108 THE CANADIAN ENTOMOLOGIST 



veloped and is quite apparent while the larva is emerging from the 

 egg. It is the principal means of attachment for the newly hatched 

 larva, which is able to hold much more firmly with it than with the 

 legs. 



In feeding, the young larvcC are gregarious as a rule and fre- 

 quently may be observed arranged in the form of a circle or arc 

 with the heads pointed outward. Larval groups of 10, 13, 11 and 

 16 were thus observ^ed. Later the larvce spread and usually feed 

 singly. The\' are found as a rule on the under surface of the leaf 

 but in confinement, and in shady places they were observed some- 

 times feeding on the upper surface. In feeding, only the epidermis 

 is eaten, and this is in contrast with the adult which consumes all 

 the tissue. At times there did not appear to be as many young 

 larvae in a group as one would expect from the number of eggs in a 

 mass, and this may be e.xplained by the fact that recently emerged 

 larvffi were observed feeding on eggs which had not yet hatched; 

 three were feeding on one egg, two on another, and two others on 

 one each. 



When full grown the larva attaches itself to the leaf, usually 

 the under surface, by means of the sucker-like disc at the anal ex- 

 tremity, and changes to the pupa which rests with its posterior end 

 within the cast skin of the larv^a from which it could be pulled with 

 little effort. This stage was found to cover a period of two to three 

 days. 



The pupa is yellowish brown in colour; meso- and metathorax 

 dark at centre; margin of wing-pads dark; abdomen with a row of 

 broad dark spots on either side of centre, giving the effect of a light, 

 medial, dorsal line; beyond these is a second row of broad, darlc^pots 

 near the lateral margin. Length 4 mm.; width 2.5 mm. 



Certain pupce were noted which did not rest close to the leaf 

 surface and upon examination these were found to be parasitized. 

 These pupse were in a more erect position and it was noted that the 

 anterior ventral surface was more or less broken, exposing from two 

 to four parasitic larvae or pupae. Parasites bred from such speci- 

 mens were kindly determined for us by Mr. Girault. of the Bureau 

 of Entomology, as Ccelopisthia rotiindiventris Girault and Pleuro- 

 tropis tarsalis Ashmead, and judging from the condition of the 

 specimens when they were bred, the former is a primary parasite 



