Thus we see the use of a sort of 'preliminary' age in societies, when trade is bad because it prevents the separation of nations, because it infuses distracting ideas among occupied communities, because it 'brings alien minds to alien shores. And as the trade which we now think of as an incalculable good, is in that age a formidable evil and destructive calamity; so war and conquest, which we commonly and justly see to be now evils, are in that age often singular benefits and great advantages. It is only by the competition of customs that bad customs can be eliminated and good customs multiplied. Conquest is the premium given by nature to those national characters which their national customs have made most fit to win in war, and in many most material respects those winning characters are really the best characters. The characters which do win in war are the characters which we should wish to win in war. Similarly, the best institutions have a natural military advantage over bad institutions. The first great victory of civilisation was the conquest of nations with ill-defined families having legal descent through the mother only, by nations of definite families tracing descent through the father as well as the mother, or through the father only. Such compact families are a much better basis for military discipline than the ill-bound families which indeed seem hardly to be families at all, where 'paternity' is, for tribal purposes, an unrecognised idea, and where only the physical fact of 'maternity' is thought to be certain enough to be the foundation of law or custom. The nations with a thoroughly compacted family system have 'possessed the earth,' that is, they have taken all the finest districts in the most competed-for parts; and the nations with loose systems have been merely left to mountain ranges and lonely islands. The family system and that in its highest form has been so exclusively the system of civilisation, that literature hardly recognises any other, and that, if it were not for the living testimony of a great multitude of scattered communities which are 'fashioned after the structure of the elder world,' we should hardly admit the possibility of something so contrary to all which we have lived amongst, and which we have been used to think of. After such an example of the fragmentary nature of the evidence it is in comparison easy to believe that hundreds of strange institutions may have passed away and have left behind them not only no memorial, but not even a trace or a vestige to help the imagination to figure what they were.