armatures in the Hymenopterous genvs CoUetes. 33 



in various ways — sometimes looking much longer than the 

 rods, sometimes not so — in some parts quite glassy, but in 

 others often more or less dark and opaque. Besides these 

 " wings " the rods may, or may not, have other lateral 

 appendages or dilatations of rather thicker and darker 

 substance, and these sometimes give the impression of a 

 sharp triangular tooth projecting over the base of the more 

 transparent " wing," 



" Characters" presented by the sagittm in this aspect are 

 — the outline and extent of the basal dilatation of the 

 " rods " ; the shape, size, colour, and foldings or twistings 

 of the membranous " wings " ; the presence or absence of a 

 secondary (often tooth-like) dilatation of the rods as above 

 described (between their hasal dilatation and the " wing ") ; 

 the length of the sagittm as a whole in proportion to that 

 of the stiiiites, etc. But all these characters are liable to 

 become — I will not say impossihle, but difficult to recognize 

 if the organs are (as often happens) acciclentally shrunk or 

 displaced ever so little, or if the armature as a whole is not 

 lying in exactly the proper position. I must repeat, there- 

 fore, that, for practical purposes, I greatly prefer the 

 characters of the seventh segment. 



As soon as we begin to examine the armature in other 

 points of view than the direct dorsal aspect, we shall find 

 that the description given above of the stipites and sagittal 

 requires a good deal of modification. Thus, looking at the 

 stipites (Plate IX) sideways, we see at once that they are 

 never really three-jointed, but composed of a single 

 curiously-folded sheet, which is nearly but not quite com- 

 pletely chitinized throughout ; the apices look no longer in 

 the least palpiform, but more or less broadly triangular ; 

 the supposed suture dividing the stipes transversely is seen 

 to be merely a superficial though often deep sulcation in 

 it, etc. Again, looking sideways at the sagittse (PI. IX), 

 we find that the dorsal view has not shown us their real 

 apices at all — the latter are so sharply deflexed, that when 

 we viewed the organ dorsally they Avere completely out of 

 focus, and in fact out of sight. We have also to correct 

 our former impression of the "wing" as something distinct 

 from the " rod," and attached to it as a sail to a mast — it 

 is really nothing but a difference in the degree of chitiniza- 

 tion which makes them look distinct objects. The truth 

 seems to be that each sagitta consists of a single sheet of 

 substance, folded and "crinkled" up in a complicated 



TRANS. ENT. SOC. LOND, 1904. — PART I. (APRIL) 3 



