GAULISH FORTRESSES ON THE COAST OF BRITTANV. al'J Hence it is pi'obable th;it tiio occupants descended into their abodes by means of ladders or wooden stairs. It is also to be remarked that each house is complete in itself, and in no case has been divided into two parts by a cross wall, as is frequently the case in similar dwellings in Wales and Scot- land. Excavations were made in ten houses, which led to the discovery of a great many objects, of v.'hich the following is an account : — I. — Twenty mill-stones, or rather large stones on which grain was crushed — of these seven are entire — the others have been more or less broken in early times. They are all of granite, and from variety of form may be divided into five classes. (1.) Two rather thin and long grinding-stones with straight sides. Their extremities are equally rounded and raised, and of the same thickness. The upper face hollowed out, somewhat like an English saddle, and the lower one smooth and rounded. The length slightly exceeds IGin. and 12 in., and the breadth is about 7 in. These mill-stones have been formed of flattened stones, the sides of which have been cut away to diminish the breadth. The depression on their faces has been produced by the friction of a mullcr, pushed backwards and forwards by the two hands. When used, the}'- were probably placed on the knees of the person crushing the grain.'^ (See fig. 1.) {■2.) Five stones, four of which are incomplete, thicker at one of their extremities than the other. The under side is slightly flattened, the upper one hollowed out, but in an oblique direction. These belong to the most common type of primitive implements. Mr. Albert Way has noticed several in his account of the objects discovered by the Hon. - Two similar millstones and one nuiller surrounded with eutreucLments, traces were found, two years ago, by M. Eriot in of which still remain. On the same es- digging up a wood on his e-tate of Kerla- tate, and within a -short distance from gattu, about two miles from Quimper. this fortified place, M. Briot, from twelve A bronze statuette of rude character, to fifteen years ago, destroyed several representing the god Mars, and many giaves, consisting of hollows of little portions of Gaulish vases enclosing burnt more than five feet long, covered with fht bonep, were also found at the same stones, the sides being formed of stones time. A few months back M. Grenot placed edgeways on the ground. These and myself found several flint chips and graves, which aie of a type common in quartz pebbles, which had served as per- l-'inistere, and which are. in fact, nothing cuflsors. The ground, in which all these but little dolmens entirely buried in the objects were brought to light, occupied ground, contained, at the time of their an elevated position, and was formerly discovery, some poUshotl stone celts. VOL. XXIX. Y Y
