620 [December 



linous Gall-gnat analogous to those of the Hessian Fly, and that what 

 were the real cocoons of inquilinous Gall-gnats were the cocoons of 

 minute Ichneumons that had been preying on the larvoe of the supposed 

 (luest Gall-gnats I 



The existence of this species, in the peculiar situation where it is 

 found, solves an interesting question mooted by Winnertz, viz : whether 

 inquilinous Gall-gnats "take the same food with their hosts or live on 

 their excrements." {Dipt. JSf. A. p. 184.) In this case the host lives on the 

 sap of the globular stem, from which all the leaves of the pine-cone like 

 gall proceed, and the guest or inquiline must live on the sap, which he 

 manaues to extract from the scales or leaves of the pine-cone. Fre- 

 quently there is a thickness of .30 — .40 inch of solid leaves between 

 the host and the guest, so that it is quite impossible here that the latter 

 can live on the excrements of the former, or interfere with him in any 

 way, except perhaps by slightly diminishing his supply of sap. 



Larva. — Dec. Srd the larva is orange-colored, a little mottled with sanguine- 

 ous, and sometimes with a broad, dorsal, dark-sanguineous or fuscous vitta ab- 

 breviated before and behind. The two tubercles of the anal joint are a little 

 larger and more prominent than usual. The breast-bone is clove-shaped, fus- 

 cous, not very distinct, and the stem of the clove is about i as wide as the entire 

 breast-bone is long. Length .03 — .04 inch, and breadtli rather less than 4 that. 

 Six specimens from cocoons under tlie scales of the gall S. strobitoides. Speci- 

 mens taken out of the cocoon and examined Feb. 20, at which time none had 

 yet gone to pui>a, were orange-color, and on April 29 the breast-bone was darker 

 and very distinct. The cocoon is oval, white, much stouter and denser than in 

 any of the preceding species, so that the included larva can only be seen by 

 holding it up to the light, and has a good deal of the white pubescence of the 

 leaves of the gall adhering to it. Length of cocoon .07 — .11 inch, breadth .03 — 

 .04 inch; 41 specimens which were all obtained from two galls Dec. 3, by which 

 time, and probably long before that, all the larvse had made their cocoons. 

 Three of these cocoons each contained a yellowish larva, uninclosed in a sepa- 

 rate cocoon, and apparently that of a Proctotrupide, one of which was found in 

 the imago state April 29 with its head protrviding from one of these cocoons, 

 and another on the same day at large under the scales of the gall. 



Pupa. — The first jjupa was noticed April 21. but the larva was noticed as late 

 as April 29, and from the first appearance of the imago, some of the insects 

 must have existed in the pupa state at least as early as the first week in April. 

 The abdomen was sanguineous ; the rest of the body, including antennse, legs 

 and wing-cases, fuscous. The horns at the base of the antennse were rectangu- 

 larly conical, terminating in a very minute, acute thorn, and divergent in an 

 angle of about ISO'-'. The thoracic bristle was slender and \ as long as the dia- 

 meter of the thorax: (in the dried specimen it is terminally fuscous and basally 



