Small quantities of alcohol are often added to chloroform as a preservative, especially to chloroform intended for pharmaceutical purposes. The British Pharmacopoeia requires 2 per cent. of absolute alcohol to be added. For quantities of this order the following method is satisfactory.

Fifty c.c. of the sample are extracted in a separator with strong sulphuric acid, first with 20 c.c, again with 20 c.c, and finally with 10 c.c. The acid extracts are run into 200 c.c of water, and distilled gently until 100 c.c have been collected. The distillate is saturated with salt and extracted with petroleum ether to remove any traces of chloroform, and then distilled again, the distillate made up to 100 c.c with water, and the proportion of alcohol determined from the specific gravity in the usual manner.

For the detection of alcohol in chloroform the same method may be used for the extraction, except that 10 c.c will usually suffice, with proportionately smaller quantities of acid and water, and the treatment with petroleum ether is not necessary. To the distillate from the acid the iodoform test is applied, supplemented if desired by any others of the qualitative tests already described. Take 10 c.c of the distillate, add 6 drops of a 10 per cent. solution of potassium hydroxide, and warm the liquid to about 50°. A solution of potassium iodide saturated with free iodine is next added drop by drop with agitation, until the liquid becomes permanently yellowish-brown in colour, when it is carefully decolorised with potassium hydroxide solution If ethyl alcohol is present, iodoform is gradually deposited at the bottom of the tube in yellow crystals.1

Nicloux's method for estimating alcohol in chloroform is given in some foreign pharmacopoeias, and is often useful, though not very accurate as described. Five c.c. of the sample are shaken with 20 c.c. of water to extract the alcohol; 5 c.c. of the aqueous-alcoholic extract (which should contain not more than 02 per cent.

1 Hager, as quoted by Baskerville and Hamor, J. Ind. Eng. Chem., 1912, 4, 502.

of alcohol) are mixed in a test-tube with 0.1 to 0.2 c.c. of a solution of potassium bichromate (19 grams per litre), and then with pure sulphuric acid (D = 1.84), added cautiously. When the acid is in sufficient quantity (4.5 to 6 c.c.) the solution is decolorised, and the liquid is now gradually titrated with the bichromate solution, being shaken and gently warmed after each addition, until the colour changes from greenish-blue to a permanent greenish colour. This change is most readily recognised in solutions containing less than 0.2 per cent. of alcohol.

It is advisable to make a second experiment in which the amount of dichromate used in the first determination, less 0.1 c.c, is run in at once, the sulphuric acid added, and the liquid heated to boiling. The contents of the tube should remain greenish-blue, whilst in another determination in which 01 c.c. of dichromate more than the first reading is added, the colour should change to greenish-yellow. The number of c.c. consumed, divided by 1,000, gives the amount of absolute alcohol in c.c. per c.c. of solution. When the proportion of alcohol is less than 01 per cent., the strength of the dichromate solution should be doubled.

As mentioned above, this method does not give very accurate results, the reason being that all the alcohol is not removed by the treatment described. This can be done, however, by shaking the chloroform with twice its volume of water, separating the aqueous extract, and repeating the operation at least ten times. The combined extracts are then made up to a definite volume, and a suitable quantity taken for the oxidation.

Another useful method for the detection of alcohol in chloroform is one proposed by Thresh.1 The sample of chloroform is shaken with an equal volume of water, the aqueous extract distilled with sulphuric acid and potassium dichromate, and the distillate tested for acetaldehyde by passing it into a tube containing a few c.c. of a strong (syrupy) solution of caustic soda. The liquid in the tube is boiled for a few seconds, and then set aside for a couple of hours. A yellow colour and a characteristic odour are developed, and if the proportion of aldehyde is sufficient, flocks of the aldehyde-resin are deposited. The general directions given by Thresh for the oxidation of dilute alcohol solutions are to use 100 c.c, 2 c.c of saturated solution of potassium dichromate, and 8 c.c of a dilute sulphuric acid (equal quantities of B.P. acid and water), distilling 20 c.c of the mixture. As applied to the detection of alcohol in chloroform, one-fourth of these quantities, or less, can be taken. Thresh made the method approximately quantitative by colori-

1 Chem. News, 1878, 38, 251.

Metric estimation of the resin produced (dissolved in warm alcohol and diluted to match standard solutions), but the method already given is preferable when quantitative results are required.