482 BRITISH LEPIDOPTERA. 



Speyer's evidence tends to show (Stett. Ent. Zeit., xxxviii., pp. 40 

 et seq.) that the July A. trij'olii, in north-western Germany, inhabits 

 marshes, and also suggests that the marsh insect emerges over a long 

 period of time, and that a little specially advantageous or disadvan- 

 tageous environment might readily make it an early or late insect, 

 i.e., that local considerations determine whether the insect in a given 

 place shall be of the early or late form. Speyer captured two speci- 

 mens of ab. trivittata on July 7th and 10th respectively. He further 

 says that in his district the insect generally commences to emerge 

 about the middle of June, the period of emergence lasting from four 

 to six weeks. In a late year, 1876, he states that the first imago did not 

 put in an appearance until June 28th, larva? being found as late as 

 June 21st. One could have wished that Speyer had given some exact 

 dates as to the relative appearance of the " large " and " small " forms 

 of A. trifolii, and a detailed difference as to their habitat, for that 

 both forms occur, and are well known in Germany, is certain from 

 his remark that a specimen of ab. trivittata, described by him, 

 " belongs to the small- winged form of this, in other ways very variable, 

 species, and is somewhat under the average size." 



Even in Britain there is no sharp line of demarcation in the time of 

 appearance (nor in the size) of what we have termed trifolii-minor and 

 pahistris (trifolii-viajor), for, in some places, the small race is regularly 

 later in its time of appearance than in others, e.g., the small form found 

 in the New Forest is regularly later than that in Kent (Canterbury, 

 Strood, etc.). Fletcher (including both forms in his generalisation) 

 says that there seems to be a constant succession of specimens 

 from different colonies from May until the beginning of August. 

 At Emsworth on the downs, and about Abbott's Wood in wet 

 meadows on the clay, it is out early in June ; in the New Forest 

 it flies early in July, at the same time as A. viciae (meliloti). Fletcher 

 found a colony in a heath bog not far from Worthing, fully 

 out on June 30th and July 1st, 1897, and bred specimens from a 

 score of pupa? up to July 18th. These last dates reach up to the 

 appearances that have been recorded for July and August. Oberthiir 

 remarks also on the variation of the time of appearance of this species, 

 and connects it with the various races (vide, p. 481). He observes that 

 in July, 1895, A. trifolii was still fresh at Biarritz more than a month 

 after it had gone over at Uriage and at Kennes. In August, 1883, in 

 the Isles of Chausey, in front of Granville (Manche), A. trifolii was 

 still flying, two months later than the appearance of the same species 

 in Ille-et-Vilaine. As bearing on this point, we suspect that the 

 difference in the time of appearance is due to a tendency for this insect 

 to become a two-year or one-year species, in some localities, according 

 to its environment, and the evidence seems to suggest, not only in this 

 species, but also in A. Jilipendulae, that a marshy habitat tends to a 

 later appearance, yet it is marvellous that some of the mildest parts of 

 the British Islands — Freshwater, Swanage, etc. — produce the latest 

 broods. Usually, the latest broods (presumably the larva? having 

 taken longer to feed) are the larger, but the Bennes form, as large as 

 our largest British form, appears in May and June, whilst smaller 

 insects occur in August, on that part of the French coast nearest the 

 Channel Islands. On these islands, too, there are two races, an early 

 form occurring in June, on the Guernsey cliffs and in Sark, and a later 



