416 BRITISH LEPIDOPTERA. 



— setli, anthyllidis ( triptolemus ) , bellidis (bellis), loti (achilleae), exulans, 

 cynarae, sarpedon. (8) Mesembrynus — brizae, scabiosae, punctum, pluto, 

 initios, erythrus, rubicundus. We are quite prepared to believe that our 

 British species are members of at least four different natural genera. 

 Bow ell has shown that the scales of A. purpuralis (inlnos) are more 

 generalised than those of the A. Jilipendidae-lonicerae group, and it is 

 well-known that A. exulans shows very marked characters in its pupa 

 and cocoon. If our British species were to be subdivided, we should 

 have the following groups : — 



1. — Mesembrynus — purpuralis (=minos, pilosellae). 



2. — Lycastes — exulans. 



3. — Thermoph ila — viciae (meliloti). 



4. — Anthrocera — lonicerae, paludis (trifolii-major), trifolii-minor, hippocrepidis, 

 filipendulae. 



Anthrocera is a most attractive genus of moths, and to one who 

 has studied the extra-British species, offers much of interest. The 

 species, popularly known as Burnet moths, have usually dark green 

 or dark purplish-blue fore-wings, with crimson spots and crimson 

 hind-wings. The males inclining to the purple, the females to the 

 greenish, hue. The interest is partially due to the generalised nature 

 of the genus, exhibited particularly in the pupa and in the neuration 

 of the imago, but still more so to the more or less undefined character 

 of some of the species, which, distinct enough in habits, habitat and 

 time of appearance, yet present many difficulties in the determination 

 of the imagines, which also, in some instances, hybridise very freely. 



When one has a short series of any one of these species, he is quite 

 prepared to give a confident opinion as to that particular species. As 

 the number of specimens examined increases, the opinion becomes less 

 confident, and when the number becomes still more largely increased, 

 the opinion is never stated as being more than probable. Whenever 

 we express any opinion as to these species, we wish it to be understood 

 that it must generally be considered as only a probable one. 



The Anthrocerid egg is ovoid, oblong, pale yellow or greenish in 

 colour, very shiny, with one pole transparent, generally smooth, but 

 rarely very faintly reticulated, the eggs of the various speices being 

 very similar. The egg stage rarely lasts more than fourteen or fifteen 

 days. 



The newly-hatched larva is very pale, and those of different species 

 are very similar. They all have tubercles i and ii (Dyar's nomenclature) 

 arranged as anterior and posterior trapezoidals on both thoracic and 

 abdominal segments, and each bears a single hair ; iii is placed as a 

 supraspiracular, iv and v being both placed as subspiraculars. After 

 the first moult these simple single-hair-bearing tubercles are modified, 

 a swollen base uniting i and ii on either side, and bearing several 

 single-haired warts ; iv and v are similarly united into a common 

 subspiracular mass, whilst iii is also of a more complex structure. 

 With the exception of changes in the depth of tint, the larva is 

 structurally identical from this time with the adult form. The species 

 hybernate as larvae, moderately small, usually feeding a little through 

 August and September, and commencing to feed again in spring. 

 Boisduval states that they only feed for a fortnight before becoming 

 torpid, but this is not so with many species — A. lonicerae reaches its 

 fourth skin, A. viciae (meliloti), A.filipendulae and A. carniolica, the third, 

 before hybernating. All these feed some 8-10 weeks before hyberna- 



