POLYOMMATUS AND AGRIADES. 121 



Polyommatus and Agriades. 



By the late Dr. T. A. CHAPMAN, F.E.S. 

 Tutt {British Lepidopteya) adopted or accepted PobjonunatiiH, Latr., 

 and Af/riades, Hb., as distinct genera, with ieanis as the type uf the 

 former and coridon of the latter. Except, however, by indicating these 

 species as types, he gave us no character by which the species could 

 be allocated to their proper genera. 



Can we maintain these as distinct genera ? This cannot be 

 discussed without some preliminary agreement as to what we mean by 

 a genus. It will probably be agreed that a genus is a group of species 

 more nearly allied to each other than to other species ; as a theoretical 

 position this is unaffected by the practical difficulty, of intermediate 

 forms that might be put in either of two genera that, nevertheless, we 

 more or less agree should be treated as distinct. I do not think we 

 can accept Tutt's contention that a separate genus is due, if a few (or 

 even one) species have distinctive characters in common. In tabula- 

 tions of species in any large genus, it is usual to find "groups" 

 distinguishable from each other by some common characters. Tutt 

 viewed each group as a distinct genus. As a matter of fact, I not only 

 assert as my own opinion, but as a nearly universal one, though 

 comparatively few persons will admit that their attitude is really one 

 of agreement in this dictum, that a genus is entirely arbitrary; so far as 

 its extent is concerned, it is a matter of convenience. It is desirable 

 if we can find characters by which to define them, to substitute two or 

 more moderate-sized genera, in place of one large one. On the other 

 hand, however, we maintain genera containing only one species, 

 because that species has nr very close allies, and we wish to observe 

 what we may call the average amount of difterences by which genera 

 are separated from each other. 



To return to Poli/oiii»iatiis : i'oli/owinatiis and Agriades comprise a 

 large number of species which it would be convenient to sub-divide 

 into two or more genera. Tutt 8eemed to see that these fell into two 

 groups Polyounnatiis and Atjriadex, which he was able to distinguish 

 and which presumably had characters separating them. One, at once 

 asks, if this be so, what are these characters. Tutt did not tell us. 

 If there be no such characters, only Tutt can tell us, beyond the type 

 species, icanis and coridon, what species belong to either genus. This 

 is an impossible position, and if we stay there the two genera must be 

 combined. It remauis, however, that Tutt did see some characters, 

 not easy to seize for description, and which he did not attempt to 

 describe because he felt that a closer acquaintance than he had with a 

 larger number of species was necessary to do so with any valuable efltect. 



Considering that the separation of these two genera, if possible, 

 was for convenience desirable, I was pleased to be able to find a 

 definite structural difference between them, which I described in Ejit. 

 Ptec, vol. xxii., p. 101. I must admit, therefore, that I am to some 

 extent the upholder of these two genera as distinct. I seriously demur, 

 however, to being asked, " Am I really prepared to found genera on 

 one character in the J genitalia." As a general question, I think it 

 possible a case might occur when one should do so, but broadly, 

 I should answer as decidedly as the questioners believe I ought to do, 

 because they have cornered me, " Decidedly not." 



July — August 15th, 1922. 



