Another cause helps. In early states of civilisation there is a great mortality of infant life, and this is a kind of selection in itself--the child most fit to be a good Spartan is most likely to survive a Spartan childhood. The habits of the tribe are enforced on the child; if he is able to catch and copy them he lives; if he cannot he dies. The imitation which assimilates early nations continues through life, but it begins with suitable forms and acts on picked specimens. I suppose, too, that there is a kind of parental selection operating in the same way and probably tending to keep alive the same individuals. Those children which gratified their fathers and mothers most would be most tenderly treated by them, and have the best chance to live, and as a rough rule their favourites would be the children of most 'promise,' that is to say, those who seemed most likely to be a credit to the tribe according to the leading tribal manners and the existing tribal tastes. The most gratifying child would be the best looked after, and the most gratifying would be the best specimen of the standard then and there raised up. Even so, I think there will be a disinclination to attribute so marked, fixed, almost physical a thing as national character to causes so evanescent as the imitation of appreciated habit and the persecution of detested habit. But, after all, national character is but a name for a collection of habits more or less universal. And this imitation and this persecution in long generations have vast physical effects. The mind of the parent (as we speak) passes somehow to the body of the child. The transmitted 'something' is more affected by habits than, it is by anything else. In time an ingrained type is sure to be formed, and sure to be passed on if only the causes I have specified be fully in action and without impediment. As I have said, I am not explaining the origin of races, but of nations, or, if you like, of tribes. I fully admit that no imitation of predominant manner, or prohibitions of detested manners, will of themselves account for the broadest contrasts of human nature. Such means would no more make a Negro out of a Brahmin, or a Red-man out of an Englishman, than washing would change the spots of a leopard or the colour of an Ethiopian. Some more potent causes must co-operate, or we should not have these enormous diversities. The minor causes I deal with made Greek to differ from Greek, but they did not make the Greek race. We cannot precisely mark the limit, but a limit there clearly is.