VAHIATION CONSIDERED BIOLOGICALLY. 181 



of the eggs of the two species : the young larvae of the hybrid, being 

 probably lai'ger than those of S. ocellatus, \von\d not have sufficient room 

 to develop in the egg of that species. Then the small percentage 

 of hybrids that arrive at maturity is noteworthy. Weismann's theory 

 of the " germ-plasm " furnishes us with a comparatively simple explana- 

 tion of the small number of larvas that hatch from the eggs ; only those 

 ova which had a large number or a majority of unaltered or ancestral 

 determinants in their germ-plasm would be able successfully to com- 

 plete their development. Thus the unmodified determinants from both 

 parents would be al)le to combine and work in unison ; and if, when 

 combined, the}^ formed a majority, they would be able to control the 

 development of the embryo and the earlier larval stages. If, on the 

 other hand, the ancestral determinants were few in number, the modified 

 determinants would assume control, but these would differ in accordance 

 with the parent they were derived from, and as it is probable that the 

 routine of development differs to some extent in each species, the 

 growth of the embryo would be checked in its early stages.* Probably 

 the embryonic stage is the most critical one, for it is then that the 

 internal organs necessary to larval existence are developed. The 

 jjupal would also be a critical stage, but the larval stages after hatching 

 would not be of nearly so much importance in this respect inasmuch as 

 the size, shape, coloration, &c. (especially of the adult larvae), as well 

 as the food, are so similar. The tendency to be double-brooded is 

 much stronger in the hybrids than in eitlier of the parent species ; this 

 may be due to reversion ; inasmuch, however, as S. tiliae, which I 

 believe to be the oldest form, has very rarely been known to emerge 

 before the winter, it is more likely to be due to physiological causes. 

 Hybrid specimens also seem to have a greater tendency to " grease " 

 than those of either of the parent species. 



So far as I know S. tiliae has never been observed to cross with 

 either of the other two. This may be due to no one having experi- 

 mented in this direction, or perhaps the difference in the genitalia 

 prevents the union. Such a cross, however, does not seem impossible ; 

 though, if fertile eggs were obtained, it would probably be difficult to 

 know what food the larvee would eat. 



It occurred to me, when reading the account of the sujjposed copula- 

 tion between two males of Bombijx qaercas {Eat. Rec, vol. v., p. 198) 

 that the best way to get crosses between different species would be to 

 2)lace the two individuals that you wish to cross in one cage, and to 

 place a $ of the same species as the ^ in an adjoining cage ; I have 

 since discovered that this was the method employed by Mr. House in 

 his successful experiments. 



Vapiatioii considered biologically.f 



{Being some Notes suggested by the Romaxks Lecture of 1894.) 

 By J. W. TUTT, F.E.S. 

 1. — The action of lNTR.\-SEr.ECTION ON THE DEVELOPMENT OF ABERRA- 

 TIONS AND Varieties. — One of the most interesting features of The 

 Romanes Lecture of 1894, was Professor Weismann's definition and 



* Roui^hly speaking there are three factors present in the fertilized ova. (I) 

 modified determinants of iS. populi; (2) modified determinants of S. ocellatas ; 

 (3) ancestral or unmodified determinants from both species. It is the latter 

 which can combine and work together to produce the hybrid. 



t Part of a Paper read befoi-e the Lancashire and Cheshire Entomological 

 Society, Jan. 13th, 1895. 



