WILLOW-GALL TINEIDS. 583 



rlwdoides Walsh. This is the insect that I think I mentioned to you as being very 

 prettily marked iu the larval state, each segment having a broad black band, and 

 the ground color being whitish. I had a single one come out last summer, but the 

 great bulk of thtim hybernated either in the larva or pupa state and came out May 8 to 

 20. They vary but little. I have beaten larvaj of very similar appearance off oak 

 trees. 



So far as I aui informed the larva is unknown to European Lepidop- 

 terists, although it is recorded that the perfect insect, prceangusta, is 

 very common among willows and poplars iu July, and may frequently 

 be observed sitting on the trunks of those trees with the anterior feet 

 put back like Bdellia and the head raised a little. 



Mr. Walsh has the honor of having made an interesting discovery 

 that puts an end to all uncertainty respecting the larva and its food- 

 X)lant. 



This larva occurred in abundance August 23, and subsequently in the Tenthredi- 

 nidous gall, S. pomum, Walsh manuscript, which grows on the leaves of Salix cor- 

 data. Each gall contained but a single larva, unaccompanied by the larva of the 

 Nematus which makes the gall, which it must consequently have destroyed or starved 

 out, either in the egg or in the larva state. 



A siugle imago came out in the autumn of the same year, but the great bulk of 

 them came out next spring, May 8 to 20, from galls kept through the winter. There 

 «an be no doubt of the correlation of larva and imago, because no other Lepidopterous 

 larva or imago occurred iu the gall S. pomum, though I had three or four hundred of 

 them iu my breeding vase. The insect must hibernate normally iu the larva state, 

 for I noticed numbers of them in the spring crawling about among the galls. Iu a 

 «tate of confinement it generally retires to the inside of the gall to assume the pupa 

 state, though I noticed one or two cocoons spun among the galls. Probably in a 

 state of nature it hybernates in the gall, comes out of it in the spring, and spins its 

 cocoons among dry leaves and rubbish. 



I also bred a single imago of this same species. May 11, from the Cecidomyidous 

 gall, S. rhodoides, Walsh, from galls kept through the winter, and I found iu the 

 spring a denuded imago of what was apparently the same species, dead and dry 

 amongst a lot of Tenthredinidous galls, S. desmodiodes, Walsh manuscript, which is 

 •closely allied to S.immum, but occurs on the leaves of a very distinct species of willow. 

 Thus we have three diiierent willow galls inhabited by the same moth, two of them 

 made by saw-tiies and one by a gall-gnat. 



I have several times beaten off black-oak trees larvie apparently very similar to 

 this Batrachedra, and with the same harlequin-like markings, but whether the two 

 are specifically identical I can not say. 



In a subsequent letter Mr. Walsh kindly supplied me with the fol- 

 lowing description of the larva : 



Larva. — Length, .20 inch. Body tapering at each end, opaque, milky-whitish, 

 with a few short, whitish hairs. The first segment behind the head with an obsemi- 

 circular, shining, glabrous, brown dorsal shield ; second segment with an interrupted 

 opaque brown dorsal band on its anterior edge, the interruption occupying about 

 one-third of the baud ; segments 3 to 12 with an interrupted opaque brown dorsal 

 baud on the anterior edge, and segment 11 with a similar band at its tip also. 

 Head yellowish. Legs and venter immaculate whitish. Legs six, prolegs ten, nor- 

 mally arranged. Spins a thread, wriggles much when disturbed, and runs backward 

 with great agility. (Clemens' Tiueina. Edited by H. T. Stainton.) 



Moth. — Forewings fuscous, with a rather broad whitish stripe, freely dusted with 

 fuscous, running through the middle of the wing from the base aud along the apical 



