The other type of trap-door is less interesting and much more elementary, consisting simply of a wafer-like sheet of silk mixed with earth and vegetable matter, but it is a curious fact that while all known trap-door nests of the cork type are simple tubes, the burrows with wafer doors are often much more complex. In some cases there is a branch tube, like that constructed by Lycosa picta, leaving the main tunnel at a depth of some three inches, and reaching the surface perhaps two inches away from the trap-door, so that the whole excavation is Y-shaped. This branch tube is permanently closed by a thin sheet of silk and earth, which, however, it would not be difficult to break through if it were urgent for the spider to escape while the enemy was exploring the main tunnel.
But a more interesting case is the occurrence of another trap-door some way down the tube. If the tube is unbranched, this forms merely a second line of defence if the outer door is forced, but in the case of a branched tube the additional door hangs at the fork of the Y, and is so shaped as to form a perfect valve, so that the spider, by holding it against one or the other side of the tunnel, can connect the bottom limb of the Y with either fork at will, leaving to the intruder a beautifully smooth-lined tube to explore, with no hint of the possibility of escape in other directions.
There are sometimes other complications in the ramification of the tube, but these need not detain us. Each species of spider adheres to its own particular type of architecture, and may safely—in a given neighbourhood—be identified by its nest.
As with the Lycosidae, the burrowing is all done by the mandibles, but here the first joint—the handle of the penknife—is of more importance than the blade or fang. Indeed the burrowing species of the Aviculariidae may be distinguished from the rest by their mandibles, which are provided in front with a rastellum, or row of teeth for digging. A trap-door spider, then, does not go to work like a rabbit, or a terrier, scratching and kicking away the earth as it digs; it laboriously dislodges particles of soil with its powerful mandibles, and carries away the loosened fragments to deposit them at a distance.
The trap-door spiders of the Mediterranean region are nocturnal creatures, and little is known of their habits. Erber relates that a species found in the island of Tinos comes out at night, fixes open the trap-door with a few threads, and spins a web near its nest to entrap passing insects, clearing away any trace of it before the dawn. In the case of some Chinese and also some Australian species observers allege that they frequently wander from their nests in the day-time.
A Californian species was able to leave its nest when the trap-door was weighted with three ounces of lead. On re-entering, it seized the edge of the door with its mandibles, and, raising it slightly, inserted its front legs. It then turned round and slipped backwards into the tube. It always resisted the forcible opening of its door to the last moment, when it let go and slid into the tube “as though going down a well.”
The larger Aviculariidae have acquired a reputation for feeding on birds, and this has given rise both to their scientific and their popular name—bird-eating spiders. Several travellers have stated that they have observed them with birds in their grasp, and there is no doubt of their ability to kill any small bird or mammal, though it is probable that they seldom have the opportunity, for they spin no snare in which birds may be caught. Even without the aid of their poison, their jaws are so large and powerful that they may easily attain the vital organs of small animals. Probably their staple food consists of the larger insects.
They live in holes in the ground or in trees, or sometimes in the fork of a tree-branch. In such hiding places they spend the hours of day-light, emerging at night in search of food. Their large size and uncanny appearance have attracted the attention of the collector, and a great many species are known, but the fact that they chiefly inhabit tropical countries has militated against any very extended study of their habits, and the few items of information we possess are best related with regard to the particular spider observed, and not taken as necessarily characteristic of the whole tribe. There is little doubt that they live for several years. McCook kept a specimen of Dugesiella hentzi in captivity for five and a half years, and he considered that when it reached him it was at least a year and a half old, and probably more. The same species has recently been made the subject of some very interesting observations by Petrunkewitch, who obtained numerous living specimens from Texas and kept them in captivity; unless carefully packed, they bore the railway journey badly, and it was above all things necessary to supply them with water.
The captives were fed on grass-hoppers, crickets, cockroaches and wolf-spiders, but they ate sparingly, one grass-hopper sufficing for three days in the summer, while in the winter hardly any food at all was taken.