While he qualitative determination of salicylic acid is so delicate and easy of execution, the quantitative estimation is unfortunately a tedious and unsatisfactory operation, in the small quantities in which it is found. This is an argument in favor of the entire prohibition of its use in foods in preference to the restriction of the quantity to be used.

The determination may be made by the same method as described above for the qualitative test, simply making the extraction complete. One hundred cubic centimeters of the beer or wine are taken, acidified with a few drops of hydrochloric acid, and extracted with three successive portions of ether of 50cc. each; these are mixed and the whole allowed to evaporate spontaneously. The residue from this evaporation is heated for an hour on the water bath to drive off volatile acids, and treated with 150cc. of pure benzine, which is allowed to stand in contact with the residue for twenty-four hours, when it is drawn off carefully and the residue again treated with 50cc. of benzine, which is added to the first portion. This 200cc. of benzine is then made up to 500cc. with absolute alcohol, and titrated directly with a decinormal solution of soda, previously standardized by operating upon a similar mixture.

The acid may also be extracted by treatment with successive portions of chloroform, which is carefully decanted, the solution evaporated, and the crystallized salicylic acid weighed.1

A. Remont2 published in 1881 a method for the colorimetric estimation of salicylic acid, which was modified by Pellet and De Grobert3 in the following manner:

A series of uniform test-tubes are prepared about 20cm. in height and 15mm. in caliber, to which are added, successively, 1cc., 0.75cc, 0.5cc., 0.4cc., 0.3cc., 0.2cc., 0.1cc. of a solution of 1 gram of salicylic acid in 1 liter of distilled water, and the volume in each tube brought to 10cc. with distilled water. To the first tube is added three drops of a dilute solution of ferric chloride (1005 to 1010 specific gravity), in the second and third two drops, in the others one drop. The last tube may be simply stirred with a glass rod which has been dipped into the iron solution. One hundred cubic centimeters of the beer or wine to be tested are now taken and shaken up with 100cc. of ether, and five drops H1S04 of 30° B., allowed to stand, carefully separated, and this operation repeated twice. The ether is quickly evaporated off in the water bath, the residue brought into a porcelain evaporating dish of about 6-8 cm. diameter, the flask washed out with a few cc. of ether and the dish placed in an air bath heated to about 50° C. to drive off the ether completely; 1.5cc. of a solution of soda is then added, of such strength that l0cc. contain 0.4 grams Na2O, which is sufficient to saturate about 0.2 grams of salicylic acid, equal to a content in the sample of 2 grams per liter. If the residue is still acid after this addition, it is due to acetic acid. By evaporating to dryness this excess can be driven off while the salicylic acid is retained, as it is able to displace acetic acid from its salts. The residue is now treated with five drops of H1S04 of 30° B., and then with 20cc. of benzine, and the whole filtered; l0cc. of the filtered benzine solution is brought into a test-tube of similar dimensions to those mentioned above; l0cc. distilled water and one or two drops of the dilute ferric chloride solution added, and the contents well shaken. If salicylic acid is present it is all taken up by the lower watery portion and the color may be compared with that of the standard tubes. If it agrees in intensity with one of these, the calculation is very simple. Suppose, for example, it agrees with the fourth tube, which contains in the l0cc. of liquid 0.0004 grams salicylic acid, then the l0cc. of benzine solution also contained 0.0004 grams, and the 20cc. from which it was taken contained 0 0008 grams in l00cc. of the wine, or .008 grams to the liter. The authors of this method found on applying it to wine to which a known quantity of salicylic acid had been added, that only 93 per cent. of the amount added was found, so their results were divided by 93 on that basis. With this modification they obtained from two samples of wine to which had been added, respectively, 0.084 and 0.126 grams per liter 0.084 and 0.123 grams.

1 These two methods are employed in the Municipal Laboratory of Paris.

2 Jour. Pharm. Chim. [5], 4, 34, Chem. Cent., 1881, 773. 3 Compt. Rend. 93, 278, Chem. Cent., 1881. 711.

M. Remont1 also devised, in 1882, what might be called an empirical method for the estimation of the quantity of salicylic acid added to wine or beer, which would doubtless prove very convenient in case its use were limited by law to a certain definite quantity, as follows:

In a liquid, similar to that which is to be tested, is dissolved a known quantity of pure salicylic acid; of this standard liquid 50cc. are taken and well shaken with 50cc. of ether, and allowed to separate; 25cc. of the ether are taken in a flat dish and subjected to evaporation at a temperature below boiling, in the presence of 10cc. of water; when the ether has disappeared, the water is poured into a graduated cylinder and its volume made up to 25cc. with the washings of the dish. This solution contains the same proportion of salicylic acid as the standard; l0cc. of the liquid to be analyzed are then taken and shaken with l0cc. of ether; 5cc. of the clear ether are taken, evaporated with 2cc. of water, and the residual liquid made up to 5cc. with the washings of the dish, as above.

Two tubes, each graduated to 30cc. are taken, and into one is introduced 5cc. of the standard aqueous solution as obtained above, and into the other the 5cc. obtained from the sample for analysis. To each tube is added the same quantity of a 1 per cent. solution of ferric chloride, avoiding an excess. The comparison of the two tubes may then be made, and the process can be made quantitative by diluting with water, or other colorimetric methods.

The author insists upon the necessity of taking, as a means of comparison, a liquid of the same nature as the sample for analysis, as the foreign matters which ether dissolves from wine, beer, or cider alter the delicacy of the color test.