LLME^"ITIS I. 



the feeding on sugar water was to make the bodies enormously fat, so that they 

 swelled out like the bodies of wasps and the msect could scarcely move about, 

 aiid in a few days died. Probably this excess of fat hindered the laying of eggs, 

 for there always were found to be a few mature eo-o-s in the ovaries. These esrsrs 

 o£ Proserjjina hatched, beginning on the ninth day, and the larvic fed well. But 

 before they had reached the hA-bernating stage, I had to guard them in a warm 

 room, and to keep branches of aspen and willow in water to supply them with 

 food. Out of doors the leaves were falling, the frosts Ijecoming severe, and it 

 was certain that not one of these larvae or of any larvie, then feeding naturally, 

 could have reached the hybernating stage. The existence of the species is in fact 

 due to the develojiment of the eggs laid in July and earljMU August. The larvtB 

 hatched 10th to 12th September, began to pass their 

 first moult, ISfli. and second on 24th. Nine lived 

 through the second moult. On 27th, the first one had 

 completed and taken possession of its case.^ The eggs 

 had been laid on willow, but after second moult I trans- 

 ferred the larvie to aspen, which they readily took to. 

 Very soon after the transfer some of them began to cut 

 out the patterns of the cases. First eating a narrow 

 canal for one ({uarter inch, the width of the head, ob- 

 liquely outward from the stem at base, a ; next a canal of 

 same length on the side of the leaf, about three fifths the 

 distance to the apex, perpendicular to the edge, b, then 

 turninij; this at a riii'ht anu-le in the direction of the first canal and cuttinu' for a 

 little distance; then crossing to the other half of the leaf and cutting similar 

 canals; after which the extremity of the leaf was cut off by an incision from I In' 

 bend in the second canal directed obliquely forward to the midrib, first on one 

 side then on the other, c ; next the first and second canals on one side were joined, 

 d, then on the other side, and there remained of the leaf but a small iiddle- 

 shaped piece, lying almost equally on either side the rib. Before and during the 

 time this work was progres.sing, the larva had taken intervals of rest from the cut- 

 ting, and had occupied itself in weaving threads from the branch to the stem, 

 and along the upper side of the leaf, thus coating with silk what was to be the 

 inside of the case. Finally, beginning at the base, it drew the edges partly 

 together for a little distance, leaving an open space between of about one tenth 

 inch, and held them in position by single threads ; then proceeded to weave a 



' The larv.-e of DislppHs, at Co;ilburn;h, pass either two ov three moults before hybernation, liut I have 

 known of only two moults in this species after hybernation. To the northward I think it possible that the 

 fall moults are limited to two, as with Arthemis. 



