air. In the wasp this cohesion is further delayed; the fluid secretion requires a longer exposure to make it harden; the separate threads meanwhile melting into each other, so as to form a membrane rather than a web.

From the earliest period, as we shall see further on, the secretion of the salivary glands is more free in the wasp than in the honey-bee, their cocoon is thicker and stronger in all its details, increasingly so to the cap, the crown of larval labours. In the honey-bee this is reversed, and after the larval period her physiological energies are turned in another direction; she now makes wax instead of silk, the secretion of the salivary glands, never very abundant, is henceforth quite secondary to that of the abdominal fat masses, and loses all power of independent cohesion whatever.

Besides the salivary glands the thorax contains portions of the general circulatory, respiratory, and nervous systems, and the oesophagus on its way to the stomach. All these will be most conveniently examined in connexion with the abdomen, where, if not their most important, at least their most prominent portions, such as would be most likely to catch the eye of a general observer, are displayed. If exception be made to this arrangement in any particular it should be in reference to the respiratory apparatus, which is so highly developed in the thorax. Large air cells fill the spaces between the muscles, trachea ramify among their fibres with a minuteness not seen elsewhere, and it is the thoracic spiracles which are adapted in wasps to be the chief organs of

voice. Still, as, without the opportunity of studying it in the abdomen, we should know very little of their respiratory system, it seems best to defer its consideration till we can trace its development there.

We proceed, therefore, here to examine the external appendages of the thorax, the legs and wings. The legs of wasps are formed on the same plan as those of insects generally. The names of the different parts have been adopted from the limbs of Vertebrata. First comes a short flat joint, the coxa or hip, by which the limb is connected to the trunk. The next joint is a still smaller piece, which is let into the articulation, as it were, and thus gives a much greater extent of movement. This is the trochanter, on which the limb turns. The femur or thigh, which is so largely developed in the grasshopper and the flea, succeeds to this. It stands out almost horizontally from the body, and is probably the first in the series which would attract attention. Then comes the shank or tibia, the long straight bone which turns down towards the ground. The tibia of the two first pair of legs have only a single spine at the end furthest from the body, the distal end as it is called; while the hind pair have two of these spines: this is important to recollect. Otherwise they are unarmed; contrasting in this particular respect with the corresponding limbs of many of the sand-wasps, which are clothed quite down to the feet with strong hairs or spines.

The articulation between the thigh and shank bones is very strong. The form of the surfaces only allows of movement in one direction, being a hinge joint. The movement by which the limbs are turned

backwards or forwards is effected in the articulations between the smaller pieces which connect the thigh to the body; and the extent of this movement is much greater in the fore-legs than in the other limbs. These parts are all comparatively rigid, but the mode of arrangement of the tarsal joints, which come next under consideration, gives the limbs the requisite amount of elasticity.

The tarsi in wasps, as in all other Hymenoptera, are five-jointed. In wasps the first and last joints are longer than the rest; but none are of very disproportionate size. This circumstance, as already noticed, supplies an important distinctive character between wasps and bees, the first or proximal joint in bees, and particularly in the honey-bee, being much enlarged, flattened out and fitted with rows of hairs, called, from their use, pollen-brushes. Poets, into whose dreams the homologies of the insect skeleton

b

Fig. 3.-Comparative view of the posterior tarsi of the wasp and honey-bee.

a, tarsal joints of the honey-bee, with the first joint disproportionately larger than all the rest, and the tarsal hooks denticulated, as in the solitary wasps.

b, tarsal joints of the wasp, with the first joint narrow, the tarsal hooks simple, and the end of the tibia armed with two spines.

never entered, have incorrectly called this over-developed joint of the foot the bee's thigh. Each of the tarsal joints consists of a wedge-shaped piece, the point of which is received between the open horns of the preceding member. The ends of the horns of the last piece are not armed with spines,

but rounded off, so as to leave room for the free play of the foot, which forms a kind of sixth tarsal joint in which the limb ends.

The most prominent feature in the foot is a large pair of claws, one on each side. These, it should be particularly noticed, are simple in the Vespidæ, while in the Eumenidæ they are toothed in their concavity. This rule has no exceptions.* Between the tarsal hooks we find the pulvilli or cushions on which the insect treads. These are covered on their plantar surface with minute hairs, some of which are hooked at their extremities, and others are bulbous, reminding one of the bulbous suckers with which the arms of the star-fish are provided. It is by means of these bulbs that the foot clings to smooth and polished surfaces; but whether by atmospheric pressure only, or by the aid of a glutinous secretion, does not seem absolutely certain.† Though the increased difficulty with which a fly makes its way over a pane of glass damped by the breath seems strongly to favour the first opinion.

The insect sucker consists essentially of a hair, with its extremity dilated into a bulb and hollowed out in the centre. This is seen in the most perfectly developed form on the fore-legs of our great waterbeetle, Dyticus marginalis; and the same form of instrument, only very much reduced in size, is set over the pulvilli of the fly's foot, where there is no great difficulty in recognizing it. It requires, however, a high power of the microscope, and a careful

* De Saussure, 'Guêpes Solitaires,' tome I, p. xix.

+ See on all this subject, Tuffen West, On the Foot of the Fly, 'Linnæan Transactions,' Vol. XXIII, p. 393.

dissection, to display these suckers on the pulvilli of the wasp, which has these parts altogether smaller and much less distinct than many of the Diptera.

Indeed, the bulbed hairs on the wasp's foot, which are meant for suckers to act on smooth surfaces, do not seem to play as important a part in her progression as the little hooked hairs, which are mixed with them, and are as obviously meant for holding to rough surfaces. Wasps do not appear at all at ease on a polished surface: they cannot walk readily up a pane of glass, as their hold is less perfect, and their bodies are heavier in proportion to the area of their feet than those of flies. They walk much better on the rough surface of their own paper nests, where their tarsal hooks and the little hooked hairs of their pulvilli can be of use, than on the windowpanes. And they do not seem to take the same quiet monotonous pleasure-happily-as our domestic flies, in walking about the windows; it is a scramble up and a tumble down, and the sash-bar proves an obstruction only to be surmounted by the aid of their wings.

The wings of the Hymenoptera supply the characters by which this large Order of insects is scientifically distinguished from this others, and is familiarly recognized. Some of the most valuable specific distinctions between the various subordinate genera are also derived from the wings. But, for this purpose the fore-wings alone are employed. Whether the arrangement of the nervures in the hind-wings is as various and as constant for each subdivision as in the fore-wings I cannot say. At least, the varieties