[Illustration: FIG. 73.--Thigh-bone of child from Grotta dei Colombi (Capellini). _a_, Cuts; _b_, Outline of corresponding thigh-bone from cavern at Cefn.] In this cave, as in those quoted above, there are no polished stone implements, or works of art, that establish that these feasts were carried on in the cave by neolithic cannibals, for the rude flint-flakes and bone articles, taken by Professor Capellini to fix its date, are common both to the palæolithic and the bronze ages. Nevertheless, since the inhabitants have left behind no trace of any metal, and since their food was wholly supplied by the existing animals, they were probably in the neolithic stage of culture, if this be taken to cover the wide interval extending from the pleistocene to the age of bronze. They are proved, by the rudeness of their implements, to have been savages of a very low order. We may gather from various allusions, and stories scattered through the classical writers, such for example as that of the Cyclops, that the caves on the shores of the Mediterranean were inhabited by cannibals in ancient times. In the island of Palmaria we meet with unmistakeable proof that it was no mere idle tale or poetical dream. But we have no proof that cannibalism was universally practised at any stage in the history of man. All the caves of Europe, explored up to the present time, merely afford some three or four examples in the neolithic and bronze ages. In the pleistocene there is no instance which is devoid of doubt. This atrocious practice is therefore to be viewed as abnormal, and it probably became ingrafted into the religious ideas of the nations of antiquity from the horror by which it was surrounded, ultimately surviving in the form of human sacrifices to the offended gods. _General Conclusions as to Prehistoric Caves._ We have seen in the fifth and sixth chapters that the prehistoric caves which are so unimportant in the ages of bronze and iron, were used in the neolithic age throughout western Europe both for habitation and burial, and that they therefore offer us most valuable materials for working out the ethnology of Europe at that remote time. The two races of men, the remains of which they contain, are represented by the modern Basque and Berber on the one hand, and on the other by the Celt, and in Russia and Germany by the cognate Finn, Sclave, and Wend. And since all the human remains described in the present chapter, those of Cro-Magnon and possibly of the Grotta dei Colombi being exempted, belong to one of other of these types, they may be referred to the neolithic age with a high degree of probability. In the present stage of the inquiry, it is much safer to put them into a distinct class, apart from those to which we can assign a relative age with tolerable certainty.