72 SMITHSONIAN MISCELLANEOUS COLLECTIONS [VOL. 50 



Apis mellffica are at the point of insertion in the tube about 12 mi- 

 crons and at the basal end about 15 microns thick. The lumen of the 

 hook, which in the Anthophilids possessing a hollow costal vein is in 

 open communication with the lumen of the vein, does not terminate 

 with a funnel-shaped extension ; at the most, such an extension is only 

 slightly indicated. This anthophorid type of hook insertion possesses 

 the decided advantage of greater strength over the tenthredinid type, 

 but this greater strength is gained at the expense of that elasticity 

 and movability which is attained in the Tenthredinids through the at- 

 tachment by means of the drum membrane and the chitin threads. 

 The distal hook of a bee or bumblebee can only be removed from the 

 wall of the costal vein, without being itself broken, by shattering the 

 costal vein to which it is attached. As a result of this, during the up- 

 ward stroke of the wings, the fore wing presses principally on the cen- 

 tral parts of the hooks in the Anthophilids, while in the Tenthredinids 

 this pressure is transmitted to the drum membranes covering the 

 pans and the chitin threads attached to the bases of the hooks. In 

 accordance with this, the distal hooks of the Anthophilids are con- 

 siderably broader, longer, and thicker than the typically ribbon- 

 shaped hooks of the Tenthredinidse. Numerous transitional forms 

 connect these extremes. The Anthophilidae type is by far the most 

 common, narrowness of the lumen of the costal vein, or even solidity 

 of it, being the rule. 



Besides the distal hooks, in many Hymenoptera another kind of 

 hook occurs — Staveley's subbasal hooks. These are situated, as the 

 name implies, near the base of the wing, and sometimes also half- 

 way between this and the distal hooks (pi. vii, fig. 3, SbH ; pi. viii, 

 figs. 14-18). They are not met with in all the families, and show 

 great diversity in regard to number, arrangement, and development. 

 In the Vespidae, Formicidse, Evaniidse, and Proctotrupidse they are 

 wholly absent. In some Apids and Fossores they can be recognized 

 as slightly curved, stump-like processes, placed midway between the 

 basis of the wing and the distal hooks, generally nearer the latter. 

 Many genera of these families, however, are entirely without them. 

 It appears doubtful whether the chitinous structures of the Cyn- 

 ipidae, Braconidae, and Tenthredinidse corresponding to the sub- 

 basal hooks of other families, should be designated as clasping 

 organs, because they are hardly at all curved. Only in a single spe- 

 cies of the last-named family, in Pamphilius hypotrophicus D. T., I 

 discovered curved subbasal hooks. These are very peculiarly bent, 

 about ten in number, and form a group near the base of the wing 

 (pi. viii, fig. 15). In Sir ex gigas L. appendages are found in the 



