232 BRITISH BUTTERFLIES. 



forewings powdered with coppery scales," may belong to this category. 

 Holmes also reports a $ of var. salmacis, from the Kendal district, 

 with the orange spots on the right side much lighter than those on 

 the left.* 



Variation. — Grum-Grshimailo says (Rom. Mem,. Lep., iv., p. 393) 

 that " amongst the Lycsenids there is no single species which varies 

 so much as this. Every corner has a form of astrarche which belongs 

 to it." This statement, we must confess, appears to us to be in wild 

 exaggeration of the facts. It is true that within very narrow limits the 

 variation of A. medon exhibits a considerable number of forms, a few 

 of which are very strking, but there is nothing approaching the same 

 scope for variation as is possessed by species of dimorphic sexual 

 colouring, such as Polyommatus icarus, or the two British species of 

 Agriades, and the invariable absence cf basal spots from the underside 

 forewing deprives this species of a fruitful source of minor aberration, 

 of which P. icarus and A. coridon, at any rate, seem to take the fullest 

 advantage. Again, the striated, and, except in the north of England, 

 the obsolete, forms are few and scarce, and the colour- variation in the 

 ground of the upperside is, comparatively speaking, very slight. So 

 far, again, from every corner having its own special form, purely local 

 races can hardly be said to exist, apart from var. artaxerxes and (if it 

 be a form of medon at all) var. chinensis, though, to some extent, nazira, 

 salmacis, sarmatis, montana, and cram era are specially identified with 

 particular districts. Courvoisier, indeed, goes so far as to assert (Ent. 

 Zeits., xxiv., p. 126) that neither the amount of red spots nor the 

 colour of the underside (the two principal points in which variation 

 takes place) has any strong connection with sex, or brood, or locality, 

 nor have these points any special connection with each other. This 

 statement is based on some 200 specimens, and our own observations, 

 based on probably ten times as many, bear out his view to a consider- 

 able extent, though it certainly requires modification. Indeed, he him- 

 self implies that it is rather the number of exceptions than the absence 

 of any general resemblance among the specimens of any given locality 

 that causes him to write to this effect. If, however, the range of 

 variation is comparatively small, it is very fully occupied. The 

 ground-colour of the upperside may be of any shade from a deep 

 black-brown to a brightish burnt-sienna, while that of the underside 

 is to be found varying from almost pure white to a rich coffee-colour, 

 passing through every intermediate shade of ochreous-grey on the one 

 hand, and of yellow-brown and cafe-au-lait on the other. The spot- 

 ting of the upperside may vary in amount from complete absence 

 (sarmatis, Gr., and allous, Hb.) to a broad, unbroken, antemarginal 

 band (cramera, Eschh.), and in colour from the most brilliant orange- 

 red (calida, Bell., etc.) to pale yellow (pallidior, Obth.), or even white 

 (f/raafii, ver Huell) ; while the usually black discoidal spot may be 

 ringed with white (albiannidata, Harr.), or even become entirely white 

 (artaxerxes, Fabr.) on the forewings, and sometimes on the hindwings 

 as well. Between the bright antemarginal band and the edge of the 

 hindwings may appear a row of black spots in every degree of dis- 



* I bave a Swiss specimen in which the outer half of the left forewing is 

 lighter than the rest of the ground-colour, and the spots at the border of the wing 

 dull yellow, whilst those on the other three wings are normal. — (G. W.) 



