170 BRITISH LEMDOPTERA. 



a useful character in their determination. The mouth-parts are 

 almost always reddish. 



The markings of the larva are important, because on these the 

 differentiation of allied larvae often has to be made. Of these the 

 following are the chief : (1) A square-shaped spot usually present on the 

 underside of the prothorax. (2) A pair of superficial skin marks on 

 the dorsurn of the prothoracic segment (these appear to be the 

 remnants of a prothoracic plate). (3) A pair of deep-seated, pear- 

 shaped, elongate, dark prothoracic marks in line, or nearly so, with the 

 posterior lobes of the head, lying just beyond the tips of the latter. 

 These are the cephalic ganglia, they are of firm texture, and can 

 readily be dissected out. Their conspicuousness depends, apparently, 

 less on their black or brown colour, than upon the contrast that the 

 latter bears to the colour of the head. When the ganglia and the head 

 are of the same colour, the former appear to be (when the head is 

 retracted) simply elongations of the head lobes. (4) A series of 

 ventral markings consisting of a chain of narrow, spindle-shaped or 

 linear marks, down the middle line of the abdomen. This is the 

 deep-seated ventral nerve cord. (5) A series of large, conspicuous, 

 square-shaped, surface spots, covering a large portion of the ventral 

 area, found, however, in but few species, of which the " anyulifasciella 

 group " offers the best example. These spots usually disappear (except 

 on the venter of the prothorax) with the last larval moult, and thus 

 allow the nerve-cord to come into view. 



The ventral nerve-chain is known by the linear character of the 

 ganglia, the three thoracic ganglia being wider or more oval than the 

 eight abdominal ones. The eleven ganglia can generally be counted 

 (by the aid of a good lens) and the bands connecting them in some 

 species, double in the abdominal, as well as in the thoracic, region ; 

 in the green-coloured larvae, however, the ganglia and bands are 

 practically invisible. Wood has observed that when the cephalic gan- 

 glia are noticeable on the dorsum of the prothorax, no trace of the 

 ventral ganglia is to be found, in some larvfe, on the venter. This 

 want of agreement in the colour of the supra- and infra- aesophageal 

 parts of the nervous system was found to be connected with the 

 position of the larva in the mine. Probably one-third of the Nepti- 

 culid larvae mine, venter uppermost, and when this is the case, the 

 ventral cord is coloured and visible, whilst, when the dorsum is upper- 

 most, the cephalic ganglia have the colour intensified. Wood con- 

 siders that light, being the most general and potent factor in the 

 production of pigmentation, has, by pouring through the transparent 

 tissues upon the nerve ganglia, in the course of generations, exagge- 

 rated and intensified their colour, so that when the head is uppermost, 

 the light falls upon the cephalic ganglia and, in the course of ages, 

 has blackened them, whilst the ventral cord, protected by the con- 

 tents of the intestinal canal, retains its primitive colouring. On the 

 other hand, when the position of the larva is reversed, the cephalic 

 ganglia remain unchanged, and the ventral cord is darkened. The 

 rule, however, is not invariable, many larvae showing both the cephalic 

 ganglia and ventral cord, but even then the darker is always the one 

 more exposed to the light. The high colouring of the nerve centres 

 in so many of the yellow larvsB, and its remarkable absence in the 

 bright green ones, is explained by Wood as being due to the green 



