138 Forty- fifth Report on tee State Museum. 



backward within its walls — its head and two or three of its segments 

 projecting. The other end is now commenced, and is extended back- 

 ward on the same plan and in the same manner as before, until the two 

 portions are united. Sometimes the joining is so accurately made that 

 it can hardly be detected; the ridges are in exact line and only a slight 

 elevation or other irregularity is visible. In others the ridges, instead 

 of aligning, are interposed at the suture, and a perceptible hump marks 

 the place. For a short time longer the caterpillar may be seen through 

 the translucent walls, actively twisting its front segments backward 

 and forward, as additional threads are being thrown over the joining 

 to unite the two portions, and the strikingly ingenious piece of insect 

 architecture is finished. 



Since the above was written my attention has been drawn to the 

 detailed account given by Mr. Chambers in the Canadian Entomolo- 

 gist, vol. xiv, 1882, pp. 15*7-160, of the building of the cocoon of 

 another species of Bucculatrix, viz., B. ambrosimfolitlla. It differs 

 materially from that above described, in that " a reticulated frame- 

 work" is at first constructed (of which a diagram is given by Mr, 

 Chambers), and " the cocoon proper " afterward spun within it. 



No specimens of the cocoons of B. Canadensisella are at hand 

 which would serve to show if they also are double, but I would expect 

 to find them single — simply reinforced with additional threads 

 within. 



I am unable to reconcile what I saw, or at least thought that I saw, 

 in the construction of the cocoon of B. Canadensisella as above 

 described, with Mr. Fletcher's account of the same operation, given in 

 his Annual Report for 1892 (loc. cit.): "A mat was first spun on the 

 surface of the leaf, then the foundations of the ribs were begun. 

 Little by little they were continued, and the meshes of an open net- 

 work stretched between them, the caterpillar all the while retreating 

 backwards as the structure advanced." 



There was certainly no "open network" in the cocoons that my 

 caterpillars built before my eyes. Strange that they should have done 

 so differently from what they ought to have done " according to the 

 books." Mr. Chambers, who was a close student of the habits of 

 Bucculatrix larvae, lias stated: "AH other known species of the genus, 

 save one, make these ribbed cocoons, and to do so, they must work 

 much as this [ambrosiwfoliella] does." 



The building of the cocoon of the common />. pomifolietta has 

 probably never been observed, as no account of it has been given to us. 



