This operation is repeated on as many points as may be necessary to obtain a mean sample of the soil of the whole field. All the samples are now collected on a table sufficiently large, and intimately mixed together. Two samples, each of about five kilograms, are then taken from the mixed material. One sample is immediately placed in bottles and carefully stoppered and sealed; the other is dried in the sun or on the hearth of a furnace. When sufficiently dry the second sample is also placed in bottles and well stoppered. While mixing the samples, pebbles, etc., of the size of a nut and larger are removed, the weight of the rejected matter being determined. The nature of the pebbles should also be noted; whether silicates, limestone, etc. The sample of subsoil is taken in exactly the same manner, using the same holes from which the samples of soil were taken. The nature, the arrangement and the appearance of the strata will indicate the depth to which the subsoil should be taken. In general, a depth equal to that of the sample of soil will be sufficient. The depth to which the roots of cultivated plants reach is also a good indication in taking a sample of the subsoil. In forests the sample of subsoil should be taken from four to five-tenths of a meter below the surface. If the soil in respect of its geologic formation, its fertility or its physical aspect presents great differences, special samples should be taken in each part in accordance with the directions given above. =70. Method of the Official Agricultural Chemists.=—In the directions given by the Association of Official Agricultural Chemists[52] it is stated that the soil selected should be as far as possible in its natural condition, not modified by recent applications of manure, or changed by the transporting action of water or wind. Surface accumulations of decaying leaves, etc., should be removed before taking the sample. To eliminate accidental variations in the soil, select specimens from five or six places in the field which seem to be fair averages of the soil, remove two or three pounds of the soil, taking it down to the depth of nine or ten inches[53] so as to include the whole depth. Mix these soils intimately, remove any stones, shake out all roots and foreign matter, and dry the soil until it-becomes friable.[54] Break down any lumps in a mortar with a wooden pestle, but avoid pulverizing any mineral fragments; pass eight to ten pounds of the soil through a sieve, having circular perforations one twenty-fifth of an inch in diameter, rejecting all pebbles and materials too coarse to pass through the sieve. Once more mix intimately the sifted soil. Expose in thin layers in a warm room till thoroughly air dry (or dry it in an air-bath at a temperature of 40°), place six to eight pounds in a clean bottle, with label of locality and date, and cork the bottle containing the soil, for analysis.