AGRIADES CORIDON. 53 



race, but this no doubt is so, the amazing difference in the tint of the 

 ground colour, as painted, being almost entirely due to the different angle 

 at which the insects have been observed during the time the colouring- 

 was done. We have before us a magnificent series of 36 Aragon and 

 Castilian $ s, 33 taken by Dr. Chapman, with full data. These divide 

 into two sets (1) 12 labelled "Avila, July, 1902," and 6 labelled 

 " Navalparai, 5. viii. '01," all exhibiting well-defined marginal borders 

 to forewinas, and the underside spotting rather weaker than that of 

 Herrich-Schaffer's fig. 495. (2) 1 labelled " Moncayo, 2-5000ft., 

 July 12th-24th, 1903," 4 " Soria, July 25th, 1903," 8 " Albarracin, 

 28. vii.-6. viii. '01," and 2 "Caenca, 9-16. vii. '01," averaging a little 

 larger, some with rather paler marginal borders, and the underside 

 distinctly more spotted, more like Gerhard's pi. xxxii.,fig. lb, but with 

 less brilliant red lunules. On the uppersicle', looked at from above, the 

 wings sloping down somewhat, the ground colour is largely of a delicate 

 blue-grey, .with pale lunular interneural patches exactly as in Herrich- 

 Schaffer's figure ; when the drawer is held vertically, and so that the 

 light falls directly on the specimens, the ground colour is largely 

 shiny metallic creamy- or bronzy-green (the blue-grey being reduced 

 to a minimum) exactly as in Gerhard's pi. xxxii., fig. la. Close 

 examination shows that the shiny creamy -green scales are below the 

 long blue-grey hair-scales, and that the essential difference between 

 the blue-grey albican* of Herrich-Schaffer (fig. 494) and the creamy- 

 green arragonensis of Gerhard (fig. la) is merely due to an excess or 

 defect of the long blue-grey hair-scales, which are ashy- white in the 

 Andalusian albicans, Bdv. The best blue-grey examples, therefore, as 

 represented by Herrich-Schaffer's fig. 494 ( = ab. caerulescens, n. ab.), 

 are merely aberrational arragonenah, with an excess of these long blue- 

 grey bair-scales. It may be further noted that the largest specimens 

 of this race are most frequently of the caerulescens form, as if increase 

 in size and increase in development of hair-scales usually occurred 

 simultaneously. In our long series we have no example with the 

 marginal band on the forewings as in Herrich-Schaffer's fig. 494, that 

 of Gerhard (fig. la) is very characteristic but too dark. As a matter 

 of fact there is considerable variation in the marginal border of the 

 forewing ; some are excellently marked w T ith ocellated spots, edged 

 internally with dark interneural lunules ( = ab. punctata), others have 

 the ocellated spots indistinctly merged in a marginal border the same 

 width ( = ab. indistincta), others have an increased border in which the 

 marginal spots are. indistinctly merged, or are represented only by the 

 inner halves of the pale rings, which m some are made up of a row of 

 curves (ab. subfasca), in others appearing as a straight dividing line as in 

 Gerhard's figure ( = ab. divisa), others, again, have wide dark borders 

 ( = ab. typica) in which the spots are practically invisible, whilst one has 

 the margin widely extended apically towards the discoidal lunule ( = ab. 

 maryinata). It is worthy of notice that the Avila examples have the 

 darkest margins and the palest and most obsoletely marked under- 

 sides, the Navalparai examples, also with pale undersides, coming 

 next ; the Moncayo example has a very wide border strongly shaded 

 internally, but there is only a single specimen. Inside the dark 

 margin, whatever its stage of development, the ground colour suggests 

 a series of pale interneural patches. The discoidal lunules are usually 

 (not always) traceable, sometimes distinct, those of the forewings and 



