THE HAY SOCIETY INSTITUTED MDCCCXLIV. This volume is issued to the Subscribers to the RAY SOCIETY for the Year 1889. LONDON: MDCCCXC. A MONOGRAPH OF THE BRITISH PHYTOPHAGOUS HYMENOPTERA, (TENTHREDO, SIREX AND CYNIPS, Linne.) VOL. III. BY PETER CAMERON. LONDON; PRINTED TOR THE RAY SOCIETY. JklDCCCXC, LIBRARY UNIVERSITY OF CALIFORNIA DAVIS PRINTED BY ADLABD AND SON, BAliTHOLOMEW CLOSE, PBEFACE. IN this volume I have described the remaining three families of the Tenthredinidse, the Cephidae, Siricidae, Oryssidse, and the parasitic Cynipidse. The fourth (and concluding) volume will deal with the inquiline and gall-making Cynipidaa, and will also contain such additions and corrections as increased knowledge may have brought to light during the period the book has been in course of publication. It only remains for me to thank my correspondents for their continued kindness in assisting me with my work. SALE, CHESHIRE; April, 1890. A MONOGRAPH OF THE BEITISH PHYTOPHAGOUS HYMENOPTERA, VOL. III. TENTHREDINID^E. Tribe CIMBICINA. Antennae 6 — 8-jointed, short, the third joint much longer than the rest, the last, the last two, or the last three forming a well-defined club, which is clearly separated from the preceding joints. Wings with two radial and three cubital cellules ; the first cubital receives both or only one of the recurrent nervures (Amasis). Transverse basal nervure received considerably before the stigma; the transverse median is usually received very near to, or is quite joined to it. Posterior wings with two median cellules. The accessory nervure is continued to the edge of the wings, and is united to the anal by a short transverse ner- vure. Head broad and swollen behind, much narrower in front, so that the eyes are generally placed more in front than at the sides ; the eyes are small and oval as a rule. Pronotum with a semicircular emargination on its hind border. Episternum of mesothorax large, and separated from the mesosternum by a deep groove, which is directed rather to the front than very obliquely downwards. Prosternum large. Abdomen broad, rounded behind, its sides acute, the back arched, the ventral surface flat, or semi- concave. Mandibles large, deeply 2- or 3-dentate. Saw scarcely, if at all, projecting, broad, the apex curved, round, and the sides bear little dentated bead-like projectings. Patellee well developed ; spurs often membranous at apex. The clypeus is either moderately large and the labium small, as in Cimbex, or it may be small and the latter larger and moveable, as in Clavellaria. The mandibles are large, especially in the males of Trichio- VOL. in. 1 2 TEIBK OIMBIC1NA. soma and Clavellaria. In the larger species (Cimbex, Trichiosoma, Clavellaria) the eyes are small, and are placed well in front, so that the head projects consider- ably behind them ; in the smaller species (Abia, Amasis, and the Brazilian genera Pacliylosticta and Plagiocera) they are longer and reach nearer to the base of the mandibles, while it is only in Abia and Zarcea that the head projects behind them. The other genera have them situated on the side. In the males of Abia they become nearly confluent at the top of the head. The thorax is broad and thick ; the sides project so that they are somewhat broader than at the top, the mesopleuraB being slightly hollowed, and separated by a more or less clearly defined border from the sternal region. The pronotum is almost horizontal in front, and has the head closely pressed to it ; it has a well-defined border above, and from this runs an oblique emargination towards the head, separating the somewhat projecting spiracular bearing part in front of the episternum from the upper portion. The form of the episternum of the mesothorax is characteristic through the deep groove which separates it from the mesosternum, as well as from not being so obliquely situated as in the other saw-flies. The mesonotum rises gradually from the pronotum to the scutellum, which is broader than long, rounded at the top, is raised above the mesonotum, and late- rally there runs from it to the tegulas a well-marked ridge. In front there is a deep hollow, while behind the post-scutellum is almost obliterated, so that it (the scutellum) presses closely on the metathorax, which is raised and semicircular in front and slopes from above downwards, a deep groove being thus formed between it and the scutellum, the cenchri being placed on the edge of the ridge so formed. The first cubital cellule is as long, if not longer, than the second. The stigma is linear, and in the Euro- pean genera is not much differentiated from the costa, but in the Neotropical genus Pachylosticta it is broad, TRIBE CIMBICINA. 