416 
Mr. Cecil W. Bafker^s 
Dec. 12th, 1891, 6 typical. Jan. 4tli, 1892, ? typical. 
Jan. 8th, 1892, c^, typical. Aug. 12th, 1892, 6 , dry 
season. 
Aug. 22nd, 1892, c?, dry season. Sept. 17th, 1892, 6, 
dry season. 
Sept. 9th, 1892, ?, dry season. 
November, 1892, H. dmdalus, dry season form still 
plentiful. 
January, 1893, wet season form of K, dsedalus fairly 
plentiful. 
June and July, 1893, observed H. dsedalus dry season 
form fairly prevalent at Black Umvelosi Eiver, 
Zululand. 
On return to Malvern — 
Aug. 29th, 1893, one ^, dry season. 
Oct. 23rd, 1893, one ?, dry season. 
Nov. 14th, 1893, one ?, dry season. 
Charaxes ethalion, as well as some others of the Natal 
Nymphalidse, modify seasonally, but sufficient examples 
of this family have been brought forward to illustrate 
the tendency of seasonal transition, according to the 
principles given above. 
Such of the Lycsenidae that appear at both seasons of 
the year show considerable differentiation in their 
summer and winter forms. Thus the winter examples of 
Lycsena lucida are smaller and have the spots of the 
underside more or less contracted or obliterated. 
Some species, vide Lycsena cissus and Lycsena harlcerij 
are modified by the spots and markings of the underside 
showing so faintly as to be scarcely distinguishable from 
the ground-colour. 
It is, however, among the genera Terias, Pieris, and 
TeracoluSj of the subfamily Pierinse, that seasonal 
differentiation is most conspicuous and has caused the 
greatest errors and confusion in the definition of species. 
The number of species recognized by Mr. Trimen, 
F.E.S., etc. , for the Extra- Tropical limit of South Africa, 
of the genus Terias is seven. 
Careful observations of their habits during a course 
of seasons, and bearing in mind the tendency of seasonal 
forms to adopt certain fixed rules of differentiation, 
would dispose me to reduce this number from 7 to 4, 
