374 Mr. G. A. K. Marshall on 



wings by which the " tails," the apparent anteiin;p, are 

 made continually to pass and repass each otlier, add very 

 greatly to this resemblance. The head-like appearance, 

 tirst observed by Dr. Arnold in Thcekt iarhas and con- 

 firmed in other species by Dr. Forsstrona, is quoted by 

 Kirby and Spence (People's Edition, ] 8G7, p. 423) : it was 

 independently observed by Mr. R. C. L. Perkins (" Colonrs 

 of Animals," London, 1890, p. 208) in Thcda ir-alhiirii,tiad 

 this keen naturalist obtained contirinatory evidence in the 

 case of the English Thccht, similar to that shown upon 

 Plate XI. My friend Dr. Richard Evans of the Museum 

 at Georgetown, British Guiana, independently observed 

 the same thing in Siam, when taking part in the Skeat 

 Expedition. My friend Professor Wyndham Pi. Dunstan, 

 F.R.S., sending me a pair of Deuduri.c antalas bred from 

 larvffi which are destructive to the pods of " Iwja dulcis " 

 at Manashi, near Cairo, wrote (July 4, 1900) that his 

 friend Mr. E. A. Floyer who sent the insects " remarks 

 that the butterfly has markings on its tail which resemble 

 the head. He consiilers this protective, as a bird is 

 uncertain which is the head and which is the tail, and the 

 insect often escapes by going otf in the unexpected 

 direction." My friend Mr. Champion B. Russell, who 

 presented to the Hope Department the beautifully 

 mutilated specimen of Sjnmlasis naiahnsis represented on 

 Plate XI, fig. 3, also indcipendently recognized the same 

 resemblance (1900), and thought that the lobes witii their 

 two tails passing and repassing each other looked like 

 jaws opening and shutting. I asked Mr. Marshall's opinion 

 on this subject and received the following reply : — 



" Salishuri/, Jitnc 11, 1901. — Mr. Russell's observations 

 on the tails of Lyavimhv are, as you say, of considerable 

 value as coming from an entirely independent source, but 

 I must confess that I am not inclined to believe that the 

 anal appendages in the wings of butterllies have been 

 modified in imitation of particular organs, for I fail to see 

 how this C(juld be effected by ordinary selection. And I 

 think a valid argument against such an idea is the great 

 diversity of foi-m shown by these appendages, not only 

 among the Lijvivaidx but other families as well. It seems 

 safer to regard these curious lobes and tails as having been 

 developed by natural selection for the purpose of attract- 

 ing attention to that part, and that the particular form 

 they take is due to congenital variations which we cannot 