3 long, and projecting. The transverse median nervure being either united to the transverse basal nervure, or received quite close to it, is a well-marked pecu- liarity of the family. So also is the structure of the accessory nervure in the posterior wings. In no other European sub-family does it run to the edge of the wings without being united directly to the anal ner- vure, nor with the other tribes is it united to the latter by a cross nervure. The larvae are twenty-two-footed, cylindrical in shape, usually greenish in colour, and dusted over (especially when young) with a white exudation. Over the spirales are situated orifices, from which are ejected, often to some distance, a greenish, acid liquid, which contains chlorophyll. They are, with one exception (Abia sericea), so far as is known, attached to trees or bushes, and spin a double cocoon, either to the branch of the tree on which they fed, or in the earth. When they quit the cocoon they do not eat away the one end entirely, but leave it hanging by one side (see Yol. II, Plate XII, fig. 5) as a lid. The tribe must be regarded as a Palaearctic and Nearctic one, many of the species, indeed, penetrating very far north (Siberia, Labrador, Hudson's Bay) ; but, as already mentioned, two genera are found in Brazil, and species of Cimbex, &c., occur in the West Indies (Hayti), South America (Brazil, Argentine Republic;, and in the Old World in Japan and North India. This tribe contains the largest and most bulky species of Tenthredinidde. As above defined, it con- tains the Palsearctic and Nearctic genera Cimbex, Trichiosoma, Praia, Clavellaria, Abia, Zar&a, and Amasis, and the Neotropical genera Pachylosticta and Plagiocera. The Australian genus Perga is included in the sub-family by many authors ; but while Perga has undoubtedly some affinity with the Cimbicina^ it yet possesses sufficient distinguishing points to warrant 4 SUB-TRIBE CIMBIOTDES. its being raised to sub-family rank. But whether it should form a family by itself or one along with Syzgonia, Laboceras, and Cephalocera (which have also been placed in the Cimbicina, or at any rate the two former), the fact of its having the radial cell appendicu- lated, there being also only one radial cellule, in its wanting the lanceolate cellule, and having the tibiae spined, the form of the saw being likewise different, are sufficient grounds for ejecting both Perga and Syzgonia (this genus having also these peculiarities) from the Cimbicina. The American genera Thulia and Acordulcera are by some authors united to the Cimbi- cina, but they also may be excluded, as they have only one radial cellule. That, however, the Gimbicides are more closely allied to Perga and Syzgonia than to the Tenthre- dina, there can, I think, be no doubt; this is shown by the similarity of the form of the antennae and thorax and by the neuration, especially of the form of the accessory nervure in the posterior wings, which is nearly the same in the three groups, only in Perga, &c., the accessory nervure is apt to become obliterated entirely. We may with Thomson divide the European species into two sub-tribes — the Cimbicides and Abiides. Pachylosticta and Plagiocera will form another sub- tribe allied to the latter rather than to the former. Sub-tribe CIMBICIDES. Body large (6—13 lines), hairy or pilose. Lanceolate cellule divided by a straight cross-nervure ; first cubital cellule receiving both recur- rent nervures. Mandibles large ; head swollen at the sides, projecting beyond the eyes, which are small, parallel, and situated well in front. Ualcana short, thick, membranous at the apex. Patellae well developed, somewhat membranous. Palpi short, thick, the last three ioints stout short. The larvae are well marked with coloured spots or markings, and are green or blue, or more rarely reddish. GENUS CIMBEX. 5 They are usually dusted over with a white exudation and feed on willows, birch or alder. Synopsis of Species. 1 (2) Blotch large ; labrum small ; clypeus incised, larger than labrum. Cimbex . 2 (1) Blotch abaent; labrum large; clypeus smaller than it. 3 (4) Posterior coxae widely separated; antennas with five joints before the club; posterior femora toothed; labrum and clypeus black. Trichiosoma. 4 (3) Posterior coxae continuous ; antennae with four joints before the club ; posterior femora simple ; labrum and clypeus white. Clavellaria. Genus — CIMBEX. Crabro, Geoff., Hist. Ins., ii, 261 (1762) ; Fourc., E. P., ii, 361, nee Fab., Auct. Cimbex, Oliv., Enc. Meth. Ins., iv, 22 (1789). Blotch large, wide. Clypeus with a very small emargination ; labium small. Antennae 6-jointed. Posterior coxae widely separated, and with- out teeth ; anterior more oval, closely continuous. The antennae are, I consider, six-jointed. The basal two joints are small, the third is nearly as long as the succeeding together, the fourth and fifth subequal. The club is without any trace of segmentation, or there may be one or even two strictures. Whether these actually represent joints I am not at all clear. In one or two specimens I have, the club has every appearance of being comprised of three joints ; but, on the other hand (and I think more often than not), the club is usually quite homogeneous. If the club is to be regarded as comprised of three joints that would make the antennas eight- jointed. The lateral sutures on the vertex are broad and deep, and they extend round behind the ocelli, where, however, the suture is not so deep, especially in the middle. A short, broad, longitudinal suture runs from it to the ocellus, which is situated in a deep pit. A suture runs from the lateral ocelli to the antennas. The front is thick and projects considerably from the (3 GENUS CIMBEX. sides. The eyes are placed at the side of this pro- jecting portion of the head. The mandibles are large, and strongly built ; the apical tooth is large, long, and sharp ; there are two sub- apical teeth. The labium is very nearly equally lobed ; the first joint of labial palpi is small, half the length of the second, and not much shorter than the fourth ; third very nearly as long as the other three together. Joints one, two, and