164 THE ENTOMOLOGIST. 



fig. 2, which was bred by heat, and in A. levana, fig. 3. The 

 latter is an apparently very primitively marked female specimen 

 from the Ussuri regions in Eastern Siberia (for hind wing, see 

 fig. 4 in the next row of figures), and not only exhibits the 

 second costal blotch divided into three parts (counting the light 

 median part, which is of the original ground colour), as in the 

 fritillaries — (fig. 11 is drawn from B. eaphrosi/ne ab. ; see pi. 56, 

 fig. 1, in Mr. South's 'Butterflies of the JBritish Isles') — but 

 also the black inner marginal spot is broken up into three parts. 

 That the second costal blotches of V. io and T''. urticcB originally 

 consisted of three such parts, as are shown in the upper side of 

 A. levana, is betrayed by the corresponding parts in the under 

 sides of these species : fig. 8 {V. io), and figs. 9, 10 (F. urticce). 

 All other "tortoiseshell" forms of the Vanessidae have similar 

 under side markings. In the Pyrameid groups the same parts 

 are responsible for the beautiful difierently shaped metallic-blue 

 markings in the atalanta-form species — itea-gonerilla (with con- 

 spicuous blue rings), indica-tammcamea, dejeani-atalanta. In 

 these species the blue colour demarcates the median part belong- 

 ing to tiie second costal blotch. In the Araschnids, and in the 

 cardu'i-form Pyrameids, these demarcations show a yellowish or 

 greyish colour, and are therefore comparatively inconspicuous. 

 They are all, however, easy to compare when drawn in " black 

 and white." Fig. 12 shows these markings in the under side of 

 P. carye, fig. 13 is drawn from P. myrinna, fig. 16 from P. cardni, 

 fig. 14 from P. ataUmta, and fig. 15 from P. indica. The beauti- 

 ful blue ring in itea-gonerilla arises in the same way as the 

 nearly ring-shaped (yellow) demarcation in the second costal 

 blotch of A. bureyana, fig. 6 (under side). The — especially 

 characteristic — markings in the basal and apical parts in the 

 fore wing under sides of Hypanartia* (closely allied to Pyrameis) 

 delius (from Sierra Leone), //. schoneia (Natal), correspond 

 especially well with the markings in the f-ame wing parts in 

 A. levana. The figures 4, 5, 7 represent under side markings of 

 A. levana and A. bureyana var. thibetana ? (fig. 7),t which latter 

 is the "under side io'' of the Araschnids (compare fig. 1) ; 

 fig. 4 is taken from a female levana from Berlin, fig. 5 from 

 a male levana from the Sayan Mountains (Irkutsk), which latter 

 local form shows an increase in size. While exhibiting further 

 variation of the median part in the second costal blotch, these 

 specimens also show how the shape of this costal marking is 

 altered as a whole — (1) when the apex becomes ocelliform 

 (figs. 1, 7), or (2) when the wing is fasciated (fig. 6). Figs. 4 

 and 5 exhibit intermediate markings. 



'•' Tlie only wliollij African species of Pyrameis — P. ahyssiriica — is in- 

 teresting lor resembling the African species of Hypanartia in its upper side. 



f The facia] differences between hureyana and var. thibetana aie so great 

 that one suspects them to be different species. 



