MORPHOLOGY OF THE LEPIDOPTEROUS PUPA. 257 



H'.M'.). A similar relationship is witnessed in Vanessa lo (figs. 20 and 21), although 

 the difference is rather in the angularity and amount of projection of the cusps than 

 in the depth of the bay. In dark varieties of this pupa the pigment is distributed 

 along the line H'.M'. (fig. 20) in such a manner as to suggest a former black-and-white 

 fringe, which is now absent from the imaginal hind margin. The dark parts are 

 diagrammatically indicated in the figure by thickenings along the line H'.M'., and their 

 position will be found to correspond with the dark parts of the imaginal fringe in 

 C. cardui (fig. 17) and V. Atalanta (fig. 19). Inasmuch as these latter species are more 

 ancestral than V. Io as regards the comparatively slight degree of indentation of the hind 

 margin, and especially as regards their markings *, it is most probable that V. Io at one 

 time also possessed a black-and-white fringe, which is still indicated in the corresponding 

 part of the dark pupse of this species. A comparison between the pupa and imago of 

 V. polychloris (figs. 22 and 23) similarly shows that the pupal line H'.M'. is rather less 

 indented than that of the imago. 



Finding these results in our common Vanesshlce, I Avas extremely anxious to test them 

 by an examination of the allied Grapta C-album, which possesses a far more jagged hind 

 margin than any other British butterfly. After searching in various directions, I 

 obtained an empty pupal shell of this species, through the kindness of my friend 

 Mrs. Luard. An examination of this pupa supported in the most complete manner the 

 conclusions abeady arrived at. It is clear that the hind margin of the imago (fig. 25) is 

 far more jagged than that of the corresponding line upon the pupa (H'.M'., fig. 24), which 

 indeed is not much in advance of the condition found in the imago of V. Io or V. poly- 

 chlorus (figs. 21 and 23). 



Hence we see that not only are the traces of lost imaginal wings preserved, but the 

 indications of ancestral forms and markings are also fixed on the surface of the pupa. 

 Furthermore, the results of this investigation suggest that further work along the same 

 lines will reveal many other interesting conclusions as to recent changes which have 

 taken place in the imago. 



In conclusion, I desire to thank those friends who have kindly supplied me with 

 material for this paper. Mr. W. White especially has lent me many specimens which 

 have been of the greatest service. Professor Meldola has carefully been through most 

 of the manuscript with me, and has offered many valuable suggestions. 



* See Dr. F. A. Dixey's admirable paper " On the Phylogenetic Significance of the Wing-markings in certain 

 Genera of the Nymphalidai;' Trans. Ent. Soc. Lond. 1890, pp. 89-129. 



