Taking the experiment with alumina cream as the true polarization, it is seen that the lead subacetate solution gives a reading 0.5 per cent too high, and the normal acetate 0.2 per cent too high. The excess reading in the second experiment is due to volume of precipitate and in the first to both volume of precipitate and precipitation of levulose. The dry lead and alkaline lead nitrate give readings in this particular experiment exactly identical with the true polarization. This might seem at first sight to indicate no precipitation of optically active bodies; such a precipitation does take place, however, as will be shown later, and the experiment only shows that in this particular instance the dextrose and levulose were precipitated together in quantities sufficient to neutralize one another.

The determination of reducing sugars in the sugar by the method of Allihn and of Munson and Walker are given in the following table:

'TABLE III.-Determination of reducing sugar in sugar as dextrose by Allihn's and by Munson and Walker's method, using different clarifying agents.

a The high determinations of Mr. Horne are probably due to a slight deterioration of the sugar through fermentation.

The highest results are obtained when no clarifying agent is used or only neutral acetate of lead; the results after clarifying with subacetate of lead solution or dry subacetate of lead are about 0.5 lower. The French chemists, and more lately the English chemists, have declared against the use of basic lead acetate in clarification for the determination of reducing sugars, and their contention seems wholly justified. To determine the extent of this error in analytical work the following series of experiments were carried out at the Bureau of Chemistry by Mr. A. H. Bryan. Solutions of dextrose and levulose were prepared, using 5 grams of sugar and 1 gram each of magnesium sulphate and ammonium tartrate. To 50 cc of this solution the precipitant was added and the volume made up to 100 cc. After filtering the excess of lead was removed with potassium oxalate and the sugar determined by Allihn's method.

TABLE IV.—Percentages of dextrose and levulose removed by different precipitating agents (A. H. Bryan).

[1 gram of magnesium sulphate and 1 gram of ammonium tartrate added to each 5 grams of sugar in solution.]

The normal acetate of lead removes practically no reducing sugar in any of the experiments; the lead subacetate, on the other hand, in any of its forms, removes very large quantities of both dextrose and levulose, the latter sugar always in much larger amounts. In our present official methods for reducing sugars in saccharine products, wines, fruits, etc., the subacetate of lead is always prescribed for clarification. In view of the manifest error involved in the use of this reagent, it is recommended that its use be discontinued by the association in the determination of reducing sugars.

The results obtained by the different methods of estimating copper show, as was indicated last year, the highest result by the method of weighing as cuprous oxid (Cu2O), and the lowest result by the Low method of titration. This is due to the contamination of the cuprous oxid with slight amounts of organic and mineral matter. With products of very low purity it is always safer to make a direct determination of the copper than to weigh the precipitated oxid.

MOLASSES.

DETERMINATION OF TOTAL SOLIDS.

The comparison of methods upon total solids made last year was again continued, the object of the cooperative work being to find a simple and rapid method which would give results agreeing closely with the method of drying in vacuo.

The results of the work are given in the following table:

TABLE V.-Determination of total solids in molasses.

In commenting upon the above, W. D. Horne remarks that the Pellet method of heating to 105° gives too high results owing to caramelization.

J. E. Halligan, in his method of drying six hours at 104° in a salt-water bath, obtained results agreeing closely with the method of drying 2 grams ten hours at 98°.

Mr. Bryan and Mr. Wedderburn by the method of drying in vacuum to constant weight obtained closely agreeing results, and the figures by this method, according to Mr. Spencer, are in best concordance with the workings of the sugarhouse.

The polarizations of the molasses by the different chemists using various methods of clarification are given in the following table:

TABLE VI.—Polarization of molasses with different methods of clarification.

In the factory method employed by G. A. Spencer in Cuba, one-half the normal weight is made up to 100 cc after clarifying with lead subacetate solution (54.3° Brix), using as little as will give a suitable solution; 50 cc of the filtered

solution are acidulated with dilute acetic acid to decompose the soluble levulosate of lead, and the volume is completed to 55 cc. The polariscopic reading, increased by one-tenth, is multiplied by 2 to give the corrected direct reading. Doctor Spencer states that the direct polarization is always too high when the excess of lead is not removed or neutralized with acetic acid.

In the use of potassium oxalate for removing lead, a number of the analysts experienced the difficulty of turbid filtration. W. D. Horne states that this difficulty can be obviated by the use of a little alumina cream.

The results confirm the observations made on the polarization of the raw sugars, namely, that higher results were secured with the wet subacetate of lead than with the dry and that the lowest results were obtained with the hydrosulphite. The question of the influence of the several clarifying and decolorizing agents has already been discussed under sugar and need not be referred to again here. The variations in the direct polarization of the molasses in each series of experiments show in many instances exceedingly wide variations, a circumstance due largely to difference in the temperature of polarization. These differences, it will be noted, are largely equalized in the calculations of the sugar by the Clerget formula, where the variations between the different chemists are less pronounced. The reverse of this was the case in the analysis of the sugar. The exceedingly large dilution necessary to secure a clear reading (in some cases the results had to be multiplied by 16), of course, magnifies greatly any slight error in the initial observations.

Regarding the use of hydrosulphite as a bleaching agent, many of the chemists reported a difficulty in securing a sufficiently clear reading with the use of this substance alone. There appeared to be an insufficient decolorization in some instances, while in other cases there was a redarkening of the bleached solution due to oxidation. Weisberg has recently studied the action of hydrosulphites, and concludes that the action is a double one, first, by means of the free sulphurous acid where the bleaching action is permanent, and secondly, by means of the nascent hydrogen which is evolved, when there is a redarkening of the solution through oxidation. The referee has found that this afterdarkening may be prevented by the use of a new hydrosulphite derivative, sodium sulphoxylate-formaldehyde, sold commercially as Rongalite C. Hydrosulphite bleaches are being used extensively in sugar work just at present, and the effect of these upon the polarization and reducing power of various sugars are points which should be thoroughly investigated by the referee next year. The gravimetric determination of the sucrose in the molasses was made by Mr. A. H. Bryan with the following results:

TABLE VII.-Gravimetric determination of sucrose in molasses by Allihn's, and by Munson and Walker's, method, using different clarifying agents.

The determinations of reducing sugars in the sample of molasses are reported in the following table:

TABLE VIII.-Determination of reducing sugar as dextrose in molasses by Allihn's, also Munson and Walker's, method, using different clarifying agents.

The results confirm the observations in the case of sugar, showing that a large amount of reducing sugars is precipitated by basic lead. The slightly higher result obtained in each instance by the Allihn method as compared with that of Munson and Walker is explained by the greater inverting action of the Allihn solution upon the sucrose.

THE EFFECT OF HYDROSULPHITE AND RONGALITE ON THE POLARIZATION OF DEXTROSE, LEVULOSE, AND SUCROSE.

By A. H. BRYAN.

Solutions of 10 grams of the pure sugars in 100 cc of water were prepared and to aliquot parts (25 cc) were added one-fourth, one-half, and 1 gram of the bleaching agents. When these salts were dissolved by shaking, the solutions were made up to 50 cc with water and polarized immediately. The tubes containing the solutions were set aside and polarized after five hours and again