INTRODUCTION. 111. 



with a i)hysioIo^ieal fact, and always with the sm/ir physiolo.s^ical fact, 

 r/:., the posnibihty of genuine hybridisation. This is, of course, 

 merely an ideal which can only be reached by an immense amount of 

 actual experiment, and for which neQ;ative evidence requires to be 

 cumulative before it can be regarded as havinif much force ; but to 

 search for characters by which a genus can be split up, simply because 

 it is large and " unwieldy " (why unwieldy ? who wants to " wield " 

 it?), or to divide genera by characteristics of varying importance, 

 is to cut at the root of ail possible development of classification 

 on scientific lines. The second is the doctrine that the first- 

 described form of an insect (or of anything else) must be regarded as 

 the type. It would perhaps be most scientific to regard the most 

 ancestral form as the type if this Avere possible, but, as in most cases, it 

 would be impracticable to determine which is the most ances'tral form, 

 the only possible alternative is to regard as typical the most generally 

 distributed form (surely in most cases a matter easy enough to deter- 

 mine), and others as varying from it. I cannot help feeling that 

 classification and nomenclature, even at the loss of some degree of 

 fixity, ought to be made subservient to scientific fact, rather than that 

 such fact should be obscured by a too pedantic adherence to the letter 

 of the law of priority. With regard to the application of these prin- 

 ciples to particular cases I am most humbly open to correction, from 

 the principles themselves I see no reason to deviate. 



While on the one hand, as Mr. Tutt and others have pointed out, 

 no linear arrangement of classification can be really satisfactory, it is 

 obvious, on the other hand, that both in books and collections no other 

 is possible, and the problem is to choose the least objectionable, i.e., 

 that which least glaringly contradicts scientific facts so far as they are 

 established. It is, of course, highly improbable tliat any arrangement 

 chosen would quite commend itself to anyone else, unless to someone 

 who had published one of his own which was strictly adhered to ; so, 

 as I have not strictly adhered to any, everyone is likely to find some- 

 thing to complain of, and, in fact, I have finally adopted the present 

 one, not because I could not find complaints against it myself, but 

 because " in the present state of our ignorance " I could find more 

 against any other. I consulted Mr. Tutt on this matter, and have in 

 most cases followed his suggestions, but the position which I have 

 assigned to Nemeobius Lucina, and Libythea Celtis will not, I fear, 

 meet with his approval. The practice of beginning with the more gen- 

 eralised forms, though the more logical, has still probably many enemies. 



With regard to the actual nomenclature employed, that of the Hes- 

 periides is the most uncertain ; I feel convinced that some at least of the 

 generic names will prove to be incorrect, but pending further investiga- 

 tion, I have applied them in accordance with the use of others more 

 competent to judge than myself. If it be necessary to ignore Billberg's 

 generic names in this case, it would appear that Lejitidia should also 

 give way to Lcptusi^, Hb. ; I have, however, retained Callophri/s, for 

 which there seems to be no alternative. One or two generic divisions, 

 e.f/., Drijas from Ariiynnia, and Xomiadcx from Cii/nda, I have retained 

 unwillingly, and I hope temporarily, but the genus Melampias (as well 

 as some others not generally adopted), I have felt it necessary to suppress. 

 Our knowledge of the earlier stages of the (inclusive) genus Krebia, is 

 far too slight and fragmentary to serve as a basis for subdivision, the 



