character in cattle is thus suppressed below its otherwise natural
standard by the influence of wild beasts, is shown by the greater
display of self-reliance among cattle whose ancestry for some
generations have not been exposed to such danger.

What has been said about cattle, in relation to wild beasts, applies
with more or less obvious modifications to barbarians in relation to
their neighbours, but I insist on a close resemblance in the
particular circumstance, that many savages are so unamiable and
morose as to have hardly any object in associating together, besides
that of mutual support. If we look at the inhabitants of the very
same country as the oxen I have described, we shall find them
congregated into multitudes of tribes, all more or less at war with
one another. We shall find that few of these tribes are very small,
and few very large, and that it is precisely those that are
exceptionally large or small whose condition is the least stable. A
very small tribe is sure to be overthrown, slaughtered, or driven
into slavery by its more powerful neighbour. A very large tribe
falls to pieces through its own unwieldiness, because, by the nature
of things, it must be either deficient in centralisation or
straitened in food, or both. A barbarian population is obliged to
live dispersedly, since a square mile of land will support only a
few hunters or shepherds; on the other hand, a barbarian government
cannot be long maintained unless the chief is brought into frequent
contact with his dependants, and this is geographically impossible
when his tribe is so scattered as to cover a great extent of
territory. The law of selection must discourage every race of
barbarians which supplies self-reliant individuals in such large
numbers as to cause tribes of moderate size to lose their blind
desire of aggregation. It must equally discourage a breed that is