It is often sufficient for qualitative purposes merely to add a drop or two of dilute, neutral solution of ferric chloride to a little of the distillate obtained from wine and contained in a porcelain dish. Any appreciable quantity of added salicylic acid will reveal itself by the violet coloration produced. Minute traces, however, are often naturally present in wine. If there is any doubt, 50 c.c. of the sample are acidified with a few drops of hydrochloric acid, and shaken gently with 50 c.c. of a mixture of ordinary ether and petroleum ether (equal volumes); the ether extract separated, passed through a dry filter, and evaporated spontaneously to dryness. On now testing with a little very dilute ferric chloride solution, a strong violet or reddish-violet colour is obtained if salicylic acid is present in sensible quantity.

To estimate the amount the following process, due to Fresenius and Grunhut,1 may be employed.

Twenty-five or 50 c.c. of the wine are acidified with dilute sulphuric or hydrochloric acid, and extracted repeatedly in a separator with a mixture of ordinary ether and petroleum ether (equal volumes). From the mixed ethereal extracts the salicylic acid is removed by extraction with two successive quantities of 10 c.c. each of decinormal soda solution, followed by two washings with water. The combined alkaline solutions are run into a stoppered flask of about 150 c.c. capacity, and the traces of dissolved ether expelled by heating on the steam-bath. The cooled liquid is then titrated in the following manner.

1 Zeitsch. anal. Ghem., 1899, 38, 292.

The solution is first neutralised with hydrochloric acid, and a definite volume of "bromine solution ' (12408 grams of KBr + 3481 grams of KBrO3 per litre) is added, the quantity depending upon the amount of salicylic acid which is likely to be present. At least twice the quantity of bromine required for the reaction (see below) should be added. Ten c.c. of strong hydrochloric acid are now run in, the flask is at once closed, shaken gently to mix the contents, and placed in a cold bath for ten minutes.

An excess of potassium iodide solution (10 per cent. strength) is then run into the flask, and after standing for five minutes the liberated iodine is titrated with decinormal thiosulphate solution and starch in the usual manner. To standardise the "bromine solution" a blank experiment is carried out in precisely the same way on a volume of water similar to that of the neutralised extract.

The quantity of iodine liberated in the sample will be less than that in the blank experiment by the amount of halogen used up by the salicylic acid; and from the equation given below, this amounts to 6 halogen atoms for each molecule of salicylic acid. Hence the difference between the volumes of thiosulphate solution required for the two titrations, expressed in cubic centimetres and multiplied by 00023, gives the weight in grams of the salicylic acid present.

Reactions: -

(1) C6H4(OH)COOH + 8Br = C6H1Br3OBr + 4HBr + CO2.

(2) C6H1Br3OBr + 2KI = C6H1Br3OK + KBr + I2; so that ultimately 6 atoms of halogen are used up by one molecule of salicylic acid.

The "bromine solution ' liberates bromine on acidifying, in accordance with the equation: -

KBrO3 + 5KBr + 6HC1 = 6KC1 + 3H10 + 6Br.

Its strength is about 0 01 gram of Br per c.c.