J. W. H. Harrison 263 



indication of filigrammaria blood ; their markings are absolutely those 

 of typical autwninata. But they possess this very important feature ; 

 their range of fluctuating variability is down to a minimum, for I have 

 never possessed such a uniform brood from any female in the genus. 

 They give one the impression of a deadly uninteresting monotony not 

 even relieved by the usual sexual dimorphism of the group.- 



In variation exhibited they are far inferior to either pure autumnata 

 or to filigrammaria, whereas if the multiple factor theory be the genuine 

 explanation of the continuance of the F^ blend here as well as in the 

 F^ lot, the range ought to have been much greater rather than much 

 less. The independent segregation of all of many unit factors in the 

 oogenesis of the ^i female should have resulted in an immense field of 

 variation when in zygotic combination with the gametes of the autumnata 

 male. Again the evidence of the cross distinctly opposes the multiple 

 factor scheme and its gametic purity; it is all in favour of lasting 

 gametic contamination or blending. 



It will thus be seien that a single back cross has sufiBced, to all 

 intents and purposes, to cause the hybrid to revert to one of the parent 

 species. 



As in other cases the hybrid insects were fertile in all directions but 

 only larvae from one mating in which they took part were reared to 

 maturity; these were the result of pairing the insects inter se. So alike 

 were the insects involved that no selection of genotypes was possible. 

 To ensure as variable a progeny as possible ten females, chosen at random, 

 were caged up with a similar lot of males, and about a tenth of the 

 mixed ova retained for breeding. In spite of these precautions, the 

 insects bred showed no signs of segregation and in fact were indis- 

 tinguishable in size or design from their parents and, furthermore, 

 displayed the same limited extent of variation. 



The Back Gross between autumnata $ and the Fj </ • 



Owing to a shortage of cage room this cross was not proceeded with 

 in 1918. During that season, however, a further ^i brood derived from 

 the same stocks as those employed in the 1917 experiments was reared. 

 From this brood an average male was taken and paired with a female 

 autumnata from the same strain as utilised before. Fertile ova were 

 duly obtained, and the larvae issuing from them reared during the 

 present season. 



Up to the present (October 23rd, 1919) two-thirds of the pupae 

 have yielded their imagines. These are of the same expressionless type 



