From the History of Alcoholometric Tables

The distillation of ethyl alcohol from products of fermentation was known already in ancient times. Owing to the extremely imperfect apparatus, the alcohol obtained was not only contaminated with various by-products of fermentation, but also contained various, and moreover considerable, quantities of water. Very soon, in buying and selling, the question had to arise of the “strength” of the alcohol, i.e. of the quantity of the active principle in the alcohol being sold and bought. Taste could not long serve as a measure of strength: side impurities, sometimes artificially added (in the taverns of Russia such artificial admixtures were practiced right up to the introduction of the monopoly), completely masked the quality of the commodity. Alchemists judged the strength of alcohol by whether it “burns,” or whether objects moistened with the given alcohol can burn. In the fifteenth century Savonarola judged the strength of alcohol “by whether it floats or sinks in vegetable oil.”

The correct path to determining the active principle in alcohol was found in the eighteenth century by Réaumur, who pointed out that samples of alcohol of different strength have different specific gravities.

The choice of specific gravity as a measure of the strength of alcohol must be considered very successful, since the specific gravity of alcohol changes unambiguously with concentration; that is, to each specific gravity there corresponds one and only one concentration (which is not observed for other properties of aqueous-alcoholic solutions, such as, for example, indices of refraction of light or heat capacity). But, as Réaumur had already shown, when alcohol is mixed with water contraction occurs, and the specific gravity of solutions of water and alcohol changes very complexly with the concentration of the solutions. The only way for the practical determination of the concentration of a solution (strength) is the compilation of detailed tables in which, on the basis of experimental data, the specific gravities and concentrations of aqueous-alcoholic solutions are calculated as densely as possible (at intervals of 1%).

For the first time such detailed alcoholometric tables were compiled in England by Gilpin in 1792–1795. Gilpin obtained absolute (as it seemed to him) ethyl alcohol, prepared 40 solutions of it with water, and determined their specific gravities at \(60^\circ\) F., referring them to water at the same temperature (i.e. Gilpin’s specific gravities are \(d \frac{60^\circ}{60^\circ}\) F.); from these data he calculated ...

fractional tables. In 1800 Dumas carried out analogous work in France; in 1811 in Germany—Tralles; in 1824 again in France—Gay-Lussac.

The fate of Tralles’s work is the most interesting. Tralles obtained absolute ethyl alcohol much stronger than that of Gilpin, but, according to his calculations, the initial spirit—Gilpin’s—contained only 98.2% ethyl alcohol,¹ and consequently Gilpin’s tables were completely incorrect. Tralles redetermined the specific gravities for several strong solutions, corrected Gilpin’s data for the remaining concentrations, and compiled tables, introducing volume percentages instead of percentages by weight. In 1847 Brix put in order the data of Tralles and Brix (specific gravities \(d \frac{60^\circ}{60^\circ}\Phi. = d \frac{12^4/9}{12^4/9}P.\)) which for a long time formed the basis of German and Russian official alcoholometry.

In view of the economic importance of accurate accounting of alcohol, in various countries, continuously since Gilpin’s time, different scientists have been redetermining the specific gravities of aqueous-alcohol solutions. It is interesting that the most important work in this respect proved to be the purely scientific work of D. I. Mendeleev of 1865 (“On the Combination of Alcohol with Water”). The center of gravity of Mendeleev’s investigation lay in obtaining unquestionably absolute alcohol. After long labors Mendeleev succeeded in this, and we may now say with confidence that Mendeleev was the first scientist to have absolute alcohol (100%) in his hands. No less careful determinations of the specific gravities of solutions of alcohol and water enabled Mendeleev to compile alcoholometric tables, which diverged from the official tables of all countries.

In the eighties a reform was carried out in Germany: Tralles’s tables were rejected, and in their place new ones were compiled—at \(15^\circ\) C according to Mendeleev’s data, and for other temperatures according to the additional data of the German Verification Commission; the normal temperature was taken as \(15^\circ\) C, and the unit as the percentage by weight. Following Germany, other countries also undertook checks and corrections of their tables, and everywhere Mendeleev’s data were taken as the basis. In the United States, the famous Morley compiled tables, again guided by Mendeleev’s work. Only in the homeland of D. I. Mendeleev, in Russia, Tralles still firmly prevails, with volume percentages, with the absurd normal temperature of \(12^4/9\) R. (\(=60\) F.) and with errors; true, the errors are small, but nevertheless, with the enormous turnover of alcohol, they are noticeable.

Despite the numerous studies in alcoholometry over the last century, despite the fact that among the scientists who have dealt with this question there are often persons with eminent names (Reaumur, Dumas, Richter, Gay-Lussac, Pouillet, Mendeleev, Young, Morley, and others), even at the present time the matter of alcoholometry is not finished. The late A. G. Doroshevsky, in his monograph Research in the Field of Aqueous-Alcohol Solutions (from which the data of our note are drawn), gives the results of his enormous, in quantity of labor, recalculations, comparisons, and corrections of the data of various scientists: it turned out that the tables for \(10^\circ\), \(15^\circ\), and \(20^\circ\) C, based on Mendeleev’s data, may be recognized as irreproachable;²

¹ According to Mendeleev’s data, 89.06%, whence it is clear that Tralles’s alcohol was not absolute.

² It is worth noting that, for laboratory purposes, the alcoholometric tables found in various handbooks (for example, those of Windisch, Landolt, and others) are compiled for specific gravities in vacuum and for temperatures according to the hydrogen thermometer. A. G. Doroshevsky compiled a very convenient table for specific gravities in air and for \(15^\circ\) according to the mercury thermometer; this table considerably facilitates the work.

for temperatures of \(0^\circ\) and \(30^\circ\), new experiments and calculations are required, which has been undertaken by the Berlin Verification Commission (according to private communications from the Commission to A. G. Doroshevsky).

The matter concerns only the precise determination of specific gravities, for the first and most important task of alcoholometry—the obtaining of unquestionably absolute alcohol—may now be considered solved and easy \(^{1}\).

In conclusion I note that A. G. Doroshevsky and M. S. Rozhdestvensky, on the basis of their data, have compiled alcoholometric tables for methyl and n-propyl alcohols. In view of the broad practical applications of methyl alcohol, the tables for the latter are of enormous importance and are now being considered by various governments. The tables of Doroshevsky and Rozhdestvensky have been officially adopted in the United States.

A. Rykovsky.