110 BRITISH BUTTERFLIES. 



Felder, candalus, H.-Sch., icarus. v. Rott., sarta, Alph., amandus, 

 Schn., myrrha, H.-Sch., isaurica, Staud., hylas, Esp., and escheri, 

 Wo., as belonging to this section. The group has also a wide distribu- 

 tion in the Indian region, and, of de Niceviile's supposed species 

 (Butts. India, iii., pp. 71 et seq.), ariana, Moore, stoliczkana, Felder, 

 sutleja, Moore, fugitiva, Butler, persica, Butler, kashgharensis, Moore, 

 yarkundensis, Moore, nadir a, Moore, and pseuderos, Moore, certainly 

 appear to belong here. Some of these, as will be seen later, we treat 

 merely as forms of P. icarus. Our restricted Polyommatus comprises, 

 therefore, de Niceviile's Lycaena, group B (Butts, of India, iii., p. 68) 

 (except devanica, Moore), which he describes at considerable length, 

 and of which he notes (op. cit., pp. 70-71): " The second group 

 has the $ s more or less blue on the upperside, the ? s brown, some- 

 times with irrorated blue scales at the base of the wings. The under- 

 side of both sexes in all the species is greyish-brown or greyish, with 

 black spots surrounded with white ; in a few species there is a white 

 discal streak on the hindwing," etc. The great differences in the 

 ancillary appendages of the species of the genus Plebeius (sens, strict.) 

 and the other genera included in the tribe Plebeiidi, have already been 

 pointed out (see preceding vol., pis. xx., xxi., xxii., andl.), whilst the 

 comparative similarity of those of Cyaniris, Polyomniatus, and Agriades, 

 is equally well shown on plates xxi and xxii. As between Polyom- 

 matus (icarus) and Agriades (coridon, thetis), comparison should be 

 made of the figures of these structures as pictured in pi. xxi., figs. 3 

 and 4, and pi. xxii., figs. '6 and 4. The comparative descriptions of 

 the ancillary appendages of Agriades and Polyommatus (op. cit., p. 157), 

 as already published, read as follows : — 



Agriades: As exemplified in coridon and thetis, it is not so very difficult to 

 separate this from Polyommatus (icizrus), but in other species that appear to have 

 to be divided between these groups, the separation is much more so. This group 

 exhibiis a dorsal process with well-developed wing; the hook has a large square 

 hase, from which the comparatively slender, rather S-shaped, curved, upper part, 

 arises at about a right angle, and ends opposite the middle of the dorsal process. 



Polyommatus : As exemplified in icarus, eros, etc., this group possesses a less 

 heavy base to the clasp, a less slender upper portion, and its curvature is simple. 

 In Polyommatus the hard process of the clasp is rather shorter than in Agriades. 

 In both there is a soft process between the two divisions of the clasp [which often 

 swells out in preparation in a balloon-like way] ; whether this be regarded as a 

 part of the soft process of the clasp, or a development of the membrane between 

 the two processes does not much matter ; but its considerable development is 

 characteristic of these two genera. 



Chapman adds that, without reference to the generic value of the 

 difference in the characters indicated, the following points are noted 

 as distinguishing the $ ancillary appendages of Polyommatus (icarus) 

 from those of Agriades (coridon and thetis): — 



1. The dorsal processes are in Polyommatus (icarus) broader, and have a 

 transverse ridge across the hollow of the scoop ; in Ayriades (coridon and thetis), 

 what is apparently the same ridge is nearly longitudinal. 



2. The hooks in Polyommatus (icarus) are, as regards the angulation, inter- 

 mediate between those of A. thetis and A. coridon, but the point is much more 

 curved in the two latter. 



'6. The serrated point of the clasps has nearly the same number of teeth — 

 A. thetis, 13, A. coridon, 17, P. icarus, 16, or thereabouts — but the head which 

 carries them is more bent backwards in Agriades, projects a little forward in Poly- 

 ommatus (icarus), and is, in the latter, rather smaller, so that the teeth are closer 

 together, and stand up as rectangular points, rather than as teeth of a saw with 

 the posterior margin five or six times as long as the other. 



