of certain species of Willow. — Part 2nd. 243 



From the above facts it follows, I think, conclusively, that these co- 

 lorational streaks exist typically throughout the whole family of the 

 Sawflies, but that in certain genera and subgenera they are broken up 

 into a series of spots which we call "bullae," located on the veins and 

 that part of the membrane of the wings which immediately adjoins the 

 veins. Similarly the typical black vittae on the Chrysonielidous elytra 

 are broken up in Gerotoma caminea Fabr., Diabrotica Vl-punctata 

 Fabr., Ghry&omela scripta Fabr., and Chr. interrupta Fabr., into 

 several series of short, black, longitudinal lines or spots; and in one 

 and the same species — Blepharida rhois Forster — some varieties occur 

 with three uninterrupted vittae upon each elytrum, while ordinarily 

 these vittae are broken up into a very variable number of minute dots, 

 and are sometimes almost entirely obsolete. It further seems to follow, 

 that the system of bullae in Ichneumonidse has been derived from that 

 of Tenthredinidse, by omitting the bullar streaks, even in the darkest- 

 winged species, (except the one passing through F and G, which in 

 many genera, i. e. Tragus, is pretty distinct, and except also a vestige 

 of the submarginal streak in certain species, which I have called a 

 •■seiui-bulla," )* and by suppressing a few of the bullae themselves. For 

 example, since both the marginal cross-vein and the 1st submarginal 

 cross-vein are obsolete throughout Ichneumonidse — just as the former 

 is obsolete in the Tenthredinidous genera Nematus, Euura, &c, and 

 the latter in the Tenthredinidous genus Emphytus — the Tenthredinidous 

 bullae M and N, which are located on those two cross-veins, are also 

 necessarily obsolete in that family. Again, A' is never met with in 

 Ichneumonidse, although in Pinipla and Ephialtes both B and B' are 

 found, which I had wrongly supposed to be attributable to the trans- 

 ference of A from one cross-vein to another. (Proc. etc., V, p. 211.) 



A Ve can now see, likewise, why the bullae C and D, which are sepa- 

 rated by a wide space in the genus Ichneumon, (Fig. 2,) are in the 

 Tchnemnonidous genus Glypta separated only by a dot and occasional- 

 ly even confluent, and in the Ichneumonidous genus Crypfus are nor- 

 mally confluent. Manifestly it is because the typical white bullar 

 streak bifurcates, in the two first genera, on the basal side of the 2nd 

 recurrent vein a little before it reaches that vein, while in Cryptns, as 

 in Tenth'redo, (Fig. 1, CD,) it bifurcates on the vein itself. 



Although the locus of the bulla} and of the bullar streaks is always, as 

 I have already stated, in certain slender folds of the wing, yet it is evi- 

 dent that they are not caused mechanically by those folds, as a piece of 



* See Proc. etc. V. p. 212. 