four of maxillary palpi are sub- equal in length ; the third is as long as these three together, fifth longer than sixth, and nearly as long as the third. The first radial cellule is longer than the second ; they are long compared to the breadth. The transverse radial nervure is a little oblique. The first cubital cellule is long, narrow, of nearly equal width through- out, and longer than the second, which is considerably wider than it, especially at the apex. The third cellule is a little shorter, and wider than the second. The first recurrent nervure is received in the basal third of the cellule, and is bullated at the junction with the cubital. The second is received a very little in front of the first transverse cubital cellule, and is often inter- stitial. The transverse median nervure is interstitial or nearly so. The legs are of moderate size. The tarsi are longer than the tibias. The patellae are very well developed, and somewhat membranous. The metatarsus is short, but is longer than the second joint, the second, third, and fourth joints become gradually shorter, the fifth is longer than the basal ; its claws are strong, curved, inflated at the base, and either simple, or with a blunt, ill-developed, subapical tooth. In the $ the coxse are much larger, thicker, longer, and more widely sepa- rated ; the metatarsus is covered with long hair on the underside, and the apex beneath is projected into a large, blunt tooth. On the inner side the coxae are hollow, their sides project into a ridge, which on the posterior side and towards the apical third, curves GENUS CIMBEX. 7 down in a round curve to the apex, forming a hollow with a blunt tooth at either end. The ridge on the other side is much less developed, and there is also a blunt ridge on the underside of the femora, which are much thickened. The anterior femora are of the normal size, and they are also grooved. The coxae are not enlarged like the two posterior pairs. The abdomen is a little longer than the head and thorax; the blotch is very wide, and extends nearly across the segment ; the segments are nearly of equal size. The ninth does not project at all on the upper side. Backwards it does not extend beyond the eighth, the saw does not project beyond the apex of the abdomen. The form of the saw and back piece is quite peculiar, being rounded and curved at the apex, not brought to a sharp point as with the other saw-flies, and they are further distinguished by having along the edge little projecting bead-like points. The insects belonging to this genus are the largest of all the TentlirediniddB. They are very variable in coloration, and this has led, on the one hand, to many so-called species being made out of the various forms, and, on the other, to all the forms being united into one species, as was done by Klug, who placed them together under the name of G. variabilis. The obser- vations of Brischke, Zaddach, and Van Vollenhoven, however, have shown conclusively that there are at least four, if not five, good European species of Cimbex, s. sir. Undoubtedly the most satisfactory way of discriminating the species is by rearing them from the larvas. Failing this crucial test, it is not always an easy matter to separate the perfect forms. In colour they vary exceedingly ; the thickness, texture, and colour of the hair vary ; the form of the antennae varies, especially in their comparative length and in the shape of the club ; the position of the nervures fluctuates : thus the second recurrent nervure is found in the second cellule, in the third, and is interstitial in all the species. 8 GENUS CIMBEX. One species (0. sylvarum) is usually distinguished by having a fuscous fascia in the first discoidal cellule, and the edges of the wings are also fuscous, while usually C. lutea is distinguished by its wings wanting entirely these fuscous clouds ; but I have seen speci- mens of sylvarum with the wings completely hyaline, and of lutea with faint indications of the fuscous tint along the edge and at the discoidal cellule. I thought once (like Zaddach, loc. cit., p. 243) that the form of the scutellum would give a good specific character, but soon found that this also varied, especially with bred specimens. Then the posterior coxas and femora vary in length and thickness in the males. In some bred specimens I have examined they are not much longer or thicker than the middle pair, while with caught specimens (of nearly the same size) they are more than one half longer and thicker. Nor is the variation in the size of the coxse and femora confined to bred examples. The males, too, vary very much in size. Thus I have one 3 of C. lutea which is 17 lines in length ; the posterior femora are 4 lines long, and nearly 2 lines broad in the middle ; another is 8^ lines long, and the hind femora 2J lines long. Synopsis of Species. 1 (2) Clypeus and pronotum clear yellow, wings yellowish, the fore half fuscous throughout ; mesonotum and base of abdomen punctured. Humeralis. 2 (1) Clypeus and pronotum reddish-yellow or black, wings not fuscous throughout in front ; mesonotum not punctured. 3 (4) Thorax, femora, the base, sides in front, and lower side of abdo- men, violaceous. <$ Apex of abdomen black or dull brownish, and with a violet iridescence. Connata. Thorax, femora, and base of abdomen not violet. 