﻿30 [Juiy> 



tenements full of " frass." T handled sundry specimens of this larva 

 before they were full-fed, and noticed no particular odour ; but when I 

 happened to take up with my finger a full-fed larva, which I had 

 watched tumbling down, I immediately dropped it in disgust, as a most 

 unpleasant odour, similar to that of Acanthia lectularia, struck my 

 olfactory nerves.* 



The larvae now bury about an inch under ground, and pass their 

 metamorphosis there in a spun stout cocoon of brown silk (not mixed with 

 any earthy particles) , and of about the size of a grain of wheat. Many 

 of the cocoons having been attached to small pebbles, presented, after 

 removal, one or more flattened sides. When the cocoon is finished, 

 the larva? have completely lost the offensive odour they previously 

 possessed ; I repeatedly made sure of this by turning larvae out of their 

 cocoons and purposely irritating them ; and at the risk of being de- 

 nounced as " dealing with hypotheses," I am tempted to conclude that 

 this odour is given to them as a safeguard during that short but most 

 critical moment of their lives, when neither the shell of the gall nor 

 mother Earth can protect them against j;heir enemies. 



The imagos made their appearance between the 13th and 15th 

 October of tbe same year. On the evening of the last-named day, at a 

 quarter-past six, I began to watch one of the cocoons, which, by sundry 

 slight movements, had given me warning of some impending change. 

 I had previously taken the precaution of placing several cocoons upon 

 a highly-polished glass. A few minutes afterwards the antennae of the 

 insect were gently pushed through a very narrow slit at the upper end 

 of the cocoon, and, with a tremulous motion, they seemed to explore 

 the surrounding spot ; five minutes later a part of the forehead made 

 its appearance, and I could now see how the little busy jaws worked 

 their way along the slit mentioned, gradually widening it, until the top 

 of the cocoon, shaped like a skull-cap, was nearly detached. The insect 

 directly pushed this easily aside, rapidly walking out of the cocoon, and 

 now busily cleaning its antennae by moving the fore-feet over them, and 

 subjecting the wings to the same process with the hind legs. It took 

 to flight immediately afterwards, and must have attained to full ma- 



• Other accounts of this faculty being possessed by larvae of saw-flies" are on record : see, for 

 instance, Westwood, Introduction to Mod. Class, of Insects, vol. ii, p. 10 t. Some species, as is well 

 known, have the power of discharging an oily liquid, others are covered with a slimy substance, &c. 

 These various protective properties may possibly, to some extent, account for the Tenthredinidte being 

 almost the only group amongst Hymenopterous larvae, of which many species parade with ornamented 

 skins in various lively colours, for it seems evident that a group thus rendered distasteful by various 

 attributes has a far greater chance of holding its own in Nature than others. This " survival of the 

 fittest" seems to have been conducive to two distinct results : a strong increase of numbers, both of 

 species and individuals, and an assuming of a more or less gay livery in that stage of development, 

 which generally lasts the longest— i.e., the larval state. The internal feeders of the group have only 

 had the advantage of numerical increase, whilst the species living exposed on various ^plants have 

 profited by the influence of light besides ; hence the former all show dull colours, and present a marked 

 contrast in this respect to the external feeders. — A.. M. 



