But the piedmont zone and the river banks of the steppe must have been occupied by agriculturists before 10,000 B. C., probably much earlier. Pumpelly's explorations seem to warrant this view. Alongside of agriculture, but at a somewhat later date, sheep-herding and cattle-raising were practised. But the nomad of these days was a less dangerous neighbor than at later times because the horse had not yet been domesticated. During these post-glacial times he would be less dangerous here than farther south around Arabia, when the dryness which finally produced the Arabian desert was making itself felt, burning up the pastures and leaving only the choice between starvation and migration in mass. Again comparing this migration with the pioneer movements of peoples in historic times, we have good reason to believe that the sheep-herders and cattle men--and they were probably both at the same time--advanced faster than the agriculturists, who were more bound to the soil. Between herdsmen and farmers there were almost certainly many intermediate grades. We may be fairly confident, therefore, that the movement or tide along this route did not take the form of a procession marching in lock-step, but of a series of waves, generally with hunters in front and along the forest flank, herdsmen in the middle, and farmers bringing up the rear and making permanent settlements at favored spots. Hunters had been spreading northward at least as early as the beginning of Upper Paleolithic times. Farming on the lowest grades of agriculture is essentially Neolithic. A town or village had risen at Susa 20,000 years ago. Neolithic civilization probably reached Crete nearly or quite 15,000 years ago. Small Sumerian cities were being founded in southern Babylonia at or before 5000 B. C. Population was increasing in density in the Iranian plateau, as almost every mountain region with its healthy atmosphere and low death-rate quickly becomes overpopulated. Our pioneer column was continually pressed forward by new recruits from the rear as well as by its natural increase. We have practically no records of the march. But our sketch is no mere invention of fancy. It applies to every great migration of peoples extending over centuries or millennia. The last illustration was the great westward movement in America beginning a century or two ago, and still far from completed. The Hungarian plain is the last extension of the great south Russian steppe far into Europe. West of this anything like nomadic life was practically impossible. Here our pioneers scattered and followed the river valleys, settling more or less permanently the loess deposits as farmers, but on less favorable soils devoting themselves more largely to cattle-raising. The latter form of life seems to have been more common on the great North German plain, though accompanied by much hunting, a genuine pioneer life. We may now turn to Europe and consider the distribution of its races and peoples.