THE EXTEENAL MORPHOLOGY OF THE LEPIDOPTEROUS PUPA. 49 



to the iniaginal wing (and also to the present outer margin of the 

 pupal wing) we have preferred to quote Poulton at length as an earnest 

 of our thanks for the paper which first stimulated our observations in 

 this direction. There is, however, one point we should like to suggest 

 in addition. It appears to us that "Poulton's line " is the boundary at 

 which the hind-margin of the imaginal wing originates. This line 

 varies in individual pupae as does the hind-margin in individual 

 imagines ; it may be, therefore, looked upon as having varied simul- 

 taneously and uniformly with the shape of the imaginal wing, and its 

 present form may be looked upon as bearing a relative difference only, 

 when compared with that of the imago, to that which it has borne 

 during the evolution of the imaginal wing in the direction of its 

 present form, and, as Poulton says, its present shape possibly 

 represents one of the phases through which the wing has passed. On 

 the other hand, we are inclined to look upon the hind-margin of the 

 pupal wing as it now exists as representing much more nearly the 

 ancestral wing from which the angulated Vanessid wing has been 

 developed and this differs little (in all the species) from that which now 

 characterises the Nymphalids as a whole that are not particularly 

 specialised in this direction. Chapman looks upon "Poulton's line" 

 as the margin of the membranous portion of the imaginal wing, and 

 considers that the outer portion represents the fringe. This view is 

 " essentially the reverse of that formulated by Poulton, since the 

 pupal margin is not the record of a past imaginal wing, both margins 

 having been always both pupal and imaginal as they are now, but, in 

 certain species, the entire margin (margin of fringe) has submitted to 

 variation in the pupal state to suit the exigencies of the pupal form." 

 Experiments on a number of pupa? at the time of the final develop- 

 ment of the wings of the imagines in Pararge egeria, P. megaera, Dry as 

 paphia, Argynnis aglaia, Brenthis euphrosyne, all the British Vanessids 

 (except Euvanessa antiopa) &c, have tended to show that the development 

 of the membranous portion of the imaginal wing begins at Poulton's 

 line, that it has not proceeded very far before it has slipped away from 

 it, crowding the hind margin of the wing down to the extreme of the 

 available space, also that, in such species as Polygonia c-album, which 

 Poulton selected as illustrating the declining size of the imaginal wing, 

 the latter, whilst developing, is really too large for its pupal space and 

 not too small. Bearing on this is an observation on the development 

 of the imaginal antennae of Gonepteryx rhamni and Pararge megaera, 

 which, occupying the whole of the pupal antennae in the very early 

 stages of their development, are, by the time they become visible from 

 the outside, much shorter than the pupal structures and become still 

 shorter before exclusion. Chapman notes, contrary to the facts 

 observed in Argynnids, Vanessids, &c, that the imaginal wing of Parasa 

 chloris, after its development along "Poulton's line," slips towards 

 the base of the wing and not outwards as in those we have ourselves 

 examined. The further relation of the pupal to the imaginal wings 

 was worked out at considerable length by Poulton. He observes 

 (Trans. Linn. Soc. London, 2nd ser., v., p. 248) that the wings of 

 b'umca nitidella ? are reduced in the imaginal stage, to minute tubercles, 

 so small as readily to escape detection, and having neither the shape 

 nor the appearance of wings. The female pupa, however, " possesses 

 small but distinct wings of characteristic structure and shape, and 



