truncate or emarginate; the maxillary palpi are shorter or scarcely longer than the maxillary lobe, the basal joints being elongate-cylindric; the terminal joints of the antennæ of the males are often recurved and hooked at the tip (fig. 87. 11. extremity of ant. of Epipone spinipes ); the thorax is short and truncate both in front and behind; the basal segment of the abdomen is more or less coarctate, in the typical species forming a long and narrow peduncle (fig. 87. 6.); the legs are not furnished with ciliæ or spines, although these insects generally construct their nests with earth in cavities of walls, old wooden palings, or in sand banks, forming a succession of cells placed end to end, in each of which an egg is placed, together with a sufficient supply of food for the entire consumption of the larva, consisting of other insects, larvæ, spiders, &c. The female then carefully closes the mouth of the cell with earth.

Réaumur (Mém. tom. vi. pl. 26. f. 1-10.) has given the history of a species of this family, which Latreille (Règne An. tom. v. p. 336.) considers to be the Odynerus (Vespa) muraria Linn.* This insect, during the early months of summer, forms a burrow in the sand to the depth of several inches, in which it constructs its cells; besides which it builds, with the grains of sand brought up whilst burrowing, a tubular entrance to the burrow, often more than an inch long, and more or less curved, the grains of sand of which it is formed being agglutinated together; each female forms several of these burrows and deposits an egg in each cell, together with a number of green caterpillars, which it arranges in a spiral direction, one being applied against the other, and which serve as food for the larva when hatched. When the store of food is secured, the insect closes the mouth of the burrow, employing the grains of sand of which the funnel was composed for that purpose.

The larvae of Odynerus are fleshy grubs, destitute of feet (fig. 87.5. magnified), with transverse dorsal tubercles serving in their stead. The body (including the head) consists of fourteen segments and a minute anal tubercle, with ten spiracles on each side. I have reared both

Mr. Shuckard (Mag. Nat. Hist. Sept. 1837) gives the muraria as identical with Odynerus (Epipone) spinipes, stating that it is the latter which constructs this trumpet-shaped tube. Such may be the case, but Réaumur's description and figure do not accord with the former species, but rather with the true Linnæan V. muraria (which belongs to a different subgenus), according to the description of the authentic specimen of the latter still in the Linnæan Cabinet, forwarded by me to M. Wesmael, and by him published in the Bull. Acad. Bruxelles. Ceramius Fonscolombei has similar habits.

sexes of one of the species from larvæ thus constructed; thus disproving Mr. Shuckard's suggestion, that the female larvæ will necessarily have one segment less than those of the male, as in the imago. (Trans. of Burmeister, p. 35.; and Mag. Nat. Hist. Sept. 1837. See my memoir "On the Apod Larvæ of the Hymenoptera, with reference to the Segmental Theory of Annulose Animals," in Trans. Entomol. Soc. vol. ii. p. 121.) In the same Transactions (vol. i. p. 78.) I have also published some notes upon the habits of Odynerus antilope, which lines its cells with mud, of which it carries small round pellets into its burrow, under the breast. It employs the green caterpillars of a Crambus? for the food of its young. Mr. Ingpen exhibited to the Entomological Society (August 4, 1834) the nest of Odynerus quadratus, which had been discovered between the folds of a piece of paper which had fallen behind some books. It was nearly six inches long and one wide, and had several openings to the cells, through which the insects, on arriving at the perfect state, had escaped; it appeared to be composed of dried mud.

H. (in Mag. Nat. Hist. No. 25.) describes some larvæ found in the healing of a book, apparently those of an Odynerus. Mr. Curtis also figures a species (O. parietinus), the cells of which were formed on the top of a book.

Bouché states that Od. parietum stores up flies, &c., as well as the caterpillars of Tortrices (Naturg. der Ins. p. 179.).

M. L. Dufour has recently communicated a series of interesting observations upon the economy of several species of Odynerus to the Académie des Sciences, but they have not yet been published. (See Comptes Rendus, No. 10. Sept. 1838; and Annals Nat. Hist. No. 8. Oct. 1838.) I am indebted to Mr. F. Smith for specimens of Odynerus lævipes Shk., and its nest, first described by Mr. Shuckard (Mag. Nat. Hist. Sept. 1837). The cells are formed of agglutinated sand, lining the cavity of a dead stick of the common bramble, from which the pith had been excavated, probably by Osmia leucomelana, and placed end to end. Mr. Sells has found twenty-five small Lepidopterous larvæ in the nest of an Odynerus; and Mr. Saunders as many as seventy-five in the nest of Epipone spinipes. Mr. Waterhouse has also discovered Lepidopterous and Chrysomelideous larvæ in the cell of an Odynerus (Trans. Ent. Soc. vol. ii. p. xviii.): I have also observed and captured many specimens of O. crassicornis Pz. near Paris, which were always loaded with the larvae of Chrysomela Populi, which

was extremely abundant in the neighbourhood. (See Vol. I. p. 389.) Mr. Kennedy has captured O. bidens with its prey, which consisted apparently of a Chrysomelideous larva. He also mentions finding in one of its cells a small dipterous insect. (Lond. and Ed. Phil. Mag. Jan. 1837.) I have described (Gardener's Mag. 1837) a singular instance of instinct exhibited by a species of Odynerus, which provisions its nest with the leaf-rolling larva of Tortrix (Argyrotoza) Bergmanniana, which it dislodges by introducing its sting into the rolls of the leaves, immediately running to the end of the roll in order to ascertain whether the larva was endeavouring to make its escape from its unknown enemy.

