For a more exact determination of the sugar, a gravimetric determination of the reducing sugars in the aqueous extract, before and after inversion, is recommended. (Weight of copper obtained after inversion) - (weight of copper obtained before in-version)=weight of copper equivalent to cane sugar present.

Determination of starch. - Mr. K. P. McElroy devised the following method for the determination of starch:

Five grams of chocolate were weighed into an Erlenmeyer flask, wet with alcohol, and 30 to 40 cc of water added. The flask was then shaken at intervals until all sugar present had gone into solution. Fifteen cc of a saturated solution of neutral lead acetate were then added and the mixture again shaken. After allowing it to settle, the clear liquid was decanted through an asbestos filter and water added to supply its place. This was in turn decanted, and so on until the filter began to clog, when just enough acetic acid was added to the material in the flask to turn it red. Decantation and filtration were then continued till the filtrate came through color-.less. Under these conditions filtrations were rapid and filtrate extremely clear. The asbestos filter and contents were then placed in the flask, water enough added to bring the whole to about 100 cc, and the flask placed on the steam bath for three hours. At the end of this time 10 cc of concentrated HC1 were added, and the heating continued three hours more. After cooling, Na2C03 was added until the contents of the flask became blackish. The whole was now washed into a 250 cc flask and made up to the mark. Dextrose was oxidized by Allihn's method, 25 cc of solution being used, and the resulting Cu20 collected on a Gooch crucible, redis-solved in HNO3, converted into CuS04, and the Cu determined electrolytically.

Five samples of commercial starchy materials were examined for the purpose of determining the amount of starch converted to dextrose by the method employed. Three grams of material were heated on the steam bath for three hours with 200 cc of water; 20 cc of HC1. (sp. gr. 1.125) were then added and the heating continued three hours longer. After neutralization with Na2CO3 dilution to a definite volume, and filtration from the slight residue, the dextrose was determined by Allihn's method, the amount of Cu20 being determined electrolytically. Duplicate samples were inverted and duplicate weighings made from each inverted solution. The results were as follows:

Percentage of starch in commercial starchy materials used in manufacture of cocoa preparations, i. e.,per cent of dextrose obtained by inversion x 0.9.

[All samples were air-dried.]

Starch determinations are only reliable when the conditions are held rigidly exact. In order that this may be made more certain each set of determinations should be accompanied by a check determination with material of known dextrose-yielding power.

Determination of fiber. - The official method of the Association of Official Agricultural Chemists 1 was adapted to this work, as follows: Two grams of material were placed in a long narrow test tube with a lip. washed with several portions of ether, decanting through a Gooch crucible. After evaporation of the ether, the material in the tube and crucible was washed into an Erlenmeyer flask with 200 cc of 1.25 per cent H2S04. After boiling a half hour with a reflux condensing tube, the solution was filtered through a linen filter, and the residue thoroughly washed with hot water; while still hot, the filtrate was refiltered through a Gooch crucible, and the slight residue again thoroughly washed with hot water. The material on the linen and in the crucible was rinsed hack into the flask with 200 cc of 1.25 per cent NaOH, and the boiling and filtration repeated. The material on the linen was then washed into a beaker with alcohol and transferred to the crucible through which the alkali extract had been refiltered. After displacing the alcohol with ether, the residue was dried at 110° C., weighed, ignited, and weighed again. The difference == the weight of crude fiber.

1 Bulletin No. 3I, Chem. Div., U. S. Dept. of Agriculture.