AT ITS SIMPLEST, forecasting is no more than the intelligent extrapolation of past trends. Such a process presumes that the underlying economic and social conditions of the past will remain sensibly unchanged in the future. Obviously in present times this could not be presumed so and other methods have to be used. Before discussing these, or indeed actually employing any forecasting technique, it is necessary to establish the present nature of the thing to be forecast; in this case surface transport in Africa. For a mixture of historical and physical reasons, surface transport in Africa is dominated by the rail and road systems. Although the waterways were important avenues for the initial penetration of the continent, with the exception of the Zaire river system and the Sudanese sections of the Nile, they have relatively little significance as a commercial means of transport. The reasons for this are mostly physical: troublesome entry to the rivers from the sea because of sand bars; extreme seasonal fluctuations in river water levels; and frequent fierce currents and unnavigable cataracts. The great inland lakes are a significant means of transport locally, and in isolated cases are important links between the rail services of different countries, but they are generally too remote either from the sources of exports or population concentrations to attract much traffic. Similarly pipelines have yet to make a significant contribution to total transport movements.