160 f''"'y- 



The saw is remarkable in liaving only aLout seven of the divisions 

 into Avliieh these saws are divided, the basal half having no dividing 

 marks, and especially in these divisions being hardly indicated along the 

 cutting-edge. Usually each section is marked off from the next by a 

 notch (giving the saw-like outline). In T. m/\rf(i there are no notches; 

 each division is divided from its neighbour by a })ale transverse line, 

 across the end of which at the margin passes a slightly curved edge 

 with very fine serration ; then a nearly equal portion of the end of the 

 division is without teeth, and is, to that extent, a little less jM'ominent 

 than the serrated portion. But practically the cutting-edge is straight; 

 the opposite margin has a strong double chitinous margin that extends, 

 without ta))ering, to the tip. 



The only species approaching this in Cameron's figures are JTr/iu/f/n^ 

 consobriiius andiV^ ribesii, but in both these, teeth, though not marked, 

 are shown at the notches, and the divisions are 12 or 18 instead of 7. 

 iV. ribesii, as alread}'' quoted, probably lays its eggs similarly to T. vti.rfa 

 Pln/matocera aterrima, of the species I know, has the teeth as little 

 marked as the species above-mentioned, and much less than most other 

 sawflies. It has to deal only with the lifting of a very slight cuticular 

 pellicle. 



I did not isolate a larva and watch it specially in order to decide how 

 many moults it had, so that in saying there are four moults (and five 

 instars) I am depending on observations on the larvae feeding together, 

 and on measurements of the crania cast Avith the skins. As the larvae 

 were not all of one age, some a moult or two before others, and as cast 

 crania are sufficiently distorted to give uncertain measurements, there is 

 room for uncertainty. There may be only three moults and four instars. 

 My measurements of crania give the following figures, measured from 

 lateralbase to apex, in millimetres: — 1st instar 0'45, 2nd OQQ, 3rd O'OO, 

 4th 1"14, 5th 1"35. There is no character to distinguish any instar from 

 another, except the last. In the previous instars the minute hairs become 

 a little less abundant, but even so are variable. In all, the skin is 

 colourless. In the last the dorsum has a dark tint and the last segment 

 has two black patches. 



April 24th. — Two larvae have hatched. They are in this 1st instar 

 (before feeding) 2 mm. long, with large black heads (about 0'4 nmi. 

 in diameter); body very pale whitish or faintly greenish from intestinal 

 contents, having, when examined, ajjparently eaten a little since hatching. 

 Bod}' looks smooth, but under lens is seen to have the segments sub- 



