two or three inches. But this is altogether a very large nest, larger than any I have seen elsewhere.

The embryo hornets' nest is somewhat different from that of the other species, it is shorter and clumsier, and my specimen, taken from the roof of a stable, wants an outer cap.

The nest of V. vulgaris (Plate VIII.) is very similar to the hornets' nest, both in the nature of the materials employed and the mode of its construction. But there is no difficulty in distinguishing the miniature from the coarse work of the larger insect. The fragments of wood are smaller, and the colour much more varied; and the whole structure affects rather a globular than an oval form. The cells of which the outer case is made are small, short, and irregular, quite unlike the long tunnels which the hornet builds. Their vaulting is made of shelly patches, the lines in which are disposed in concentric curves of a short radius, with the concavity mostly directed downwards, arching across the open ends of the vaults. These nests are friable to the last degree, and cannot bear any handling. There is nothing in wasp architecture more beautifully symmetrical than a large comb of V. vulgaris, nor more prettily varied than the case of the nest. And I am sorry to add that there is nothing of this kind more perishable.

The hornet prefers to build in a roof, or in the hollow of a rotten tree or post, above- or, sometimes, under-ground; V. vulgaris has a more varied taste, for, though she is an underground wasp by preference, yet she will build in roofs, and in the most out-of-the-way places, among which a dovecote, a

Professor

pump, and a sugar-loaf, are included. Edgeworth,* who has made this species his particular study, finds that' ninety per cent. of its nests are built in close proximity to those of the Bombus terrestris or agrestis. The neighbours are said to live on excellent terms. For whose benefit is the relation maintained? Does the wasp court the society of the bee, or the bee seek in the neighbourhood of the wasps, protection from her enemy the field-mouse?† A somewhat similar relation is said to exist between the mocking birds and the pasteboard wasps; the birds building over the wasps' nests to secure their young from the attacks of the monkeys.

V. germanica (Plate IX.) makes her nest either in the ground or hanging from a rafter. Underground, of course she adapts her work to the necessities. which projecting roots or stones involve; but, when she is at liberty to build in any direction the nest takes the form of a sphere flattened at the poles. The materials employed are vegetable fibres, which are felted together into a dull grey texture, wanting all the elegance of colour and arrangement of the nest of V. vulgaris. The same shelly patches indeed are to be seen, but in the mode of their arrangement they bear no constant relation to the whole work. They seem to be laid on anyhow, merely to cover space; and instead of springing in arches they lie nearly flat. Many of the patches are complete circles. These are all made from without inwards, in concen

*Notes on Irish Vespidæ, 'Annals and Magazine of Natural History,' Vol. XIII, p. 466. June, 1864.

+ Darwin 'On the Origin of Species.' p. 74.

tric lines, and a little hole in the middle marks the spot through which the last artificer withdrew her mandible when she put the finishing touch to her work.

I am writing with a singularly clean and beautiful specimen of the work of this wasp before me, which was taken from a cottage rafter at Warbleton. But this is quite an exceptional specimen, for the cases of all the other nests of V. germanica which I have seen dug out of the ground are most uninviting: they are rugged, dirty, without any attempt at arrangement of the work; and the dull grey hue of the paper, instead of being relieved by white lines, is generally made still deeper and duller by earthy soils and stains. The combs are not made so rigidly straight as those of V. vulgaris, but they are not curved as gracefully as those of the tree-wasps.

V. rufa (Plate X.), though a ground-wasp, builds her nest on the laminar, not on the cellular plan. But the structure has characters of its own, by which it may readily be distinguished from the tree nests which are built in this way. The successive layers are not perfectly free from each other, but tacked down here and there at the edges. And these edges are very numerous, marking the whole surface with frequent ridges, like the lines of sheep-tracks on a Down hill-side seen from a distance. I have only one specimen of this kind of nest, for which I am indebted to the kindness of Mr. F. Smith. This is in the form of a flattened sphere, pressed in equally at the top and bottom, though not quite to such an extent as the nest of V. germanica. The entrance is lateral,

single, oval, the edges very neatly finished off and not prolonged into a horn. The material is vegetable fibre.

The marginal cells of the comb diverge from the vertical direction to a slight extent, but the comb looks stiff and heavy, and this nest does not give an impression of any great constructive ability. The general plan is less bold, and the texture less close and firm than the work of the other species which build on the same plan. Beyond the fact that V. rufa is a ground-wasp, I cannot say where its nests are likely to be found. The barrenness of my cabinet is indeed a proof that I have never looked in the right place for it. It is said to be very common at Bournemouth. In Ireland it is rare. It is an early wasp, for while the other species are still in full activity, this wasp has done its work for the year and abandoned its nest.*

The nests of the tree-wasps, like those of the ground-wasps, have each their own distinctive character. For beauty of appearance none of them can be compared to the nest of V. vulgaris; but in their elegance of form and perfect adaptation to their purposes they are quite unequalled.

The queen of paper-makers is V. britannica (Plates V?, XI. XIII. XIV.). Her nest is distinguished from those which we have already noticed by the fact of the paper being made in large sheets, with long, nearly parallel, stripes running horizontally round the nest. It resembles in this respect the nest of * Edgeworth,' op. sup. cit. p. 472.

V. sylvestris, which also presents broad, smooth sheets. But the nest of V. britannica is of a thicker structure and more compact: the sheets are more numerous, closely overlying one another, and the edges are pressed down on, and often attached to, the sheet beneath them. From the specimens which I have seen, I am inclined to believe that the strong contrast of colour of the different stripes, indicating a very varied source of supply of materials, is highly characteristic of the work of this species. They are most determined builders; the remains of a swarm will immediately set about to replace the nest of which they have been robbed; and the neighbouring twigs, under these circumstances, garnished with scraps of paper, attest the misdirected energies which could not be repressed till the new nest was far enough advanced for all the wasps to find space to work at, or an end to join on to.

This species generally builds in low bushes or hedgerows, though I have one specimen from the top of an apple tree. The outline of the nest is pear-shaped, that is to say somewhat drawn out at the lower part where the entrance is situated. This is single as usual, and lateral, and instead of a simple hole there is often a little horn here, a kind of porch in which the sentinel, whom it is the habit of this species of wasp to post, mounts guard. Dissimilar as the mature nest is to that of V. germanica, yet in the embryo state the nests of the two species cannot be distinguished from each other. It is made of vegetable fibres.

The tree-wasps are less common and have not so