252 Mr. Curtis on the Nests of two Hymenopterous Insects of Brazil, 



evidence of the fact, I should be fearful to lay the present materials before the 

 Linnean Society. The Bees and Wasps are well known for the skill and in- 

 stinct which they exhibit in the structure of their hexagonal cells, one forming 

 them of wax, the others of wood, and resembling coarse paper or pasteboard ; 

 but in this Saw-fly is a union of the woolly cocoon of the Bomby.r and the 

 cells of the Wasps. I think it has been intimated that the Tenthredinidce 

 might form a distinct order from the Hymenoptera % and those who enter- 

 tained such a view, based on the ground that the larvae were totally different 

 in structure from the rest of the Hymenoptera, and feed, like those of the 

 Lepidoptera, upon the leaves of plants, might perhaps with justice have con- 

 sidered the present as additional evidence of the connexion which this family 

 holds between those orders. Caterpillars of the solitary Saw-flies, especially 

 the larger ones, form singly oval cocoons of a very tough and leathery mate- 

 rial upon the twigs of bushes and trees ; and those which are gregarious, as 

 the Lophyri, do the same, placing them irregularly, and having no common 

 and united design, each one, as it becomes full-fed, forming an oval case of 

 silk and gum, in which it undergoes its transformations ; and it is the same 

 with the social species of Nemati : but D. Ellisii is evidently gregarious, and 

 the caterpillars unite and form upon the branch of a tree an oval or elliptical 

 case (fig. 3), which is narrowed at the top, and four or five inches long. It is 

 very uneven and roughish outside, of a dirty whitish-ochre colour, resembling 

 in texture the gummed side of the cotton wadding employed in ladies' dresses ; 

 but it does not shine, yet I doubt not it is impervious to wet. The side next 

 the tree (fig. 4) partakes of the form of the bark or portion to which it is 

 strongly attached, being sometimes concave, at others flat ; this surface is 

 woolly, so much so, that it looks precisely like the coat from our sheep {h) ; 

 along the centre of this are indistinctly defined the cells (/), placed trans- 

 versely, and amoimting to 13 in the smaller specimen examined; there were, 

 however, 38 in all, as shown in fig. 5, which is a longitudinal section divided 

 at right angles with the branch of the tree. These cells were piled one upon 

 another, but all placed horizontally ; they were unequal in size and irregular 

 in form, those next the tree being quinquangular, the central ones hexagonal, 

 and the outer ones the most irregular, some of them being nearly round or 

 oval. In one of these cells I found a dead female saw-fly, and most of them 



