592 [December 



of including the wart in his cocoon, had had the remarkable foresight to con- 

 struct his cocoon entirely above the wart, and was thus compelled to make 

 another diaphragm just above the wart, besides the usual one near the beak, 

 and to lie in a much smaller compass than usual between the two. 



Described from 10 living- specimens and 27 old and dead ones, all 

 from S. humilis. Rather rare near Rock Island. Varieties of S. ha- 

 tatas n. sp. occur, which externally can scarcely be distinguished from 

 «S'. siliqua ; but on cutting into them they are seen to be not hollow, 

 but filled with a spongy substance containing several of the cells which 

 are inhabited by the Cecidomyia of that polythalamous (xall; and more- 

 over, the terminal bud is not beak-like and tubiliform. 



Specimens found on S. cordata in November differ as follows: — \st. 

 The average dimensions are about i smaller, the length in 4 living spe- 

 cimens and 41 dry and dead ones being .45 — .85 inch and the breadth 

 .17 — .28 inch. 'Ind. Out of three of the living galls where the Ceci- 

 domipa was present, there was in two a double diaphragm both at top 

 and bottom of the hollow, instead of the single diaphragm at the top 

 only; but in the other one the diaphragm was single and normal, 'ird. 

 The number of buds on the external surface of the 45 galls is 1 — 3 in- 

 stead of 2 — 5. \t]i. The terminal beak in i of the above 45 specimens 

 is conspicuously recurved, whereas it is never recurved in those that 

 grow on 8. humilis, thovigh it is sometimes a little oblique and in a 

 single specimen is at right angles to the axis of the gall. A gall found 

 August 1 had the beak so much recurved as to touch the side of it, like 

 the tongue-case of the pupa of vSphinx 5-maculata Haw. — From my 

 having in two successive seasons found the old dead and dry galls on 

 both the above two willows at least 8 or 10 times as numerous as the 

 green ones, and from the very weather-worn appearance of many of 

 them, and the fact that a few of them were overgrown and almost obli- 

 terated by the twigs that surrounded their base, I infer that they hang 

 on the twig for several years. 



A single living gall gathered on S. discolor in November differed from 

 the living ones found on S. humilis as follows: — Is^. The woody matter 

 composing the outer shell is much thinner than in any one of 14 green 

 specimens off S. humilis and 3 green specimens off S. cordata that I 

 have cut into, being to a much greater extent medially interrupted by 

 a layer of brown spongy matter, so that the gall was rather crushed by 



