J. W. H. Harrison 83 



factors for melanism in both of the original parents which were piled 

 up in the F^ insects to an extent quite impossible in either of the pure 

 strains or in the F^ broods. 



As, however (even were the truth of that hypothesis admitted), the 

 parthenogenetic male could only at the most possess the same number of 

 such factors as its parent, and very possibly actually less, the multiple 

 factor theory of transgression in F^ broods is not supported. With the 

 removal of this further prop, coupled with the destructive evidence 

 offered by the dissimilar results of the two sets of work just described, 

 it seems to me that a very shrewd blow has been dealt to the applica- 

 bility of the hypothesis to the critical case of Castle's rats. 



Let us now examine the matter from a somewhat different angle. 

 Opinion^ has been settling to the view that the differences between 

 species are, at bottom, chemical, and that diverse cytoplasms reacting 

 and interacting with diverse enzymes located in the chromatin of the 

 cell nucleus are responsible for opposed, specific characters. Apply this 

 concept to the present problem, and one readily perceives that the in- 

 troduction of enzymes foreign to the cytoplasm of an egg might inter- 

 fere very profoundly with the regular course of development ; in our 

 case the intensity of the melanism might be so influenced as to suggest 

 a very significant departure from the usual Mendelian anticipation. Two 

 very forceful arguments serve to destroy this idea. Firstly, the disturb- 

 ance (if any), if due to the difference in interacting chemical substances 

 and to no other circumstances whatever, should be at its maximum in 

 the Fi generation which, as we know, is not correct. And secondly, since 

 the amounts of cytoplasm supplied by ovum and spermatozoon are so 

 unequal, if there were any truth in it, obvious differences should exist 

 between the results of reciprocal crosses between the two species. This 

 is not so for the two lots of insects are substantially alike ; whether the 

 melanism is brought in by a crepuscularia male or female the broods 

 agree inasmuch as the melanism acts in the F^ lots as a simple Men- 

 delian dominant. Only in F2 insects and those of similar origin do the 

 abnormalities make their presence felt, and this at a time when one 

 would have anticipated that cytoplasmic values, chemically speaking, 

 would have attained their mean. 



From this position only one logical path of escape lies open, and 

 that is to admit that during the progress of hybrid development the 

 enzyme responsible for melanism, whether by stereoisomerism and the 



* Beichart has admirably summarised the facts concerning the chemical distinctness of 

 species, etc. in Science, Nov. 6, 1914. 