5 (6) Pubescence on head, thorax and base of abdomen longish, black, or dark fuscous ; wings with a fuscous fascia in the basal cellule, and a fuscous border at apex. Head and thorax A M -D v rarely yell°W' Sylvarum. t> (5) .Pubescence on head, thorax and base of abdomen woolly, close, thick, griseous, pale, or yellowish ; wings not marked with fuscous; head, thorax, and abdomen for the greater part, OIMBEX SYLVARUM. 1. ClMBEX SYLVABUM. Vol. II, PI. V, fig. 1, ? ; la, Head ; 1 b, Antenna ; 1 c, Tarsus. Vol. II, PL XII, fig. 8, Larva. Tenthredo sylvarum, Fab., E. S., ii, 105, 4; Pz., F. G., Heft 88, pi. xvi. Cimbex sylvarum, Fab., S. P., 16, 4; Voll., Tijd. Ent., xriii, 33, pi. iii. Tenthredo femorata, Pz., 1. c., Heft 26, pi. xx. — tristis, Fab., Iter. Norw., 334 ; E. S., ii, 106, 7. Cimbex tristis, Fab., S. P., 17, 7. varians, Leach, Z. M., iii, 105 ; Ste., 111., vii, 6, 2. — europ&a, Leach, Z. M., iii, 105, 4. — femorata, $ , Fall., Acta Holm., 191, 3 ; Mon., 8, 2 ; Lep., Mon., 31, 83; F. Fr., Hym., 5, 7 ; Drewsen, Ann. Ent. Fr., iv, 169 (1835) ; Ste., 111., vii, 6, 1. — variabilis, var. femorata, Klug, Berl. Mag., 1, 78 ; Htg., Blattw., 6, 3. — biguetina, Lep., Ann. Soc. Ent. Fr., ii, 455. — betulae, Br. and Zad., Schr., Ges. Konig., ii, 233 and 249, pi. ii, f. 1 (lar.). — fagi, Br. and Zad., 1. c., 233, 251. — lutea, Thorns., Hym. So., i, 19, 1. — siberica, Kirby, List of Hym., i, 4, pi. 1, f . 1, $ , 387 ; pi. xvi, f. 3, J. Black ; antennae and tarsi luteous, body covered with a black or fuscous, longish pubescence ; wings hyaline, a blotch along the trans- verse basal nervure, and the outer margin fuscous. Length 7— 13| lines. Ab. — a. Black; antennae and tarsi luteous (tristis). b. Black ; antennae and tarsi luteous, the three or four middle segments of the abdomen red. ? and $ (sylvarum). c. As in b, but abdomen entirely red $ , rarely ? . d. Black ; antennae, tibiae and tarsi luteous, the three or four middle segments for the greater part pallid testaceous, or more rarely the middle segments are entirely pale testaceous (varians, biguetina, Lep.). e. The antennas, the head, except the vertex, pro- notum, the fore part of mesopleurae, the middle lobes of mesonotum at the sides, scutellum, legs, and abdo- men, reddish-yellow ; wings yellowish ; the pubescence is shorter and paler than usual. This is the most variable and commonest of the 10 CIMBEX SYLVARUM. species of Cimbex. The variations in the coloration are very numerous in both sexes. Commencing with quite black specimens save the antennae and tarsi, we meet with specimens having one yellow band on the abdomen, then with specimens with two or three bands, which lead into specimens with the middle of the abdo- men entirely yellow, the tibiae being also yellow, and from this we reach the extreme yellowish form de- scribed above as Ab. e. We find the same variations with the red banded forms. The larva feeds on birch, and is found from July to September. Its head is pale yellow, granular, and with black eye-spots, the body is green, more or less yellowish at the skin folds and along the back, where there is a black stripe, which is bluish along the centre, and is widest towards the middle of the body, it being very narrow along either end. The skin is wrinkled, especially on the sides, which are beset with numerous white warts, particularly over the legs. The feet are white, with brownish claws ; the spiracles black, longer than broad, and broader below than above ; they arise from a raised part of the skin. When young the larva is greenish-white, dusted over with white powder. It wants the black dorsal stripe entirely. Brischke and Zaddach (1. c., p. 237, pi. ii, fig. 2) describe and figure a larva which was found by Drew- sen and by themselves on beech ; this is described as being the largest of all the Cimbex larvae. The ground colour of the body is a beautiful bluish-green, the dorsal stripe is clear blue or violet ; the head is larger and more cubital than with sylvarum, and the white tubercles are smaller and more pointed. According to Brischke the characters presented by this larva are constant, and if that be so the chances are that it per- tains to a good species ; but unfortunately Drewsen and Brischke and Zaddach bred only males, which cannot be separated apparently from G. sylvarum, and like the males of that species would seem to be very CIMBEX CONNATA. 11 variable. The above-named German authors name the beech-feeding form C. fagi. The following parasites have been reported to feed on the larvae of C. luteasmd C. sylvarum : — Campoplex argentatus, Rtz. ; G. holosericeus, Rtz. ; C. pubescens, Rtz. ; Cryptus cimbicis, Rtz. ; C. incubitor, Gr. ; Hemi- teles dispar, Rtz. ; Mesochorus cimbicis, Rtz. ; M. splendidulus, Gr. ; Mesoleptus rufus, Gr. ; Paniscus glaucopterus, L. ; P. testaceus, Gr. ; Pezomachus cur- sitans, Gr. ; Tryphon mesoxanthus, Gr. ; T. rufus, Gr. ; T. sorbi, Sax. ; Monodontomerus obsoletus, Fab. The species is of universal distribution in Britain, the flies appearing in June. Continental distribution : General in Europe, also in Siberia. 