50 [February, 



fcnlled by Thomson rather oddly nervus tramversus ordinarius, and by Mr. Cameron, 

 •e. /;., see his Tables of Species for Emphytus, the transverse median) and t x the 

 anal nerve. Of these the former is the more important, its direction and the 

 point at which it is received in the cell above it giving several useful characters. 



We have now, I believe, dealt with all the nervures which are 

 reijjularly present in the upper-wing of a Saw-fly. But certain others 

 which appear onJj/ in particular Genera or Faniilies are for that very 

 reason especially useful for "determinations." 



Thus — between the costs and subcosta may lie a 6th longitudinal nervure, and 

 this may ultimately either run simply into the subcosta somewhere near /, or be 

 forked at its apex into two branches, one joining the subcosta, and the other the 

 costa. Or, in the same region (the intercostal field), there may be a transverse 

 nerve stretching from the subcosta to the costa, either before or after the point 

 where the former receives the discoidal nerve. Or, as in Arge, the costa, instead of 

 lying wholly on the margin of the wing, may quit it just before its apex and bend 

 down to meet the radius, thus cutting off from the rest of the radial area a 

 little subtriangular apical cell {cellula radialis appendicuJata). Lastly, in the 

 AMme?*n? area (" lanceolate cell") we have a number of important characters for 

 determining Genera depending partly on the presence or absence of transverse 

 nerves in that field, and partly on certain modifications in the structure of the 

 humerus itself, especially in its basal part. These characters we have now to 

 examine. 



It will be seen by reference to Fig. 3 that the humerus (quite near its base) 

 shows a strong inclination to unite with the brackius, long before it ultimately does 

 soat^f. Grenerally this inclination is, if we may say so, suddenly checked — the 

 humerus, thougli approaching very near the brachius, starts off at a tangent and 

 gradually recedes to a respectful distance from it, before taking the final curve by 

 which it ultimately reaches it. Sometimes, however, the inclination is not checked ; 

 the humerus continues its approach to the brachius till it actually reaches it, and so 

 the two veins for a while coincide — it maybe only for a moment or for a considera- 

 ble distance — butalwaysseparate again, so as to enclose a spindle-shapedspace between 

 them before their final point of union. Again, sometimes the humerus without quite 

 reaching the brachius at the (sub-basal) point alluded to above, all but does so, and 

 throws a " short perpendicular " nerve across the narrow interval which separates 

 it from its companion vein. Yet again the humerus, soon after its origin, seems to 

 vanish and presently to re-appear emerging from the brachius, though it has never 

 been seen to join the latter! Lastly, sometimes — though never, I believe, in con- 

 junction with any of the phenomena described above —an " oblique cross nervure " 

 runs over (the apical portion of) the humei-al area, cutting it into two divisions, the 

 apical one " completely enclosed," and very much smaller than the other. 



Hence arise six distinct forms of " humeral cell," extremely useful for generic 

 determinations, which may be enumerated as follows : — 



(1) Fig. 4, a. The humerus, though approaching the brachius near its base, 

 keeps quite apart from it till they meet at g. This, the simplest but by no means 

 the commonest form of humeral area, has been called " Cellula lanceolata aperta " 

 (Thomson). 



Example : Selandria. 



