76 Inheritance of Melanism in Tephrosia (Ectropis) 



^ by the existence of the fact but by the approximation of the percen- 

 tages of each sex and colour they display, the matter must be submitted 

 to further investigation. 



Returning now to the question of the inheritance of the melanism, 

 since the hybrids were perfectly fertile, matings were made between 

 homozygous type males from brood VII and females of the same source 

 and form, as well as between heterozygous melanics chosen from the 

 same batch. Furthermore, males from family III and, in one case, a 

 wild second brood male bistortata were caged up with typical females 

 from brood VII. Just as in the case of homozygous recessives extracted 

 in the course of the experiments detailed in Section II nothing but 

 types (16 ? $ and 34 (/(/) were obtained, and the same holds good with 

 the back cross with T. bistortata where the results are 32 ? ? and SSf^J*. 

 Here again we see nothing antagonistic to Mendelian anticipations. 



Widely opposed to this was the evidence submitted by the other 

 two crosses. In the second cross (family IX), where Fi heterozygous 

 melanic insects were mated together, so chaotic were the F2 imagines 

 (22$ $ and SQc/^/*) in appearance, and so wide was their range of 

 variation that neither description (except as individuals !) nor classifica- 

 tion of the imagines seems possible. In addition to this, an F3 generation 

 bred from these nondescripts simply reproduced the parental condition 

 and equally defied classification. Strangely enough, careful study of 

 the exact tendencies of these broods suggests that they lean most 

 strongly to paleness; captured wild amongst first brood T. bistortata very 

 little attention would be attracted by their patterns. 



Nor are matters greatly improved with the other cross (family X) of 

 parentage {crepuscularia % x bistortata cT) ? x {bistortata ? x crepus- 

 cularia delamerensis </) cT which gave rise to 21 $ $ and ^*l ^ ^. Of 

 these the following is an attempted classification on a colour basis : 



(a) One male was as dark as its male parent. 



(6) Eight (all males), varying amongst themselves, whilst not black, 

 are dark enough to excite little or no comment. 



(c) Two (both females) were palish, but nevertheless powdered with 

 dark scales. 



id) Thirty-seven (four males and thirty-three females) were so 

 variable and transitional as to baffle description. As far as their general 

 coloration is concerned, however, none transgress beyond the range of 

 variation shown by North Kent bistortata; they are simply separated 

 from section (e) because of their variation in other respects. 



