The material which passes through the finest sieve employed will contain all the true clay in the material; that is to say, the coarser portion will, as already mentioned, be devoid of the ordinary characteristics of clay. At the same time, this very fine material will seldom consist exclusively of clay, but will usually contain a considerable proportion of silt, extremely fine mineral particles and, in the case of calcareous clays, a notable proportion of calcium carbonate in the form of chalk or limestone particles. Only in the case of the purest clays will the material now under consideration consist entirely of clay, so that it must be again separated into its constituents. This is best accomplished, as first suggested by Schoene, by exposing the material to the action of a stream of water of definite speed. H. Seger (7) investigated this method very thoroughly and his recommendations as to the manner in which this separation by elutriation should be carried out remain in use at the present time. Briefly, all material sufficiently fine to be carried away by a stream of water flowing at the rate of 0·43 in. per minute was found by Seger to include the whole of the clay in the samples he examined, but, as was later pointed out by Bischof, it is not correct to term the whole of this material 'clay substance,' as when examined under the microscope, it contains material which is clearly not clay. Processes of decantation of the finest material obtained after elutriation still fail to separate all the non-clay material, and Vogt has found that when the material has been allowed to stand in suspension for nine days some particles of mica are still associated with the clay. It would thus appear that no process of mechanical separation will serve for a complete purification of a clay; indeed, there are good reasons for supposing that extremely fine particles of quartz and mica render physical characteristics an uncertain means of accurately distinguishing clays from other rock dust. When chemical methods of investigation are employed the problem is not materially altered, nor is its solution fully attained. It is, of course, obvious that any chemical method should be applied to the product obtained by treating the raw material mechanically as above described, for to do otherwise is to create needless confusion. Yet by far the greater number of published analyses of 'clays' report the ultimate composition of the whole material, no attempt being made to show how much of the various constituents is in the form of sand, stones or other coarse particles of an entirely non-argillaceous character.