When the clay has once been converted into the granular condition, considerable difficulty is experienced in restoring it to the state in which it is capable of prolonged suspension in water. The results of the studies herewith reported may be summed up as follows: 1. The Berlin-Schöne method of elutriation gives fairly correct separations with sandy soils containing little clay or matters finer than 0.01 millimeter diameter, but on soils of fine texture, as loams rich in humus and clays, it gives results which are grossly inaccurate, the error on single grades amounting to from eight to fourteen per cent. 2. In respect of rapidity, economy of time, and ease of operation, the Schöne elutriation has no advantage over the beaker method. 3. Schloesing’s method on its mechanical side makes no satisfactory separations, and the chemical treatment it employs is liable to alter seriously the texture of the soil. 4. The determination of clay from a single subsidence from any conventional depth or volume of water, or for any conventional time, is not a process certain to effect even a roughly approximate separation of the finest quartz grains from true clay. 5. Boiling with water must be rejected as a treatment preliminary to mechanical analysis, because it not only abrades and reduces the coarser sediments, but may dehydrate and coagulate the true clay and thus alter essentially the texture and grain of the soil. =239. General Conclusions.=—The methods of Hilgard and Osborne have been given in detail and largely in the descriptive language used by the authors. The other methods of elutriation in use in other countries have also been described. For practical use the methods of Hilgard and Osborne are to be preferred to all others. For simplicity and speed the Osborne method has the preference over the Hilgard. For rigid control of the work the Hilgard method is to be preferred. The effect of long boiling on clay pointed out by Osborne would suggest that the boiling process preliminary to the Hilgard method be made as short as possible. It would seem that the churn attrition in the Hilgard method might well be regarded as a substitute for the soft pestling of the Osborne process, and any prolonged boiling in the former method might be safely omitted. When carefully carried out, the results of the Hilgard and Osborne method are fairly comparable. =240. Distribution of Soil Ingredients.=—The determination of the distribution of the soil ingredients in the sediments obtained in silt analysis is illustrated by the following table:[159]