1903. 



13 



the other segments, joined to the following segment by a pellucid extensible mem- 

 brane, but as it were " soldered " to it immediately, so that tlie two segments, 

 dorsally viewed, look like one. The ventral plate of segment 8 is represented by a 

 pair of " wires " like those of segment 7, bearing attached to them the so-called 

 "supports" or back ])ieces of the saws, and also bj' the chitinous and pilose 

 (bivalvcd) sheath which encloses laterally both the saws and their supports.* Tiie 

 9th dorsal plate already mentioned forms the superior apex of the abdomen ; it has 

 no chitinized ventral plate corresponding to it, but from under its apical margin 

 (low down on each side) spring two pilose and palpiform (ventral?) processes 

 which are called the cerci. 



The S abdomen at first sight appears to consist of seven pretty simple " annuli," 

 on dissection, however, it will be found that the apparent last dorsal plate is the 

 7th and the last ventral the 8th. The true 7th ventral plate is concealed, and, like 

 most concealed plates, modified in form ; it consists of two lateral subtriangular 

 lobes united by a narrow basal band, in other woi'ds, it is excised to such an extent 

 as to be practically bisected. The 8th (apparent 7th) or apical segment is very 

 large and convex externally, forming a sort of saucer under the genitalia, and com- 

 pletely hiding them from below. The dorsal 7th plate is not really apical, though 

 it looks so. Under it lies, nearly but not quite concealed, the true apical plate 

 bearing cerci as in the $ ; and also, quite out of sight, the vestiges of yet another 

 dorsal plate, the 8th, so that the plate with the cerci attached is, as in the y , the 

 9th. This 8th plate, though I have repeatedly dissected specimens, escaped my 

 notice till I had read Zander's paper of 1900, but I have now satisfied myself by a 

 fresh investigation that it is really present in Tenthredo, and doubtless in the other 

 genera. In T. viesoinela, $ , what I take to be the remains of dorsal plate 8 appear 

 as two lateral subtriangular scraps of very deep green chitinous substance, united 

 by a frail belt of transparent almost colourless membrane. f Dorsal plate 9 is 

 chitinized along its lateral margins, but becomes membranous (though pilose) at the 

 apex ; its base and centre are also thin and transparent. The cerci are formed and 

 placed as in the i , but are smaller and less conspicuous. These two plates (8 and 9) 

 seem to be very much the same in Tenthredo as they are in the genera which Zander 

 has examined, viz., Cimbex and Sirex. 



The " neuratiou " characters of both wings must of course be 

 mastered by any one wishing to determine Tenthredinidce, but they are 

 often a great stumbling block to beginners through their variability. 

 This, however, only applies to certain of the nervures. Those whose 

 direction is lonrfitudinal seem to be thoroughly constant ; and even of 

 the transverse nervures those nearest to the bases of the wings seem 

 to vary little, if at all. The more apical transverse nervures, however, 

 are very liable to appear and vanish abnormally in particular speci- 

 mens, and in this respect irregularity is more common in certain 

 species than in others. It is unlucky that the nervures whose presence 



♦The "sheath," together with dorsal plates 8 and 9, forms a complete annulus of chithi 

 round the apex of the insect's body when tlie saws, &c., are retracted. 

 t cj. the structure of ventral plate 7 described above. 



