AGRIUS CONVOLVULI. 343 



tit., xxxiv., p. 282) that, in 1901, when he found several larvae in the 

 neighbourhood of Dovercourt, in potato fields, choked in places with 

 masses of Convolvulus arvensis, Polygonum aviculare and Chenopodium, 

 the larvas often climbed the potato stems to rest, and he adds that the 

 larvae have, OAving to this habit elsewhere, occasionally been mistaken for, 

 and recorded as, larvae of Manduca atropos. The nearness of the larvae can 

 usually be discovered by their frass, which is rather elongated and small- 

 er at one end than the other, and thus differs from that of M. atropos 

 which is square or brick-shaped ; moreover, the frass of the latter lies 

 in a mass under the plant upon which the larva has been feeding, 

 whilst that of the former is found at intervals upon the ground ; by 

 following this the larva can be traced. Brazenor notes (Ent., xxxii., p. 

 16) that the larvae of A. convolvuli bear no resemblance in form, colour 

 or behaviour to those of Sphinx ligustri. Contrary to the action of the 

 latter, the larva of A. convolvuli would, on being touched, or even 

 the food shaken, turn the head sharply round until it met the last 

 segment and then drop from the plant. Mathew also further notes 

 (Ent., xxxiv., p. 283) that the larva, which is hard-feeling to the 

 touch, when annoyed, curls itself into a ring or violently jerks itself 

 from side to side. Bell-Marley observes (Ent., xxxi., p. 67) that the 

 young larvae spin long webs over their food, and that, after the 1st moult, 

 the larvae resemble those of S. ligustri in miniature, only the stripes 

 are very much paler. Hatching on September 14th, 1897, the moults 

 of the larvae in his possession occurred on October 14th, October 27th, 

 November 8th, two larvae finally going underground on December 1 2th 

 and pupating on December 16th, without having undergone a 4th 

 moult and having taking 86 days from the time of hatching to pupa- 

 tion. Only one pupa appears to have resulted. Comparatively few 

 larvae (or pupae) have been taken at large in the British Islands. The 

 following are the records that we have so far observed* : October 14th, 

 1859, nearly fullfed at Alphington (D ; Orville) ; October 5th, 1859, near 

 Freshwater (Zool., p. 6788); October 8th, 1859, at Penzance (Hay ward); 

 5 pupae were found at Chatteris in early September, 1859 (Fryer) ; 

 several larvae in 1859 in the Carlisle district (Armstrong); in 1859, 

 several pupae dug up in potato fields in the South of England (Newman); 

 3 larvae from Deal, on September 7th, 1859, imago emerged September 

 15th, i860")" (Syme, Zool., pp. 6788, 7269) ; 7 larvae, taken on small bind- 



* Apparently the earliest record of rearing the imago from the larva in Britain 

 is that by JDavies [Zool. Journ., v., p. 142 (1830)], who states that he obtained larvae 

 which in due course became pupae, Irom which he reared 2 % s for the collection of the 

 Portsmouth Philosophical Society, several $ s assembling to the ? s on their 

 emergence, and entering the room in order to reach them. 



t The only instance that we know of a larva pupating in autumn in Britain and, 

 going over the winter as pupa, producing an imago normally the next year, is the 

 following : " Air. Syme exhibited a ? specimen at the meeting of the Ent. Soc. ot 

 London, on October 1st, i860, which had emerged from pupa on the 15th ult., 

 and had been produced from a larva found in a potato field at Deal, in the autumn 

 of last year, and had remained nearly a year in the pupal state " (Zool,, p. 7269). 

 This record must be read in connection with a note by Hellins (E.M.M., 

 vii., 139), which states that " Syme had three larvas brought him on September 7th, 

 1859, that they went down on the 12th and 13th of the month, two producing 

 perfect pupae, one of which produced an imago, as far as Syme can recollect, about 

 the end of May, i860, the pupa having- been kept in a room without a fire in 

 it during the winter." We do not understand the difference in these two accounts 

 of the rearing of what is evidently the same moth. Mathew notes (in litt.J that a 

 pupa, which assumed this stage in the autumn of 1901, was turned out of its cocoon 

 alive in April, 1903, but that it died directly afterwards. 



