significance of wing-markings. 117 



with the three similarly lettered in V. polychloros (figs. 

 32, 33; II. 6, 7, 8). The same three occur throughout 

 the genus Grajjta ; in G. c-aureum, as has been seen, side 

 by side with spots undoubtedly representing series III. 

 (fig. 34). In Grapta we also get a representative of the 

 spot at the anal angle of the fore wing in V. polychloros, 

 which we have identified as III. 8. So far there is no 

 difficulty. Now turning to A. levana, we may, I think, 

 satisfy ourselves on a careful comparison with any 

 Grapta, that II., 6, 7, and 8 are present, as well as III., 

 6, 7, 8 ; moreover, that there exist in A. levana two 

 additional black areas in the angles between the sub- 

 median and median nervures, and between the 1st and 

 2nd median nervules, i.e., in the roots of the 7th and 8th 

 interspaces. These we may regard as in series with II'. 

 1 — 5, and may conveniently call 11'. 7, and II'. 8 (fig. 40). 

 The latter does not occur as a distinct spot in any 

 Grapta with which I am acquainted ; the former, how- 

 ever, is found in many species (e.g. G. interrogationis), 

 also constantly in V. r-allmm, and sometimes in V. 

 jjolychloros (as in a specimen in the Hope Collection, 

 fig. 33). Having thus met with a tolerably generalised 

 type in A. levana, and having used it to interpret this 

 part of the series in the Graptas and Tortoiseshells, we 

 shall find that we are able to trace from the same type 

 two diverging lines, the first leading towards the genus 

 Pyrameis, the other backwards, as we may suppose, 

 towards the Argynnids. The best species to compare 

 with A. levana for this purpose is P. carye (fig. 41), in 

 which II. 6 and 7 will be found, forming a continuous 

 band across the wing, connected in front with the 

 anterior portion of series II., and behind with III. 8 ; 

 while ir. 7 is seen to be similarly connected with II. 8 

 vertically ; and also with II. 7 horizontally by a lateral 

 streak following the course of the 1st median nervule. 

 II. 8 is represented by a distinct but narrow patch 

 connected with 11'. 7, while a dark and not distinctly 

 defined area at the root of the 8th interspace may 

 stand for 11'. 8. The passage from P. carye to P. huntera 

 and P. myrinna, and through P. kershaivi, P. cardui and 

 P. callirrhoe, to P. atalanta and P. gonerilla is so easy that 

 it is not necessary here to trace it in detail ; the main 

 features being the merging of II. 6 in the general dark 

 ground colour, the paring down of II. 7 until it becomes 



