orse and develop the
characteristic culture of the nomad type. Yes, but why did man tame
the horse later rather than sooner? And why did the American redskins
never tame the bison, and adopt a pastoral life in their vast prairies?
Or why do modern black folk and white folk alike in Africa fail to
utilize the elephant? Is it because these things cannot be done, or
because man has not found out how to do them?

When all allowances, however, are made for the exaggerations almost
pardonable in a branch of science still engaged in pushing its way
to the front, anthropo-geography remains a far-reaching method of
historical study which the anthropologist has to learn how to use.
To put it crudely, he must learn how to work all the time with a map
of the earth at his elbow.

First of all, let him imagine his world of man stationary. Let him
plot out in turn the distribution of heat, of moisture, of diseases,
of vegetation, of food-animals, of the physical types of man, of
density of population, of industries, of forms of government, of
religions, of languages, and so on and so forth. How far do these
different distributions bear each other out? He will find a number
of things that go together in what will strike him as a natural way.
For instance, all along the equator, whether in Africa or South America
or Borneo, he will find them knocking off work in the middle of the
day in order to take a siesta. On the other hand, other things will
not agree so well. Thus, though all will be dark-skinned, the South
Americans will be coppery, the Africans black, and the men of Borneo
yellow.

Led on by such discrepancies, perhaps, he will want next to set his
world of man in movement. He will thereupon perceive a circulation,
so to speak, amongst the various peoples, suggestive of interrelations
of a new type. Now so long as he is dealing in descriptions of a detached
kind, concerning not merely the physical environment, but likewise
the social adjustments more immediately corresponding thereto, he will
be working at the geographical level. Directly it comes, however, to
a generalized description or historical explanation, as when he seeks
to show that here rather than there a civilization is likely to arise,
geographical considerations proper will not suffice. Distribution is
merely one aspect of evolution. Yet that it is a very important aspect
will now be shown by a hasty survey of the world according to
geographical regions.