to mention all the caves in which human remains have been found in association with, and apparently belonging to the same period as, those of the extinct Mammalia. We will only call attention to a few of those which have been most carefully studied, and in which the conclusions appear to be most satisfactorily established.

It is unnecessary to say that a great number of caves present evidence of having been inhabited during times long subsequent to those which we are now considering; but for the later stone age, as well as for all subsequent periods, we have, as has been already mentioned, other sources of information, and more satisfactory evidence than any which can be derived from the examination of caves.

Some writers, indeed, have gone so far as to question altogether the value of, what may be called cave-evidence. They have suggested that the bones of extinct animals may have lain in the caves for ages before the appearance of man, and that subsequently remains belonging to very different periods may have been mixed together and deposited where we now find them. This was indeed the conclusion arrived at by M. Desnoyers, even so recently as the year 1845, in his article on Bone-caves.* We trust, however, to be able to show that a satisfactory reply may be made to this argument.

During the last year M. Lartet, in company with Mr. Christy, has been examining with great care a number of small caves and rockshelters in the Dordogne, some of which had already attracted the attention of archaeologists.† These caves are particularly interesting, because, so far, at least, as we can judge from the present state of the evidence, they belong to M. Lartet's Reindeer period, and tend therefore to connect the later or polished stone age with the period of the river-drifts and the great extinct mammalia; representing a period about which we had not previously very much information. Those which have been most carefully examined are ten in number, viz., Laugerie, La Madelaine, Les Eyzies, La Gorge d'Enfer, Moustier, Liveyre, Pey de l'Azé, Combe-Granal, and Badegoule, most of which we have ourselves had the advantage of visiting.

The rivers of the Dordogne run in deep valleys cut through calcareous strata, and while the sides of the valleys in chalk districts

* Recherches Geologiques et historiques sur les cavernes, particularment sur les cavernes à ossements. Dictionnaire Universel d'Histoire Naturelle.

† De l'Origine et de l'Enfance des Arts en Périgord. Par M. l'Abbe Audierne. N.H.R.-1864.

2 E

are generally sloping, in this case, owing probably to the hardness of the rock, they are very frequently vertical.

Small caves and grottoes frequently occur; besides which, as the different strata possess unequal powers of resistance against atmospheric influences, the face of the rock is, as it were, scooped out in many places, and thus "rock-shelters" are produced. In very ancient times these caves and rock-shelters were inhabited by men, who have left behind them abundant evidences of their presence. But as civilization advanced, man, no longer content with the natural, but inconvenient, abode thus offered to him, excavated chambers for himself, and in places the whole face of the rock is honey-combed with doors and windows leading into suites of rooms, often in tiers one over another, so as to suggest the idea of a French Petra.

In the troublous times of the middle ages many of these no doubt served as very efficient fortifications, and even now some of them are still in use. At Brantôme we saw an old church which had been cut in the solid rock, and reminded us somewhat of the celebrated rock-cut temples of India. It is now used as a wine-cellar, and close to it is a whole row of cottages, to which the rock serves as a back and roof.

Apart from the scientific interest, it was impossible not to enjoy the beauty of the scene which passed before our eyes as we dropped down the Vezére. As the river visited sometimes one side of its valley, sometimes the other, so we had at one moment rich meadowlands on each side, or found ourselves close to the perpendicular and almost overhanging cliff. Here and there we came upon some picturesque old castle, and though the trees were not in full leaf, the rocks were in many places green with box and ivy and evergreen oak, which harmonised well with the rich yellow brown of the stone itself.

But to return to the bone caves. Remains of the cave-bear have been found at the Pey de l'Aze, of the cave-hyæna at Moustier, and separated plates of the molars of Mammoth have occurred at Moustier and at Laugerie, accompanied at the latter place by a piece of a pelvis. As regards the two first species, MM. Christy and Lartet evidently regard them as belonging to an earlier period than the human remains found in the same caves. The presence of the pelvis has been regarded as an evidence of the contemporaneity of the Mammoth with the Reindeer hunters of Laugerie, and it is certainly difficult to see why they should have brought a fossil bone into their

cave, more especially as the bones of elephants, from the looseness of their texture, are not well adapted for implements. Still MM. Christy and Lartet do not commit themselves to any opinion, having as they say, laid down "une loi de ne procéder dans nos inductions que par évidences incontestables."

As regards the Felis spelæa, a metacarpal bone belonging probably to this species, and bearing evident marks of knives was found in the cave of Les Eyzies.

Still, so far as the positive zoological evidence is concerned, the antiquity of the human remains found in these grottoes, rests mainly on the presence of the reindeer, as regards which the evidence is conclusive. The bones are all broken open for the marrow; many of them bear the marks of knives, and at Les Eyzies a vertebra was found which had been pierced by a flint flake. We can hardly believe that this can have been done during the life of the animal; but MM. Christy and Lartet are quite satisfied that the bone must have been fresh when it was thus transfixed. As we shall presently see, there is still more curious evidence that Man and the Reindeer were contemporaneous in this locality.

