The huts must have opened on to the platform by doors. Did they possess windows? Nothing is known as to this point. But in all probability there was an opening at the top of the roof, through which the smoke of the fire made its way. To avoid any fear of conflagration, a stone fire-place was placed in the middle of each dwelling. The daylight must have come in through the hole in the roof in a quantity almost sufficient to cause the absence of windows to be not much felt. In each habitation, there was, no doubt, a trap-door in direct communication with the lake, such as those which existed in the dwellings of the Pæonians described by Herodotus. Under this trap-door there was a reservoir made of osiers, intended for the preservation of fish. As the inhabitants of the lacustrine villages only lived upon the water with a view of increasing their security, it would be absurd to suppose that they would construct a large number of bridges between their aquatic settlement and the banks of the lake. There must have been, in general, but one bridge for each of these lake villages. How were the huts constructed, and what were their shape and dimensions? These questions certainly seem difficult to answer, for, as may be well imagined, no specimen of these ancient dwellings has been preserved to our days. Nevertheless, a few relics, insignificant in appearance, enable us to reply to these inquiries in a way more or less satisfactory. Everything seems to indicate that the huts were formed of trunks of trees placed upright, one by the side of the other, and bound together horizontally by interwoven branches. A coating of earth covered this wattling. It has been fancied, from the imprint left by some of the branches which were used in building these huts, that it might be inferred that they were circular, like those which historians attribute to the ancient Gauls. This was Troyon's opinion, and at first Dr. Keller's also. This author has even sketched a circular hut in a plate representing a restored lacustrine habitation, which accompanies one of his memoirs. Sir C. Lyell, also, has reproduced this same plate in the frontispiece of his work on the 'Antiquity of Man.' But Dr. Keller has subsequently abandoned this idea, and in another of his memoirs he has supplied a fresh design showing nothing but huts with flat or sloping roofs. From this latter plate, taken from Dr. Keller's work, we here give a representation of a Swiss lacustrine village (fig. 149). [Illustration: Fig. 149--A Swiss Lake Village of the Bronze Epoch.] The suggestions for this reconstructive sketch were furnished to Dr. Keller not only by various scientific indications, but also and especially by a drawing made by Dumont d'Urville among the Papuans of New Guinea. According to Dr. Keller, during the last century there still existed on the river Limmat, near Zurich, some fishermen's huts built in a similar way to those of the lacustrine villages.