A single subsidence of the clay water for twenty-four hours will free it from all particles of sand having a diameter greater than 0.005 millimeter, but in many cases a considerable amount of finer sand will remain in suspension for many hours or days. On the other hand, the sediment formed during the twenty-four hours subsidence will not be free from clay, as may be easily seen by suspending it in water a second time and allowing it to stand again for twenty-four hours. Both Hilgard and Schloesing direct attention to these defects, but assume that they do not usually influence the results to a sufficient extent to deprive them of value. In many cases this is undoubtedly true, as, for example, in such soils as that from the garden of the Experiment Station, at New Haven, in which there is but little clay and fine sand; but in soils of the opposite character, as in the North Haven brick clay where exact separations are most desirable, a very considerable error is thus inevitably encountered. =238. Effect of Boiling on the Texture of Clayey Soils.=—Most investigators who have worked upon mechanical soil analysis advise boiling with water in order to detach clay and sand from each other and make a good separation of the several mechanical elements practicable or possible. In general, however, the instructions as to the time and manner of boiling are rather indefinite, and no definite research as to the effects of this treatment has been undertaken. The practice of Hilgard, to boil twenty-four hours or even longer in case of adhesive clays, according to Osborne[157] appears to be objectionable in view of the dehydration and change of physical properties known to occur in case of many hydroxids, especially those of iron and aluminum, which may be present in the soil. It is a familiar fact that the hydroxids above named and many other amorphous substances when precipitated from cold solutions are more bulky and less easily washed upon a filter than when thrown down hot. It is also well known that their properties are considerably changed by warming or boiling with water. Heating with water to boiling for some hours or days gradually converts the bulky brown-red ferric hydroxid, which when precipitated cold and air-dried for eighteen days, contains thirty-eight per cent of water, into a much denser, bright red substance containing but two per cent of water. St. Gilles has also observed the partial dehydration of aluminum hydroxid from Al₂O₃.5H₂O to Al₂O₃.2H₂O by prolonged boiling. The hydrate of silica and the highly hydrated silicates are most probably affected in a similar manner, and if such be the case, boiling would evidently change the constitution of clay in a very essential degree. The following experiments throw light on this subject: