AGKIAbES C0Rl£>ON. 17 



elongate forms produce well-defined submedian cuneate streaks, some- 

 times of considerable length ( = ab. extensa), and, occasionally, join 

 the discoidal externally, whilst the basal spots in addition may be 

 elongated toward, and join, the discoidal internally, making very 

 complicated striate forms (ab. striata, Tutt). The discoidals vary con- 

 siderably in size and shape, on the forewings from a faint grey 

 obsolete lineation to a large black kidney- shaped, or 8-shaped, spot 

 divided medially by a pale nervure (discoidalis-duplex) , on the hindwing 

 it may be quite lost in the ground colour in the most albescent forms, 

 whilst in the best developed it may have a well-defined black kernel, 

 although usually a mere white blotch. The basal spots vary con- 

 siderably, and may consist of 0, 1, 2, 3, or 4, to which forms Courvoisier 

 applies the names impuncta, unipuncta, type, tripuncta, and quadri- 

 puncta respectively. In the case of the unipuncta form, the top or 

 bottom spot may be absent, sometimes a different one on the two wings, 

 in the tripuncta it may be the upper or lower one that is duplicated, 

 equally commonly the extension is in the form of a line, so that we get a 

 dot and line ( — elongata, Courv.) with the dot top or bottom, whilst in the 

 quad rip uncta form the extension frequently takes on linear development 

 and the spots become a line with two dots or two lines (bilineata) ; the two 

 sides are frequently asymmetrical in the basal spotting. There is con- 

 siderable difference in the size of the ocellated spots, and, on the whole, 

 those of the $ s are smaller than those of the £ s, but this is, in some 

 instances at least, in part due to the ground colour. The spots normally 

 consist of a black kernel surrounded by a fairly wide white ring, con- 

 spicuous enough when the ground colour is grey, fuscous, or brown, but 

 lost when the ground colour is white, hence, in the $ s, from the Trafoi- 

 thal, and other localities of the albescent, i.e., pallida, form, the black 

 spots, robbed of their rings, look quite small, although probably not really 

 less than other specimens whose spots look much larger because of the 

 broad white circles contrasting against the darker ground colour. 

 The spots, however, may be, and frequently are, much reduced in size 

 (usually with some sign of accompanying obsolescence), and these 

 small-spotted forms have been named parvipuncta, Kebel ; on the other 

 hand, the spots, together with their encircling white rings, may be 

 exceptionally large, when they are known as crassipuncta, Courv. South 

 figures (Ent., xx., pi. i., fig. 7) a very finely-spotted $ underside, but 

 our 2 examples from Bourg St. Maurice, Versoix, Fontainebleau, Susa, 

 the Riviera, etc., seem almost or quite as large ; indeed, large-spotted 

 ° s appear to be almost the rule in many localities. Blachier notes 

 (in lift.) that, in the Swiss and Savoy specimens, the spots on the fore- 

 wings are, on the whole, small, and often, because of the pale ground 

 colour, scarcely appear to be surrounded by white ; on the hindwings 

 the spots are strongly circled with white, the black kernels relatively 

 small, the white rings relatively wide. Beverdin observes (in litt.) 

 that the increase and decrease in size is not always uniform or pro- 

 portional, and, especially in some cases of parvipuncta, the top and two 

 bottom spots often seem to be more reduced proportionally than the 

 others. He further notes that the white rings appear to be more 

 effaced in mountain races, less in those of the plains, but observes that, 

 whilst in specimens from Arolla the rings are almost absent, in 

 those from Berisal they are well-marked, and whilst those from 

 Brides-les-Bains are well-defined, those from Allevard, at the same 



