significance of icing-markings. 121 



which, while all the dark series keep their normal places, 

 an entirely new effect is produced by the still further 

 paling of the ground colour in certain situations, espe- 

 cially between II. and III., or that region of the wings 

 which we have denoted by C. In this way definite 

 light-coloured bands are produced, which may readily 

 suggest the source whence the groups represented by 

 L. Sibylla and A. iris respectively may have derived 

 their characteristic pure white markings, the corre- 

 spondence of which with those in A . sagana can scarcely 

 be accidental.* In the neighbourhood of the brown 

 Argynnids we find the slightly divergent form CiMntherata, 

 which in the breaking up of the patch of original ground 

 colour Iir. marks a point of specialisation not reached 

 by any true Argynnis, while its somewhat dull ground 

 colour, and the peculiar accentuation of the costal por- 

 tion of C and D, relating it to y1. ralesina and A. cliana 

 ? , would seem to indicate its early separation from the 

 Argynnid stem. This accentuation of C and D is carried 

 to a much greater extent in A. nijjJte 2 , one of those 

 transitional species in which we see the old and new 

 ground colour fighting for the mastery. In this con- 

 spicuous and highl}? ornamental species we seem to have 

 reached the point at which those butterflies called in the 

 wide sense Vanessas quit the parent stem.f 



* W. Miiller {op. cit., p. 622) considers that the group including 

 Limenitis (which he names " Rippenbauenden ") must find its 

 root either among the VanessincB or the DiademincB. Tliis would 

 remove Limeintis a step finther off the Argj-nnids than my own 

 suggestion would wai'rant, though it accords with my conclusion as 

 to their relative antiquity. The position of Aj^atura Miiller owns 

 to be a puzzle (Ibid.). He inclines to remove it from the gi'oup of 

 " Rippenbauenden," and somewhat doubtfully suggests its enrol- 

 ment among the Vanessina, in the neighbourhood of the genus 

 Hyioanartia {Eurema). In favour of this it may be mentioned that 

 (according to Scudder) the eggs oi Apatura bear more resemblance 

 to those of the VanessidcB and even of the Argynnldce than to 

 those of Limenitis, and also that Ai^citura corresponds with many 

 of the Vanessidce in having 2 rows of spines instead of 4 on the 

 under side of the last tarsal joint. (Scudder, ojj. cit., vol. i., 

 pp. 227, 8). 



■\ The above statement as to the relative antiquity of Argy^inls 

 and Vanessa is in accordance with Milller's conclusions derived 

 fi'om a study of the larvae. " A multiplication of the rows of 

 spines from 6 [as in Argynnis] to 9, and the addition of the Ds 

 and Ped rows has led to the forms that the VanessincB now show 

 us." [Op- cit., p. (321). 



