336 BRITISH BUTTERFLIES. 



1909, when the season was again very late, and A. thetis only just 

 emerging towards mid- September at Folkestone, the ? s were very 

 dark in ground colour, and well scaled with blue. Adkin reports 

 (Proc. Sth. Lond. Ent. Soc, 1906, p. 69) that, in June, 1906," the 

 species was common near Eastbourne, that most of the $> s showed 

 some blue coloration, without being extreme in this direction. He 

 had compared them with $ s of the spring brood captured at 

 Folkestone, with which they agreed closely in, the amount of blue 

 scaling, but that ? s from Reigate, and other inland localities on the 

 North Downs, were much more strongly scaled with blue ; further, he 

 observed that the spring 2 s appeared to be infinitely more scaled with 

 blue than those of the autumn brood, but, in September, 1908, also 

 near Eastbourne, he noted (op. cit., 1908, p. 93) that the $ s were 

 more heavily scaled with blue than he had before observed them, either 

 in spring or autumn, and he came to the conclusion that neither 

 locality nor time of emergence accounted for the prevalence of the blue 

 form. Certainly it would seem that the brownest, and least scaled 



with blue, $ s are those that occur comparatively early (August) after 

 hot summers in Britain, e.g., 1887, 1893, etc. Similarly, there is no 

 trace of blue in the $ s of the second-brood taken at Gresy-sur-Aix, 

 Digne, and other localities in southern France in July and early 

 August. Oberthiir records the most brilliantly-blue $ s from western 

 France, from La Vendee to Bordeaux, along the littoral, and not occur- 

 ring for any distance inland ; this form has been named coelestis. There is 

 a great amount of variation in the size of the specimens in both sexes, 

 in both broods. The smallest examples (doubling from apex of wing- 

 to centre of thorax) in our British collection, measures about 24mm., the 

 largest about 36mm., $ , 38mm. $ ; the smallest $ in our Continental 

 collection reaches 30mm., the largest $ s 40mm. -42mm. ; the smallest 

 5 32mm., the largest ? s 38mm. -40mm. Reverdin states that the 

 largest $ in his collection (out of 65 specimens) measures 33mm. 

 (doubling the length of the wing), the largest $> (out of 27 specimens) 

 34mm. ; the smallest $ measures 28mm., and the smallest ? 30mm. 

 Turner records {Proc. Sth. Lond. Ent. Soc, 1908, p. 64) the capture of 

 two small 5 s at Ranmore, the smaller only 22mm. in expanse. 

 Examples 26mm. and less we call ab. minor, those more than 38mm. 

 in expanse, ab. major. As will be noted above there is some difference 

 in size racially, e.g., our British race averages much less in size than 

 those of southern France. The localities that give the largest average 

 are Draguignan (May, 1905) and Digne (August, 1905). The general 

 ground colour of the underside of the <y s varies from whitish- 

 and pale grey, to blackish- or brownish-grey, always a little 

 darker and browner on the hind- than on the forewings ; 

 Blachier notes (in litt.) that some of the specimens of the 

 spring emergence, have the underside of the hindwings as grey 

 as the forewings, but slightly darker ; some ^ s of the autumn emerg- 

 ence (Folkestone, 1909) are quite blackish-grey. The ground colour 

 of the ? s is more brown-grey than the $ s, that of the hindwings 

 more fulvous or brownish than the forewings, although in some 

 examples the difference is slight. The undersides of both sexes may 

 be roughly grouped as (1) Whitish-grey, the hindwings tinged with 

 yellowish = ab. pallescem, n. ab. (2) Dark grey, the hindwings strongly 

 browned = ab. fuscescens, n. ab. (3) Blackish-grey, the hindwings 



