J. W. H. Harrison 275 



died one by one before revealing characters sufficiently striking to 

 admit of study. 



Whether by mere coincidence or founded on some genuine physio- 

 logical basis, it is observable that in all four back crosses, when the 

 female in the back cross is the same as that entering into the original 

 cross, then no ova hatch ; and, on the contrary, if the female of the back 

 cross is of the same species as that providing the male in the first cross, 

 then the eggs hatch and the larvae progress more or less favourably. 

 That the matter is more than mere coincidence the separate and con- 

 firmatory evidence of several broods seems to show. 



(3) Miscellaneovs Pairings. 



For an exceedingly long time I have striven to obtain the necessary 

 pairings for introducing 0. dilutata into the autumnata-filigrammaria 

 series of hybrids, but only to be thwarted year after year by the exaspera- 

 ting way in which almost invariably 0. flligrammaria and any hybrid in 

 which it took part were over before dilutata appeared : the latter species 

 is immovable in its preference for the last days of September and for 

 October as the period of its imaginal activities. However, 1917 put me 

 in possession of a brood of 0. dilutata which produced a small proportion 

 of its members in September and these most luckily coincided in their 

 emergence with stragglers of several of my autumnata-filigrammaria 

 broods as well as with belated examples of filigrammaria itself. I was 

 therefore able to enclose dilutata females with Fi {filigrammaria % x 

 autumnataf^) males, Fi (filigrammaria $ x autumnata ^) females with 

 dilutata males, dilutata females with F^ males of the same cross as well 

 as to attempt reciprocal crosses between dilutata and filigraTnmaria ; 

 what these matings produced will be discovered below. 



The Pairing o/dilutata $ and the F^ (filigrammaria $ x autumnata</ ) 

 Male. 



In this cross over a hundred ova were laid, and although I did not 

 observe the pairing such had evidently taken place, for out of 114 ova 

 seventeen changed colour from green to pink four days after deposition. 

 This proved their fertility but, nevertheless, when spring made its 

 influence felt, instead of turning to the dark steel blue of ova developing 

 normally they collapsed and assumed a rusty hue and produced no larvae. 



The Reciprocal Pairing to this. 



Pairs caged up simultaneously with the last lot yielded ova, none of 

 which changed colour. They were thus infertile — a fact that might 

 Journ. of Gen. ix 18 