2. CIMBEX CONNATA. Tenthredo connata, Schrank, Beitr. z. Naturg., 83 (1776). Crabro maculatus, Fourc., E. P., ii, 361, 2. Tenthredo montana, Pz., F. G., Heft 84, pi. xii. — femorata, De Geer, Mem., ii, 2, 943. Cimbex decemmaculata, Leach, Z. M., iii, 106; Curtis, B. E., i, pi. xli; Ste., 111., vii, 7,4. ornata, Lep., Mon., 28, 77 ; F. Fr., Hym., 4, pi. i, f. 2. maculata, Ste., 111., vii, 8, 7. connata, Voll., Tijd. Ent., vii, 50, pis. i and ii; Br. and Zad., Schr. Ges. Konig., vi, 254, pi. ii, f. 6 (lar.); Andre, Species, i, 26; Cat., 1*2. pallida, Ste., 111., vii, 7, 3, teste Kirby, List of Hym., i, 2. — violascens, Thoms., Hym. Sc., i, 20, $ . — Humboldtii, Ratz., Forst., Ins., iii, 135, 46. Fuscous-violaceous ; the head, pronotum, pleurae, tibiae, and tarsi, brownish-testaceous, coxae and femora violaceous; abdomen bright luteous, the basal two segments entirely, the third broadly in the middle, apical segments at the sides, and the greater part of the ventral surface, violaceous. Antenna} luteous. Head, thorax, and base of abdo- men covered with a griseous pubescence, which is darker on the meta- thorax. Wings hyaline, more or less spotted with fuscous at base, middle, and apex. The (£ is violet-black, the apex of abdomen dull fuscous, antennae and tarsi luteous, tibiae brownish ; head, pronotum, and pleura dull brown. The wings are coloured as the $ , but have a much more clearly defined steel-bluish-violet iridescence. Length 10 — 12 lines. 12 CIMBEX CONNATA. With the darker specimens the thorax is almost en- tirely black, the tibiae are also black or dark fuscous, and the middle of the abdomen on the back and along the sides is violaceous, as in G. decemmaculata, Leach ; with the lighter aberrations the head, pleurae, the breast in part, pronotum, and the edges of the lobes of the mesonotum are brownish-testaceous ; the abdo- men is only violaceous at the base, and on the seventh and eight segments at the sides; the legs too being brownish-testaceous, save the femora, which are always violaceous. The distinctions between connata and lutea lie in the darker-coloured thorax and base of the abdomen in the former, which are always for the greater part fuscous, for the ground colour, and most distinctly violaceous, the same being the case with the coxse, trochanters, and femora ; the wings are darker at the median cellule and at the apex, while they have also a more distinctly marked violet iridescence. The pubes- cence on the thorax and base of abdomen is shorter and thinner, the spurs are longer and thinner ; gener- ally too it is a larger insect than lutea. The larva is of a beautiful clear green colour. The head is also greenish, with black eye-spots. The dorsal stripe is bluish-black, with paler spots. It commences on the first segment, continues to the penultimate, and is of a uniform breadth throughout. On each side it is bordered by a yellowish line. Not far from this yellow line is a roundish black dot ; the spiracles are also of this colour. On the sides are numerous white warts. When young (i. e. before the first moult) the larva is uniformly clear green, the head shining, the spiracles black, and there are on each segment three small black spots, which form three rows of spots, the uppermost being on the head. It feeds on alder. Campoplex argentatus, Etz. ; C. holosericeus, Rtz. ; Mesochorus splendidulus, Gr., and Paniscus glaucop- terus, L. are its parasites. CIMBEX LUTEA. 13 Rare in the south of England. Sussex, Windsor, Plymouth (Bignell). Continental distribution: Sweden, Germany, Hol- land, France, Italy, Austria, Greece. 3. CIMBEX LUTEA. Vol. II, PL XII, fig. 6, Larva after Vollenhoven, 5, Cocoon. Tenthredo lutea,, Lin., S. N., i, 655, 2; F. Su., 388; De Geer, Mem., ii, pt. 2, 943 ; Don., Brit. Ins., vii, pi. ccxxiv. — femorata, Lin., S. N"., i, 555, 1 ; F. Su., 388. Cimbex lutea, Pz., F. G., Heft 105, pi. xiv. Crabro lunulatus, Fourc., Ent. Par., ii, 362. — annulatus, Fourc., 1. c. Cimbex femorata, Voll., Tijd. Ent. (2), v, 64, pi. iii ; Ent., No. 151, 3, 1876. — Griffinii, Leach, Z. M., iii, 107 ; Ste., 111., vii, 7, 5 ; Lep., Mon., 30, 80. — Schaefferi, Lep., Mon., 26, 74; F. Fr., 3, pi. i, f. 1. — variabilis, King, Berl. Mag., i, 72 ; Htg., Blattw., 63 ; Ratz., Forst., Ins., iii, 63, f. 10, ? . — luteola, Lep., Mon., 28, 78. — saliceti, Br. and Zad., Schr. Ges. Konig., iii, 251, pi. ii, f. 3. — brevispina, Thorns., Hym. Sc., i, 21. 3. Luteous ; head and thorax darker coloured than the abdomen and legs, covered with a close, silky, white or yellowish pubescence. Wings pellucid, sometimes with a yellowish tint. The £ is covered with a longer pubescence than the ?, and is fuscous or blackish in colour. Length 7—15 lines. Ab. — a. Dull luteous ; the mesonotum, sternum, head, and legs, brownish-testaceous ; the edges of the abdo- minal segments marked with black. $ darker coloured all over, and the base of abdomen broadly fuscous- black. The wings in both sexes have a yellowish tinge. b. The breast, the greater part of mesonotum, base of abdomen, and femora above, black. Abdomen pale luteous, the segments at base black. ? c. Thorax fuscous, the ground colour, however, being dull testaceous; basal half of abdomen black, apex 14 CIMBEX LUTE A. pale testaceous, femora broadly marked with black above. $ d. Black; antennae for the greater part and tarsi reddish-luteous ; wings clear hyaline. $ e. Yellowish-testaceous ; the head, breast, back of thorax, and base of abdomen, fuscous. The luteous forms of lutea are readily separated from the normal forms of sylvarum, which have always the head, thorax, and the greater part of the legs bluish- black ; but the lighter-coloured aberrations of sylvarum come very near the darker-coloured specimens of lutea. These, however, may be always separated from sylva- rum by the pubescence, which in