Meadows and woodlands generally show the highest humus-content in their surface soils, gradually increasing while in that condition; while when taken into cultivation the humus-content gradually decreases, owing to the free aeration and consequent “burning-out” caused by tillage. Hence the humus must be from time to time replaced by the use of stable manure, or green-manure crops, to prevent injurious changes in the tilling qualities of the land. Not only humus as such, but according to Schloesing also the insoluble colloid humates, produce in the soil a loosening effect or tilth (Germ. Bodengare), which apparently cannot be brought about by any other substance.[48] [48] The decrease of humus from wheat culture in the soils of Minnesota and North Dakota has been studied by H. Snyder and E. F. Ladd, respectively. In the prairie lands of the latter State the total organic matter in the first six inches of soil ranges from 15 to as much as 26%, and the humus alone from 4 to 7.8%. _Humates and Ulmates._—That the insoluble humates of lime, magnesia, iron, manganese and alumina are present in most soils is conclusively shown by the composition of the solution obtained by the extraction of soils with weak acid, as above mentioned in connection with the quantitative determination of humus according to Grandeau; since these bases are almost always extracted by the weak acid. When the brown solution of alkali humate obtained in this process is carefully neutralized with sulfuric or hydrochloric acid, or is mixed with solutions of the above bases, flocculent, insoluble precipitates are formed, while the solution is discolored. Similar precipitates may be obtained with other metallic solutions, notably with that of copper, which precipitates the humus-acids most completely. Doubtless these compounds contribute greatly to the conservation of the humus-content of soils, protecting it to a certain extent from oxidation, and also preventing excessive acidity. The brown tint of certain subsoils in the northern humid regions have been shown by Tollens and others to be due not to ferric hydrate, as had been supposed, but to calcic, magnesic and aluminic humates. _None of the mineral bases or acids present can be detected in the humic solution by the usual reagents._