When the proportion of alcohol is not too small, the following table, due to Meker,2 maybe used to ascertain the proportion from the density of the mixture. The latter, if not anhydrous, should be dehydrated by treatment with potassium carbonate.

1 Analyst, 1916, 41, 245.

2 Calvet, " Alcools," p. 343.

Absolute alcohol and ether; proportions by volume. Alcohol, D15 = 0.795; Ether, D15 = 0.720.

For alcohol of 95 per cent. strength the densities are as follows: -

When the amount of alcohol is too small to be determined by means of the density the following process, due to P. Szeberenyi1 may be adopted. It is based upon the fact that alcohol is readily oxidised by a moderately acid solution of potassium dichromate, whilst ether is but slightly affected.

Five c.c. of the sample are diluted with water to 100 c.c. Thirty c.c. of this are placed in a 500 c.c. flask, into which have already been introduced 10 c.c. of iN/2.dichromate, 20 c.c. of dilute sulphuric acid (1: 1), and 40 c.c. of water. After boiling for fifteen minutes under a vertical condenser, and cooling, the excess of dichromate is titrated iodometrically. Anything above 0.9 c.c. of dichromate used represents 0.64 mg. of alcohol for each c.c. The exact quantity of dichromate used up by ether must be determined by experiment.

For somewhat larger amounts of alcohol, 5 c.c. of the aqueous solution are taken with 20 c.c. of the dichromate, 55 c.c. of water, and 20 c.c. of the dilute sulphuric acid. In such cases, no correction need be made for the dichromate used up by the ether.

If the ether is also to be estimated, a 2 per cent. aqueous solution is prepared, and 10 c.c. are treated for alcohol, as described. Another 10 c.c. are then placed in a stoppered 500 c.c. flask with addition of 40 c.c. of N/2 .dichromate and 40 c.c. of sulphuric acid previously mixed with 10 c.c. of water, and cooled. After waiting for twelve hours, the contents of the flask are diluted with 400.500 c.c. of water, and the excess of dichromate is titrated as before.

After deducting from the result the dichromate used up by the alcohol (multiplied by 1.1), the proportion of ether is found by multiplying the number of c.c. reduced by the factor 4 6.

1 Zeitsch. anal. Chem., 1915, 54, 409.11.

The following approximate method for the determination of alcohol and ether in a mixture of the two is described by J. Fleischer and H. Frank.1

When 10 c.c. of a mixture of alcohol and ether are treated with 5 c.c. each of benzene and water, separation takes place, the alcohol going into the aqueous layer and the ether into the benzene layer. The proportions of alcohol and ether can then be determined approximately by the increase in volume of the two layers.

If the alcohol-ether mixture already contains water, the specific gravity of the mixture must first be ascertained. The volume of the ether is.then determined in the foregoing manner. The specific gravity D of the aqueous alcohol in the original mixture is calculated from the formula

where d = the specific gravity of the original mixture, a = the number of c.c. of ether found in the 10 c.c, and 0729 = the sp. gr. of ether. From the value of D the percentage of alcohol is obtained from the tables as usual.

For determining small amounts of water and alcohol in "anaesthetic" ether a process has been worked out by Mallinckrodt and Alt.2 It depends upon the removal of the water by means of a weighed quantity of potassium carbonate, any alcohol remaining with the carbonate being eliminated by washing with absolute ether. The operations are carried out in a Regnault pyknometer as used for estimating the specific gravity of solids, and the increase in weight of the potassium carbonate gives the quantity of water in the ether treated.

To determine the alcohol, another portion (100 grams) of the sample is dehydrated by treating it in a stoppered flask for fourteen hours with freshly-dried potassium carbonate (40 grams). The specific gravity of the anhydrous mixture of ether and alcohol is then ascertained, and the proportion of alcohol read off from a curve which has been constructed with data given by mixtures of ether and alcohol in known amounts. For these data and the details of manipulation the original paper should be consulted.

A simpler procedure has been described by R. L. Perkins, who determines both alcohol and water from the specific gravity of the sample, taken before and after dehydration with potassium carbonate.3

1 Chem. Zeit., 1907, 31, 665. 2 J. Ind. Eng. Chem., 1916, 8, 807. 3 Ibid. 1917, 9, 521.

The specific gravity of the ether is taken at 25°/25°. From 100 to 200 c.c. are then placed in a flask, 30.50 grams of (dry) potassium carbonate added, and the mixture allowed to stand for twenty.four hours, with occasional shaking. The specific gravity of the treated ether is then again determined. By reference to a graph constructed from the undermentioned data the percentages of both alcohol and water are obtained.

Pure anhydrous ether has the sp. gr. 0.70968 at 25°/25°. Mixtures of this with known quantities of "absolute" alcohol (0 to 4 per cent.) and water (0 to 1 per cent.) were made, and gave the specific gravities shown in the table below. The alcohol contained 0 8 per cent. of water by weight, thus introducing the small quantities of water (0 006, 0 012, etc.) shown in the second column. The percentages are by volume.

Ether, alcohol, and water mixtures (Perkin).

The value of the specific gravity obtained for the dehydrated ether is referred to the lowest curve in the accompanying diagram. This gives the percentage of alcohol in the dehydrated ether, which is practically that in the original sample. (A correction could be made on account of the water removed, but it would always be less than 0 05 per cent.)

The intersection of a vertical line through the point thus found with the horizontal line representing the specific gravity of the original sample indicates the quantity of water in the sample. If the point of intersection falls on one of the curves, the percentage of water is shown by that curve. If it falls between two curves, the percentage is obtained by interpolation of the vertical distance between them.