The refraction reading of the distillate is taken with the Zeiss immersion refractometer. This reading will be somewhat higher than that of dilute ethyl alcohol of the same specific gravity (see table at end of chapter). The difference is multiplied by the factor 11, and subtracted from the percentage of proof spirit corresponding with the specific gravity of the distillate. (The factor 1.1 is an experimental result.)

Example: Suppose the specific gravity to be 0.98690 and the refraction reading to be 32.0.

From the table, sp. gr. 098690 = 170 per cent. of proof spirit, the refraction value of which is 28.7.

Then (320 - 287) X 1.1 = 36, the subtractive correction to be applied.

Hence the percentage of proof spirit in the final distillate is 17 0 - 3.6 = 134.

This is also the percentage (by volume) of proof spirit in the sample, since the volume taken (75 c.c.) was the same as that of the final distillate.

In a series of mixtures, of known composition, containing from 5 to 25 per cent. of ethyl alcohol at proof strength with varying proportions of higher alcohols, the mean refractometer correction was 4 3 (lowest 3.38, highest 5.76). If, therefore, no refractometer is available, an approximation to the correct result may be obtained by deducting this mean correction from the quantity of proof spirit indicated by the specific gravity of the final distillate.

As a rough "sorting out" method, which does not pretend to estimate the ethyl alcohol, but which is useful for showing whether such alcohol in a fusel oil approaches the 15 per cent. limit or not, the following process may be employed.

Mix in a separator 100 c.c. each of the fusel oil, water, and petroleum ether. Shake vigorously for a few minutes. Let stand five minutes. Run off the aqueous layer into a graduated cylinder, and determine its volume and specific gravity. If the total quantity of proof spirit in the aqueous layer, calculated on the sample, is less than 15 per cent., it is probable that the proportion of ethyl alcohol, at proof strength, is also less than 15 per cent. But if anything more than a rough idea of the amount is wanted the longer method should be employed. A little emulsion sometimes forms; the volume of the aqueous liquid containing this can be determined separately, after the main clear portion has been run off for the specific gravity determination.

Refraction.values of aqueous solutions of higher alcohols, compared with those of proof ethyl alcohol (" proof spirit "). zeiss immersion refractometer. temperature 15.6°.

Several approximate methods have been suggested for the estimation of ethyl alcohol in fusel oil, such as that of G. L. Ulex,1 but they are more troublesome than the foregoing short process, and do not give very accurate results. Ulex's process is as follows: -

Distil 100 c.c. of the sample until 5 c.c. have been collected. Shake this distillate with an equal volume of saturated salt solution, and let the mixture separate. If the fusel oil layer amounts to one-half the volume of the distillate or more, it may be taken that the sample contained less than 15 per cent. of spirit. If no fusel oil separates out, or if the volume is less than one-half that of the distillate, it may be assumed that there is at least 15 per cent. of spirit in the sample. To ascertain the quantity more nearly, shake 100 c.c. of the sample with an equal volume of saturated salt solution, separate the latter, and distil it. Make up the distillate to 100 c.c, take the specific gravity, and from this determine the amount of alcohol present, assuming that the distillate contains ethyl alcohol and water only.

1 Neu. Jahrb. Pharm., 1873, 39, 333.