338 BRITISH LHPIDOPTERA. 



to contain 160 well-developed eggs; these were carefully extracted. 

 On September 26th two larvae appeared from the eggs laid between 

 September 8th-i2th, an oval period of something less than three weeks.* 

 None of the other eggs obtained produced larvae. From these 

 facts D'Orville came to the conclusion that the imago does not 

 emerge from the pupa with oval fully developed j, but in a very 

 unformed state, and that they become gradually matured in the body 

 of the female — perhaps after impregnation has taken place. Hellins 

 adds that, as the egg of A. convolvuli is not more than two-fifths of 

 the size of the egg of S. ligustri, a $ with her full num- 

 ber (somewhat between 200 and 250) ready for extrusion, 

 would by no means show so stout a figure as a 2 S. 

 lignstri in similar circumstances. Pode obtained eggs from a 2 

 captured at Slade, near Ivybridge, on August 26th, 1887. The 2 

 was enclosed in a large box and two eggs were laid on the evening 

 of August 27th (these were preserved) and 18 on the 29th, all upon the 

 paper bottom of the box, near together, but not in contact, being scat- 

 tered over an area 18mm. long by 10mm. wide. The eggs were sent to 

 Poulton who records that the first larva emerged on the evening of Sep- 

 tember 5th, the second on the morning of September 7th ; on the 8th, 

 examination showed that the larvae were fully formed in the remaining 

 ova, which appeared somewhat wrinkled. Being afraid that the larvae 

 would die before hatching, he placed the ova in a room with a fire in 

 order that the moderate warmth might hasten emergence. By the 

 next morning (9th) eleven larvae had emerged and the remaining 

 four appeared a few hours later, in the afternoon, thus making 

 seventeen larvae altogether. One of the eggs did not undergo 

 development, but subsequently changed colour and dried up. The 

 length of time occupied in development within the ovum was as 

 follows : 1 larva — 7 days ; 1 larva — 8-J days ; 1 1 larvae rather over 

 10 days; 4 larvae rather under 11 days. Hence the development 

 is at about the same rate as that of 6". ligustri, which occupied 

 8 — 10 days. Hellins, as we have noted above, on D'Orville's 

 authority, puts the time of development for A. convolvuli at rather 

 less than 3 weeks.* This Poulton considers is probably a mistake 

 (Trans. Ent. Soc. London, 1888, pp. 516-517). As bearing out 

 the idea of a shorter oval period, one notes that a 2 , captured on 

 September 14th, 1897, in Piccadilly, laid 30 eggs, of which 13 

 hatched on September 21st, after an oval stage of 7 days, although 

 kept in a cold room (Bell-Marley, Ent., xxxi., p. 68); whilst another 2 , 

 taken on July 18th, 1898, at Brighton, had, by the 20th, deposited 25 

 eggs in confinement on Convolvulus arvensis, and laid 8 more on the 

 21st; the moth died on the 22nd and dissection of the body disclosed 

 some hundreds of eggs, of a bright emerald-green colour and much 

 smaller than those of Sphinx ligustri ; the eggs hatched on July 

 27th and 28th, so that the egg stage lasted only 7 days (Brazenor, 



* In 1859, D'Orville states that a ? in his possession laid a single egg, 

 and that this produced a larva on the 12th day, at Alpliington (Zool., p. 6818). 



t There is no doubt that this is the explanation of Newman's statement that 

 most of the ? s of this species that appear in the autumn are barren [Ent., viii,. p. 

 275). A high temperature and food appear to be absolutely necessary for the 

 development of the ova. Syme dissected a ? bred September 15th, i860 (E.M.M., 

 v., p. 139), and found the eggs in a very undeveloped state, and had no doubt that 

 the ? s wanted feeding before the ova would be developed. 



