■2-2 i LEPIDOPTEBA INBICA. 



of species, and in their last paper* they put them into two divisions, (1) those whose 

 larvae feed on dicotyledonous plants, and (2) those whose larvae feed on monocotyledonous 

 plants, and separate the divisions into groups in accordance with the habits of the 

 imagines, larvae and pupae, and these on the w^hole run very naturally on Watson's 

 lines, with a few important alterations. In this work we have adopted Watson's 

 arrangement, modified somewhat after a careful study of Davidson, Bell and Aitken's 

 most instructive papers on the life history of the species they bred, and of the habits 

 of the perfect insects, and a careful examination of their structural characters. We 

 put the Ismeneinse first, as Davidson, Bell and Aitken do, because they are more akin 

 to the other Rhojxdocera that have gone before than are any of the other sub-families 

 of the Hespei'iidse, and Watson himself says, though he puts them last, they are a well- 

 marked group of closely allied genera which show no close relationship with any other 

 genera of the Hesperiidse, and he suggests that they might, with advantage, be treated 

 as a distinct sub-family. 



We cannot, however, follow Watson in ignoring to a great extent the importance 

 of male secondary sexual characters in generic divisions ; both in this family and in 

 the last (the Lycsenidse) the sex marks we believe to be of very great importance in 

 the division of genera ; in this we are in accord with many distinguished lepidopterists, 

 such as Scudder, Moore and de Niceville. We do not deny the inconstancy of the 

 male character in certain genera, but even for the sake of convenience it is impossible 

 to ignore the male-mark distinctions between Baoris and Chapra, and the want of 

 them in Parnara and Caltoris ; to put all the species of such a very large group into the 

 genus Baoris is very inconvenient, the arrangement of putting them into sections is 

 very confusing, it is so confusing, no author ever adopts it ; in referring (for instance) 

 to Chapra matliias, no author ever puts it as Baoris [Section III.) mathias, it is 

 invariably entered by every author who understands his subject as Baoris (Chapra) 

 mathias, it is so much more easy to recognise the name than it is to recognise the 

 signification of the sectional number. Both Moore and de Niceville erected many 

 genera on the male marks alone, and we are of opinion that they were right in doing 

 so ; in a family in which the sex marks are usually so very distinctive, aud the venation 

 so constant throughout and useless for generic generalisation, the secondary sexual 

 characters must be of the greatest generic importance. Even with reference to such 

 closely allied genera as Hasora and Parata, de Niceville says,f "For my own part, I 

 think Parata might with advantage be used as a sub-genus, as its ' male-mark ' is a 

 very useful character by which to separate ofi" certain species, such as chromus and 

 alexis, from typical Ilasoras, such as badra, coulteri, hadria, anura, and cliahrona, which 

 do not possess this ' male-mark.' " 



* Jouvn. Bo. Nat. Hist. Soc. xi. 1897, p. 22. f Id. ix. 1895, p. 407. 



