He has shown us that even on the steppe the cultivation of cereals precedes the domestication of sheep and cattle. The nomadic life follows instead of preceding agriculture. The pioneers in this region cultivated the zone of steppe, into which rivers poured from the mountains. When cattle and sheep and goats had multiplied, the herdsmen drove them farther and farther on the rich pasturage of the boundless steppe. Thus nomads gradually appear. There are also different varieties of nomadism. Nomadic tribes were far less active and dangerous neighbors even after the domestication of the camel than when, about 2000 B. C., they had domesticated the horse. The first herdsman may have differed from the latter nomad almost as much as the most pacific sheep-herder of our Western plains differs from the liveliest cowboy. Pumpelly's time-estimates have been criticised by Doctor H. Schmidt, of Berlin.[74] He makes the rate of growth far more rapid than Pumpelly thought and shortens the periods. In determining length of periods he relies far more on artifacts and less on probable rate of accumulation. The criticisms seem hardly well founded. Pumpelly's estimate of rate of increase was based upon a careful and broad comparison of accumulations in the deserted city, Anau, in Merv, and other localities. They seem conservative, but we must recognize that such estimates are always only approximate. His estimates result in a series of dates generally in close agreement with those of most students of oriental archæology. In the Third Culture Epoch there was found "copper, with sporadic appearance of low percentage of tin." This describes well the close of the Copper period or the beginning of the Bronze Age, the rest of which is not represented at Anau, the settlement being deserted, probably because of aridity. Pumpelly thinks that the last strata deposited before the desertion comes down to the Bronze Age, and, assuming the latest possible date for the beginning of this period, places it about 2200 B. C. This is almost surely much too late. Obermaier dates the beginning of the Bronze period at 4000 B. C.[75] (If we substitute the later date, 2750 B. C., for Sargon's region, the Bronze period would begin about 3000 B. C., the date accepted by Montelius.[76]) Pumpelly places the beginning of the Copper Epoch at 5000 B. C., again agreeing with Montelius. His estimates seem generally somewhat too conservative, as he doubtless intended they should be; the earliest remains may be considerably older than he thought. Investigations made during the last twenty years seem generally to lead us to believe that the beginnings of Neolithic culture are far older in western Asia than we had supposed, while in middle and northern Europe they are probably somewhat younger than we had thought. In this connection we may well remember that Evans found eight metres of Neolithic remains under the palace at Cnossus, in Crete, and estimated their age at about 14,000 years.