270 BRITISH LEPIDOPTERA. 



ised oblique stripes, in the Amorphids, and asserts that the tendency 

 is for the larvae to lose this subdorsal line as the development of 

 the oblique stripes proceeds, and he suggests and fairly proves 

 that the replacement of the subdorsal by the oblique lines is a 

 comparatively recent feature in the evolution of the Amorphid (and 

 Sphingid) markings. Poulton sees (Trans. Ent. Soc. Lo?id., 1884, 

 p 32), in the remains of the subdorsal line, a more efficient means 

 of protection when a larva is at rest in the Sphinx attitude, for 

 then, the subdorsal line, following the curved anterior .segments, 

 becomes approximately parallel to the oblique stripes (loc. cit., pi. i., fig. 

 4). So many Sphingid (sens, restr.) larvae still have traces of the sub- 

 dorsal line that he considers that its entire absence in others must be 

 looked upon as a very recent line of development, e.g., Agnus 

 convoivuli, Manduca atropos, &c, retain this line throughout their 

 ontogeny, and Sphinx ligustri has it well-developed in its early stages. 

 The shagreen-spots, as we have already pointed out, are, in reality, 

 tiny patches of pale colour developed at the bases of the secondary 

 hairs, and are characteristic of many Sphingid larvae — Sesiids, 

 Phryxids, &c. — and not at all restricted to those of the Sphinginae. 

 Usually, these appear in the earlier stadia in their simplest forms 

 and become modified into special markings in the later stadia, 

 but, in the Sphinginae, as exemplified in Sphinx ligustri and Agrius 

 convoivuli, the shagreening does not appear until the second instar, 

 although the pale specks on the new skin may be seen through 

 the old skin just before the 1st moult. It is quite possible that the 

 smooth 1st instar of the Sphinginae is evidence that the larvae of 

 this subfamily are less specialised than those of the Amorphids, where 

 the shagreening has been pushed back into stage 1, and really 

 develops on the larva before it leaves the egg. The spiculation 

 of the caudal horn on the Sphinginae in the 1st instar may, of course, be 

 the remains of a previous general spiculation in that stage (in which 

 case the above view ot the Amorphids being more specialised would 

 be erroneous), or it may be a highly specialised part of the larva 

 which has received spiculation from the following stage, but which 

 the rest of the surface has resisted, but which has been submitted 

 to in the more specialised Amorphids. The tiny little mammillae 

 bearing the setae that arise from these pale specks are generally known 

 as shagreen-tubercles and must not be confused (as Poulton has evi- 

 dently done, anted, p. 269, footnote) with the primary tubercles, i, ii, iii, 

 iv, v, vi, &c, which are chitinous-based in their most simple forms. The 

 importance of these shagreen-spots in the development of the larval 

 markings has already been referred to, and is readily followed by a study 

 of the ontogeny of the Phryxid, Sesiid, Amorphid and Sphingid larvae. 

 According to Poulton {loc. cit., 1885, p. 302) the coloured borders to 

 the oblique and other lines in Sphingid larvae seem always to be 

 formed of modified ground-colour. He states that "dots are either 

 absent from the borders, or, when present, are very small," and 

 surmises that "the first trace of a border arose in the diminishing 

 size of the dots, which alone would make a relatively dark 

 stripe .... the effect afterwards being increased by a darkening 

 of the ground-colour, and, in some instances (e.g., Sphinx ligustri), 

 by a change of colour altogether." It may be well to notice here 

 the great similarity between the markings and general appearance 



