362 BRITISH LEPIDOPTERA. 



the variable development of the red spots, and (3) in the relations of the 

 latter to the borders of the stripes. The association of a bluish- 

 green ground colour with the absence of pigment and a general 

 Smerinthine appearance, is very striking, when compared with the 

 dark markings and greater peculiarity of the yellowish larvae. 

 Nevertheless, there is a very gradual transition from the one type 

 to the other, and this can be clearly seen in the series of four larvae 

 represented in figs. 10-13, although there were other larvae inter- 

 mediate between some of these. The stages of this transition, 

 he considers, probably represent the steps by which the dark variety * 

 has arisen from the green one. These stages are : 



(1). A green Smermthus-\ike larva with a red spot on the border to each stripe. 



(2). The increase in the dark part of the borders both in size and in depth 

 of tint. The red spots elongate along the border, and at the same time distinct 

 dark reddish patches appear above and below the elongating spots. Dark patches 

 and bands appear and increase elsewhere. 



(3). The borders become still darker, and their upper part tends to spread 

 anteriorly along the upper edge of the subdorsal, thus shutting off a band of dorsal 

 ground colour. 



(4). The ground colour thus shut off first becomes much darker than the 

 rest by a deepening in the tint of the ground-colour itself, finally nearly black, from 

 the gradual intrusion of the black borders of the band. 



The further continuation of this process of darkening will be 

 seen in the following stage. It must be remembered that the purplish- 

 brown colour of the markings connected with the borders, together 

 with the elongation of the red spots and the red colour of the 

 additional dark marks on the borders, seems to clearly support 

 Weismann's conclusions that the red spots are connected with the 

 coloured borders to the stripes. These facts also serve to connect 

 together Sphinx and Smerinthns in a most interesting manner 

 (Poulton, Trans. Ent. Soc. London, 1888, pp. 535-536). Poulton further 

 details the variation in the 4th and 5th stadia and submits a most 

 interesting series of figures {Joe, cit., pi. xv-xvi) which give one a 

 very clear idea of the evolution of the dark forms ; the five figures 

 (pi. xvi., figs. 1-5) of larvae at the end of the 4th stadium are 

 particularly interesting, whilst those in the 5th stadium (figs. 6-10) 

 show a certain amount of uniformity in adult coloration, that explains 

 why green adults are comparatively rarely, in some districts, seen in 

 nature. The chief variations are in the shade of ground colour which 

 passes from brown to black, in the prominence of the subdorsal, sub- 

 spiracular and oblique stripes and their borders, the subdorsal being 

 often represented by a single spot on each segment posterior to the first 

 abdominal. In the lightest varieties there is a dorsal stripe above 

 the subdorsal, also tending to form spots (above the subdorsal 

 spots indicated in figs. 6-7). The alternation of light and dark 

 varieties of the larvae of A. convolvuli in the successive stages was 

 very remarkable and very marked in many larvae in all the stages 

 except the last in which all the larva? became dark varieties. 

 Poulton also gives (loc. cit., p. 552) a diagram showing 

 in a graphic manner the way in which the larvae oscillated 

 backwards and forwards, between the dark and light varieties during 

 the successive stages. He says : " It must be understood that the dark 



* Mathew notes that none of the larvae that he found in the Mediterranean 

 district, in 1897, assumed their brown coats until after the last change of skin [Ent., 

 xxxi.,p. 115). 