the former is thick, moderately long, and alway griseous, pale or yellowish, never fuscous as in sylvarum-, which has it somewhat longer, but not so close. The hair on the abdomen in sylvarum is also longer, more scattered, and not so silky as in lutea. The wings in lutea are more yellow- ish, and there is no fuscous cloud in the median cellule, nor has the apex a fuscous border. The reddish-banded males of sylvarum are easily separated from those of lutea, which have always the abdomen black, except with Ab. a; but that again has sufficient distinctive characters in the colour of the head, thorax, and wings. The only characters that I can find to separate the black males of sylvarum from lutea are, that the median cellule of the former has a fuscous cloud, and the apex of the wings is always fuscous, and the club of the antennae is shorter and thicker, being not much longer than the two preced- ing joints ; in lutea they are not so distinctly pyriform, and are considerably longer than the two preceding. The wings in lutea g appear to have a well-marked steel-blue iridescence, but I do not know if this is a constant characteristic. The larva feeds usually on Salix alba and S.fragilis. Its body is bluish-green, the skin is in folds, and over the legs and along the sides are small white tubercles. Along the back is a dark bluish, central stripe, which C1MBBX LUTEA. 15 usually commences on the second body segment, and gets very narrow towards the tail. This stripe is darker at the folds, so that it appears to be made up of lighter squares separated by darker transverse bars. The head is of a pale greenish or greyish-yellow. The spiracles are black, lighter in the centre, and expanded on the lower side. The legs are pale green, with darker claws. When young the larva is greenish-white, and, unlike the young of sylvarum, has a small, short dorsal stripe. Besides the larva described above, Brischke (1. c., p. 239) found others feeding on Salix aurita, which Zad- dach and himself refer to 0. lutea, notwithstanding that they differed considerably from the Salix alba feeding form. The ground colour of the body with these goat willow-feeding larvse is ochre-yellow, orange, or red- dish, and marked with yellowish transverse stripes as well as with greenish-yellow spots. The head is ochre- yellow or reddish. The dorsal stripe is violet and, as in the normal form, is divided into clear spaces. The arrangement and form of the white tubercles and stigmas is the same as in the ordinary type. When young these larvae are clear bluish-green, reddish-yellow on the back, the head is white, and the dorsal stripe is small and dark blue. These larvae eject the acid liquid from the lateral pores much more copiously than the others, and they are very difficult to rear, Brischke having only suc- ceeded in rearing two males and one female during fifteen years. One of the males was identical with the black-coloured male of lutea, the two other specimens differed from any form of lutea seen by Zaddach, both being much paler coloured than usual ; but whether they represented merely extreme varieties of lutea or distinct species is a question which cannot be decided from the limited information at hand. If they truly belong to lutea it is certainly a most remarkable case of dimorphism in the larvae. The pale-coloured pair bred by Brischke are to be referred to the var. pollens. 16 CIMBEX HUMEJiALlS. I have one British specimen which agrees pretty closely with it, but the ground colour is darker. The form from Salix alba has been shown by Brischke (Schr. Natur. Ges., Danzig, vii) to be par- thenogenetic, but he did not rear the larvae. Not common. Perthshire, South of England. Continental distribution : Sweden, Germany, Hol- land, France, Italy. 4. CIMBEX HUMERALIS. Crabro humeralis, Fourc., E. P., ii, 361. Tenthredo connata, de Vill., Linn. Ent., iii, 84, 13. — axillaris, Jur., Hym., 48, pi. 6, f. 1. Cimbex axillaris, Pz., F. G., Heft 81, pi. vi; King, Beri. Mag., i, 84 ; Htg., Blattw., 68, 2 ; Spin., Ins. Lig., ii, 152, 5 ; Yoll., Tijd. Ent., v, 49, pi. i, Zool., s.s. 9263. — humeralis, Lep., Mon., 30, 81 ; Ste., 111., vii, 8, 8 ; Andre, Species, i, 24; Cat. 1*, 1. — scapularis, Stein, S. E. Z., xxxvii, 53 (1876). Black; antennae yellowish, the club more or less fuscous; clypeus and pronotuiii clear yellow; the mandibles, labrum, sides and vertex of head, pleurae, scutellurn, and legs, yellowish- testaceous; abdomen clear yellow; the basal two segments, the third in the middle, the fourth, except at the sides, the edges of the following segments, and a small triangular mark in the centre, black. Head and thorax covered with a pale fuscous, longish pubescence ; mesonotum, pleurae, and base of abdomen punctured. Wings yellowish, the fore half fuscous, ner- vures testaceous ; the second recurrent nervure not interstitial. Length 7—8 lines. As with the other species, humeralis exhibits a cer- tain amount of variation in coloration, but it is very small in comparison with that shown by sylvarum and lute a. The larva of this species feeds on hawthorn and on Prunus padus. The body is bluish- white, somewhat yellowish at the sides. On the back is a broad, bluish central stripe, divided as usual into clear and darker spaces alternately. On either side of this stripe is a line of small, transverse, black and orange spots. On each segment are: a little wedge-shaped black line, close against it an orange spot, broadest above, after GENUS TEIOHIOSOMA. 