But in its negative aspect the zoological evidence is also very instructive. No remains have been found which can be referred to domestic animals. It is true that bones of the ox and horse occur, but there is no evidence that they belonged to domesticated individuals. Remains of the boar also are very rare, and if these animals had been domesticated we might have expected to find them in greater abundance. The sheep and goat are entirely wanting, and what is still more remarkable, even the dog is absent.

A glance at the collections made by MM. Christy and Lartet, or that of M. le Vicomte de Lastic from Bruniquel, will show that a very large proportion of the animal remains consist of teeth, lower jaws, and horns. Other bones do indeed occur, but they form a very small fraction of the whole. Yet we cannot attribute this to the presence of dogs, partly because no remains of this species have yet been discovered, partly because the bones which remain have not been gnawed, but principally because dogs eat only certain bones and parts of bones, as a general rule selecting the spongy portions, and rejecting the solid shafts.

Mr. Galton tells us, however, that some of the savage tribes of Africa, not content with the flesh of the animals which they kill, pound up also the bones in mortars, and then suck out the animal

juices contained in them. So also, according to Leems, the Danish Laplanders used to break up with a mallet all the bones which contained any fat or marrow, and then boil them until all the fat was extracted. The proportions of the different bones afford us, I think, indirect evidence that a similar custom prevailed among the ancient inhabitants of Southern France.

Passing on now to the flint-implements found in these caves, we must first call attention to their marvellous abundance. Without any exaggeration they may be said to be innumerable. Of course this adds greatly to the value of the conclusions, but it need not surprise us, because flint is so brittle, that implements made of it must have been easily broken, and, in that case, the fragments would be thrown away as useless, especially in a chalk district, where the supply of flint would, of course, be inexhaustible. Many implements, no doubt, would be left unfinished, having been rendered useless, either by some misdirected blow, or some flaw in the flint. Moreover, we should naturally expect that in a bone-breccia of this nature, the flint-implements would be relatively more abundant than in a Kjökkenmödding. Each oyster furnishes but a single mouthful, and the edible portions evidently form a greater proportion of the whole, in the Mammalia than in the Mollusca. The Kjökkenmöddings, therefore, would grow ceteris paribus, more rapidly than the bone-breccia, and supposing the flint-implements to be equally numerous in both cases, they would, of course, be more sparingly distributed in the former, than in the latter.

The objects of stone found in the bone caves which we are now considering, are flakes, both simple and worked, scrapers, cores, awls, lance-heads, cutters, hammers.

The simple and worked flakes are, of course, very numerous, but they do not call for any special observations. They present the usual varieties of size and form.

Though less numerous than the flakes, the scrapers are still very abundant. On the whole they seem to me longer and narrower than the usual Danish type. Some of them were probably intended to be used in the hand, as both ends are fashioned for scraping. These may be called double-scrapers. Others were apparently fixed in handles, as the end opposite to the scraper is broken, sometimes on one side, sometimes on both, so as to form a tapering extremity,

Account of Danish Lapland, by Leems, Copenhagen, 1767. Translated in Pinkerton's Voyages, Vol. I., p. 396.

which may have been fixed in a handle either of wood, bone, or horn. Perhaps, as no trace of such a handle has yet been discovered, wood was the material used for this purpose. Many of the flakes are also nipped off at one end, in the same manner.

Of course, where there was a manufactory of flint flakes, the cores or nuclei, from which they were struck, must also be present. I was, however, astonished at the number of them; during my short visit, I myself picked out more than ninety of them.

Awls and saws are very much less frequent, but some few good specimens have been found. For hammers, the reindeer hunters seem to have used round quartz stones, a good many of which occur in the caves. These may, however, have served also as heaters. We know that the North American Indians, having no pottery but only wooden vessels, and being, consequently, unable to put them on the fire, used to heat stones, and then place them in the water which they wished to boil. Three or four flint hammers have also been met with.

These, the commonest sorts of flint-implements, are found indiscriminately in all the grottoes, but there are some other types which appear to be less generally distributed. Thus, at Laugerie and Badegoule, fragments of leaf-shaped lance-heads, almost as well worked as some of those from Denmark, are far from uncommon. If, therefore, we were to attempt any classification of the grottoes, according to the periods of their occupation, we might be disposed to refer these to a somewhat later period than most of the others. On the contrary, to judge from the flint-implements, the station at Moustier would be the most ancient. Though it would perhaps be premature to attempt any such classification, there can be no doubt that Moustier presents some types not yet found in the other caves.

One of these peculiar forms has one side left rough, and apparently intended to be held in the hand, while the other has a cutting edge, produced by a number of small blows. Some of these instru