Statistical data so scattered as this cannot be used as the basis of an inquiry into the rate of soil exhaustion. Where the normal variation from place to place and from season to season is as great as it is in agriculture, the material from which averages are constructed must be unusually extensive. So far as I know, no material in this field entirely satisfactory for statistical purposes is accessible at the present time. There is, however, one manor, Witney, for which important data for as many as eighteen seasons between 1200 and 1400 have been printed. A second suggestive source of information is Gras's table of harvest statistics for the whole Winchester group of manors, covering three different seasons, separated from each other by intervals of about a century. The acreage reported for the Winchester manors is so extensive that the average yield of the group can be fairly taken to be the average for all of that part of England. Moreover, Witney seems to be representative of the Winchester group, if the fact that the yield at Witney is close to the group average in the years when this is known can be relied upon as an indication of its representativeness in the years when the group average is not known. The average yield for all the manors in 1208-1209 was 4-1/3 bushels per acre; for Witney alone, 3-2/3. In 1396-1397 the yield of the group and the yield at Witney are, respectively, 6 and 6-1/4 bushels per acre.[49] Table III shows the yield of wheat on the manors of the Bishopric of Winchester in the years 1209, 1300 and 1397. If it could be shown that these were representative years, we should have a means of measuring the increase or decrease in productivity in these two centuries. Some indication of the representativeness of the years 1300 and 1397 is given by a comparison of prices for these years with the average prices of the period in which they lie. The price in 1300 was about 17 per cent below the average for the period 1291-1310,[50] an indication that the crop of nine bushels per acre reaped in 1299-1300 was above the normal. The price of wheat in 1397 was very slightly above the average for the period;[51] six bushels an acre or more, then, was probably a normal crop at the end of the fourteenth century. This conclusion is supported also by the fact that the yield in that year at Witney was approximately the same as the average of the eleven seasons between 1340 and 1354 noted in Table V. The price of wheat in the year 1209-1210 is not ascertainable. Walter of Henley's statement that the price of corn must be higher than the average to prevent loss when the return for seed sown was only three-fold[52] is an indication that the normal yield must have been at this time at least three-fold, or six bushels, so that the extremely low yield of the year 1208-1209 can hardly be considered typical. This examination of the yield in the three seasons shown in the table gives these results: