I will now allude to a diagram (fig. 35) which represents by a curve the yearly produce of barley, in bushels per acre, grown continuously on the same plots of ground for forty years, but with this difference, that one plot (represented by the upper curve) received 14 tons per annum per acre of farmyard manure, while the other, represented by the lower curve, has been unmanured continuously. This diagram has been constructed from figures given by Sir John Lawes and Sir Henry Gilbert in the 'Transactions of the Highland and Agricultural Society of Scotland' for 1895. I have replaced fractions by the nearest whole figure. The fluctuations of both these curves are very great, and it will be noticed that they are exactly parallel to each other. This teaches us that weather is the most important factor in agricultural success, and shows the extreme danger to the farmer of 'placing all his eggs in one basket,' as has been done by the so-called farmers of the Far West, who have attempted to grow wheat _only_ by the process of scratching the prairie, without returning any dung to the soil, and many of whom have been financially swamped by the first bad season. Taking the average of the forty years, it will be found that the produce of the manured land averaged 49 bushels per acre per annum, while the unmanured land gave only 16½ bushels. I might have added to the diagram a third curve showing the produce of that plot of ground which, of all those manured with artificials, gave the highest yield. The yield of this plot for the whole forty years averaged 46 bushels, or only 3 bushels short of the average yield of the plot treated with farmyard manure. If, however, we take the average yield of the three plots for each of the four decades comprising the forty years, the value of the organic matter becomes very manifest. Thus the yield for each decade was with Farmyard dung 44·9 51·5 50·0 51·6 Artificial manure 48·7 49·4 42·8 41·5 Unmanured 22·2 17·5 13·7 12·6 It will be observed that the yield from artificial manuring only exceeded the yield from the farmyard plot in the first decade, when it showed an excess of 3·8 bushels. In the other three decades it was deficient by 2·1, 7·2, and 10·1 bushels. The deficiency of the unmanured plot in each decade, as compared with the farmyard plot, was 22·7, 34·0, 37·3, and 39·0. These figures are very convincing, and as practical agriculturists seem to be now agreed that farming is hopeless without an adequate amount of live-stock to furnish dung, no more need be said upon this head.