A standard procedure was accordingly adopted, flour and water in the proportions of 1 to 5 being mixed, at a temperature of 25°C., and the flour maintained in suspension for 60 minutes. The mixture was then whirled in the centrifuge and the decantate pipetted.

H

It appears from the data in Table 1 that the PH of the original extracts increases slightly with the ash content; that is, the higher grades yield an extract with a slightly higher hydrogen ion concentration than do the lower grades. The differences are hardly great enough or sufficiently uniform, to make the Р a satisfactory index of grade. Moreover, this may change appreciably as the flour ages. The buffer action varies widely, however, and increases with the ash content. It appears that the data secured through the addition of 20 and 30 cc. of N/50 sodium hydroxid to 100 cc. of extract are of the most value in indicating the relative grade.

TABLE 2.

Hydrogen ion concentration in water extracts of flour before and after the addition of acid and alkali.

First middlings.. 0.44

6.07 4.48

Fifth middlings 0.61 6.31

5.04

Third break...

0.67 6.22

5.19

Fourth middlings 1.17 6.42

5.85

First break.
1.34 6.34 5.86
Fourth break.. 1.62 6.36 6.02

3.60 3.11 2.87 7.48 9.28 9.89 Second middlings 0.45 6.10 4.65 3.64 3.25 2.96 7.27 8.72 9.79 Third middlings. 0.55 6.22 4.94 3.96 3.45 3.04 7.22 8.59 10.00 Second break 0.58 6.25 5.00 4.11 3.45 3.15 7.12 8.52 9.53 3.53 3.16 7.15 8.38 9.62 3.62 3.46 6.96 7.89 9.25 4.46 4.06 6.86 7.29 7.91 5.15 4.56 4.04 6.77 7.17 7.66 5.43 4.78 4.26 6.75 7.08 7.49

10.62

10.32

10.35

10.21

4.14

10.22

The data reported in Table 2 were obtained by the use of the hydrogen electrode. A colorimetric procedure can be evolved, however, which is useful in this connection. Since phenolsulphonephthalein (phenol red) is one of the most delicate indicators, it is recommended that the treatment be such that the resulting preparations are rendered sufficiently alkaline to have a PH within the limits of color change of this indicator (PH = € 6.8 to 8.4). This can generally be brought about by the addition of 10 cc. of N/40 sodium hydroxid to each 100 cc. of the extract. The resulting Pн will be much higher in the case of patent flour extracts than in the extracts from the clears, owing to the lower buffer action of the former. The color of the preparation, after phenol red has been added, can be matched against a series of standards, and the PH thus determined. In the case of patent flours containing less than 0.44 per cent of ash,

H

it is necessary to use somewhat less than the quantity of alkali mentioned, otherwise the mixture will have a Pн greater than 8.4. It is suggested that with such flours 10 cc. of N/50 sodium hydroxid be added to each 100 cc. of extract, and a separate graph (or formula) be employed to compute the results into terms of ash content.

REPORT ON WINES1.

By J. M. HUMBLE2 (U. S. Food and Drug Inspection Station, U. S. Custom House, Cincinnati, Ohio), Referee.

In accordance with the recommendations of 19163, it was decided to confine the work to a study of the Rothenfusser method for glycerol. Due, no doubt, to war conditions and also to lack of interest in the subject coincident with prohibition, great difficulty was incurred in securing men willing to collaborate. The following collaborators were finally secured:

B. G. Hartmann, U. S. Food and Drug Inspection Station, 1625 Transportation Building, Chicago, Ill.

A. R. Todd, State Dairy and Food Commission, Lansing, Mich. (assisted by Ruth Hoare, analyst).

E. M. Meyer, City Chemist, Cincinnati, Ohio.

Wm. J. McCarthy, U. S. Food and Drug Inspection Station, 411 Government Building, Cincinnati, Ohio.

