86 



4, i88o, which he had found the day before in Wicken Fen, 

 and on the 12th a few more, laid on Peucedanum palustre, and 

 eventually three of these proved infertile. 



Of course what follows is really the personal history of the 

 individuals which I watched, and though for convenience sake I 

 shall generalize, and sometimes use the present tense and not the 

 past, I wish it to be understood that I speak only of what I was 

 aware I saw; I know I made one omission, which will be noticed 

 in its proper place. 



The eggs hatched June I3th-I5th, the larvae in every case 

 making their first meal on the empt)- shell, and for a day or two 

 I supplied them with garden carrot, but after that they were fed 

 entirely on Angelica sylvestris; from first to last each larva was 

 kept separate, and its changes noted in a separated record. 



The larva, on first turning its attention to its food plant, 

 scoops out a round cell on the surface of a leaf, but after a few 

 hours takes the bolder course of eating quite through from the 

 edge of the leaf; it does not roam, but continues at the same part 

 till the third or fourth day, when it moves off to some distance 

 and, on a stalk or leaf, spins a few silk threads for a foothold; there 

 it waits from two to three days for the first moult, and when this 

 is accomplished eats the cast skin all to the head-piece, and soon 

 after goes — apparently by design — back to the spot where it was 

 previously feeding, and attacks the leaf again: at this stage I no- 

 ticed if a larva found a speck of grass on its food, it would pick it 

 up in its jaws, stretch out its body, and somehow project the grass 

 away from the plant: again, after feeding three or four days it re- 

 tires as before, and prepares for and accomplishes its second moult, 

 which happens on about the twelfth day of its life ; similarly the 

 third moult comes on the sixteenth or seventeenth day, and the 

 fourth (the last) from the twentieth to the twenty-third day, the 

 cast skin being always eaten ; after the last moult the larva feeds 

 on for ten or twelve days, consuming a great quantity of food and 

 making very rapid growth. I may here note that its usual attitude 

 in repose is from the very first much like that of a Sphinx, with 

 the neck arched, and the head bent down. The earhest age at 

 which I noticed the curious horns of the second segment was when 

 I touched the \d.xv2.jnst after its third moult\ they were then much 

 longer and thinner than they became after the fourth moult; but 

 there accompanied their protrusion a drop or two of clear greenish 

 liquid, and a most penetrating odor, which reminded me of an 

 overkept decaying pine-apple : after the fourth moult the horns 

 were of a shorter and stouter character, but I noticed that when I 

 was holding a larva between my finger and thumb it had the power 

 to lengthen one horn at the expense of the other (which became 

 shorter), so as to manage to touch my finger with it ; the horns are 

 extremely soft and flexible. When full grown the larva ceases 

 feeding, and rests for a while, and then commences its prepara- 



