360 



delicate, somewhat twisted tubes, quite incapable of altering 

 the shape of the wings which bear them. It seems much more 

 probable that the straightening is due to the turgidity of the 

 cells of the great ectodermal extension into the channels, and 

 that the wings remain firm, later, as the result of the action 

 of the air on substances contained in the chitin of the 

 nervures. 



In connection with the small ''stigmal-vein" of the fore- 

 wing, a remarkable structure is developed, the interpretation 

 of which is very difficult. On the distal part of the veins are 

 developed four rounded globules (fig. 39), the distal pair 

 rather smaller than the other two. They are well known to 

 workers on chalcid wasps, and are frequently used in classi- 

 fications. But if a stained wing is examined under a very 

 high power each of these globules is seen to contain a heavily 

 staining sphere (figs. 39, 40) attached to a small conical piece 

 of protoplasm, the base of which is in turn attached to a long 

 fibre. The fibre from each globule passes inwards, and 

 becomes lost in the substance of the stigmal vein. It seems 

 probable that the whole structure is one cell, of which the 

 process and the conical portion represent the cytoplasm, while 

 the sphere is the nucleus surrounded by a very delicate layer 

 of cytoplasm. The long process, of course, immediately sug- 

 gests a sensory structure; the nerve fibre being, as usual, 

 almost impossible to detect, and lying within the weakly 

 staining process, which acts perhaps as neurolemma. If this 

 interpretation be correct, then it is not impossible that the 

 structures concerned should act as speed-de teeters. Increased 

 speed of flight would be produced by increased rate of vibra- 

 tion of the wings; this would result in a greater centrifugal 

 pull on the free (spherical) part of this remarkable structure; 

 or, what is more probable, it would result in a greater fre- 

 quency of the striking of this body against the walls of the 

 globule, and it is conceivable that this would affect the nerve 

 fibre which terminated in it. 



The Mouth Appendages. — The general features of the 

 histogenesis of the mouth appendages are so similar to those 

 of the thoracic, that a brief description will suffice' here. 



The imaginal rudiments of these structures are clearly 

 visible in the first larval instar (fig. 47) ; in their condition of 

 development. the labrum and mandibles are already much in 

 advance of the condition in which we find the thoracic append- 

 ages; i.e., they are no longer merely ectodermal thickenings, 

 consisting of embryonic cells, but have now become invag- 

 inated well into the cavity of the head. 



The imaginal discs of the labrum are a pair of solid ecto- 

 dermal ingrowths, situated at either angle on the fore-part of 



