OF THE SAW-FLIES (TENTHREDINID^). 127 



web, like the larvae of the lackey-moth, whilst others ( Cladius 

 vim'tnalis, Brischke, pi. 2, fig. 1) feed exposed, but in rows, 

 side by side, like the larvae of Bombyx processioned, and 

 others more irregularly (e. g. Lophyrus socius, Fintelmann, 

 pi. 25, fig. 13-16). 



Another species of which I noticed the transformations in 

 the garden of my late residence at Hammersmith (which 

 afforded me so many excellent observations), ingeniously 

 rolls up the leaves of the rose in a spiral manner, somewhat 

 similar to that adopted by some of the water- Pyralidce ; 

 others, again, secrete themselves beneath a fetid exudation 

 from the pores of the body, which gives them the appearance 

 of small black slugs, and a few have the body armed with 

 numerous short spines, occasionally recurved at the tips. 

 Some, again, when at rest, remain with the hinder parts of 

 their bodies elevated in the air in all kinds of curves, whilst 

 some conceal themselves on the under surface of the leaves, 

 coiled up like a Julus; and, lastly, another species burrows 

 into the young stems of roses, consuming the pith, and causing 

 the destruction of the shoot. 



To conclude the general and comparative description 

 of these larvae, I may add the following concise state- 

 ment. 



The adult larvae of the Tentliredinidce agree with the 

 caterpillars of Lepidopte7*a, in the following particulars : — 



Both are plant-feeders. 



Both have a nearly cylindrical, elongated body, divided 

 into distinct segments, and provided with many pairs of legs, 

 some of which are cartilaginous, and others membranaceous ; 

 the former, six in number, are terminated by a hook, and 

 both are furnished also with a horny head, armed with 

 strong horny denticulated mandibles. 



Whilst the following short comparative diagnosis will 



