82 Inheritance of Melanism in Tephrosia (Ectropis) 



deductions from the experimental evidence to assert that changes in 

 the same direction are assuredly possible at other times no matter how 

 induced, be it by crossing or otherwise. 



Had the indications of my histortata-crepuscularia hybrids stood 

 alone, to the Neo-Mendelian they would simply have afforded a further 

 illustration of the truth of Nilsson-Ehle's hypothesis of cumulative 

 factors. He would have triumphantly pointed out the uniformity of 

 the F^ lots g.nd the great spread of variation in the F^ imagines and 

 heralded them as proof positive of his theory, so beautifully do they 

 harmonise with it. To him the sole import of the experiment would 

 have been to demonstrate that melanism, far from depending on a single 

 factor, arises through the cumulative action of many. 



However, its behaviour in the intraspecific crosses confutes this 

 emphatically; then it comports itself without any shadow of doubt 

 as a unit character. This effectively removes the power from the 

 same far-fetched (although logical from the Neo-Mendelian standpoint) 

 counter-argument considered so destructive in the analogous case of 

 Castle's rats. Hoodedness, we were informed, in that case simulated 

 a unit character merely because coexistent with it was a definite 

 " character " without which its manifestation was impossible. Its ap- 

 pearance thus being interdependent on the effect of this single suppo- 

 sitious "prompter," its behaviour must be that of a unit character. 

 Once it was permitted to appear the play of varying numbers of multiple 

 factors would admit of variation in the degree of hoodedness, and there- 

 fore render possible selection in respect to the character. 



No such " prompter " can be dragged in to entangle my results in 

 the web of the multiple factor theory. The disparity between the 

 enormous fluctuation of the melanism, even to its actual disappearance, 

 in the F^ crepuscularia-bistortata hybrids, and the uniformity of the 

 melanics in the F^ crepuscularia-delamerensis mongrels is too striking 

 to allow it. The only feasible scheme capable of accounting for the 

 facts is to declare that species-hybridisation has modified permanently 

 a unit factor for melanism and has very possibly been responsible in 

 certain cases for its loss. 



Simply invaluable in furnishing further proof of this is the condi- 

 tion of the insects bred parthenogenetically fi'omthe Fi crepuscularia- 

 bistortata melanic female. Here one of the insects transgressed the 

 degree of melanism of its parents (and even its blackish grandparent). 

 Had this happened with zygotes produced by normal fertilisation, its 

 existence would have unhesitatingly been assigned tp the presence of 



