ON THE VARIATION OF INSECTS. 20l 



If truly a sexual character, this is doubtless a primary one, and 

 Mr. E. B. Poulton has suggested that it represents the blind 

 termination of the ducts of the sexual glands. 



(4.) Geographical variation. — The typical larva of Ceratomia 

 quadricornis, according to \V. V. Andrews (Can. Ent., 1876, p. 40), 

 is green, while a more exceptional form is deep brown. But at 

 Rochester, in New York State, the brown variety is the prevalent 

 one, and the green, so-called, type the exception (Can. Ent., 1876, 

 p. 120). In Britain, the typical larva of Vanessa atalanta is grey- 

 green, while black and also dingy white forms occur. But at 

 Coalburgh, in America, Mr. W. H. Edwards found that nearly all 

 the larvffi of atalanta were black in tlie last stage, with the lateral 

 stripe usually macular, and greenish yellow. A small percentage 

 was yellowish green, and a few were mottled black and yellow, 

 but the dingy white form was not to be found (Can. Ent., 1882). 

 Other such cases have been recorded, but it is to be noted that 

 generally the variation is rather in the proportions in which 

 certain forms occur, than in their actual occurrence. In cases 

 where the larva presented any noteworthy difference in different 

 regions, the perfect insects have also generally been distinguish- 

 able, and the two forms have been regarded as distinct species. 

 Such, for instance, are Papilio rutulus and turnus, and Pieris napi 

 and oleracea. Possibly, when larvae are further studied, specific 

 characters based on larval differences will in many cases prove 

 fallacious, owing to the fact that the larvse are far more variable 

 and intergrading than has been supposed. 



(5.) " Spontaneous " variation. — I apply this term to those 

 variations which appear to occur without reference to external 

 conditions, and which are possibly to be traced to some ancestral 

 tendencies, the cause of which is either now obscured or lost sight 

 of. Yet in contemplating the frequent colour-changes of larvae, 

 and especially the light and dark or green and brown forms of 

 certain species, it is impossible not to think that after all these 

 peculiar variations of colour may simpl}^ be due to the instability 

 of the colouring-matter, and its readiness to assume new forms 

 under influences so slight that we have not yet appreciated them. 

 And, indeed, as stated above, a green larva of Biston turned red- 

 brown on the application of heat, this being the normal colour of 

 another variety of the same species. To enumerate all the recorded 

 " spontaneous " variations of larvae would be a gigantic task, but 

 the following may serve as examples : — Painlio ajax, larva varies 

 in ground colour from dark smuky brown to grey, blue-green and 

 pale green, each form with peculiar markings. Vanessa gonerilla, 

 larva varies considerably (see Entom. xvi. 218). Deilephila 

 spinifascia, four vars. of larva are described (Ent. Mo. Mag., 

 1881, p. 132). Philampelus achemon, larva varies from green to 

 pale straw or reddish brown. Sphinx quinquemaculata, larva 



