AGRIUS CONVOLVULI. 367 



its sculpturing ; a combination of pits with longitudinal wrinkling, 

 not regular enough to say the pits are in rows down the furrows of the 

 wrinklings, but with some such tendency, especially on the 3rd segment. 

 This would appear to be a very similar development to the callosities of 

 the metathorax. The rest of the segment is much less strongly marked 

 and has both pits and wrinkles, but not so definitely related 

 to each other. In places, as subdorsally on abdominal segment 3, the 

 wrinklings have a flowing combed look and vary much in direction : 

 ventrally 4, 5 and 6 are very smooth and polished, showing, however, 

 the same wrinklings and pits as if nearly polished away ; 7 and the 

 following segments are well pitted ventrally. The slit of the 1st 

 spiracle has the usual posterior lip, dark, smooth and projecting 

 well forwards ; just within the anterior lip is a row of very fine 

 shining points or spicules ; the slit is about 3mm. long ; the ab- 

 dominal spiracles are rather wide slits with sharp raised margins in 

 an oval area, which is at the bottom of a hollow, more marked on 

 the latter segments. The scars of prolegs are slightly depressed, 

 smooth, with the wrinklings puckered to them as usual, so far as 

 the nearly effete sculpturing here can be made out. There is a 

 little pit (5 or 6 times as wide as the pits of the surface sculpture), 

 inside and posterior to the scar in some specimens, as well as another 

 outside it and a little more forward (larval tubercles?). The pre- 

 spiracular flanges of 5, 6 and 7 are somewhat complex. They are 

 4mm. in length, their ventral end level with that of the spiracle and 

 fully 1 mm. in front of it. The surface passing forward from the 

 spiracle, instead of curving inward to the suture, maintains the 

 full width of the segment, or rather more to the first ridge of the 

 flange. This ridge is sharp, and the surface in front of it passes 

 directly inwards (in a transverse plane) for about imm. ; thence the 

 segment continues its normal curvature, but is occupied first by two 

 (smaller) sharp transverse ridges, returning into each other at the 

 ends and with a hollow between them due to the height of the 

 ridges. In front of these again are two more ridges, that are little, 

 if at all, raised, but are (rather perhaps than ridges) the margins of 

 a rather deep hollow that lies between them. These are practically 

 identical on all three segments, but larger and more marked on 5, 

 least so on 7. The caudal horn scar is, as usual, variable ; a small 

 smooth area with pits round it, smaller the nearer they are to it, and so 

 giving a radiating effect, seems normal. In one specimen, the smooth 

 area is larger, and there is an actual radiation of raised ridges. 

 The anal spike varies much in size and boldness in my few 

 specimens. Seen dorsally, the regular conical tapering of 8, 9 and 

 10 is continued to end of horn ; seen laterally, the end of 10 is 

 truncate for its ventral half and the spike, rising from the dorsal 

 half, proceeds directly backwards, so that its apex is in line with 

 nearly the centre of its base ; both ventrally and dorsally there is 

 a curve (concavity ventral). In two specimens its length is — in 

 one 4mm., in the other 2mm. It finishes (not in a bifid point 

 but) with two quite distinct minute points ; the dorsal wrinkling is 

 coarse, basally longitudinal, for the rest very labyrinthine, ventrally 

 less coarse, labyrinthine tending to be longitudinal. The anal 

 scar is wide posteriorly, narrowing forward, a slight eminence on 

 either side, longitudinally folded ; the male tubercles are well raised, 



