1897.] BUTTEErLIES COLLECTEB Ilf NATAL. 851 



The dry, intermediate, and wet phases of this species have been 

 described as distinct species^ T. hrigitta being the female of the 

 dry phase, T. candace intermediate, T. zoe wet. The capture of 

 all three phases in August looks somewhat curious, but the dry- 

 season female is a good deal worn and the intermediate male a 

 little broken. 



89. Terias mabshalli, Butler. (Plate L. figs. 8, 9.) 

 Terias desjardinsii, Trimen, nee Boisd. 



2 2 ■> Malvern, 10th and 13th August, 1896. 

 The specimens belong to the dry-season form of the species. 

 T. regularis is a very distinct species. 



90. Teeacoltjs johnstoni, Butler. 



Tugela River, near Weenen, 2500 feet, 22nd October, 3rd, 

 1.5th, and 16th Xovember, 1897. 



This is the T. eris of Trimen and Marshall, but not of King. Tt 

 differs from the North-African type in its slightly more elongated 

 primaries and consequently greater elongation of the white area 

 on these wings as well as of the ochreous streaks on the apical 

 patch, and in the abrupt termination of the black costal border of 

 the secondaries, which does not emit a transverse streak to the 

 radial nervure as in T. eris-, the internal black stripe on the 

 primaries of the female is also constantly narrower. 



From the Eastern T. opalescens, and the Western T. inaimuna, 

 the differences are even more marlced, and nobody looking at a 

 long series of each could hesitate for an instant respecting the 

 local constancy of the characters which distinguish these four types 

 of the T. eris group. 



91. Teeacoltjs BirxToifi, Butler. 



(S cJ , wet-season ( = natalensis, Staud.), Tugela, near Weenen, 

 2500 feet, 10th and 15th November, 1896. 



Mr. Marshall labels this " T. plilegyas" but, in my opinion, it 

 occupies a central position between the latter and T. ione. The 

 wet-season form nearly approaches this butterfly on the upper 

 surface, whereas the uuder surface much more nearly resembles 

 the wet form of T. imperator. It must, however, be borne in 

 •mind that (in his " Notes on the Synonymy of Teracolus ") 

 Mr. Marshall has not distinguished between T. imperator and 

 T. pJilegyas, the absence of any examples of T. natalensis in our 

 collection, when he examined it, having perhaps led him to believe 

 that the latter was indistinguishable from wet-season examples of 

 T. imperator. The two differ much as our Oanoris rapce does from 

 G. hrassiae ; and as they do uot occur together, I am satisiied to 

 regard them as distinct species. 



92. Teeacolus ione, Godart. 



cJ , dry-season, Malvern, 800 feet, 10th August, 1896. 

 This is the form to which I gave the name of T. johina. 



