TERACOLUS. 63 



underside of the hind wing becomes pure white without any marking's whatever. 

 Such specimens, however, seem to be rare, as there is nearly always some trace of the 

 oblique dusk}^ discal ray from costa, whicli is so characteristic of the group. It was 

 on a dr^'-season specimen of this variet}- that Dr. Butler founded his T. jaloiie, which 

 has the underside of secondaries of a pinkish tinge with a faint discal ray. This 

 again merges gradual!}" into T. impcralor (Central East Africa) both in the develop- 

 ment of the discal ray and in the tendency to assume a sixth spot in the purple apical 

 patch, wliich is generally present in that species, but is very variable in size, and 

 sometimes absent. '/'. ioiie. Trim, (part), and 7! ja/o/ie, var. i/nluleinsi's, Stand., 

 represent the Natal form of the sjjecies, which differs from T. ini/jerator in its smaller 

 size, the absence of the sixth spot in apical patch, and the rather stronger development 

 of the inner black edging of the purple ; but specimens from Mashoua Land and the 

 Transvaal show every intergrade between the two forms. 



" In the cjuasi-tropical coast-belt of Natal another variation occurs, in which the 

 purple patch is slightly reduced owing to the broadening of the inner black edging, 

 and the ends of nervules on the underside of hind wings ai'e strongly blackened, and 

 often terminating in spots on the hind-margin. T. hudioiii is the normal dry-season 

 form of tlie species in Soutli Africa, the Central African specimens being noticeably 

 larger. Although the males of this species are so variable, the females are even 

 worse and the variations are not so localised. Not only does the ground-colour vary 

 from white to bright yellow, but even the discal black markings are apt to be very 

 much reduced, and the apical patch may be either red or black ; in the latter case it 

 contains a row of small spots, which may be either white, yellow, or red. Tiie tints of 

 the underside also vary much, and there seems to be sporadic tendency to l)lackening 

 of nervules. 



" T. /jlile//i/as ranges tliroughout East Africa from Natal to Abyssinia, and in the 

 Southern Tropic it extends westwards to Damara Land and Ovampo Land." 



I3r. Butler writes : — " T. jjhl(;(/j/as, in all its forms, can be distinguished from 

 T. im/jeralor by its somewhat inferior size, the whitisli scaling in spots upon the 

 apical border of the males, and the transverse bar on the under surface of the 

 secondaries being usually more broken uj). The females are much less heavily marked 

 on the upper surface. 



" T. coViafjenes, which Mr. Marshall regarded as linking the T. erk and T. faitsta 

 groups, is certainly nothing more or less than the wet-season female of the typical 

 male of T. phlefjym ; the female which I described is the dry-season type, and there- 

 fore is that sex of 7\ Jf/loi/c." 



