420 BRITISH LEPIDOPTERA. 



shades, with no trace of red in it, and the horn ivory-coloured in- 

 stead of brown. This must have been similar to that referred to 

 by Poulton (Tra?is. Ent. Soc. London, 1886, p. 144) which had the 

 ground colour of the thoracic segments pink, whilst that of the ab- 

 dominal segments was brown. Inchbald says that, of 12 larvae taken at 

 Hornsea, in 1889, 7 were of a brownish-olive colour, with the anterior 

 segments white, the white being well-defined and conspicuous. 

 Balding records a form of larva with the thoracic segments pure white, 

 with a broad dorsal marking of deep black coming ■ down on 

 either side like a double saddle ; the remaining segments without the 

 oblique markings on the sides, but with a broad chain of diamond- 

 shaped lines and purple tracery down the back. Anderson notes that 

 several larvae, that he had in 1895, were of the dusky variety, which, in- 

 stead of being of a beautiful apple-green tint with violet and yellow 

 streaks, is of a dingy brown hue with a latticed pattern of dull blackish- 

 purple on the sides and back, and a sprinkling of dirty white spots, 

 the anterior segments being white with two stripes broken up into 

 spots a little darker than the general colour of the body; both type 

 and variety, he says, have a similar whitish warty caudal appendage. 

 Sharp notes that, in 1896, when the larvae were abundant at 

 Cambridge, three well-marked varieties (not including the usual dark 

 form) came under his observation : 



1. The usual greenish -yellow form with moderately distinct violet-purple 

 oblique stripes. 



2. A pure green variety, with the stripes very faint, reminding' one at first 

 sight of the larva of Sphinx ligustri. 



3. Yellow, with the oblique stripes very extensive, definite and deep- 

 coloured.* 



It appears not to have been generally recognised tbat the 

 dark forms of the larva of M. atropos, not only differ entirely 

 from the yellow and green forms in colour, but that the markings 

 are also quite different. The ground colour of the thoracic 

 segments is usually paler than that of the abdominal segments 

 in the darker forms, and thus far they agree with the green 

 and yellow varieties. Poulton, however, deals with this matter at 

 length (Trans. Ent. Soc. London, 18S6, pp. 149-15 1), basing his 

 remarks on two dark examples, of which he says : " The ground 

 colour of the thoracic segments is light brownish or of a dirty white, 

 and resembles the other (normal yellow) variety in being much lighter 

 than that of the rest of the larva, but it is entirely different in 

 the possession of distinct and very dark markings. There is a 

 broad median dorsal band of a very dark brown colour, which 



* Sharp notes (Ent., xxix., p. 284) : "This individual exhibited a remarkable 



peculiarity, inasmuch as, on the 2nd abdominal proleg of the left side, there was no 

 trace whatever of any of the clasping hooks ; the leg seemed, in other respects, quite 

 normal, and all the other legs had their full complement of hooks, these structures 

 being in this species numerous and large. This specimen had received some injury 

 near the anus, and was attacked with diarrhoea and died, so that I had no oppor- 

 tunity of ascertaining whether there might be a repair of the deficiency at the last 

 moult, which possibly (but not certainly) had not taken place in this individual. 

 Whilst examining this larva after death (or after apparent death), when flaccid, 

 discoloured, and shrunk to about hall its normal size, and lying on its side on blotting- 

 paper, an apparent corpse, and without reaction to pressure by the fingers, a house- 

 fly alighted on it and walked over it, whereon the skin was thrown into violent 

 contraction, a striking instance of the independence of the life of parts of the insect 

 organisation." 



