302 Mr. Poulton's /Wrf//r/- ttota^ }()>()ii tJie 



the origin of certain marks bj^ the gradual crowding and 

 coalescence of the dots in the former species, and the 

 special coloration and fusion of two dots upon a variable 

 number of segments in the case of D. hijypophaes (see 

 pp. 204 and 221 of the English translation of the Essay 

 on the markings of Caterpillars). 



8. The origin of the coloured borders to the 

 OBLIQUE AND OTHER LINES IN Sphingid.e. — Tliese borders 

 seem to be always formed of modified ground colour. 

 Dots are either absent from the borders, or, when 

 present, very small. It seems probable that the first 

 trace of a border arose in the diminishing in size of the 

 dots. This alone makes a relatively dark stripe, for 

 the colour of the larva elsewhere is due to the ground 

 colour modified by the closely-set light-coloured dots. 

 After this the effect must have been increased by a 

 special darkening of the ground colour, and in some 

 instances (e.g.. Sphinx ligustri) by a change of colour 

 altogether, The dark superior border to the subdorsal 

 of M. stellataruin has a similar history. This theory of 

 the origin of the dark borders is borne out by the 

 ontogeny of those species which I have been able to 

 observe. 



9. Some characters of the horn in the primitive 

 Sphinx larva. — It has been seen that the horn of 

 S. ligustri is distinctly bifid in the first stage (see figs. 

 1 and 2, Plate VII.), and less markedly so later (see 

 fig. 3). The same was true of S. occllatns, and this 

 character persists through a considerable part of the 

 ontogeny. (This summer, 1885, I have found that the 

 horn of the young larva of <S. populi is also forked). 

 The same structure is described by Weismann in the 

 first and second stages of Anccryx pinastri. The fork 

 is so marked in this species that I have no doubt that it 

 really exists in more advanced stages, but requires a 

 lens for its detection. Weismann also figures, in his 

 Essay on the markings of Caterpillars (plate v., fig. 38;, 

 the margined larva of Deilepliila eup)]iorhi(e shortly after 

 emergence from the egg, with the horn terminated by 

 two diverging bristles. Mr. Meldola also, in an editorial 

 appendix to his translation of Weismann's essay, quotes 

 from Mr. Pkoland Trimen the fact that the caudal horn of 

 the young larva of Lophu>itctJius Dumolim is forked at the 



