on colour -relation. 239 



meanwhile the third ecdysis of the second, third, and 

 fourth larvpG occurred. In all three, red spots were found 

 in the spiracular row only, exactly as in the first larva, 

 both as to number and position ; the red spot on the 

 supra-anal plate was also present. The change in the 

 horn was also exactly as in the first larva. 



Sept. 3rd.— The fourth ecdysis of the first larva 

 occurred. The ground colour was bright yellowish- 

 green ; the stripes and shagreen dots pale primrose- 

 yellow. The character of the spiracular spots was 

 slightly changed ; whereas in the fourth stage the spi- 

 racle itself only showed as an orange- red spot outlined 

 with a deeper green than the body-colour (PL XI., fig, 9) ; 

 each spiracle was now distinctly margined with orange- 

 red, the red area having increased, but being still out- 

 lined with deep green (PI. XL, fig. 10). The most 

 striking change was the development of an upper row of 

 large bold red spots, seven in number, on abdominal 

 segments 1 — 7, one anterior to each stripe, the second 

 largest being those in front of the third, fourth, and fifth 

 stripes, counting the most anterior stripe as the first. 

 The last spot was the largest, those in front of the first 

 and second stripes much smaller, but distinct ; the spot 

 before the sixth stripe was a mere trace. The oblique 

 stripes now took a slight upward bend at the places 

 where the upper spots occurred, and the third, fourth, 

 and fifth stripes were here slightly suffused by the spot, 

 on which the shagreen dots showed up as on a back- 

 ground. The spots showed a slight tendency to lengthen 

 vertically, but kept strictly to the lines of the rings of 

 the body, of which there are eight in each segment. 

 The spots in front of the third, fourth, and fifth stripes 

 covered three rings in width, the last spot four rings, 

 the first and second spots only two rings. The two 

 sides of the larva corresponded exactly. The black 

 dotting of the caudal horn entirely disappeared ; im- 

 mediately after ecdysis the upper surface was pure green, 

 the under surface was yellow, and the red line up the 

 sides was longer and more distinct. The length of the 

 larva was nearly 4|^ cm. 



Sept. 4th.— The horn of the first larva changed to 

 bluish on the upper surface. 



Sept. 5th. — Having been absent on the day the second, 



TRANS. ENT. SOC. LOND. 1892. — PART III. (NOV.) T 



