The animal has the instinct to bury itself in the mud of the bottom, on the approach of the droughts, penetrating to a depth of several feet. There it coils itself into a ball, with the tail folded over the nose, but so as to leave the nasal apertures uncovered; and, probably by its wrigglings, it forms a cavity or chamber in the clay, which becomes lined with a membranous slough thrown off from its body. Meanwhile the water evaporates, the mud dries, bakes, and cracks under the torrid heat of the dry season, thus allowing air to penetrate down to the retreat of the torpid mud-fish, in sufficient quantity for its very sluggish respiration. Here it lies inactive for five or six months, until the wet season again sets in, and the returning floods cover the old beds, soften the baked clay, revivify the imprisoned animal, and restore it to liberty and aquatic locomotion. To meet these strange conditions of life, the _Lepidosiren_ is furnished with a twofold apparatus for respiration; the one aquatic, consisting of gills, ordinarily contained in a branchial chamber, (but in one species, at least, external,) suited for the separation of oxygen from the water, and the other aerial, consisting of true lungs, closely resembling those of serpents, though manifestly only a modification of the well-known swim-bladder of many fishes,--by means of which the animal breathes atmospheric air, during its periodic captivity. The same emergency is met by other species in another way. It does not appear that the _Lepidosiren_ has the power of voluntarily forsaking the water, or of travelling on land, notwithstanding its twofold respiration; but some of the fishes of the tropics certainly resort to this mode of evading the fatal contingency of being baked out by the evaporating power of the periodical dry season. Theophrastus, the contemporary of Aristotle, mentions fishes found in the Euphrates which in the dry seasons leave the vacant channels and crawl over the ground in search of water, moving along by fins and tail.[81] Pallegoix gives three kinds of fish in Siam, which leave the tanks and channels and travel through the grass;[82] and Sir John Bowring states that in ascending the river Meinam to Bangkok, he was amused with the sight of fish leaving the stream, gliding over the wet banks, till they disappeared among the trees of the jungle.[83] The _Hydragyræ_ of Carolina in like manner leave the drying pools, and seek the nearest water in a straight line, though at a considerable distance. And Sir R. Schomburgk tells us that certain species of _Dora_ in Guiana have the same habit, and are occasionally met with in such numbers in their terrestrial travels that the negroes fill baskets with them.[84]