f)2 THE CANADIAN ENTOMOLOGIST. 



facts pointing toward these objects, I only ask the intelligent student to 

 select a group for study, and when he concludes his investigations, let us 

 have the matter fully explained. But before he can investigate cocoons 

 or other nest forms of any particular genus, it is necessary that a collection 

 be made of the various structures that contain either the living pupa or 

 nymph. These should be preserved in separate boxes, and those speci- 

 mens which his knowledge leads him to suppose belong to identified genera 

 should be kept in such condition as to accord with the position in which 

 he originally found them. This is the proper course to follow in inves- 

 tigating nest-structures of insects, which I claim will lead to the correction 

 of many errors in our present classification, and place in their proper 

 position many species that are now arranged under wrong genera. If 

 this plan is carried cut, the student will be rewarded with instructive 

 lessons and discoveries of the greatest interest to science. This was the 

 system commenced and partly worked out by the late lamented Uenjamin 

 1). Walsh, of Illinois, whose investigations of insect life were of the highest 

 order; indeed, much of the present advanced state of the science in 

 America is due to him. When my few illustrations and descriptions of 

 insect architecture appeared in the Canadian Naturalist^ he was the first 

 to notice the matter and send me additional information regarding the 

 species ; and as I consider his remarks of value. I give them here as an 

 addenda to said descriptions. 



"No. i (see "Canadian Naturalist and Geologist," Dec. 1865, p. 461), 

 except in being slightly smaller, strongly resembles the nest of Eumenes 

 fratemus Say — a very common insect here. I have bred the female 

 imago from the nest, and some that I broke open in the summer con- 

 tained numerous green caterpillars, enough, I should judge, to feed the 

 larvae to maturity. L do not believe any wasps that are not social feed 

 their larvae after they are hatched out. The use of the short tube, which, 



Chinese species named Selene by Dr. Leach ;" and regarding the Po'yphemus, which is our 

 most common species, he says that " it is remarkable that two insects which are so similar 

 in their preparatory states that their larva? differ only by slight and unimportant marks, 

 and their cocoons cannot be distinguished from each other, still come to be so unlike each 

 other in their perfect state as is Polyphemus and Luna. These facts show that the meta- 

 morphoses of the insects of this order are not so accurate a guide to their systematic arrange- 

 ment as many have assumed them to be." 



I have some reason, on another ground, to divide Promethea from Cecropia on cocoon 

 form alone ; and no doubt when the American species constituting the Linnsean genus 

 Attacus are properly studied, great differences will be discovered, not only phytophagi- 

 callv but also in the internal structure of their larva'. 



