140 BRITISH BUTTERFLIES. 



brown spiracles ; this is almost certainly the lateral tracheal trunk 

 showing through the skin and subjacent green fluid. The waved 

 flange below this, in some lights, or, perhaps, rather at some angles, 

 looks definitely paler ; this is when one looks through its margin rather 

 than into the depths of the larval structure. This appearance is more 

 noticeable on the prothorax and on the last segment, where there seems 

 to be a really lighter tint, or the light passing through it is affected by 

 the differently tinted head and anal plate. The prothorax has a central 

 transverse incision, the portion in front of this is convex and freer 

 from skin-points, but carries one row of hairs. The brown spiracles, 

 laterally, are very large, three or four times the diameter of the 

 abdominal ones, except the last, which is again very large.- but smaller 

 than the prothoracic ; it is a very distinct terra-cotta brown spot, visible 

 without a lens, and is much more dorsal than the others ; all the 

 spiracles stand up at the summit of brown chitinous tubes, like tailors' 

 thimbles. The black points already alluded to are not skin- points, 

 but hairs; each is a small round black button with a minute hair at 

 top. They are about six or seven times their diameters apart, the 

 distance apart being about 0*06mm., and their diameters about 0-Olmm. 

 The true skin-points are colourless, or nearly so, and present six or 

 eight between any two of the hair-points. The subsegmentation 

 (dorsally) of the abdominal segments is into seven portions, of which 

 the first is nearly as wide as three of the others, the others about equal, 

 but the last narrowest ; from front to back of the largest (front) one, there 

 may be perhaps eleven, of the others three, hair-points, but as their align- 

 ment is rather oblique when there is any, this is more of an estimate than 

 correct counting. On the 2nd and 3rd thoracic segments there are 

 three subsegments, of which the middle one is very wide laterally, very 

 narrow dorsally, the other two disappear below laterally but are wider 

 dorsally, with curved margins against the middle subsegment. The 

 1st abdominal seems to have fewer subsegments (the posterior small 

 ones) than those following, but they may be sunk in the incision. At 

 the spiracular level the subsegmentation is evanescent — each spiracle 

 is a little in front of the middle of the segment, which is here rather 

 prominent. A little way below the spiracles is the lateral flange, 

 which almost hangs down like a curtain, and has a waved margin, 

 varying a little with attitude, usually a small wave in front and a 

 larger behind below the spiracle, but this one may be again divided ; 

 posteriorly the flange is rounded, less curtain-like, and less definitely 

 divided. Below the lateral flange is another, half-way between the 

 upper one and prolegs, sometimes very obvious, at others invisible, 

 according to attitude. There are numerous lenticles ; the most 

 conspicuous are two about two- thirds down the anal plate. There are 

 also two marks near the base of the plate which have, however, no 

 lenticular structure. On the prothorax is one well below the spiracle, 

 and one well below and behind the spiracle is on each abdominal 

 segment (=iv '?). There is one above and in front of spiracle ( = iii ?), 

 and one on the 4th subsegment higher up ( = ii ?). If these identifica- 

 tions mean anything, then tubercle i is unrepresented. On the 8th 

 abdominal, iii is in front of, not above, the spiracle, the latter being so 

 much raised. The 1st, 2nd, 7th, 8th and 9th abdominals have ventral 

 lenticles apparently in place of prolegs ; they arc however, on what cor- 

 responds with fchesegment just aboveproleg. On the 2nd and 3rd thoracic 



