=59. In= the method of soil sampling adopted by the German Experiment Stations[42] it is directed that the samples of soil should be taken according to the extent of the surface to be sampled, in three, five, nine, twelve or more places at equal distances from each other. They should be taken in perpendicular sections to the depth turned by the plow; and for some studies of the subsoil to a depth of sixty to ninety centimeters. The single samples can be either examined separately or carefully mixed and an average portion of the mixture taken. =60. Method of the Official French Commission.=—The official French commission[43] emphasizes the fact that the sample of soil taken for analysis, should represent a layer of equal thickness through the total depth of its arable part. An analysis of the subsoil taken in the same way, will often be useful to complete the data of the soil study. First of all, according to this authority, it is necessary to determine the point of view from which the sample is to be taken. If the object is a general study, having for its aim the determination of the general composition of the soils of a definite geologic formation, the sample should be taken in such a way that the different characteristics of the soil alone should enter into consideration without paying any attention to its accidental components, which have been determined by local causes, such as are produced by continued high cultivation, the application of abundant fertilizers, or the practice of a particular line of agriculture. The samples of soil therefore, with such an object in view, should be taken on parts of the earth which are beyond the reach of the causes mentioned above and which tend to modify the nature of the primitive soil. In such a case it is the soil which has not been modified, or better still, virgin soil, such as is found in the woodlands and prairies, which should be taken for a sample, choosing those places in which the geologic formation is most perfectly characterized. In such a case a soil taken in one spot corresponding to the conditions before mentioned, would be the best for the purposes in view. The sample would thus represent a true type to be studied, not one of a mean composition got by taking samples from different localities and mixing them into a homogeneous parcel. This last method of proceeding could introduce into the sample earth modified by culture or by influences purely accidental. However, it would be wise, in a region characterized by the same geologic formation, to take a certain number of samples in different localities, and examine singly each one of them in order to be assured that there is a uniformity of composition in the whole of the soils.