AGRIUS CONVOLVULI. 355 



obtained (omitting the 9th in the division of 10 pale larvae): 





BLUISH-GREEN. 



INTERMEDIATE. 



YELLOWISH - 



GREEN. 



The 10 pale larvae 



1 . 



2 / 



3 

 10 / 



- 4 |i] 



— — 



M\ 



The 7 dark larvae 



I 



! 2 l _ _ 

 Is) 



4 ( 6 ( 



7 



Fourth stadium* : (I) The six lightest and bluest lan>ae of the 

 3rd stage. — Of these larvae three were large (about 38mm. long) and 

 three small, about 30mm. They are arranged in the order of 

 relative- darkness, the lightest larva being described first : 



f. The lightest and greenest larva was one of the small ones mentioned 

 above ; it was painted October 10th, and is represented in pi. xvi., fig. 1 ( x 2 ). The 

 arrangement of the dark markings and relation to the green ground-colour is 

 sufficiently indicated in the figure. This was the lightest of all the larvae in the 

 4th stage, and it is seen to be transitional from the darker larvae of the last stage, 

 while it is also connected with the darker larvae of this stage by a very complete 

 series of gradations, the chief of which are indicated in figs. I, 2, 3, 4, 5, of 

 plate xvi. The larvae were still covered with shagreen-tubercles as in Smerinthus 

 and Sphinx ligustri ; these are not shown in the figures. 



2. All the five remaining larvae were dark varieties; the least dark of 

 these was 2, one of the small larvae, which possessed distinct stripes and 

 borders, together with faint traces of the green ground-colour, while the sub- 

 dorsal became distinct in its anterior part. This larva is represented in plate 

 xvi., fig. 3 ( x 2), and it is seen to afford a beautiful transition towards the darker larvae. 



3. One of the larger larvae possessed a very distinct subdorsal line, and 

 a fairly distinct subspiracular, both these lines remaining light and contrasting 

 strongly with the dark larval surface. The red spots could be made out plainly, 

 but they were very dark, almost black. 



4. Another of the large larvae was similar, only the two longitudinal lines 

 (subdorsal and subspiracular) were not so bright and distinct. 



5. The third of the larger larvae, in which the stripes and red spots were so 

 dark that they could hardly be distinguished from the general surface of the 

 larva. The two longitudinal lines were distinct, and of a bright yellow colour, 

 as in plate xvi., fig. 4, which represents one of the larvae of division III. 



6. The third of the smaller larvae- was much like the last only still darker, 

 although the stripes were much more distinct. 



The difference between the 6th and the other larvae decreased, 

 and that between the 2nd and the others increased, subsequently, 

 for the dark colour becomes rather less deep during growth. The 

 real line of demarcation was, however, between the first larva and 

 all the rest ; this great difference dwarfed all the minor differences 

 between the latter. The horn in this stage is slightly curved, and 

 possesses the shape shown in pi. xvi., figs. 1-5. It will be noticed 

 that the curve is now the reverse of that seen in the 1st stage. 

 The surface is shining, although still rough, with comparatively small 

 tubercles, which represent the far larger thorns present in the two previous 

 stages. The colour is black, with a dark reddish-brown patch on 

 each side of the base, especially bright in the green variety (1), and 

 hardly present at all in the case of 4 and 5. There was a very faint 



* On October 8th all the larvae were carefully compared except the single one 

 in division III, which was still in the 3rd stadium, 



