CELASTRINA ARGIOLUS. 455 



Wight, in August, 1895, that these pupated in due course, one pupa 

 producing an imago of the third brood, on September 30th, 1895, the 

 rest of the pupae going over until the following spring. Eaynor states 

 that some third brood imagines were reared, in September, 1906, from 

 larvae that pupated from August 28th onwards, at Hazeleigh, and that 

 he found fullfed larvae apparently of this brood from October 5th-7th. 

 Burrows also reared third, brood imagines on September 14th ( $ ), 

 September 15th (<? ), September 20th ( ? ), September 22nd (<? ), 1906, 

 from larvae taken at Mucking, and that pupated August 28th, 

 1906, onwards. These records show distinctly that, in very favourable 

 seasons, the species may be even partially triple-brooded. One suspects 

 the same in Central France, for Oldaker bred, on October 17th, 1907, 

 an imago from a pupa that was formed on September 17th, from a larva 

 taken a few days before at St. Cloud, the rest overwintering. Dennis 

 records that he found the species in all stages — egg, larva, pupa and imago 

 — at Earl's Colne, Essex, on October 9th, 1902. Burrows, however, further 

 notes that, even at Mucking, the species may be single-brooded, for, in 

 April, 1906, he bred imagines from larvae of April, 1905, none emerging 

 in the autumn of the latter year. Of its probable single-broodedness in 

 Ireland, Kane observes that he has not seen the second-brood in Ireland, 

 but, as the species has been met with at Killarney, and in Wicklow, in 

 early May, and in Ulster from the beginning of May to the beginning of 

 June, there appears to be no reason why, after an early, genial, spring, 

 an occasional second emergence should not take place in the autumn. 

 In the north of England, Forsythe states that the species occurs at 

 Witherslack by the end of May, but appears to be single-brooded there ; 

 similarly, Moss also thinks it is single-brooded in the Windermere and 

 Kendal districts ; in Sutton Park, in Warwickshire, Wainwright 

 believes it is only single- brooded, and Bree asserts that it certainly is so 

 at Allesley, yet, at Wolford, Wheeler says that both broods appeared in 

 1896. Tetley also observes that, although the spring brood was some- 

 times abundant in Montgomery, he never observed a specimen of the 

 second-brood there, and Arkle affirms it to be single-brooded in 

 Denbighshire. The occasional third brood examples obtained in 

 this country appear to be a memory of its continuous-broodedness in 

 its more southern haunts, for Walker says that imagines have been 

 taken on the Rock at Gibraltar, as early as January 12th, resulting 

 probably from the September- October (or even later) larvae and pupae 

 (as third or fourth brood) ; these January imagines would lay 

 eggs that would no doubt produce imagines, at latest, in March- April, 

 i.e., at about the same time as the over-wintering pupae disclose their 

 imagines on the French and Italian Riviera, and these, in turn, must 

 be the parents of the very abundant brood, that Walker says occurs at 

 Gibraltar in June and July. His actual dates for Gibraltar are : 

 January 12th, 28th, 1888, near San Roque ; March, 1887, June 23rd, 

 1887, July 6th, 1887, and on throughout the month, July 14th to 

 August, 1888, and so on. The same observer also records the species 

 as common throughout the summer, 1878, at Port Baklar, near 

 Gallipoli ; his dates are: April 28th, June 2nd-14th, July 20th (quite 

 a fresh brood), whilst de la Garde says that it certainly occurs until 

 October in Corfu; W T alker also records it at Vigo in N.W. Spain, 

 September 11th, 1878. Meade- Waldo's records for Morocco are not 

 unlike those of Walker for Gibraltar, for he says that it first 

 appears in February, then at intervals throughout the summer, e.g., 



