Value of Soil Analysis.—Professor E. W. Hilgard, in a paper in the "American Journal of Science," on "The Objects and Interpretation of Soil Analyses," accepts as correct the principle that, other things being equal, productiveness is, or should be, sensibly proportional to the amount of available plant-food within reach of the roots during the period of the plant's development, provided that such supply does not exceed the maximum of that which the plant can utilize, when the surplus simply remains inert. For finding the exact value of the soil from analysis, it is necessary, however, not so much to find the actual amounts of the constituents in the soil, as to find the amounts which are accessible to and assimilable by the plants. The problem is, then, to find a solvent which shall as nearly as possible represent the action of the plant itself. Analyses of European soils fail because virgin soils do not exist in Europe, and no generalizations can be drawn from the examination of any spot. In the United States we still have perfectly natural soils in nearly every part of our territory, with the original vegetation, which reveals so much to the farmer, still growing upon them. Professor Hilgard's method of analysis starts from the observation of the productive qualities of the soil as indicated by the native growth. He then tries to ascertain what are the peculiarities of the soil that favor this kind of growth, as distinguished from some other growth on some other soil. As a rule, a soil showing a high percentage of plant-food is fertile; but the converse is not always true; a soil having a low percentage is not necessarily poor. A loose soil, by enlarging the sphere of expansion of the roots may enable them to reach as large quantities of food, even when it is more widely scattered, as they can find in a more highly charged but more compact and less penetrable soil. Hence mechanical conditions should always be taken into account. The analyses so far instituted prove that, other things being equal, the thriftiness or present productiveness of a soil is measurably dependent on the presence of a certain minimum quantity of lime.
The evidence on this point is "overwhelming." The lime operates by effecting the more rapid transformation of vegetable matter into active humus, by retaining the humus against the oxidizing influence of hot climates; by rendering minute percentages of phosphoric acid and potash effective; by a tendency to secure the proper maintenance of the conditions of nitrification; and, physically, by promoting the flocculation of the soil. After that of lime, the proportion of phosphoric acid seems to be the most important factor in the productiveness of soils. A certain percentage of potash is required, but it is present in most soils; and Professor Hilgard infers, generally, that "potash manures are not among the first to be sought for after the soils have become 'tired' by exhaustive culture." Iron, in the shape of ferric hydrate finely diffused, appears to be an important ingredient, valuable on account of its physical, and partly also of its chemical qualities. It has a high absorptive power for gases, and soils in which it occurs resist drought better than others; and the universal preference given by farmers to red lands shows the results of experience in this respect. The efficiency of the hydrate depends essentially upon a state of fine division; and when merely incrusting the sand-grains, or aggregated into bog-ore grains, it exerts little or no influence, although the analysis may show a high percentage. On the other hand, ferruginous soils are the first liable to damage from imperfect drainage, overflows, etc.
The Eocene Strata of Alabama.—Mr. Angelo Heilprin has communicated to the Academy of Natural Sciences of Philadelphia an effort to arrange the Eocene deposits of Alabama, determine their depth, and fix their relative position in the geological system. For this purpose he has given comparative reviews of the examinations of various exposures, including the Claiborne Bluff and sections on Bashia Creek, Clarke County, and at points on the Tombigbee River. The Claiborne Bluff was formerly considered to be near the base of the Eocene system, but it appears to be underlaid by the strata which crop out at other points, indicating a thickness of at least 200 feet, if not more, of Eocene de-
