DEIDAMIA INSCRIPTA 93 



rough with black granules. It was depressed or bent 

 backward till nearly horizontal. 



The third molt followed in four days, and then the 

 caterpillars were an inch and a quarter long, with the 

 third and fourth segments slightly thickened. The 

 head was bright green, with two bright yellow face- 

 lines, continued as subdorsals over the bright green 

 body. The venter was bluer green than the rest of the 

 bod}', and not dotted. The body had yellow trans- 

 verse lines instead of dots, the lines being broken in 

 the stigmatal region. The horn was short, almost 

 triangular, greenish at base, rough with brown gran- 

 ules in the middle, and yellow at tip and on the sides, 

 up which extended the seventh pair of obliques. The 

 anal plate was edged with yellow. The spiracles were 

 conspicuous for the first time, being white with a blue- 

 black crescent on each side. The other lines were the 

 same as before. 



The caterpillars ate and grew for five days, and then 

 surprised us by turning pink on the dorsum, losing 

 their grasp of the stems, and stopping eating. All 

 this meant approaching pupation ; but they had not 

 molted the regulation " four times," though they were 

 about two and a quarter inches in length. 



On looking over three years' records of each box we 

 found several variations. Some had but seven obliques, 

 while the rest had eight on each side. Some had no 

 yellow dots below the stigmatal line, while others were 

 dotted almost to the ventral line. Some caudal horns 

 were marked up the front and down the back with a 

 continuation of the dark dorsal lino ; some had no green 

 at base, and two were wholly yellow; a few were pink 



