Geographical Journal (Journal of the Royal Geographical Society): Volume VIII, No. 5 [‎44v] (91/154)

494
RAILWAYS IN AFRICA.
We niayj however, safely conclude that we must as far as possible avoid both
dense forests and sandy and rocky wastes in the construction of our first railways.
Then, as to the lines of communication, considered as a whole, rail and river
combined, we must obviously, if any capital is to be expended, make them in the
directions most likely to secure a profitable traffic. In considering this part of the
question, it will be seen that there are several different problems to be discussed:
( 1 ) trade with the existing population in their present condition; ( 2 ) trade with
the native inhabitants when their countries have been further developed with the
aid of European supervision; and (3) trade with actual colonies of European settlers.
To many minds the last of these problems will appear to be the most important,
and in the end it may prove to be so. But the time at my disposal compels me
to limit myself to the consideration of trade with the existing native races within
the tropics, with only an occasional reference to the influence of white residents.
We must, no doubt, carefully consider which are the localities most likely to attract
those Europeans who go to Africa with the view of establishing commercial inter
course and commercial methods in the interior; and there can be no doubt that
considerations of health will play a prominent part in deciding this point. More
over, as the lowest types of natives have few wants, the more primitive the inhabi
tants of the district opened up, the less will be the probability of a profitable trade
being established. For both these reasons the coast districts are not likely in the
end to be as good a field for commercial enterprise as the higher lands in the
interior; for the more we recede from the coast, the less unhealthy the country
becomes, and the more often do we find traces of native civilization. To put it
simply, we must consider both the density of the population and the class of
inhabitant in the districts proposed to be opened up. Of course, the exact nature
of the products likely to be exported, and the probability of demands for European
goods arising amongst the natives of different districts, are vitally important con
siderations in estimating the profits of any proposed line of railway; but to discuss
such problems in commercial geography at length would open up too wide a field on
an occasion like this.
If the importance of considering the density of the population in the different
districts in such a preliminary survey is admitted, we may then simplify our
unquiry by declining to discuss any lines of communication intended to open up
regions where the population falls below some fixed minimum—whatever we may
like to decide on. Of course, the question of the greater or less probability of a
locality attracting white temporary residents is very important, but unless there is
a native population ready to work on, there will be little done for many years to
come. Politically, it may or may not be right to open up new districts by railways
for the sake of finding outlets for our home or our Indian population; but here I
. am considering the best lines for the development of commerce, taking things as
they are. What, then, shall be this minimum of population? The population of
Bengal is 470 per square mile; of India, as a whole, about 180; and of the United
States, about 21 or 22. If it is remembered that the inhabitants of the United
iStates are, per head, vastly more trade-producing than the natives of Africa, it will
be admitted that we may for the present exclude from our survey all districts in
which the population does not reach a minimum of 8 per square mile; it might be
right to put the minimum much higher than this. On the map now before you,
the uncoloured parts show where the density of population does not come up to
this minimum, and we can see at a glance how enormously this reduces the area
to be considered. The light grey indicates a population of from 8 to 32 per square
mile, and the darker grey a denser population than that. Of course, such a map,
in the very imperfect state of our knowledge, must be very inaccurate, as I am