In the Insect Architecture (p. 26-29.) is an account of the proceedings of a species of Odynerus which forms its burrow in a brick wall, and which was infested by a Tachina, which deposited an egg in the nest; and the grub hatched from it, after devouring one of the wasp grubs, formed itself into a cocoon, as did the other undevoured grub of the wasp-both of which cocoons are presented as of an oval form and placed side by side in a large oval flask-like cell. Such habits are, however, at variance with the recorded economy of the Odyneri.

Geoffroy (Hist. Ins. Paris, tom. ii. p. 378. pl. 16. fig. 2.) has described a species of Eumenes (V. coarctata Linn., Panz. (fig. 87. 6.) which differs somewhat in its habits from the rest of this family. This species constructs, upon the stems of plants, especially heath, small spherical nests (fig. 87. 7.) formed of fine earth: at first a hole is left at the top, through which the parent fills the cell with honey, and deposits a single egg therein; the hole is then closed up, and the larva, when hatched, feeds on the honey, undergoes its metamorphosis, and makes its escape through a hole which it forms at the side of the cell, which contains but a single insect.

The habits of E. Saundersii Westw. MSS., an Indian species allied to E. petiolata, not described by Fabricius, observed by W. W. Saunders, differ from those of E. coarctata; this insect making its cells of mud, attaching them under projections, each cell being somewhat smaller than a pigeon's egg, and sometimes two or three are placed side by side; they are well stored with green caterpillars. Mr. Saunders observed the Eumenes repeatedly at work, and found one nest in a keyhole, and another beautifully constructed inside an old flute. (Trans. Ent. Soc. vol. i. p. 63.) One of these nests produced a Pælopæus, evidently a parasite upon the Eumenes.

The genus Ceramus Latr. (Gnatho Klug) is very interesting on account of its upper wings not being folded longitudinally, and having only two submarginal cells, as in the Masarides; the labial palpi also, as in those insects, are larger than the maxillary, which, according to Latreille and Fonscolombe, have only four joints. Klug figures them as 3-jointed, but I can only observe two joints in C. Lichtenstenii Kl. M. Fonscolombe (Ann. Soc. Ent. de France, 1835) has given a detailed description of C. Fonscolombii Latr. (a species inhabiting the South of France, for which I am indebted to M. Serville), which forms cylindrical tubes at the entrance of its burrows (like Odyn. muraria), which it destroys after it has completed the construction of its nest. The genus comprises two European and two South-African species. Paragia Shk. has also only two submarginal cells; but its eyes are oval (not reniform), and its mandibles resemble those of the social species, whence Mr. Shuckard considers it is the Australian representative of Vespa.

Some anomalous species of wasps, forming Latreille's family MASARIDES, agree with the Eumenidæ in having the fore wings longitudinally folded and the eyes notched, but differ in the antennæ being inserted wide apart, and composed apparently of only eight joints (fig. 87. 12. ant. Cælonites apiformis ), the last being large, rounded at the tip, and forming a solid mass with rudimental articulations, indicating the terminal joints; the labium is terminated by two very long setæ, which when at rest are retractile within the basal tube of the mentum; the upper wings have only two perfect submarginal cells; the clypeus is emarginate in the centre, with the labrum inserted into the notch; the maxillæ are short and obtuse, with the maxillary palpi very small, consisting of two, three, or four joints. The insect described by Fabricius (Ent. Syst. vol. ii. p. 284.) from the collection of Desfontaines, and figured by Coquebert (Illust. Iconogr. tab. 15. fig. 4.) under the name of Masaris vespiformis, is evidently a inale, from the elongated antennæ and abdomen*, which is armed on the under side with two tubercles near the base. The insect figured in the great work on Egypt, which Latreille regards as forming a distinct subgenus, is a female. I have received it from Dr.

* Fabricius nevertheless says, "aculeo recondito punctorio," perhaps only from analogy.

*

Klug, under the name of Masaris Hylæiformis Klug MSS., from Egypt; the antennæ are very short, as they are in both sexes of Celonites, the males of which have the abdomen terminated by two deep notches.

The second family, VESPIDE (Polistides St. Fargeau), is here restricted to those species (including the common wasp and hornet) which live in temporary societies, consisting of males, females, and workers or neuters (fig. 88. 1. V. vulgaris ♂, 2—6. details of ditto;

fig. 88. 7. female, 8—16. details of ditto, 17. and 18. neuter†). The upper lip is concealed by the clypeus; it is very much narrowed in front into a tongue-like piece (fig. 88. 10.); on the under side of it is attached a membranous plate rounded in front, which Savigny terms the epipharynx (fig. 88. 11.). The clypeus is nearly square, and but slightly produced in front; the mandibles are not longer than broad; their extremity is broadly and obliquely truncate and toothed; in V. vulgaris (fig. 88. 2. ♂, 12. f, 18. 9) they are formed alike in all the individuals; they are, however, differently coloured in the male, and clothed with much longer hairs. The maxillæ and labium are but

My specimen is also a female. Should the male agree with Celonites in having short antennæ, the species will require a distinct subgenus for its reception.

† The individuals here figured and dissected were obtained from the same nest in a bank. The observation of M. Perrot, recorded by Huber (Nouv. Obs. tom. ii. p. 443.), and quoted by Kirby and Spence ( Int. vol. ii. p. 108.), of a kind of females not larger than the workers, and which produce only male eggs, requires confirmation.