AGRIUS CONVOLVULI. 353 



represented in fig. ir, form a very perfect transition from this to the lighter larvae. 

 The very different and darker ground colour of the area above the subdorsal is very 

 striking-. 



9. The ninth larva was the darkest obtained except one (the 7th of the next 

 division),- it was the larva which died September 27th. 



10. The tenth larva was the one which was separated in the 2nd sta^e, 

 because of the especial development of brown spots (see fig. 8). Under these circum- 

 stances it might have been expected that the larva would be among the dark 

 varieties of the succeeding stage, but, as in other cases, the reverse took place. 

 This larva was examined and compared September 29th ; it was a distinct bluish- 

 green variety, ranking with the 1st, 2nd and 3rd larvae already described. The 

 red spots were distinct, becoming larger and more distinct anteriorly, the last was 

 far the smallest, and the last but one intermediate between this and the others. 

 The stripes and borders were very distinct and Smerintkus-tike, as in the 2nd 

 larva. The subdorsal line was very distinct anteriorly. The larva was, therefore, 

 a very light variety, and it would rank between the 2nd and 3rd of those described. 

 On October 2nd it was again examined. There was no dark shade on the head, very 

 little on the claspers ; the thoracic legs and the horn were as in the other light 

 varieties. There was a very slight shade on the anal claspers, and hardly any on 

 the anal flap. The whole appearance was extremely Smerinthine. The stripes 

 were of a bright yellowish-green colour (somewhat deep in tint as compared with 

 other larva?). 



(II) The seven dark larvae of stadium 2. — Upon the whole these 

 larvae were somewhat darker than the ten larvae just described, and 

 hence the effects witnessed in the last stage have, in this case, 

 passed on into the succeeding stage ; but it will be seen that the 

 differences are very small. As before, the bluish lighter larvae will 

 be described first. The larvae were compared October 2nd. 



1. A bluish-green larva, about equal to the lighter of those described. 

 The shade upon the head was almost confined to the ocellar area, an almost 

 imperceptible cloud extending upwards from the latter. The increase in the size 

 of the red spots anteriorly was well seen, the last spot being very minute. The 

 first and last two spiracles were light brown, and the surrounding patches were 

 in all cases very small. The amount of dark colouring elsewhere was very slight, 

 although distinctly traceable in the usual positions. There was hardly any dark 

 pigment between the 1st and 2nd spiracles. 



2 and 3. These two larvae, although bluish-green, would rank with the 

 5th and 6th of the previous divisions, lor there were distinct red marks upon the 

 borders to the stripes, below the red spots, in both larvae, and also above the 

 spots in the 3rd larva. The 2nd larva had a slight transparent zone in the horn, 

 and a very dark head ; while the 3rd had the zone even less marked, but the head 

 lighter. Both larvae were fairly dark around the anus, but the flap of the 3rd was 

 much the darker. The stripes and dark green borders were quite distinct in the 

 2nd, fairly so in the 3rd, larva. 



4. Then followed a larva with a ground colour of an intermediate tint. 

 (This larva is represented in fig. 11 X 3). The condition of the borders made this 

 larva decidedly darker, for the dark red shade was not only present upon the 

 borders above and below the red spots but, in the former locality, a dark smoky 

 shade extended from the central red mark, stretching along the upper edge of 

 the subdorsal line, and forming- a very conspicuous feature. In addition to this, 

 there was a highly conspicuous oblique dark patch above the subdorsal, upon each 

 of the 2nd and 3rd thoracic segments, and upon the 1st abdominal (see figure). 

 The degree of development of the dark markings in other parts of this larva is 

 shown in fig. 1 1. This larva would be quite as dark as the 7th of the previous division, 

 but it was not of the same distinct yellowish-green colour. The elongation of 

 the red spots along the borders (see fig. 11) is a very interesting feature, indicating 

 the relation of these characters to the coloured borders of Sphingidae. It should 

 also be noted that the transparent zone in the horn was well-marked. 



5. This larva was very similar and an intermediate variety, although with 

 a slightly yellower tint of ground colour. 



6. This larva was also an intermediate variety. In darkness it would 

 be intermediate between the 7th and 8th of the other division. These last three 

 larvae were very similar, although their relative darkness is expressed by their order. 



7. This was the darkest of all the larvae in the 3rd stage. It was a yellowish- 



