The difference between the power to accumulate nitrates in the soil after various rotations has been determined by incubating the soils at the termination of these rotations, and measuring the nitrate formed between January and May, a period of 4 months. The test has been repeated during 8 successive years, and the mean result is presented. The previous treatments, each of which lasted 3 years, consisted of (a) a ley grazed off each season by sheep, (b) a crop of lucerne which furnished two or three crops of hay each yer, and (c) and (d) an arable rotation of potatoes, wheat, and either a hoed crop of kale or a mixed ryegrass and clover crop of hay.In each case, the ammonia capable of distillation with magnesia was reduced during the incubation to an amount which differed little in all the soils, while the nitrate made an almost steady increase—most regular, however, after the previous cropping with lucerne—at least for a period of 4 months. The soil after the previous grass and clover hay crop started at a lower level of nitrate than the others and, even after the longest incubation period, never quite reached the accumulation which occurred in the other cases. Comparing the accumulation which occurred after the two 3-year ley crops (grazed ley and hayed lucerne) with that after the arable cropping, there was very little difference.