A copy of the Rothenfusser method, submitted herewith, was sent to the above collaborators; also, two synthetic solutions, Nos. 1 and 2, containing 0.50 and 0.75 gram, per 100 cc. of glycerol, respectively, and of the following composition:

ROTHENFUSSER METHOD FOR THE DETERMINATION OF GLYCEROL.

The method is based upon the following points:

(1) Removal of sugars and organic acids (lactic acid excepted) with basic lead acetate and strong ammonium hydroxid.

(2) Removal of lactic acid with stannous chlorid.

(3) Removal of lead and tin with sodium phosphate and potassium carbonate.

(4) Oxidation of glycerol in a sodium carbonate solution with potassium permanganate in the cold.

(5) Precipitation of calcium oxalate.

(6) Titration of oxalic acid with potassium permanganate.

Method I.

(Applicable to white and red wines containing up to 1 per cent of sugar.) Treat 50 cc. of wine with 30 cc. of sodium carbonate (1 to 5) and 5 grams of crystalline stannous chlorid and, after mixing thoroughly, allow to stand for a few minutes. Make to 250 cc. with water. Filter through a double filter1 and treat 200 cc. of the filtrate with a freshly prepared mixture of 40 cc. of lead subacetate (sp. gr. 1.24) and 20 cc. of 10% ammonium hydroxid. After shaking thoroughly, make up to 300 cc. with 10% ammonium hydroxid. Shake, filter and to 200 cc. of the filtrate in a nickel dish (sugar dish) add 5 cc. of a 10% solution of sodium phosphate and 6 grams of pure sodium carbonate. Evaporate in a water bath to a volume of about 60 cc. Transfer the evaporated solution to a flask with about 40 cc. of water, add 10 grams of potassium carbonate and, after cooling to about 25°C., oxidize in the cold with 2 grams of powdered potassium permanganate. Allow to stand in the cold for 45 minutes, shaking occasionally. Then add, with constant shaking, 3% hydrogen peroxid until the supernatant liquid is colorless1. Transfer to a 250 cc. flask with water, make to volume and filter3. Transfer 220 cc. of the filtrate (23.46 cc. of wine) to an Erlenmeyer, mix with 50 cc. of 30% acetic acid and bring to a boil. Treat the clear solution with 3 cc. of a 30% calcium chlorid solution and boil for about 5 minutes. Allow to stand for a short time, filter through a Gooch charged with asbestos and wash with water.

Dissolve the calcium oxalate on the felt in 200 cc. of boiling sulphuric acid (1 to 8), using very weak suction at first. Transfer the filtrate to an Erlenmeyer, heat to about 80°C. and titrate with potassium permanganate solution. Prepare the potassium permanganate solution by dissolving 8 grams of potassium permanganate in 2500 cc. of water. Allow to stand for 2 days and standardize with N/20 oxalic acid.

Calculation. The glycerol is calculated according to the following formula:,
TXNX 4.26 grams of glycerol per 100 cc. in which

=

glycerol titer;

N = cc. of potassium permanganate solution required for oxidation.

1 Disregard any turbidity which may subsequently form.

2 To hasten filtration carefully sweep the wall of the filter with a policeman.

Disregard the white precipitate formed.

4 An excess of hydrogen peroxid is not harmful.

Use a fluted paper.

Amount of wine used.

Assuming that 24.8 cc. of potassium permanganate are required to oxidize 50 cc. of N/20 oxalic acid, 50 × 0.00315 = 0.1575.

Assuming that it requires 31 cc. of potassium permanganate to oxidize the oxalic acid formed from 23.46 cc. of wine, then

(Applicable to sweet wines containing from 1 to 6 per cent of sugar.)

