I was at Rothamsted in 1879, and witnessed the injurious effect of an excessive rainfall, in washing out of the soil nitrate of soda and salts of ammonia, which were sown with the wheat in the autumn. It was an exceedingly wet season, and the loss of nitrates on all the different plots was very great. But where the nitrates or salts of ammonia were sown in the spring, while the crops were growing, the loss was not nearly so great as when sown in the autumn. The sight of that wheat field impressed me, as nothing else could, with the importance of guarding against the loss of available nitrogen from leaching, and it has changed my practice in two or three important respects. I realize, as never before, the importance of applying manure to crops, rather than to the land. I mean by this, that the object of applying manure is, not simply to make land rich, but to make crops grow. Manure is a costly and valuable article, and we want to convert it into plants, with as little delay as possible, which will, directly or indirectly, bring in some money. Our climate is very different from that of England. As a rule, we seldom have enough rain, from the time corn is planted until it is harvested, to more than saturate the ground on our upland soils. This year is an exception. On Sunday night, May 20, 1883, we had a northeast storm which continued three days. During these three days, from three to five inches of rain fell, and for the first time in many years, at this season, my underdrains discharged water to their full capacity. Had nitrate of soda been sown on bare land previous to this rain, much of it would, doubtless, have been lost by leaching. This, however, is an exceptional case. My underdrains usually do not commence to discharge water before the first of December, or continue later than the first of May. To guard against loss of nitrogen by leaching, therefore, we should aim to keep rich land occupied by some crop, during the winter and early spring, and the earlier the crop is sown in the autumn or late summer, the better, so that the roots will the more completely fill the ground and take up all the available nitrogen within their reach. I have said that this idea had modified my own practice. I grow a considerable quantity of garden vegetables, principally for seed. It is necessary to make the land very rich. The plan I have adopted to guard against the loss of nitrogen is this: As soon as the land is cleared of any crop, after it is too late to sow turnips, I sow it with rye at the rate of one and a half to two bushels per acre. On this rich land, especially on the moist low land, the rye makes a great growth during our warm autumn weather. The rye checks the growth of weeds, and furnishes a considerable amount of succulent food for sheep, during the autumn or in the spring. If not needed for food, it can be turned under in the spring for manure. It unquestionably prevents the loss of considerable nitric acid from