One or the other of these modes of construction was employed according to the nature of the bed of the lake. In lakes with a muddy bottom, the first plan could be easily employed; but when the bed was rocky, it was necessary to have recourse to the second. This is the reason why on the northern shore of the Lake of Neuchâtel, where the banks of limestone come very close to the surface, a comparatively large number of ténevières may be observed.

These are the facts as generally noticed, especially in wide and deep lakes; the edifice, however, was not always constructed in this mode. In marshes and small lakes, which have now become peat-bogs, another system was frequently applied, a remarkable instance of which is furnished by the peat-moss at Wauwyl. In this locality were found several quadrangular spaces very distinctly enclosed by piles, between which were raised as many as five platforms one above the other. These piles are naturally very long, and some are buried as much as seven feet in the solid ground—an operation which must have required an enormous amount of labour. The intervals between the platforms are filled up with boughs of trees and clay, and the floors themselves are made in nearly the same way as those we have before mentioned. The lowest rested directly on the bed of the lake, and on the upper one the huts were placed.

It is sometimes the case that these heaps of stones rise above the water; they then form perfect artificial islands, and the habitations which covered them are no longer, properly speaking, dwellings on piles. Of this kind is the station on the Lake of Inkwyl in Switzerland; of this kind, also, are the crannoges of Ireland, of which we shall subsequently make special mention. Some of these artificial islands have braved the destructive action of ages, and are still inhabited at the present time. M. Desor mentions the Isle of Roses in the Lake of Starnberg (Bavaria) which has never been known to have been unfrequented by man; it now contains a royal residence.

Let us revert to the mode of construction of the aquatic dwellings of Switzerland.

In all probability the stones used were conveyed to the required spot by means of canoes made of hollowed-out trunks of trees. Several of these canoes may still be seen at the bottom of Lake Bienne, and one, indeed, is still laden with pebbles, which leads us to think that it must have foundered with its cargo. But it is very difficult to raise these canoes from the bottom, and it is, besides, probable that when exposed to the open air they would fall to dust. Nevertheless, one of them is exhibited in the Museum at Neuchâtel.

In the Museum at Saint-Germain there is a canoe very similar to that of Neuchâtel. It is made out of the trunk of a hollow tree. A second canoe, very like the first, but with the bark still on it, and in a bad state of preservation, lies in the entry of the same Museum of Saint-Germain. It was taken out of the Seine, as we stated when speaking in a previous chapter of the first discovery of the art of navigation during the Stone Age.

It may very easily be explained how the constructors went to work in felling the trees and converting them into piles. M. Desor has remarked that the pieces of wood composing the piles are cut cleanly through round their circumference only; the central part shows inequalities just like those which are noticed when a stick is broken in two by the hand after having been cut into all round the outside. The builders of the lacustrine villages, therefore, when they wanted to fell a tree must have acted much as follows: having cut all round it to a depth of 3 or 4 inches, they fixed a cord to the top, and broke the tree down by forcibly pulling at the upper part. They then cut it through in the same way with stone or bronze hatchets, giving it the requisite length, hewing it into a point at one end so that it should more easily penetrate the mud. Sometimes a fire applied to the base of the tree prepared for, and facilitated, the effect of the sharp instruments used. A great number of the piles that have been found still bear the marks of the fire and the cuts made by stone hatchets. In constructing the ténevières, the labour of pointing the piles was needless, as the latter were thoroughly wedged in by the accumulation of stones of which we gave a representation in fig. 148.

When the piles were prepared, they had to be floated to the spot fixed upon for the village, and to be fixed in the bed of the lake. If we consider that, in many cases, the length of these piles reached to as much as 16 or 20 feet, some idea may be formed of the difficulty of an undertaking of this kind. In the construction of the ténevières much thicker piles were used, and the labour was much less difficult. For instance, in the more ancient ténevières of the Lake of Neuchâtel piles are found made of whole trunks of trees which measure 10 to 12 inches in diameter.

The mind is almost confused when it endeavours to sum up the amount of energy and strong will which the primitive population of Switzerland must have bestowed on constructing, unaided as they were by metal implements, the earliest lacustrine settlements, some of which are of very considerable extent. The settlement of Morges, one of the largest in the Lake of Geneva, is not less than 71,000 square yards in area. That of Chabrey, in the Lake of Neuchâtel, measures about 60,000 square yards; another, in the same lake, 48,000 yards; and, lastly, a third, that of La Tène, 36,000 yards. There are many others which are smaller, although of respectable dimensions.