 the sexes reversed, and there is, therefore, a far greater likelihood of 

 fertile ova resulting. I am aware that there are two assumptions in 

 the above explanation that requii'e proof : (1) That *S'. ocellatm is the 

 older or more generalised species ; of this I have little doubt, 

 but I should like the opinion of more experienced lepidopterists than 

 myself. (2) That the female is the more conservative sex. This has 

 been put forward in Messrs. Geddes and Thompson's Kvulution of Sex, 

 and has, I believe, frequently l)een alluded to in magazine and review 

 articles, but I do not tliiuk the question is yet considered as definitely 

 settled. 



It seems just possible that the in-and-in breeding of the two 

 species for several seasons before attempting to cross them, might 

 produce better results than attempting to pair wild moths. 1 do not, 

 however, feel at all certain of this, as the loss of vitality might either 

 prevent the pairiug or result in a very weakly brood. 



Description of Plate III. — Fig. 1 : Genitalia of # Smerinthus tiliae. Fig. 2 : 

 J S. occllalus. Fig. 3 : j S. popiili. Fig. 4 : s Hybrid .S'. ocellatus ( cT ) X 6'. 

 2)opiiU ( ? ). Fig. 5:28. ocelUUus. Fig. 6 : 9 -S. populi. Fig. 7 : ^ Hybrid S. 

 ocellattis ( c? ) X S. populi ( ? ). 



