10 Mr. E. G. Josepli on the Rhopalocera 



thaUa, Linn., is the only representative of this subfamily that 

 finds a place in Buichi IPs collection. The individuals of that 

 species, however, exhibit a very large amount of variation 

 both as regards size and pattern. 



We believe that none of the above-mentioned variations 

 are of specific value, although a fair proportion of the 

 specimens answers very well to Godman and Salvin's descrip- 

 tion of A. anfeas, Doubl. & Hew.^, and a few others possess 

 the rectangularly bent median black bar of the hind wing 

 that forms so consj)icuous a feature of A. pellenea, Hiibn. 

 But seeing that a careful examination of Burchell's material 

 shows every intermediate grade between tiie patterns of 

 A. thalia and A. anteas on the one hand, and of A. thalia 

 and A. pellenea-Yxko, specimens on the other, — considering, 

 furthermore, that the whole series was captured in S.E. Brazil, — 

 we conclude that, as far as this collection, at any rate, is con- 

 cerned, the above-mentioned differences are only of varietal 

 value. It is important, however, to confirm this opinion by 

 examination of structural characters in fresher and more 

 perfect specimens than those of the Burchell collection, by 

 the records of individuals observed in coitu, and, above all, 

 by the paramount test of breeding. 



I have, moreover, studied the descriptions and drawings of 

 these three species by recent and older writers, and, after 

 careiully comparing them with BurchelFs specimens, have 

 been unable to find any permanent characters by which 

 A. thalia, A. anteas, and A. pellenea may be separated — a 

 conclusion already arrived at, as regards the two first-named 

 forms, by Godman and Salvin, Staudinger, and other autho- 

 rities. Attention will be directed to any marked variations 

 that appear in the series, under the respective numbers. It 

 gives me much pleasure to acknowledge the kind help 

 rendered by Mr. Harry Eltringham, M.A., of New College, 

 Oxford, in the endeavour to ascertain the correct sex of the 

 Burchell specimens — many of them in a very dilapidated 

 condition. 



The whole of the geographical data and other notes beai'ing 

 on the scene of BurchelPs botanical and entomological opera- 

 tions in Brazil has now been transcribed by Professor Poulton 

 from the Catalogue of Brazilian Plants, containing the 

 numbers 701-10108, and has been utilized for the purpose of 

 this paper. It is a record of nearly five years* continuous 



1 ' Biologia Centrali-Americana,' Insecta, Lepidoptera Rhopalocera, 

 vol. i. p. 141 (1879-1901). 