Treat 50 cc. of wine with 15 cc. of sodium carbonate solution (1 to 5), add 2.5 grams of stannous chlorid, and, after shaking, allow to stand for about 5 minutes. Make to 250 cc. with water and filter. To 220 cc. of the filtrate add 30 cc. of 5% ammonium hydroxid and 200 cc. of the lead subacetate ammonia mixture. The lead should be in excess. Filter a small portion of the mixture and test the filtrate with a drop of ammonium sulphid. Add more lead ammonia mixture, if necessary, and make to 500 cc. with 10% ammonium hydroxid. Filter, and to 250 cc. of the filtrate in the nickel dish add 5 cc. of 10% sodium phosphate and 6 grams of sodium carbonate. Evaporate on the water bath and proceed as in Method I.

Treat 50 cc. of wine with 15 cc. of sodium carbonate (1 to 5) and 2.5 grams of stannous chlorid and allow to stand for 5 minutes. Add 100 cc. of water, heat to 50°C. on a water bath, make to 250 cc., and filter. To 220 cc. of the filtrate add 50 cc. of 5% ammonia and a freshly prepared lead ammonia mixture (180 cc. of lead plus 70 cc. of 10% ammonia). Determine whether an excess of lead is present. Add 20 cc. of 10% ammonium carbonate solution, mix thoroughly, make to 600 cc. with 10% ammonium hydroxid and filter. To 300 cc. of the filtrate (22 cc. of wine) in the nickel dish add 10 cc. of sodium phosphate and 6 grams of sodium carbonate and evaporate to about 80 cc. on a water bath. Wash into a flask with about 20 cc. of water. Cool to 25 cc., add 10 grams of potassium carbonate and 1 gram of potassium permanganate. Allow to stand for 30 minutes, shaking occasionally, and proceed as in Method II.

Method IV.

(Applicable to excessively sweet wines.)

Dilute to a sugar content of about 20% (grams per 100 cc.) and proceed as in Method III.

1 Amount of wine used.

Method V.

(Applicable to wines containing sucrose.)

Wines containing sucrose should be inverted before attempting the determination of glycerol by one of the above methods. Neutralize before proceeding with the inverted solution.

Instructions were to follow the method for white and red wines containing up to 1 per cent of sugar, for the analyses of Solution No. 1, and to follow the method for sweet wines containing 1 to 6 per cent of sugar for analyses of Solution No. 2.

Results obtained by each collaborator, as well as by the referee, were so erratic as to hardly warrant a detailed report.

On Solution No. 1, containing 0.50 gram per 100 cc. of glycerol, analyses show a general tendency to give high results as follows: 0.74, 0.75, 0.85, 0.86, 1.26, 1.21 and 1.26, grams per 100 cc. No particular comment was made by the collaborators on Method I. The details of the method work out nicely and can be carried through in a comparatively short time, but the results obtained are entirely too high and show a wide variation.

With respect to Solution No. 2, containing 0.75 gram per 100 cc. of glycerol, and 3 per cent of sugar, on which Method II was followed, the collaborators seem to have experienced more or less difficulty. Hartmann reports "Precipitates so bulky, not considered worth while to carry analysis to completion".

E. M. Meyer "Could not secure results that would check".

The referee experienced much the same difficulty as Hartmann with respect to bulky precipitates, but finally overcame this by using the centrifuge. However, the results obtained were not consistent.

Analytical results on this sample show an even wider variation than on Solution No. 1, as follows: Glycerol present, 0.75 gram per 100 cc.; glycerol reported varied from 0.29 to 1.78 grams per 100 cc.

It would seem that the results obtained by the use of this method on wines are inaccurate. In this connection, it is interesting to note the comment of John R. Eoff, Fermentation Chemist, with William F. Jobbins, Inc., Aurora, Ill., "My experience with the Rothenfusser method does not inspire me to recommend it for anything short of fairly pure glycerol solutions". However, not enough work has been done to determine where the trouble is.

With the advent of prohibition, the question of wines becomes of rather minor importance, but it may be found that this method can be adapted to the determination of glycerol in grape juice and vinegar. The short time in which a determination can be made by this method