If the particles of a 'clay' which are sufficiently small to be carried away by a stream of water with a velocity of only 0·43 in. per minute are analysed, it will be found that their composition will vary according to the origin of the clay and the subsequent treatment to which it has been subjected during its transport and deposition. If the clay is fairly free from calcareous material and is of a white-burning nature it may be found to have a composition like china clays. Red-burning clays, on the contrary, will vary greatly in composition, so that it becomes difficult to find any close analogy between these kinds of clay. This difference is partly due to the extremely fine state of division in which ferric oxide occurs in clays, the particles of this material corresponding in minuteness to those of the purest clays and so being inseparable by any mechanical process.
In 1876 H. Seger ([7]) published what he termed a method of 'rational analysis,' which consisted in treating the clay with boiling sulphuric acid followed by a treatment with caustic soda. He found that the purer china clays (kaolins) and ball clays were made soluble by this means and that felspar, mica and quartz were to a large extent unaffected. Later investigators have found that this method is only applicable to a limited extent and that its indications are only reliable when applied to the clays just named, but the principle introduced by Seger has proved invaluable in increasing our knowledge of the composition of clays. By means of this so-called rational analysis Seger found that the purer clays yielded results of remarkable similarity and uniformity, the material entering into solution having a composition agreeing very closely with the formula Al2O32SiO22H2O which is generally recognized as that of the chief constituent or constituents of china clay (kaolin) and the purer ball clays. This crude substance, obtainable from a large number of clays by the treatment just described, was named clay substance by Seger, who regarded it as the essential constituent of all clays.
Red-burning clays when similarly treated do not yield so uniform a product, and the ferric oxide entering into solution makes the results very discordant. Moreover, even with the china clays or kaolins a small proportion of alkalies, lime and other oxides enter into solution and a number of minerals analogous to clay, but quite distinct from it, are also decomposed and dissolved. For these reasons the 'rational analysis' has been found insufficient; it is now considered necessary to make an analysis of the portion rendered soluble by treatment with sulphuric acid in order to ascertain what other ingredients it may contain in addition to the true clay present.
As the china clays (kaolins) and ball clays on very careful elutriation all yield a product of the same ultimate composition, viz. 39 per cent. of alumina, 46 per cent. of silica, 13 per cent. of water, and 2 per cent. of other oxides, they are generally regarded as consisting of practically pure clay with a variable amount of impurities. Many years ago Fresenius suggested that these non-clayey constituents of clays should be calculated into the minerals to which they appeared likely to correspond so as to obtain a result similar to that obtained by Seger without the disadvantages of the treatment with sulphuric acid and as supplementary to such treatment in the case of red-burning and some other clays. More recent investigators have found that if a careful microscopic examination of the clay is made the results of estimating the composition from the proportion of the different minerals recognizable under the microscope and by calculation from the analysis of the material agree very closely and are, as Bischof ([28]) and, more recently, Mellor have pointed out, more reliable than the 'rational analysis' in the case of impure clays. If care is taken to make a microscopical examination identifying the chief impurities present the calculation from the analysis may usually be accepted as sufficiently accurate, but it is very unsatisfactory to assume, as some chemists do, that the alkalies and lime in the clay are all in the form of felspar and that the silica remaining in excess of that required to combine with the alkalies, lime and alumina is free quartz. Some clays are almost destitute of felspar but comparatively rich in mica, whilst others are the reverse, so that some means of identifying the extraneous minerals is essential. When this is not used, the curious result is obtained that German chemists calculate the alkalies, etc. to felspar whilst the French chemists, following Vogt, calculate them to mica; English ceramic chemists appear undecided as to which course to follow, and some of them occasionally report notable amounts of felspar in clays quite destitute of this mineral!
A statement of the composition of a 'clay' based on a mechanical separation of the coarser ingredients followed by an analysis of the finer ones and a calculation of the probable constituents of the latter, as already described, is known as a proximate analysis in order to distinguish it from an ultimate analysis which states the composition of the whole material in terms of its ultimate oxides. A proximate analysis therefore shows the various materials entering into the composition of the clay in the following or similar terms:
| Stones | per cent. |
| Gravel | " |
| Coarse sand | " |
| Medium sand | " |
| Fine sand | " |
| Silt | " |
| Felspar or mica dust | " |
| Silica dust | " |
| 'True clay[12]' | " |
| Moisture | " |
| Carbon | " |
| Other volatile matter | " |
[12] In analytical reports a note should be appended stating that the figure under this term shows the proportion of the nearest approximation to true clay at present attainable.
For some purposes it is necessary to show the proportion of calcium, iron and other compounds as in an ordinary ultimate analysis.
A comparison of the foregoing with an ultimate or 'ordinary' analysis of a clay ([p. 16]) will show at once the advantage of the former in increasing our knowledge of the essential constituent of all clays, if such a substance really exists. Its absolute existence is by no means proved, for, as will have been noticed, its composition is largely based on assumption even in the most thorough investigations, particularly of the admittedly less pure clays.