The yield of the soil in single seasons at widely separated intervals is a piece of information of little value for our purpose. These tables reveal other facts of greater significance. The yield for the year gives almost no information about the normal yield over a series of years, but the area planted depends very largely upon that yield. The farmer knows that it will pay, on the average, to sow a certain number of acres, and the area under cultivation is not subject to violent fluctuations, as is the crop reaped. The area sown in any season is representative of the period; the crop reaped may or may not be representative. Land which, over a series of years, fails to produce enough to pay for cultivation is no longer planted. If the fertility of the soil is declining, this is shown by the gradual withdrawal from cultivation of the less productive land, as it is realized that it produces so little that it no longer pays to till it. Table IV shows that in fact this withdrawal of worn out land from cultivation was actually taking place. The area sown with wheat on the twenty-five manors for which the statistics for both periods are available was reduced by more than fifty per cent between the beginning and the end of the thirteenth century. A similar reduction in the area planted with all of the other crops, mancorn, rye, barley and oats, took place. A process of selection was going on which eliminated the less fertile land from cultivation. If six bushels an acre was necessary to pay the costs of tillage, land which returned less than six bushels could not be kept under the plow. The six bushel crop which seems to be normal in the fourteenth century is not the average yield of all of that land which had been under cultivation at an earlier time, but only of the better grades of land. Plots which had formerly yielded their five or six bushels an acre had become too barren to produce the bare minimum which made tillage profitable, and their produce no longer appeared in the average. Even with the elimination of the worst grades of land the average yield fell, because the better land, too, was becoming less fertile. At Witney (Table V) the area planted with wheat fell from about 180 acres in 1277 to less than 140 acres in 1340; but, in spite of this reduction in the amount of land cultivated, the average annual yield after 1340 was less than 6-1/2 bushels, while it had been about 8-1/2 bushels per acre in the period 1277-1285. This withdrawal of land from cultivation took place without the occurrence of any such calamity as the Black Death, which is ordinarily mentioned as the cause of the reduction of arable land to pasture in so far as this took place before 1400. It affords an indirect proof of the fact that much land was becoming barren.