markings and attitudes of lepidopterous larvcE. 325 



Fig. 4, X 2 diam. — A larva of the same species at the close of 

 the third stage, comfortably extended at rest. The markings are 

 very similar to the last. The subdorsal is now indistinct anteriorly 

 and absent elsewhere. The large white spots are still present 

 upon the anterior segments, and contribute to their markings. 

 There is still a trace of the bifid termination of the horn in some 

 individuals, but this feature requires higher magnification for its 

 detection than has been employed in this figure. The side of the 

 head is now often darkened by a black cloixd, which spreads from 

 above downwards, behind the (yellow) marginal line. There is 

 great variability in this character, and it is often absent, as in the 

 individual figiu'ed. The spiracles are faintly ochreous, the thoracic 

 legs red, and the claspers dark purplish. The ground colour is 

 often dark green below, which extends upwards as a border to the 

 lower third of the while stripes in the darker individuals. During 

 growth in this stage the shagreen dots of the oblique stripes 

 become enlarged, fuse, and form continuous lines. The annulation 

 of the segments keeps the dots separate for some time (as each dot 

 in a stripe is on a separate annulus, and there is a deep furrow 

 between adjacent anuuli). In the 8th abdominal segment there is 

 no annulation, and here therefore the dots coalesce much earlier 

 than elsewhere, and the stripe is more distinct than the others. 

 The anal flap is boi'dered with white. The purple borders to the 

 oblique stripes appear in this stage. They are never present at 

 the beginning, and there is extreme irregularity in the time at 

 which they appear and the extent to which they are developed. 

 Sometimes they are not present at all in this stage. The borders 

 are linear, and they are first seen in front of the middle of the 1st 

 and last stripe, afterwards upon the intermediate ones. The 

 shagreen dots are absent from the purple borders, or, if present, are 

 very small, the borders being modified from the ground colour. 

 At this stage the SjMnx attitude is often assumed during rest ; 

 in previous stages it was also seen, but far less commonly. In the 

 present stage, and in the second, there is the greatest resemblance 

 between this larva and that of Smcrintkus ocellatus. The purple bor- 

 ders of the individual figured are rather more developed than usual. 

 Fig. 5. — Slightly over the natural size. Larva of 8. ligustri 

 at the close of the fourth stage, at rest in the Sphitix attitude, 

 which is more marked in this stage than in any other. The larva 

 at first exactly resembles the appearance in the last stage, but the 

 pui-ple borders are broader. The remnant of the subdorsal and 

 line above it, with the large dots, are at first present, but disappear 

 later. The "8th stripe" remaitis. The stripes are white where 

 they are bordered with purple, yellowish above this, while the 

 purple is replaced superiorly by dark green. The ground colour 



