AGRIUS CONVOLVULI. 349 



(i). Eleven of the larvae without dark markings in stage I. — Dark pigment 

 (black in the extreme varieties, smoky in others) is now present in some of the 

 larvae, on the head, prothorax, thoracic legs, claspers, near the anus, and round the 

 spiracles. The four lightest larvae had the dark pigment only slightly developed 

 on the claspers, thoracic legs (in one of them the claspers were free from dark 

 pigment, and the thoracic legs comparatively pale), round the spiracles and along 

 the lower edge of the furrow running below each lateral margin of the anal flap ; 

 from this furrow the dark colour tended to spread downwards. There was also 

 a very slight dusky shade over the region of the ocelli in two of these larvae. The 

 remaining seven larvae were much darker ; in only one of them was the pigment 

 on the head confined to the ocellar area. The darkening below the anal flap was 

 strongly marked, and, in a few of the darkest larvae, it extended on to the flap 

 itself, although chiefly developed upon the edges of the latter. The previously 

 described dark marking on the dorsal surface of the prothorax was distinct in one of 

 the darkest larvae and just indicated in one or two others. One of these larvae 

 is represented in pi. xv., fig. 7 (x 5"8). 



(2). Six of the larvae, of which five possessed the dark markings in stage I. — 

 These larvae were much lighter than those described above. Only one possessed 

 the dark shade (and this not strongly marked) on the side of the head, other than 

 on the ocellar area ; only one (not the same larva) showed some slight indications 

 of the prothoracic darkness which was marked in five oi these larvae in the last 

 stage. There was a slight cloud over the ocellar area of four larvae. The lightest 

 of all the seventeen larvae was included in this division ; in it, the dark colour was 

 absent from the region of the anus and the claspers, and was very slight on the 

 thoracic legs and round the spiracles. All the larvae except this, and one from the 

 former division, had the dark shade more or less marked on the claspers. The 

 transparent zone on the caudal horn was especially distinct in this division and in 

 the lightest larvae of the other division. 



The larvae were now (September 20th) rearranged in two new 

 divisions, according to the presence of dark pigment in this stage. 

 There was an obvious contrast between the seven darkest larvae of 

 the first division and all the others ; the new division of ten light 

 larvae was, of course, constituted by adding the four palest of the 

 old first division to the old second division. On September 21st, I 

 observed the presence of a row of brown spots on each side of the 

 larva in the light division, which possessed traces of the prothoracic 

 darkening. There was a spot on each of the abdominal segments 

 1-7 (inclusive), situated just above the spiracle and below the lower 

 margin of the subdorsal line. This marking could not have be- 

 longed to the next stage, as I thought at first, for the larva was 

 still feeding, and continued to feed for two days. On September 

 22nd, the subdorsal line became much more distinct on all 

 the largest larvae. It had also become very broad by ex- 

 tending upwards, being thus gradually transitional into the 

 broad green dorsal band. Its lower margin remaining sharply 

 defined, it follows that the ground-colour of the body is somewhat 

 sharply divided into a dorsal and ventral shade. The brown-spotted 

 larva was carefully examined ; it had a dusky tinge in the region 

 of the ocelli, from which a faint cloud extended up the side of 

 the head, there was a fair amount of dark pigment on the thoracic 

 legs, which became red toward their extremities, on the claspers, 

 and below, and (by this time) upon the anal flap. The 7th 

 white stripe which terminates upwards at the base of the caudal 

 horn, could now be just made out in its upper part, in this and 

 most of the other larvae. [This stripe was always the first to 

 become conspicuous in Sphinx ligustri and in Smerinthiis?^ An ex- 

 cesssively faint brown border to this stripe was continued downwards 

 and forwards from the 7th brown spot. Similar indistinct 



