414 BRITISH LEPIDOPTERA. 



i3 , 5mm., size of head 6mm. in height, 5*5mm. in width, length ot 

 head and thorax 25mm. The length of the caudal horn differs 

 scarcely at all in the two larvae, and is 7mm. It will be seen that the 

 smaller (and presumably younger) larva has the larger head ; the larvae 

 also differ in that the head of the larger one is markedly notched at 

 the crown, whilst that of the other is not notched at all. The head is 

 a rounded trapezoid in outline, and appears much taller and narrower 

 than it really is, owing to the broad black stripe down either side. 

 Its colour is bright yellow, the surface corrugated or convoluted rather 

 coarsely, although the surface of these convolutions is quite smooth 

 and polished-looking ; in addition to the broad black bands down the 

 sides of the head, a narrower black streak runs across from the bases 

 of antennae to the mouth-parts, which are, for the most part, black, 

 although the skin at the articulation shows white ; the bases of the 

 antennae are pale, the two upper joints black. The ocelli are clear and 

 glassy-looking, with a white area beneath (but this is probably not 

 the colour when the larva is alive). The body is of fairly even 

 thickness, tapering quite suddenly to the head. The prothoracic 

 scutellar plate is structurally apparent, but not differentiated as regards 

 colour ; the thoracic segments are very distinctly different from the 

 abdominal, both in colour and structure ; the subsegmental divisions 

 being larger on the former, and not so apparent owing to the complete 

 absence of the coloured areas surrounding the bases of the shagreen- 

 hairs, which are so strongly marked a feature on the dorsal area of the 

 abdominal segments. The spiracles are very distinct, black- or dark- 

 crimson-centred surrounded by white. The true legs are black, as are 

 also the strong fringes of hooks on the prolegs. The subsegments, as 

 shown by the purple dots at the base of the shagreen- hairs on the 

 dorsal area, are six on the 1st abdominal, and eight on the following 

 segments, 2 to 6, the 7th abdominal only shows five rows, although there 

 are quite eight subdivisions. There is doubtfully a seventh structural 

 subdivision in front of the marked ones on the 1st abdominal, 

 but this is not very clear in the larger larva, and does not show up at 

 all on the smaller one; the 1st abdominal segment is shorter than the 

 others (Bacot. June 27th, 1903). Chaumette describes (Zoologist, ix., 

 p. 3242) the green and the dark forms very fairly ; Newman also 

 describes the full-grown green form of larva (E?itom., ii., p. 281), and 

 gives (Joe. cit., p. 282) a short diagnosis of the dark form, which, he says, 

 was also described by Fuessly, three times figured by Hiibner and three 

 times described in the Zoologist (pp. 1658, 6282, 6788). 



Comparison of the larvae of Manduca atropos and Sphinx 

 ligustri. — Final instar : The larva bears the closest resemblance 

 to that of Sphinx ligustri. The shape of the head is exactly similar, 

 with the same arrangement of the black marginal line relatively to 

 the face, except that the black borders very nearly meet at the apex 

 of the head in S. ligustri while they do not come so near to complete 

 fusion in M. atropos. In the larvae of both species the black bands 

 attenuate considerably towards the apex of the head. The most striking 

 resemblance, however, is afforded by the appearance and position, 

 during the resting attitude, of the thoracic segments, which, in both 

 species, are swollen and transparent-looking, and without a trace of 

 marking. All the larval markings end abruptly at the anterior edge of 

 the first abdominal segment, except the border of the anterior stripe, 