17 this are one or two longer black lines, then a smaller one, and, lastly, three of nearly equal size, all black. The last segment has three black spots in the centre, and on either side one orange and three black lines. Stigmata black. Legs white, with black claws ; at the end of the penultimate joint is a long black stripe ; the claspers are greenish-white. It appears in early summer, and becomes full-fed in July. As parasites there have been recorded : Mesochorus cimbicis, Ratz., Mesochorus splendidulus, Gr., and Paniscus glaucopterus, L. Rare. Salisbury (Leach). Continental distribution : Germany, Austria, Hol- land, France, Pyrenees. Genus — TRICHIOSOMA. Trichiosoma, Leach, Z. M., iii, 108 (1817). Blotch absent. Labrum large, clypeus with a distinct emargination . Posterior femora large, long, toothed beneath. Coxae widely separated ; anterior coxaa slightly apart. Body covered with long hairs. The neuration and the form of the antennae agree with Oimbex, as does also the form of the head. The clypeus is much smaller, the labium is much larger, and it projects considerably. The mandibles are large, and have two blunt teeth in the middle ; they are much larger and gaping with the <$ . The labium has three equal lobes, and is rather short and thick. With the labial palpi the second joint is the longest, the first the smallest, and the third is longer than the fourth. The maxilla is long and narrow, its inner lobe is shorter than the outer. The basal four joints of maxillarj^ palpi are thicker than apical. Joint one is a little longer than second, the third is as long as the basal two together, fourth a little shorter than the fifth ; the sixth is the longest. VOL. in. 2 18 GENUS TRIOHIOSOMA. On the underside the femora are grooved, and bear a blunt tooth on either side at the apex of the groove, but the inner tooth is very slightly developed ; and this is pretty much the case with both in many females. At the apex of the COXSB beneath there is a saddle-like depression, with a tooth at either end as in Cimbex. The metatarsus has a curved depression on the under- side covered with hair as in Cimbex, but there is no distinct spine at the apex as in that genus. The cal- caria have a membranous loJbe at the apex ; the claws are simple. As with Gimbex, the males have the legs very much more developed than the females. The scutellum is flatter and longer than in Gimbex, and at the base it projects forwards a little, so that the depression at its base has a slight curve. The larvas are green, and when young are covered over with a white powder. In form they agree quite with those of Cimbex. One species feeds on birch, another on Sorbus aucuparia, one on willows, and another on hawthorn. Trichiosoma agrees very closely with Gimbex, and by many authors the two are not considered generically distinct. I certainly think the genus is a natural one, and is sufficiently distinguished from Cimbex by the blotch being absent, by the larger size of the labrum compared to the clypeus, by the femora bearing teeth, and by the body being covered with long hair. The genus is not so widely distributed as Cimbex, and it contains fewer species. Four European species are known, and three from North America. Possibly one or two of the latter may not be distinct from the Old World forms. Considerable variation is shown by the species, and as the specific distinctions between them at the best are not great, they are not always easily separated, unless they have been bred, the Iarva3 here, as in many other cases, affording the best criterion of specific distinctness. TBIOHIOSOMA VITELLIN./E. 19 Synopsis of Species. 1 (2) Tibiae fuscous. Betuleti. 2 (1) Tibiae testaceous or yellowish. 3 (4) Femora clothed with black hairs, distinctly violaceous ; abdomen clothed at base with yellowish hairs, with short fuscous hair at the middle, and with long yellowish or greyish hair at apex. Scalesii. 4 (3) Femora clothed with pale hair. 5 (6) Abdomen dull black, covered uniformly with long grey hairs, the apex and sides of abdomen sometimes reddish ; the tibiae not darker than the tarsi. Lucorum. 6 (5) Abdomen bronzy black, the hair on base longer than on middle or apex; the apex, sides, and belly usually ferruginous, rarely black ; tarsi paler than tibiae. Vitellinse. 1. TRICHIOSOMA VITELLIN^I. Vol. I, PI. X, figs. 14 and 15, Larva. Tenthredo vitellinse, Lin., F. Su., 386. Trichiosoma laterale, Leach, Z. M., iii, 109 ; Curt., B. E., i, pi. xlix; Ste., 111., vii, 9, 2. Cimbex vitellinse, Br. and Zad., Schr. Ges. Konig., iii, 259, Taf. iii, f. 1 (lar.). Trichiosoma salicis, Htg., S. E. Z., i, 19. vitellinse, Thorns., Hym. Sc., i, 24, 3 ; Cam., Proc. N. H. S. Glas., iii, 205 (lar.), Fauna, 46, 2; Andre, Species, i, 28; Cat., 3,* 4. Tenthredo amerinse, De Geer, Mem., ii, 2, 948, 3 (non L.). Black, somewhat bronzy; tibiae reddish-testaceous, tarsi testaceous, wings yellowish, the apical margin of the wings fuscous, costa testa- ceous, stigma bluish-black ; the apex of abdomen, the sides, and more or less of the belly, ferruginous. The hair on the head is black or fuscous; on thorax and abdomen pallid or yellowish. It is thick and long on the base of the abdomen, but becomes shorter and thinner towards the apex, where it is silky. Length 7—12 lines. Ab. — Abdomen quite black. $ and £ . The amount of reddish colour on