Having determined the tanning, and soluble non-tanning matters of the materials, it remains to determine the moisture, and the insoluble which make up the whole. To determine moisture, a quantity, not exceeding two or three grams of dry solid materials, or half a gram of moist or liquid extracts, is weighed into a basin, and dried in the same way as has been described for the residues, only that a considerably longer time will be required before constancy is attained. The object of employing so small a quantity of liquid extracts is to abridge this time, and the consequent oxidation, as much as possible, as the extract soon forms a hard skin on the exterior, which renders further drying very tedious. It is advantageous to add a little alcohol to liquid and semi-liquid extracts, and so dilute them that by inclining the basin they can be distributed in a thin layer over its sides, while at the same time the alcohol facilitates the evaporation of the water. The weight of the dried residue in the basin is the “total solids,” while the loss is the “water”; and these can be converted into percentages by multiplying by 100 and dividing by the weight of substance originally taken. An alternate method, which is frequently convenient with extracts, is to pipette off 50 c.c. (in duplicate) of the dissolved and well-mixed extract-solution _before filtration_, and dry exactly in the same way as for “total soluble.” The sum of the two residues in milligrams, divided by the weight of extract taken for analysis, gives the “total solids”; subtracting this from 100 leaves the “water,” while the difference between the “total solids” and the “total soluble” is the percentage of insoluble matter. Two further points must be noted. If the total solids are determined by the first method, and the total soluble in the ordinary way, in an extract which contains no insoluble matter, it frequently happens that they differ by 0·1 or 0·2 per cent., owing either to the difficulty of driving off the whole of the water, or to slight oxidation of the total soluble residue. On the other hand, if the second method is adopted, a small amount of “insoluble” is invariably found, even in perfectly soluble extracts, which is due to the absorption of tannin or colouring matter by the filter paper. On the correction of this error, see Collegium, 1902, pp. 145-158, and App. A, p. 477. As the value of a tanning material often depends very much on the paleness of its colour, it has become customary to specify in contracts the intensity of colour of a solution of it containing one-half per cent. of tanning matter (as measured by the I.A.L.T.C. method of analysis), in a glass cell of one centimeter thick, by comparison with standard coloured glasses in the tintometer. On the method of making the measurement see L.I.L.B., p. 131.