THE EARLY STAGES OF COLXAS EDUSA, 173 



ment, showed at the base a cell-like appearance. When first 

 deposited, the eggs are of a creamy-white colour, which changes to 

 a delicate pink on the third day. A depression appears in the side, 

 and the pink gradually merges into a decided red. A day before 

 hatching the colour changes into a dull metallic black. 



Lakva. — The young larva emerges from a point about one-sixth of 

 the distance from the apex, and proceeds to eat a portion of the egg, 

 but rarely eats even one-half of it. The larvte as they emerged were 

 placed, some upon Trifulium repenx, some upon T. pratense, and some 

 upon Lutits cornicnlatti^i. The larva described in the following notes 

 fed on T. prnteme. Without exception, the larvag fed readily on the 

 plant oftered to them, and, at any period of their existence, if removed 

 from Trifoliniii to Lutus would continue to feed without cessation. 

 One point I consider worthy of note is that they irutdd eat ;irass. By 

 chance, in digging the foodplants, I brought in some grass. I was 

 astonished to find it being eaten, so I put some larvae on the grass, 

 and found that they ate it readily enough. The larva, chosen to 

 describe, hatched on October 2nd, and was placed on clover on the lower 

 side of a leaf, but it immediately, without eating, proceeded to the 

 upperside. This each larva did, utterly refusing to feed on the under- 

 side. The reason of this I afterwards discovered. 



First instar : The larva soon after hatching was l-5mm. long. 

 In colour the body was of a bright brown. No traces of hairs were 

 visible. The head was jet black in colour, and was small in com- 

 parison with the body. The larvte fretted out irregular holes in the 

 upperside of the leaf but did not eat the epidermis on the underside. 

 They fed constantly during the day. When done feeding, they went 

 to the midrib of the leaf and stretched themselves at full length along 

 it. When the clover leaf closed for the night, the small larv* were 

 safely folded up in it and thus protected. On October 8th my particu- 

 lar care ceased to feed, and attached itself to the leaf, remaining 

 motionless for two days. It moulted on October 10th and entered its 

 second stadium. 



Second uistar : The larva was now 8mm. in length — exactly twice 

 its original length. A change in the body coloration had now taken 

 place. Although the head was still jet black the body was no longer 

 brown. Ventrally it was decidedly green, but dorsally it was green 

 with a suggestion of brown, giving it a rusty appearance. No 

 spiracular lines were visible, but the larvae had now distinct hairs, and 

 each segment could be distinguished as being subdivided. The larvte 

 still continue to have the same habits of feeduig and rest as in the first 

 stage. They feed by day and rest in the folded leaf at night. This 

 continued until October 16th, when my first larva again ceased to feed, 

 and rested again for two days, casting its skin on October 18th. 



Third imtar : It was now 4-5mm. long. There again had been an 

 increase of l-5mm. in the length. It was now uniformly of dull sage- 

 green. The head, if anything, was a lighter green than the rest of the 

 body. One individual, and only one, kept its head of a shining black 

 even to this stage. The body, as also the head, was distinctly hairy, 

 and each segment was divided into six subsegments. The larva 

 ta,pers both to the head and the last abdominal segment, from the last 

 thoracic segment. A faint indication of a lighter spiracular line was 

 visible as the green shaded oft' almost imperceptibly. In this 



