J. W. H. Harrison 67 



Practically, as the form is looked upon merely as a fluctuation of 

 var. delamerensis, and would be called by that name by all British 

 lepidopterists, and, further, as this popular conception of the insect, 

 when viewed in the light of genetical fact is certainly sound, it will 

 be the uniform practice throughout this paper to refer to the insect 

 by the latter (and prior) name. Thus the use of redundant names with 

 which modern entomological literature is so congested will be avoided. 



For ova of T. crepuscularia type I have mainly relied on my friend 

 Mr Goodwin of Wateringbury, Kent, save in the rare instances when I 

 obtained typical females from- Cheshire and Buckinghamshire ; on the 

 contrary, my ova of the melanic variety were derived from various 

 sources, Birmingham, Sutton Coldfield, Delamere Forest, Wakefield, 

 Doncaster and York, all having contributed their quota via different 

 friends. Similarly, Mr Goodwin has always been good enough to pro- 

 vide me with ova of ordinary double-brooded T. bistortata, whilst I 

 myself have captured the females which supplied ova of the single- 

 brooded race, peculiar to northern and subalpine habitats, in larch 

 woods in the Cleveland District of Yorkshire, and in North Durham. 



Nevertheless, except casually, the offspring of only three melanic 

 females has been utilised in the progress of this work. Two of these 

 females proved, from the nature and genetical behaviour of their 

 progeny, to have been homozygous, and the third heterozygous, for 

 melanism. 



Almost at their inception the experiments seemed doomed to failure, 

 so difficult did it appear to rear Tephrosia crepuscularia of either form^; 

 with exasperating uniformity the larvae either contracted some disease 

 or the resulting pupae were few in numbers and dwarfed in size. More- 

 over, no amount of care and cleanliness seemed capable of removing the 

 obstacle which at first was considered to depend on the long drawn-out 

 larval period. Since, however, the single-brooded race of T. bistortata 

 lived an equally protracted length of time in th^t stage and betrayed 

 no such constitutional weakness, this possible cause was ruled out of 

 court. Juggling with various food-plants was then attempted to 

 neutralise the defect but Salix caprea, S. aurita, S. fragilis, S. alba, 

 Quercus pedunculata, Betula alba and Crataegus oxyacantha were all 

 quite ineffective, and even in one case, Salix alba, appeared to increase 



1 In view of the experience of myself and others with the superior vigour of melanic 

 varieties of Tephrosia consonaria, Boarmia consortaria, Oporabia dilutata, and O. autum- 

 nata it is advisable to state here that, if any difference was perceptible in the two varieties 

 of T. crepuscularia, the type form was the more robust ; at any rate, I always seemed to 

 rear larger broods and more vigoroas imagines from type stocks. 



6—2 



