174 THE entomologist's record. 



stage the larval habits underwent a decided change. No longer did 

 the larva feed by day on the upperside of a leaf. By day it rested on 

 the petiole of a leaf head downwards, and fed only about sunrise and 

 sunset. I do not think I observed one to feed by day after this stage. 

 When they did feed, they ate at the edge of a leaf as large larvae 

 usually do. They also ate the stalks and stipules of very young leaves. 

 On October 26th, the larva observed again set for a moult, which was 

 successfully accomplished on October 28th. 



Fourth imtar : The larva was now 7mm. long. It is of a dull 

 green colour dorsally. Ventrally it is of a bluish-green colour, in 

 fact, it is quite glaucous. The head is decidedly lighter than the rest 

 of the body. The body is very rough, but under the lens is found to 

 owe its roughness to the subsegments mentioned above. On these 

 transverse divisions are black dots. Each dot emits a short stiff hair, 

 giving the larva the appearance of being bristly. This is particularly 

 noticeable immediately after a larval moult. Around each of the black 

 dots is a ring of a light green colour. The head likewise, is covered 

 with black dots emitting hairs, but the black dots are not surrounded 

 with lighter coloured rings. The green of the dorsum shades oft' to the 

 spiracular line, which is of a white colour. The larva still tapers to 

 last segment, and to the head from the last thoracic segment. On 

 November 5th it again set for a cast, which took place on November 

 7th. 



Fift/i instar: Its length was now 10mm. Generally, in all its 

 coloration, hairs, etc., it is the same as in the last stadium. The 

 spiracular line, however, is diti'erent. It is yellowish in colour on the 

 last segment, and of a pure white on the others, and generally is more 

 distinct than before. On November 14th, when the larva was l-6cm. 

 long, the spiracular line was now blotched with yellow upon every 

 segment, between the 8rd and 4th transverse divisions. This yellow 

 mark was not noticed until seven days after the moult, so must have 

 developed m the meantime. On November 19th, when l-9cm. long, 

 the spiracular line again underwent a change. Although the blotch 

 remains yellow on the prothorax and on the last abdominal segment, 

 yet on all the others it is orange-red, shading oft' to yellow on the 5th 

 subsegment. Immediately below each red spot is a very conspicuous 

 black dot. The last moult took place on November 25th, the larva 

 having prepared itself on November 22nd. 



Hii-th imtar : It was then 2-Ocm. in length, but was not described 

 until full-grown. It reached its full size on Dec. 22nd. Its length 

 then was 2-8cm, (measured when at rest on the foodplant) ; width of 

 broadest segment, 4mm. ; width of head, 2-75mm. ; width of pro- 

 thorax, 3-26mm. ; width of last segment, 3mm. The 2nd abdominal 

 segment is the largest. Thence the larva tapers to the head and last 

 segment. Dorsally each segment is divided into six divisions or sul> 

 segments. The first of these six is the broadest, being almost of the 

 width of two others. For this reason it does not seem so rounded as 

 the others. These divisions are not so recognisable on the ventral 

 portion. In colour the body is of a deep velvety-green dorsally, but 

 ventrally it is quite glaucous. The head is also green, but is of a 

 lighter shade than the rest of the body. Under the lens, the body is 

 found to owe the deepness of the green to black dots, arranged upon 

 the subsegments, each giving rise to a short hair. Upon the first of 



I 



