1886.] W. Doherty — A List of Butterflies taJcen in Kiimaon. 107 



For the last three years I have been studying the eggs of Indian 

 butterflies, in the hope of thereby throwing new light on the grouping 

 of these beautiful insects. In this hope I have not been disappointed. 

 In most cases a genus is well and easily defined by its venation, but 

 that venation only very imperfectly expresses the relationship in which 

 it stands to other genera. To define families and subfamilies, one 

 must, I think, study the ' ovation '. For example, the group called 

 by Kirby the subfamily of the Nymphalincs now presents a confused 

 mass of many genera the relations of which intsr se are extremely 

 vague and uncertain. When the ovation is studied, these genera fall 

 into convenient groups, defined by constant and peculiar forms of egg. 

 I am aware that this classification is not likely to become a popular one : 

 the student will always prefer to separate his genera by an artificial key 

 based on the venation. But that these groups of mine are natural ones, 

 and indicate in most cases the true line of descent, I do not doubt. 

 All considerations of markings, shape, habits, geographical range, and 

 sexual variation indicate the affinity of the genera here placed together. 



Unfortunately my work is still very imperfect. The study of the 



eggs of butterflies is a difficult one. I may state as an example of this, 



that I examined the bodies of more than forty females of Melanitis leda 



before I obtained a single egg. I was for two years almost equally 



unfortunate with the much rarer females of Gharaxes. My method was 



in all cases to obtain the egg by pressure from the abdomen of the female. 



Now the egg of some butterflies, including many of the Satyridcd, is 



marked with highly characteristic lines of tubercles appearing some 



time after it is laid. These I have hitherto been unable to study, and 



am unable to say whether they extend to the closely allied groups of 



the Morphidce and Elymniadce. On account of the deficiencies of my 



early notes, I am not able to place such common genera as Curetis and 



Loxura, though from my vague descriptions these genera seem to belong 



to the Deudorix group. Many of the rarer forms I have not yet obtained ; 



the eggs of Calinaga, of Neiirosigma, and of the true Gerydus are 



still unknown to me. Owing to these great imperfections, I have 



delayed, and shall delay for another year or two, the publication of my 



studies of the eggs of eastern butterflies. But I take the opportunity 



to give a short synopsis of the results hitherto obtained, as a kind of 



prodromus or precursor of the perfected work, defining each group as 



briefly as I can. I omit certain points not yet thoroughly tested, such 



as the distinction between the Vanessa and Ajpatura groups of the 



ApaturidcB. My division of the Hesperiadce is a tentative one. The two 



last groups, though typically very distinct, show a tendency to coalesce, 



and I include them under the head of the single subfamily Baorince, for 



