246 BRITISH LEPIDOPTERA. 



distance. The larvae of the common Tenthredinid, Croesus septen- 

 trionalis, are of about the same size as these small lepidopterous larvae, 

 feed in similar small groups when large (when small the groups 

 contain tar more individuals), and also often frequent the birch. 

 Experiments have shown that Tenthredinid larvae are, as a whole, 

 far more unpalatable than lepidopterous larvae, so that the resem- 

 blance would be of advantage to the Endromis larvae." White and 

 Griffiths had previously shown (Joe. cit., 1888, p. xix) the resemblance 

 that these larvae have in their third stadium to the catkins and leaves 

 of their foodplants ; the latter had noticed " the habit of the young 

 larvae to congregate at the ends of the twigs of birch with their 

 heads nearly always in the direction of the end of the twig, and in 

 their favourite resting attitude, in which the fore-part of the body was 

 elevated at a curve, they bore a great similitude to the young catkins ; 

 larvae had also been noticed to eject from the mouth a greenish fluid, 

 which was assumed to have a protective value, and to be produced 

 under sudden alarm ; it was also a common habit for them to swerve 

 their heads sharply round in a threatening manner when the anal 

 hump was touched." Bacot notes that "the larvae in the 1st and 

 2nd stadia rest in little batches of 6 to 12, at or near the extremity 

 of a twig or branchlet, and, if possible, will feed without crawling. 

 The whole batch, however, will crawl slowly forward by stages along the 

 leaf stalk, one or two eating and the others resting." He suspects that 

 they eat by turns, but they move as short a distance as possible ; if 

 a more extensive movement becomes necessary, the whole batch, or a 

 portion of it, if large, migrates together, and, if disturbed, the larvae 

 will wander singly about the leaf stems or twigs, but always come to 

 rest again in little groups. He cannot say, however, if these groups are 

 always composed of the same individuals or not, but observes that 

 " the larvae sometimes rest with their heads and thoracic segments 

 curved backwards in the characteristic manner, at others stretched 

 at length. They do not all moult together, so that a group may 

 be composed of individuals in two different stadia. In the 3rd 

 stadium the habit of resting in groups (rarely more than 3 — 6 in 

 a group) is still maintained, but is not so marked as hitherto, and 

 the position not so crowded, rarely more than two couples on adjacent 

 leaf-stalks (of course their size now prevents some 5 or 6 getting 

 on a single petiole). In the 4th stadium they are usually solitary, 

 although occasionally they rest in couples, probably the result 

 of being kept in a confined space, and they would possibly 

 scatter widely if opportunity offered ; they still rest with the 

 backward curve already mentioned, as characterising younger larva?. 

 Before pupating the larva changes to a livid hue, pinkish or purplish 

 on the dorsal area ; it shrinks considerably, and the skin becomes 

 moist as though the larva was in a profound sweat. The excrement 

 is mixed with a dark-green fluid, just before they leave off feeding ; 

 in this, the species resembles Saturnia pavonia and some of the 

 other large cocoon-spinning species. 1 have always connected this 

 trait with silk-spinning, but am by no means sure that there is really 

 any connection." Jenvey gives (/./.) the following notes as to larval 

 habits — eggs laid April 7th, 1900; 19 larvae emerged April 24th and 

 25th, young larvae entirely social, forming one colony only during 

 the whole of the 1st stadium; larvae rested 36 hours and then moulted 



