Analysis of Sugar.—The examination of sugar is ordinarily confined to the estimation of the water, ash, and determination of the nature of the organic matters present. The proportion of water contained in a sample is found by drying it for about two hours in an air-bath, at a temperature of 110°. Moist and syrupy sugars, such as muscovadoes, are advantageously mixed with a known weight of ignited sand before drying. The ash is determined either by directly incinerating a few grammes of the sugar in a tared platinum capsule, or by accelerating the process of combustion by first moistening the sample with a little sulphuric acid. In this case the bases will naturally be converted into sulphates, and a deduction of one-tenth is usually made from the results so obtained, in order to reduce it to terms of the corresponding carbonates. The proportion of ash in raw cane sugar varies somewhat, but it should not much exceed 1·5 per cent. Its average composition, as given by Monier, is as follows:—
| Calcic carbonate | 49·00 |
| Potassium carbonate | 16·50 |
| Sodium and potassium sulphates | 16·00 |
| Sodium chloride | 9·00 |
| Alumina and silica | 9·50 |
| 100·00 |
Insoluble mineral adulterants are readily separated by dissolving a rather considerable amount of the sample in water and filtering. In this manner the presence of sand, terra alba, and foreign pigments may be recognised.
The determination of the character of the organic constituents of commercial sugars is effected, either by chemical or by physical tests, and, in some instances, by a combination of these methods. The presence of such adulterants, as flour or starch, is very easily detected upon a microscopic examination of the suspected sample.
If cane sugar, containing grape sugar, is boiled with water, to which about 2 per cent. of potassium hydroxide has been added, the solution acquires a brown colour.
Upon mixing a solution of pure cane sugar with a solution of cupric sulphate, adding an excess of potassium hydroxide, and boiling, only a slight precipitation of red cupric oxide takes place. Under the same conditions, grape sugar at once produces a copious green precipitate, which ultimately changes to red, the supernatant fluid becoming nearly or quite colourless. A very good method for the quantitative estimation of grape sugar when mechanically mixed with cane sugar, is that of P. Casamajor. It is executed by first preparing a saturated solution of grape sugar in methylic alcohol. The sample to be tested is thoroughly dried, and then well agitated with the methylic alcohol solution, in which all cane sugar will dissolve; any grape sugar present remains behind, and upon allowing the mixture to remain at rest for a short time, forms a deposit which is again treated with the grape sugar solution, and then collected upon a tared filter, washed with absolute methylic alcohol, and weighed. Glucose and invert sugar are usually quantitatively determined by means of Fehling’s solution.
As this preparation is liable to decompose upon keeping, it is advisable to first prepare cupric sulphate solution by dissolving exactly 34,640 grammes of the salt in 500 c.c. of distilled water, and then make up the Rochelle salt solution by dissolving 68 grammes of sodium hydroxide, and 173 grammes of Rochelle salt in 500 c.c. of water, the solutions being kept separate. When required for use, 5 c.c. each of the copper and Rochelle solutions (corresponding to 10 c.c. of Fehling’s solution) are introduced into a narrow beaker, or a porcelain evaporating dish, a little water is added, and the liquid brought to the boiling point. The sugar solution under examination should not contain over 0·5 per cent. of glucose. It is cautiously added to the hot Fehling’s solution from a burette until the fluid loses its blue colour (see p. [37]). The number of c.c. required to completely reduce 10 c.c. of Fehling’s solution, represents 0·05 gramme of grape sugar. The foregoing volumetric method is sometimes applied gravimetrically by adding a slight excess of Fehling’s solution to the sugar solution, collecting the precipitated cupric oxide upon a filter and weighing, after oxidation with a few drops of nitric acid; or, it may be dissolved, and the copper contained deposited by electrolysis, in which case the weight of copper obtained, multiplied by 0·538, gives the equivalent amount of glucose. The proportion of cane sugar in a sample of raw sugar can be determined by first directly estimating the proportion of invert sugar contained by means of Fehling’s solution, as just described. The cane sugar present is then inverted by dissolving one gramme of the sample in about 100 c.c. of water, adding 1 c.c. of strong sulphuric acid, and heating the solution in the water-bath for 30 minutes, the water lost by evaporation being from time to time replaced. The free acid is next neutralised by a little sodium carbonate, its volume made up to 200 c.c., and the invert sugar now contained estimated by Fehling’s solution. The difference in the two determinations represents the glucose formed by the conversion of the cane sugar; 100 parts of the glucose so produced is equivalent to 95 parts of cane sugar.
Commercial cane sugar is, however, generally estimated by the instrument known as the saccharimeter or polariscope.
In order to convey an intelligent idea of the physical laws which govern the practical working of the polariscope, it will first be necessary to refer to the subject of the polarisation of light. The transformation of ordinary into polarised light is best effected either by reflection from a glass plate at an angle of about 56°, or by what is known as double refraction. The former method can be illustrated by Fig. 1, Plate X., which represents two tubes, B and C, arranged so as to allow the one to be turned round within the other. Two flat plates of glass, A and P, blackened at the backs, are attached obliquely to the end of each tube at an angle of about 56°, as represented in the figure. The tube B, with its attached plate, A, can be turned round in the tube C without changing the inclination of the plate to a ray passing along the axis of the tube. If a candle be now placed at I, the light will be reflected from the plate P through the tube, and, owing to the particular angle of this plate, will undergo a certain transformation in its nature, or, in other words, become “polarised.” So long as the plate A retains the position represented in the figure, the reflected ray would fall in the same plane as that in which the polarisation of the ray took place, and an image of the candle would be seen by an observer stationed at O. But, suppose the tube B to be turned a quarter round; the plane of reflection is now at right angles to that of polarisation, and the image will become invisible. When the tube B is turned half-way round, the candle is seen as brightly at first; at the third quadrant it disappears, until, on completing the revolution of the tube, it again becomes perfectly visible. It is evident that the ray reflected from the glass plate P has acquired properties different from those possessed by ordinary light, which would have been reflected by the plate A in whatever direction it might have been turned.
PLATE X.