SPHINX LIGUSTRI. 309 



and without definite arrangement.* Upon the thoracic segments 

 the hairs are arranged so as to continue the four lines described 

 above, but they are less numerous, at any rate in the case of the two 

 upper rows, where there is only one hair to each row on each segment, f 



The head is thinly covered with similar hairs The 



mandibles are green, with the cutting edges black. The ocelli are 

 black and distinct. The spiracles are hard to make out, and the 

 dorsal vessel is visible but not distinct. The head is rounded and 

 of a generalised larval shape and not Sphinx-like. The horn is 

 flexible and movable ; it soon becomes straight, and may be held at 

 an angle or parallel with the back. . . . The dorsal vessel becomes very 

 distinct after the larva has fed for a short time, and the tracheal 

 system is visible. The larva becomes greener, and the first trace 

 of marking appears with the formation of a white circular patch 

 round the base of each long hair in the four rows ; the patches are 

 especially distinct on the dorsal rows, and the larva, therefore, shows 

 four spots on the back of each abdominal segment in front of the 

 8th, and two spots on the back of each thoracic segment. (Weismann 

 describes and figures similar spots on the thoracic segments of Smerinthus 

 populi in first stage.) These white patches are the first readily seen 

 shagreen-dots| ; then minute white spots appear all over the body, which 

 are the white areas round the bases of the smaller hairs § ; then by 

 a linear arrangement of the white spots (in which both large and 

 small take part) the essential markings are established (see loc. cit., 

 pi. vii., fig. 2). The subdorsal becomes distinct before the oblique 

 stripes, and the latter are rather faint during the whole of this stage. 

 The white subdorsal is perfectly normal, extending the whole length 

 of the body and bending gently upwards anteriorly in the 1st thoracic 

 segment, and posteriorly into the base of the horn (being continuous 

 with the 7th oblique stripe). The subdorsal may be seen to be 

 formed of white dots, which are also present over the whole surface || 

 but are not conspicuous ; so also with regard to the oblique stripes, 

 but here there is an interesting relation to the larger spots. The 

 anterior dorsal pair on each segment always forms part of the stripes, 

 but the latter, converging posteriorly, fall within the hinder pair ; so 

 also the anterior spots \ only of the lateral rows on each segment 

 fall into the oblique lines. There is a very minute and faint 8th 

 stripe in front of the 1st oblique stripe on the 1st abdominal segment; 

 it does not extend downwards far enough to meet the subdorsal. 

 The effect of the series of large dorsal dots on the thoracic segments 



* This is not so ; the setae that are present are easily homologised. 



t This might suggest that what Bacot considers (supra p. 308) a possible 

 abnormality in the specimen he described is really a normal condition, i.e., the seta 

 of ii on the meso- and metathorax is largely atrophied, but Chapman (who has 

 specially examined mounted specimens) states that this is not so (see 1st footnote on 

 p. 308). 



% These pale patches are at the bases of the primary setae and so can in no 

 sense be described as shagreen-dots, which are connected with secondary hairs. 



§ There are no smaller hairs in this instar. 



|| These white dots, if, indeed, any are present in this instar, are at most 

 only showing through skin from beneath. 



H Does Poulton here mean the prespiracular by " the anterior spots of the 

 lateral rows ? " As a matter of fact the lines go through iii as well as v (see antea p. 

 307, line 41), so that either might from this standpoint be his "anterior " spots. 



