--------------------------------+-----------+-----------+-----+----- It will be noted that the greater fineness of grain in the Washington dust soil induces a higher absorption of moisture than occurs in the sandy soil from Mississippi, although the latter contains more clay. Comparison of the figure for the Mississippi pipe-clay and clay soil with the ferruginous soils, from the same state and from Oahu, indicate plainly the influence of the ferric hydrate in increasing absorption; although in the latter case the clay determination was not made, because of the excess of ferric hydrate. The influence of humus is plainly shown in the case of the marsh muck and soil, neither of which contain any appreciable amount of either clay, or ferric hydrate in the finely diffused condition. The relatively slight difference in the absorptions of muck and soil is due to the only partial humification of the organic matter in the former, while in the soil the humification is sensibly complete, and the sand forming the body of the material serves to render it more loose. These data, referring to natural materials, while not as complete as could be desired, are sufficient to prove the facts, and seem preferable to any artificially devised imitation of their kind. _Influence of Temperature, and Degree of Air-Saturation._—The amount of moisture absorbed varies materially both with the temperature, and with the degree of saturation of the air to which the soil is exposed. Schübler, Knop and other earlier observers, operating with earth exposed to air only partly saturated, and with soil layers of considerable thickness (in watch glasses), found that the absorption decreased as the temperature increased, according to a law formulated by Knop. The writer found that under the conditions established in the experiments of Knop and others, the air was not nearly saturated,[70] so that these determinations are marred by ineliminable faults, the more as the soils used are only designated in general terms, as “garden soil,” “loam,” “peaty land,” etc., without any definite indication of their actual physical or chemical constitution. The writer therefore undertook to correlate these coefficients, determined with respect to completely saturated air, with the physical composition of certain soils, as determined by means of the methods heretofore described. [70] It should be understood that it is by no means easy to insure full saturation in any considerable volume of air.