BRITISH LEPIDOPTERA. 



irly, 



nts: 



which is attached to the prothorax. The legs separate very irregularly, 

 but all the parts remain attached together by the internal dissepiments ; 

 even the thoracic coverings separate, to some extent, by the stretching 

 of the intersegmental membranes. The spiracles are partially closed 

 by very fine dendritic processes surrounding the orifice. 



FOOD-PLANTS. Trifolinm medicayo, Lotus corniculatus. [Also re- 

 puted to feed on Spiraea filipendula (Linne), Plantayo, Taraxacum, 

 Veronica, Hieracium pilosella, Btiza media (Kaltenbach), Leontodon, 

 Hypericum (Snellen).] 



PABASITES. Campoplex decipiens, Gr., Cryptus filipendulae, Bore, 

 Cryptus fwnipennis, Gr. (Perkins), Mestogtetttu obnoxius, Gr. (Bennett, 

 Bignell, etc.), Heiniteles fureattu, Tasch. (Bignell), //. fulripes, Gr. 

 (Bignell), Anomalon tenuitarsum, .Gr. (Weston), Lihoyas tricolor, Spin. 

 (Jenkins), Apanteles zygoma/nan, Marsh. (July 21st, 1885, Bignell), 

 A. diffidlis, Nees von' Esenbeck (July 21st, 1885, Bignell), A. 

 juniperatae (Bignell), Macrocentrus linearis, Hal. = M. abdominal^, 

 Fall. (Bignell), Exorista vulyaris, Fall. (Bignell), and Tachina larva nun, 

 Linne (Bignell). Besides these, Hemimachus instabilis, Forst. = 

 H. rufocinctus, Gr. (bred July 10th, 1885, Bignell), and PezomacJnts 

 analis, Forst. (Grigg), have been bred as hyperparasites upon Apanteles 

 zyrjaenarum. The cocoons of Mestostenus obnoxius are found within the 

 cocoons of Anthrocera filipendulae in winter, the imagines emerge 

 during the last week of May or first week in June, one parasite only 

 to each cocoon (Watkins) ; the imago of M. obnoxius does not appear 

 until the larvas of A. filipendulae are nearly mature, emerging from 

 June 19th-July 3rd (Bignell). These two observations give a period 

 extending from May to July for the emergence of this parasite. 

 Anthrocerid cocoons, said to contain living pupas that go over the 

 winter, have probably been parasitised by this or an allied species. 



HABITS AND HABITAT. This species appears to be able to accommo- 

 date itself to almost every possible kind of habitat. Waste ground, 

 hillsides, downs, sloping cliffs near the sea, coast sandhills, marshes, 

 and even fenland are recorded as its haunts in Britain. It has the 

 widest distribution of all our British species in these islands, extending 

 from Sutherland to Cornwall, and the extreme west of Ireland, and 

 appears to be more or less abundant in most of our English and Irish 

 counties, less so, however, in Scotland. Sometimes one slope of a 

 hill will produce the species more abundantly than another, <>.//., on 

 the south escarpment of the chalk-hills, at Guildford, the species 

 abounds, whilst on the other slopes it is comparatively rare (Groverj ; 

 the stretches of shingle on the Sussex coast are a favourite haunt 

 (Fletcher), whilst the marshy sides of the ditches scattered over the 

 sandhills near Deal produce it in abundance (Tutt). Hodgkinson records 

 it as abundant in hay-fields all over north Lancashire, whilst Robson 

 says that it is particularly abundant along the coast of Northumberland 

 and Durham, although uncertain in its appearance, abounding one 

 year and very rare the next, and Fenn notes a similar uncertainty in 

 its appearance, the insect abounding at St. Margaret's, in 1889, very 

 rare in 1890. On August 5th, 1891, their cocoons were so common 

 on the stems of rough grass on Helpston Heath (nr. Peterborough) 

 that they made quite little white patches in places when seen from a 

 distance (Morley). Clarke notes it as abundant on the coast, and 

 widely distributed throughout the inland districts of the Isle of Man, 



