If we now go over to positive evidence, we see that the earliest traces of man, dating from the glacial or the early post-glacial period, afford unmistakable proofs of man having lived even then in societies. Isolated finds of stone implements, even from the old stone age, are very rare; on the contrary, wherever one flint implement is discovered others are sure to be found, in most cases in very large quantities. At a time when men were dwelling in caves, or under occasionally protruding rocks, in company with mammals now extinct, and hardly succeeded in making the roughest sorts of flint hatchets, they already knew the advantages of life in societies. In the valleys of the tributaries of the Dordogne, the surface of the rocks is in some places entirely covered with caves which were inhabited by palaeolithic men.(4) Sometimes the cave-dwellings are superposed in storeys, and they certainly recall much more the nesting colonies of swallows than the dens of carnivores. As to the flint implements discovered in those caves, to use Lubbock's words, "one may say without exaggeration that they are numberless." The same is true of other palaeolithic stations. It also appears from Lartet's investigations that the inhabitants of the Aurignac region in the south of France partook of tribal meals at the burial of their dead. So that men lived in societies, and had germs of a tribal worship, even at that extremely remote epoch. The same is still better proved as regards the later part of the stone age. Traces of neolithic man have been found in numberless quantities, so that we can reconstitute his manner of life to a great extent. When the ice-cap (which must have spread from the Polar regions as far south as middle France, middle Germany, and middle Russia, and covered Canada as well as a good deal of what is now the United States) began to melt away, the surfaces freed from ice were covered, first, with swamps and marshes, and later on with numberless lakes.(5) Lakes filled all depressions of the valleys before their waters dug out those permanent channels which, during a subsequent epoch, became our rivers. And wherever we explore, in Europe, Asia, or America, the shores of the literally numberless lakes of that period, whose proper name would be the Lacustrine period, we find traces of neolithic man. They are so numerous that we can only wonder at the relative density of population at that time. The "stations" of neolithic man closely follow each other on the terraces which now mark the shores of the old lakes. And at each of those stations stone implements appear in such numbers, that no doubt is possible as to the length of time during which they were inhabited by rather numerous tribes. Whole workshops of flint implements, testifying of the numbers of workers who used to come together, have been discovered by the archaeologists.