THE EARLY STAGES OF THESTOR BALLUS. 239 



only on Lotus hispidas, but Kambur says the larva lives on 

 " Herbaceous Legiiiiiinosce," which would, I suppose, imply that 

 T. ballus feeds on all low-growing papilionaceous plants. The 

 statement, notwithstanding, seems somewhat loose. The same 

 writer says that the caterpillars " devour each other." 



Immediately on arrival, the eggs changed from their normal 

 colour to a dull blackish, and some larvae hatched out on the 

 same day, and continued to emerge daily for about a week. 

 These caterpillars, in captivity, have a tendency to wander from 

 their food-plants, which makes them rather difficult to rear ; and 

 for some unaccountable reason, even when on their plants, many 

 of them seem to die, while others do well. When very young, 

 the larva of hallus is blackish, and is thickly covered ivith rough 

 hairs ; spiracles black ; but it afterwards changes in colour, a 

 fact which Guenee, who described the species, seems to have 

 ovei'looked. 



The caterpillar which I am now describing was born on 

 March 29th last, and was kept from the first entirely in a corked 

 bottle, in a room with partially opened windows facing south, and 

 averaging about 65° Fahr. It was fed exclusively on Medicago 

 denticulata. 



On. April 22nd it changed its skin for the first time. During 

 the periods of moulting the larva remained quite stationary, 

 generiillj' at the bottom of the bottle, without eating. This 

 operation usually lasted for three days. 



Up to April 24th it only consumed the chlorophyl, eating 

 holes in the middle of the leaves. 



The second moult was undergone on May 1st, and the cater- 

 pillar now ate not only the chlorophyl of the plant, but also the 

 parenchyma itself, devouring the sides as well as the centre of 

 the leaves. 



On May 12 th it cast its skin for the third time. It would be 

 interesting to know whether these changes are accelerated in a 

 state of confinement, or vice versa. 



On the 27th ult. I transferred the larva, which now measured 

 exactly half an inch in length, to a cage ; but it was afterwards 

 returned to the forcing-glass; and on the 31st inst. it ceased 

 eating, and wandered restlessly about. At this time the cater- 

 pillar began to assume a pale pinkish colour ; it also emitted a 

 reddish-coloured fluid, and on the 6th of June it settled on a 

 piece of paper to pupate. On looking for the larva at 9 a.m. on 

 the morning of the 1 1th ult., I found that the transformation had 

 already taken place, although it was still in the caterpillar state 

 late on the previous evening. 



The time thus occupied, from the birth of the larva to the com- 

 pletion of the chrysalis, was seventy-four days ; the pupa hyber- 

 nates. The chrysalid of Thestor ballus is not attached at the 

 head and tail by silken threads, and the discarded skin of the 



