322 BRITISH LEPIDOPTERA. 



white secondary hairs numerous, hairs on tubercles not halt width of 

 segment in length. Colour black with orange lateral line. One or two 

 have this only, most have further markings. The orange band has the 

 subspiracular tubercle in its centre, the spiracle being immediately 

 above it. The orange line extends to the thorax in all specimens. 

 The variation in the others would require a description for each 

 specimen, but the general line of progress seems to be an orange 

 circle round each tubercle first occurring on iii and afterwards on i, 

 first as a lunule round lower margin, then as a complete circle. 

 Then the circle of iii unites with the lateral line, afterwards with that 

 round i. At the same time the lateral line becomes broader, and orange 

 marks appear at bases of prolegs. These changes are most advanced 

 on abdominal segments 3 — 6, although the junction of the orange circles 

 of i is very rare on the abdomen but usual on the thorax. The next 

 step is the appearance of a spot which is greenish-yellow rather than 

 orange at the posterior margin of segment, level with the spiracle, 

 and another above it at the level of (the absent) ii. On one specimen 

 these two spots have united into a vertical stripe at posterior margin 

 of segment, of a quite green tint, and just touching the orange stripe 

 enclosing i and iii. In this specimen the orange stripe is still 

 discontinued across dorsum on abdominal segments 1 — 6. The 

 spiracle is a black spot in this stripe. The lateral band extends 

 downwards, uniting with the orange stripes at bases of prolegs. but 

 includes two black islets on each segment. On abdominal segments 3 

 and 4 the lateral stripe has a greenish tendency. The tubercles are 

 still black; looking directly down on one where the ground colour 

 is still black, it shines with its 7 black spines (or hairs), then 

 round it is a dull black circle, and the black round this shows up 

 the white secondary hairs as a circle of white rays radiating from 

 the tubercle. When the larva is fullfed in the 3rd instar the orange 

 areas become rather green than orange ; what is curious is that 

 the median zone of the segment, surrounding the tubercles, is the 

 first to become green, yet is the one that so often remains black 

 in last instar. This anomaly to some extent remains, but is less 

 difficult to understand when we more carefully examine the colours 

 at this (the third) instar, and find that the black tubercles, ap- 

 parently surrounded by a yellow ring, are not simply the 

 tubercles, but that the black centre includes also a black 

 circle round the tubercle proper. Fourth instar: In this instar 

 we find a very definite change of coloration, we have not now a black 

 larva with orange (latterly becoming greenish) markings, but a 

 larva of very strong and vivid green, with black markings, and 

 though there is still great variety in the relative amounts of green 

 and black (and no orange) the tubercles are nearly black and the head is 

 usually more green than black. The moult then from 3rd to 4th instar 

 is critical in green appearing as the dominant colour. Using green 

 in this instar as synonymous with orange in the 3rd, we may note 

 that the lateral line persists but tubercle v usually has a black ring 

 round it, which was previously wanting. The semilunar mark under 

 iii actually, but not always, persists, that under i is wanting or pushed 

 backwards ; a continuous dorsal black stripe is common, and at its 

 greatest reduction is a black line connecting the black circle round i on 

 each side, at least on forward abdominal segments. Green is the most 



