When we read some of the more recent works on anthropology, the primordial savage seems to be not unlike one of those hideous india-rubber dolls that can be squeezed into every possible shape, and made to utter every possible noise. There was a time when the savage was held up to the civilised man as the inhabitant of a lost paradise—a being of innocence, simplicity, purity, and nobility. Rousseau ascribed to his son of nature all the perfection which he looked for in vain in Paris and London. At present, when so many philosophers are on the lookout for the missing-link between man and beast, the savage, even if he has established his right to the name of man, cannot be painted black enough. He must be at least a man who maltreats his women, murders his children, kills and eats his fellow-creatures, and commits crimes from which even animals would shrink. This devil-savage, however, of the present anthropologist is as much a wild creation of scientific fancy as the angel-savage of former philosophers. The true Science of Man has no room for such speculations. Sometimes the history of a name can take the place of its definition, but this is hardly so in our case. The Greeks spoke of barbarians rather than of savages, and the Romans followed their example, though they might possibly have called the national heroes and sages of Germany and Britain not only _barbari_ but _feri_—that is, savages not very far removed from _feræ_, or wild beasts. Our own word _savage_, and the French _sauvage_, meant originally a man who lived in the woods, a _silvaticus_. It was at first applied to all who remained outside the cities, who were not _cives_, or civilised, and who in Christian times were also called _heathen_—that is, dwellers on the heath.