In all cases it would prove useful to take a sample of the subsoil which is far from playing a secondary rôle. The rootlets bury themselves deeply in it and seek there a part of their nourishment. The subsoil, therefore, furnishes an important addition to the alimentation of the plants. For taking a sample of the subsoil a ditch is dug of sufficient depth, say one meter, and the arable soil carefully removed from the top portion. Afterwards pieces are taken from the four sides of the hole at variable depths, which should always be indicated, and which should extend in general, from six to eight-tenths of a meter below the arable soil since it is demonstrated that the roots of nearly all plants go at least to this depth. The analysis of the subsoil, however, is less important than that of the soil, properly so-called, because the agronomist does not act directly upon it and takes no thought of modifying it and enriching it as he does the layer of arable soil. But the composition of the subsoil is a source of information capable of explaining certain cultural results and capable sometimes, of leading to the correct way of improving the soil, as in cases where the subsoil can be advantageously mixed with the superficial layer. =73. Wolff=[57] suggests that a hole thirty centimeters square be dug perpendicularly and a section from one of the sides taken for the sample. To the depth of thirty centimeters the sample shall be taken as soil and to the additional depth of thirty centimeters as subsoil. The thickness of the section taken may vary according to the quantity of the sample desired. For analytical purposes, five kilograms will usually be sufficient. When culture experiments are also contemplated a larger quantity will be required. =74. Method of Wahnschaffe.=—The method of sampling advised by Wahnschaffe[58] is but little different from that of Wolff already mentioned. A square sample hole is dug with a spade having its sides perpendicular to the horizon. The soil which is removed is thrown on a cloth and carefully mixed. From the whole mass a convenient amount is next removed care being taken not to include any roots. In a similar manner it is directed to proceed for the sample of subsoil. At first the subsoil should be removed to a depth of two to three decimeters. The number and depth of subsequent samples will depend chiefly upon the character of the soil. Where samples are taken to the depth of two meters the use of a post-hole auger is recommended. The samples taken should not be too small. In general from two to three kilograms should remain after all preliminary sampling is finished. =75. Method of König.=—The directions given by König[59] for taking soil samples are almost identical with those prescribed by Wahnschaffe and do not require any further illustration.