They used barbed spearheads of bone that were later supplanted by beautifully precise harpoons with three or more barbs and a perforation through the butt; and the nearest parallels to these harpoons of the Nile Valley occur at sites in the Wadi Azaouak, two thousand miles to the westward across the grim Sahara that we know today. Even as late as the third millennium large numbers of cattle are known to have found grazing in lower Nubia where, as Arkell says, "desert conditions are so severe today that the owner of an ox-driven water- wheel has difficulty in keeping one or two beasts alive throughout the year." And anyone who has traveled in these dusty latitudes will have noticed how the wilderness of sand and rock that lies to the west of the Nile, far out upon the empty plains, is scored with ancient wadi beds which must once have carried a steady seasonal flow of water, but are now as dry as the desert air.

The immediate reasons for this long and ruthless desiccation, which is still going on, are unknown; they belong, clearly enough, to the same grand order of events which pushed the tropics southward through the ages, governed the advancing and retreating ice, and set the course of storm and cyclone in prehistoric times. The important point, at any event, is that the Sahara began to offer a major barrier to human passage some five or six thousand years ago - - at about the same time, that is, as black peoples began to move and multiply, and North Africa began to develop settled agriculture. In thus stretching a barrier to human contact between the lands to the north and the lands to the south, this Saharan wilderness would deeply influence the course of human development in Africa.

North of this worsening desert there was intense and seldom interrupted contact between all the developing societies and civilizations of North Africa, the Middle East, and the Mediterranean. South of the desert there was more or less unrestricted movement throughout the continental mainland, so that Negro or Negroid people are found today in almost every part of it. But the south and the north were increasingly divided from one another. They developed apart. They developed differently.

This broad truth is subject to reservations. Contact between north and south was never completely severed. Raiding and trading and migration routes led southward from the Fezzan to the Niger, or coastwise down the Red Sea and round the eastern horn of Africa. Carthage traded down the western coast, though Phoenician secrecy has stopped posterity from knowing how much or how far. Horses and chariots were common in the Sahara for several centuries after about 1200 B.C., and later on there was the camel. Yet the trails across the desert were forbiddingly hard to follow, and long and hazardous. Even medieval Arabs, riding across the Sahara from established well to well, would need two months for the journey, and many who set forth would never reach their goal.

This is not to say, of course, that without the desiccation of the Sahara the growth of human society in continental Africa would have followed a Mediterranean pattern. This vast and various continent must always and in any case have developed irregularly, unequally, some peoples being ahead of others, for the nature of the country, its forests and its plains, healthy uplands and malarial swamps, abundance of some forms of vegetation and complete lack of others, would always have imposed irregular and unique patterns of development.

Yet desiccation of the Sahara is none the less important in the record. North of the desert the civilizations of the Fertile Crescent were free to act and react upon one another, piling invention upon invention, exerting ever- renewed pressures of rivalry upon themselves and their neighbors until, in the course of centuries, they moved from primitive beginnings to the hierarchical splendors of an age of bronze. South of the desert there penetrated even to the peripheral peoples of that time no more than the faint and puzzling echo of this northern ferment: the echo was lost, the effect was vain.

To ask why early civilization should have appeared in the Nile Valley, the Near East, Mesopotamia -- and not in northern Europe or in southern Africa- may be an interesting exercise in speculation. At this stage in knowledge, it can be little more. River-valley cultivation seems to be the clue. Early civilizations all took their rise in great river valleys; and these, no matter how much they otherwise differed, all had the peculiar characteristic of natural irrigation and soil renewal. Annually these rivers offered new soil, exceptionally rich soil, for cultivation. They enabled nomadic man, then discovering the possibility of growing food rather than merely collecting or hunting it, to turn from his nomadic life. In so doing -- in settling in one place for several years at a stretch -- he was faced with the technical problems of regular cultivation. And in solving these problems -- precisely where river annually offered new soil -- he also solved the problem of growing a surplus of food.

And with the emergence of this hitherto unknown phenomenon of surplus food, there emerged the foundations of commerce. But trade was the foundation, in turn, for permanent settlement; and permanent settlement meant specialization, the division of labor, the growth of cities. And the growth of cities meant civilization, the development of central government -- of the autocratic and often divine rule that was peculiar to Bronze Age Egypt and other ancient civilizations. And with that the conditions were present when calculation was required, if only to count the goods that priests of the Pharaoh piled in his granaries and storehouses; and it was early means of calculation which led in turn to means of writing. Much of this complex and anything but automatic machinery of growth has been charted by archeologists over the past fifty years of revolutionary discovery. If the exact procedures are still in question, the general nature of the process is accepted.

South of the desert, largely severed from communication with the civilizations of the ancient world, things went differently. Conditions of river valley settlement that were decisive in the Middle East, India, and China appear to have failed in continental Africa. Not only that, so vast was the land that need for a surplus of food at any one place was absent too. Early peoples, running short of game, simply moved elsewhere. And whenever agriculture and, later on, metal age technique produced a greater density of population than any given area of land would support, the same thing happened again: sub-tribe hived off from parent tribe and marched away to new land.

Often enough it would move to virgin land. Sometimes it would collide with earlier migrants or earlier nomads, and then the shunting process would begin once more, until waves and tremors of new migration would slowly ripple out across the forests and the plains. To this simple picture there were large and obvious exceptions. Yet it is a picture that is worth holding in mind because it helps to explain both means and motive in the peopling of historic Africa. Many tribal histories are known by now; invariably they include the story of migration and new settlement. Often enough they tell of movement from the northward or eastward, and the general trend of migration was very probably from north to south. Thus the picture south of the desert is one of restless ever-quickened movement across a continent where no great mountain ranges or unflankable deserts ever interposed a lasting check. Even dense forests that wall the Congo basin witnessed this onward penetration of nameless tribes in times beyond memory. They moved like the unseen armies of the stars, southward, westward; then, as time passed, returning on their tracks, eastward, northward, in hidden orbits that we do not know.