the water in a porcelain mortar, and thoroughly rinsed out with the remainder into a flask. The cocoa is left in contact with the water for a definite number of hours, and then a sufficient quantity filtered to allow 40 or 50 cubic centimetres being taken for evaporation to dryness, and from the result the total amount soluble in cold water is easily calculated. The dry residue is ignited to obtain the amount of ash. In the analysis of a Trinidad cocoa, the amount of ash in the cold water solution was 2'44 per cent., and that obtained from a prepared cocoa was o'90 per cent. The calculation of the proportion of cocoa in the latter sample from the above data is therefore as 2'44: 0'90 :: 100:37. This result is below the true quantity, judging from the direct estimation of the sugar and starch in the sample. The determination of the ash affords a useful but rather rough test.

Sugar. To directly estimate the sugar in commercial cocoa, two grams are taken and exhausted with alcohol of 70 per cent. strength; the spirit is evaporated, acetate of lead added, and the precipitate separated; the filtrate is freed from lead with sulphuretted hydrogen, and the solution boiled with a few drops of sulphuric acid to invert the cane-sugar, and the amount present is then determined by a standard solution of copper salt.

Starch.-It has already been stated that cocoa contains from 4 to 5 per cent. of starch; but there is no difficulty in distinguishing and identifying the starches commonly employed for mixing with cocoa from the starch natural thereto. The granules of the starch of cocoa differ in size and shape, as well as in other characters, from those of the starches which are used in the preparation of some commercial cocoas. The starches employed for this purpose are limited in number, and usually consist of arrowroot, sago, tous-les-mois, and sometimes potato.

Arrowroot.-The microscopic appearance of this starch is represented in Fig. 24. The granules are more or less oblong

in shape, and generally marked with a series of delicate concentric rings.

Sago. The appearance of the granules of this starch under the microscope is given in Fig. 25. The granules are elongated,

%

FIG. 25.-SAGO.

round at one end, truncated at the other, and altogether so distinctive in size and shape that they are readily identified.

Tous-les-Mois.-The granules of this starch are given in Fig. 26They are ovate in form, obtuse at one extremity and somewhat pointed at the other, and are marked by a series of regular and distinct concentric rings. The hilum, which is circular, is situated near the pointed extremity.

Potato.-The granules of this starch are represented in Fig. 27. They vary greatly in size and form; some being triangular or oyster shape, some ovate, and others round, especially the smaller ones. The granules are marked by well-defined concentric rings, and bear a hilum which is sometimes circular and sometimes stellate.

A general idea of the quantity of added starch may be arrived at by the microscope, by comparing the appearance of the mixed cocoa with equal portions of cocoa containing known percentages of the kind of starch identified in the sample.

The amount of added starch in a cocoa may also be determined by chemical analysis. In making this analysis the residue left after treatment with alcohol for the estimation of sugar, is boiled with water for ten minutes with the addition of five drops of sulphuric acid. By this means the starch is rendered soluble and can be readily filtered from the albuminous matter, whilst the cellulose remains practically unaffected. The soluble starch is removed by filtration, and, after the addition of ten drops of sulphuric acid, boiled for six hours to convert the starch into glucose. The colouring matter is removed as usual with subacetate of lead, and the amount of sugar estimated by the copper solution, from which the percentage of starch is calculated. As cocoa contains from 4 to 5 per cent. of starch, an equivalent deduction on this account must be madefrom the total amount of starch obtained.

Chicory. The presence of chicory is indicated by the high colour of a cold-water extract, and proved by the subsequent detection of the characteristic tissue by the aid of the microscope. It might be estimated by one of the methods described under Coffee.