the abdomen varies considerably ; sometimes the apical third above, the sides and the belly from base to apex are reddish, some- times only the sides and apex, and more rarely the reddish colour is entirely absent. The latter is the case with all the specimens ( ? and $ ) I have bred from the larvae figured on PI. X, Yol. I, figs. 14 and 15) these specimens being also much smaller (7 lines only) and more shining than usual. In the 3 the hair on 20 TRICHIOSOMA VITELLIN^l. the thorax and abdomen is usually more reddish or yellowish in tint compared to the ? . The middle of the antennae is generally testaceous ; seldom is it entirely black. This is the largest species in the genus, although very small specimens are also met with. As a rule, the reddish colour of the sides, apex, and lower side of the abdomen readily separates it from lucorum, but as this is not a constant character, and as the same colour occurs with lucorum, the most reliable point of dis- tinction lies in the abdomen of vitellince being much more shining and bronzy than that of lucorum, besides which the hair on the middle and apex is much shorter, the colour of the hair on the thorax and abdomen having also a redder tinge. The larva has the head white, with a fuscous mark on the apex, not touching the black eye-spots ; the lower part of the clypeus is black in the centre. The body is light green, the folds of the skin whitish; along each side of the dorsal vessel is a row of white tubercles, and there is another row over the legs ; the dorsal vessel itself is pale green. The spiracles are dark reddish, and over each is one or more small reddish marks. The anal segment is clear of tubercles. Legs clear white, claws blackish. The above is the description of the larvae I find in Scotland feeding in .August and September on Salix aurita and 8. caprea. Brischke (1. c.) bred the species from two different larvas. One he found in August and September on Salix caprea. Its body was clear yellowish -green beset with numerous white tubercles; the head shining, yellowish, granular in texture, and with black-eye spots ; the spiracles elliptical, reddish, the claws brown. The other form he found in July on Salix viminalis and S. caprea, and was smaller than the autumnal one. The ground colour was bluish-green, the dorsal line (which was free from white warts) of a darker green, and the whole body (except the last seg- ment) beset with raised, often confluent, tubercles and TRICHIOSOMA SCALESII. 21 dots. The spiracles were reddish-brown, and over each (except the first and last) was a small dot of the same colour, which is not found on the other larva. The claws were brown, and upon the yellow shining head was, between the eyes, a brownish spot. The only variation in coloration I have noticed was in some specimens which had the green of a bluish tinge. As with the other species, the young larvae are dusted all over with a white powder. Seemingly not very common in England and Scot- land, but is widely distributed. Continental distribution : Sweden, Germany, France. America : Vancouver's Island and the Rocky Moun- tains (Kirby, List of Hymen., i, p. 10). 2. TRICHIOSOMA SCALESII. Vol. II, PI. V, fig. 2, $ ; 2 a, Hind Leg ; 2 b, Head of ositor are formed as in Sirex, only they are finer and inch longer, being nearly double as long as the abdo- len. In repose they are not exserted. I am not quite certain how they are kept in position in the abdomen when retracted, not having a ? to dissect ; but in one specimen the apex was apparently firmly inserted into the basal abdominal segment. The c? has all the tarsi five-jointed, but the ? has only three joints in the anterior ; the front tibiae have only one spur. Claws simple. The four posterior tibiae bear blunt irregular tubercles on the outer side. Genus — OEYSSUS. Oryssus, Latr., Precis, p. Ill ; Klug, Mon., 3 ; Westw., Int., ii, 121. 1. OEYSSUS ABIETINUS. SpJiex abietina, Scop., Ent. Cam., 296 (1763). — vespertilio, Fab., E. S., ii, 129, 19 (1793) ; Klug, Mon., 7, pi. i, fs. 1-3. Oryssus vespertilio, Htg., Blattw., 366. — coronatus, Fab., E. S. Suppl., 218; Curt., B. E., x, pi. ccclx ; Thorns., Hym. Sc., i, 333. — albopunctatus, Gimm., Bull. Mosc., ix, 434. Black ; the lower orbits of the eyes above the clypeus, the greater part of the third, fourth, and fifth antennal joints beneath, the knees, and the greater part of the outer side of tibiae, white. Tarsi reddish ; the abdomen bright red from the second segment. Wings hyaline, the apex smoky, except the extreme tip and a broad stripe at the end of the stigma, extending to the cubital nervure from the stigma. Head and thorax and basal segment of abdomen coarsely punctured. The <$ has a broad yellow mark on the apical segment above. Length 5 — 6 lines. Kare. Daren th Wood, Devonshire (Stephens). 140 FAMILY CYNIPIDJ). Family CYlSttPHLE. The name " Cynips," as a generic term, originated with Linnaeus. Under it he ranged diversely related species, part only really belonging to the Cynipidde, the others being parasites or gall-flies not connected with the Hymen- optera. Scopoli followed Linnaeus and described six species, but one of them was not a true Cynips. The same is to be said of Schrank, who enumerated ten species, but of the same mixed character. In 1764 Geoffroy split up the Cynips of Linnaeus into two sections — Cynips and Diplolepis, — the distinction between them lying in the antennae,