An examination of the spinnerets throws some light upon the operation. The posterior pair are very long and mobile, and the hair-like spinning-tubes are distributed on their under surface. The cephalothorax and abdomen are far more rigidly connected in Agelena than in the Orb-weavers, but its length of leg and the length and mobility of its posterior spinnerets enable it to give a wide lateral sweep as it walks along, strewing fine silken threads upon the foundation-lines already laid down. Some hours elapse before even a moderately stout web is constructed, and for long afterwards the spider devotes odd moments of leisure in going over the ground again and strewing new silk upon the gradually thickening web. At one corner a silken tube of similar structure is formed, and in this the spider awaits the advent of any insect which may alight upon the sheet, when it immediately rushes forth and seizes it.

The webs of the Dictynidae are very similar in general appearance to those of the Agelenidae, consisting of a closely-woven sheet with a tubular nest. They are to be found, moreover, in similar situations, stretching across the angles of walls in cellars or outhouses, though some species prefer an out-door existence. Crannies in rock form convenient sites for such snares, but the family is not without its representatives in still more open situations. The web, though so similar to that of Agelena, is, however, constructed in a different manner. In the Dictynidae neither the legs nor the spinnerets are unusually long, and they do not strew the foundation-lines by a swinging motion of the body, but the operation is effected by a special apparatus. These spiders are cribellate, and in front of the six ordinary spinnerets there are a pair of perforated plates connected with a large number of additional minute spinning glands (see Fig. [182], p. 326). In conjunction with this, the female possesses on the last joint but one of each hind leg a curious comb-like arrangement of spines, the “calamistrum.” The animal constructs a sort of skeleton web by means of its ordinary spinnerets, and when this is completed it combs out silk from the cribellum by means of the calamistrum, using each hind leg alternately, and distributes it with a curling motion upon the scaffolding prepared for it, a nearly opaque web being the result. The silk from the cribellum is of an adhesive nature, and renders escape from the web very difficult.

Spiders’ Nests and Retreats.—All Spiders construct some description of nest, and often display great ingenuity in building them. Perhaps none are more curious than those of the burrowing Aviculariidae, a family which includes the interesting “Trap-door Spiders.” They are nocturnal in their habits, about which, consequently, little is known, but their nests have been carefully studied, especially by Moggridge, who found them in considerable abundance in various districts of the Riviera.

The jaws of these spiders are especially adapted for digging, and with them a hole is excavated in the ground to the depth of several inches, and wide enough to allow the animal to turn. This is carefully lined with silk which the spider throws against the sides from its long and upturned posterior spinnerets. But the chef d’œuvre of the whole structure is a lid or door which protects the entrance to the tube. There are two types of door which find favour with different species—the wafer and the cork type, as Moggridge has named them. The former consists of a thin circular or oval sheet of silk which flaps down loosely over the tube-entrance, with which it is connected by a hinge-like attachment. A trap-door of the cork type is a more complicated structure, being of considerable thickness and having a bevelled edge, so that it fits into the tube like a plug. Like the wafer door, it possesses a silken hinge.

To form the wafer door, the spider covers the entrance to the tube with a closely-woven layer of silk, which it afterwards bites away at the edge, except at the point where the hinge is to be. Doors of the cork type consist of alternate layers of silk and earth. After weaving a covering of silk, the creature brings earth in its jaws and lays it on the top, binding it down with a second layer of silk, and the process is repeated until the requisite thickness is attained.

The nests are exceedingly difficult to detect, as the spiders take the precaution of attaching leaves, moss, or small twigs to the outer surface of the doors. This does not appear to be the result of intelligence, but a mere instinctive habit; for if a door be removed and the surrounding earth denuded of moss, the spider will render the new door conspicuous by bringing moss from a distance, and thus making a green spot in the bare patch of earth.[[278]]

The cork doors fit with great exactness, and there is always to be found on their under surface a notch by which they are held down by the fore-legs of the spider against any attempt to open them from without.

Many nests with trap-doors of the wafer type are found to have a second and more solid door within the tube. This serves to shut off the lower part of the nest as a still more secure retreat. This second door opens downwards, and the Spider, getting beneath it, is effectually shielded from an enemy which may have mastered the secret of the outer barrier. The nests of some species present still further complications in the way of lateral branches from the main tube. In one case (Nemesia congener) the burrow becomes Y-shaped, and the second door hangs at the fork of the Y in such a manner as to connect the bottom chamber either with the entrance or with the branch, which does not reach the surface, but ends blindly.

Trap-door Spiders are greatly attached to their tubes, which they enlarge and repair at need. They begin burrowing very early in life, and their tiny tubes resemble in all respects those of their parents. Their habits are nocturnal, and little is known of them; an observation, however, on a species inhabiting the island of Tinos in the Grecian Archipelago (Cteniza ariana), by Erber,[[279]] must not be omitted. This spider leaves its tube at night and spins a web near at hand and close to the ground. It carries captured insects into its tube, and in the morning clears away the net, adding the material of it, M. Erber believes, to the trap-door.

No true trap-door Spider has as yet been found in this country, but the allied Atypidae are represented by at least one species, Atypus affinis, which has been discovered in colonies in some localities in the south of England, notably near Ventnor in the Isle of Wight, and on Bloxworth Heath in Dorsetshire. This spider, like its continental cousins, excavates a hole in the earth, generally near the edge of a heathery bank, and lines it with a tube of silk of such firm texture that it may be removed intact from the earth in which it is embedded. The silken tube projects some two inches above the ground, either erect among the roots of the heather, or lying loosely upon the surface. Its extremity is always found to be closed, whether from its own elasticity or by the deliberate act of the proprietor is uncertain, and it seems probable that the animal spends almost the whole of its existence in the tube. Simon believes that it feeds almost entirely upon earth-worms which burrow into its vicinity, and which it, therefore, need not leave its nest to catch; but the remains of beetles and earwigs have been found in the tubes at Ventnor.