Damara land; if less gregarious, they would be too widely scattered
to keep a sufficient watch against the wild beasts.

I now proceed to consider more particularly why the range of
deviation from the average is such that we find about one ox out of
fifty to possess sufficient independence of character to serve as a
pretty good fore-ox. Why is it not one in five or one in five hundred?
The reason undoubtedly is that natural selection tends to give but
one leader to each suitably-sized herd, and to repress superabundant
leaders. There is a certain size of herd most suitable to the
geographical and other conditions of the country; it must not
be too large, or the scattered puddles which form their only
watering-places for a great part of the year would not suffice; and
there are similar drawbacks in respect to pasture. It must not be
too small, or it would be comparatively insecure; thus a troop of
five animals is far more easy to be approached by a stalking
huntsman than one of twenty, and the latter than one of a hundred. We
have seen that it is the oxen who graze apart, as well as those who
lead the herd, who are recognised by the trainers of cattle as
gifted with enough independence of character to become fore-oxen.
They are even preferred to the actual leaders of the herd; they dare
to move more alone, and therefore their independence is undoubted.
The leaders are safe enough from lions, because their flanks and rear
are guarded by their followers; but each of those who graze apart,
and who represent the superabundant supply of self-reliant animals,
have one flank and the rear exposed, and it is precisely these whom
the lions take. Looking at the matter in a broad way, we may justly
assert that wild beasts trim and prune every herd into compactness,
and tend to reduce it into a closely-united body with a single
well-protected leader. That the development of independence of