F. B. Wright

Alcoholmetry is the name given to a variety of methods of determining the quantity of absolute alcohol contained in spirituous liquors. It will readily be seen that a quick and accurate method of making such determinations is of the very utmost importance to those who are engaged in the liquor traffic, since the value of spirit depends entirely upon the percentage of alcohol which it contains. When alcoholic liquors consist of simple mixtures of alcohol and water, the test is a simple one, the exact percentage being readily deducible from the specific gravity of the liquor, because to a definite specific gravity belongs a definite content of alcohol; this is obtained either by means of the specific gravity bottle, or of hydrometers of various kinds, specially constructed.

All hydrometers comprise essentially a graduated stem of uniform diameter, a bulb forming a float and a counterpoise or ballast. The hydrometers may either be provided with a scale indicated on the neck or else with weights added to sink the hydrometer to a certain mark. The first instruments are called hydrometers of “constant immersion,” the others, of “variable immersion.”

At the latter end of the last century, a series of arduous experiments were conducted by Sir C. Blagden, at the instance of the British government, with a view to establishing a fixed proportion between the specific gravity of spirituous liquors and the quantity of absolute alcohol contained in them. The result of these experiments, after being carefully verified, led to the construction of a series of tables, reference to which gives at once the percentage of alcohol for any given number of degrees registered by the hydrometer; these tables are invariably sold with the instrument. They are also constructed to show the number of degrees over-or under-proof, corresponding to the hydrometric degrees. Other tables are obtainable which give the specific gravity corresponding to these numbers.

The measurement of the percentage of absolute alcohol in spirituous liquors is almost invariably expressed in volume rather than weight, owing to the fact that such liquors are always sold by volume. Nevertheless, the tables referred to above show the percentage of spirit both by volume and weight.

Fig. 52.—Syke’s Hydrometer.

In the United States the standard liquor, known as proof spirit, contains 92.3 per cent. by weight and 94.9 per cent. by volume, of absolute alcohol; it has a specific gravity of .9186 at 60° F. A proof gallon contains by measurement 100 parts of alcohol and 81.5 parts of water. The strength and therefore the value of spirituous liquors is estimated according to the quantity by volume of anhydrous spirit contained in the liquor with reference to this standard. Thus the expression “20 per cent. overproof,” “20 per cent. underproof,” means that the liquor contain 20 volumes of water for every 100 volumes over or under this fixed quantity, and that in order to reduce the spirit to proof, 20 per cent. of water by volume, must be subtracted or added, as the case may be. Any hydrometer constructed for the measurement of liquids of less density than water may be employed. That known as “Syke’s” is most commonly used for alcoholometric purposes. It is shown in Fig. 52 and consists of a spherical brass ball A, to which is fixed two stems; the upper one B is also of brass, flat, and about 3½ in. in length; it is divided into ten parts, each being subdivided into five, and the whole being numbered as shown in the figure. The lower stem C is conical, and slightly more than an inch long; it terminates in a weighted bulb D. A series of circular weights, of the form shown in the figure, accompany the instrument; these are slipped upon the top of the lower stem C, and allowed to slip down until they rest upon the bulb D. The instrument is used in the following way: It is submerged in the liquor to be tested until the whole of the upper stem is under the surface, and an idea is thus gained of the weight that will be required to partly submerge the stem. This weight is added, and the hydrometer again placed in the liquor. The figure on the scale to which the instrument has sunk when at rest is now observed, and added to the number on the weight used, the sum giving, by reference to the tables, the percentage by volume of absolute alcohol above or below the standard quantity.

In exact estimations, the temperature of the liquor tested must be carefully registered, and the necessary corrections made. In Jones’s hydrometer, which is an improvement upon Syke’s, a small spirit thermometer is attached to the bulb, and by noting the temperature of the liquor at the time of the experiment, and referring to the tables accompanying the instrument, the strength is found at once without the need of calculation.

Dica’s hydrometer is very similar to Jones’s instrument above described. It is of copper, has a stem fitted to receive brass poises, a thermometer, a graduated scale, etc.

In Europe, Gay-Lussac’s hydrometer and tables are chiefly used for alcoholometric testing. This instrument is precisely similar in construction to those of Twaddle and Baume. On the scale, zero is obtained by placing it in pure distilled water at 59° F., and the highest mark, or 100, by placing it in pure alcohol at the same temperature, the intermediate space being divided into 100 equal divisions, each representing one per cent. of absolute alcohol. The correction for temperature, as in the above cases, is included in the reference tables.

Another hydrometer, used in France for alcoholometric determinations, is Cartier’s. In form it is precisely similar to Baume’s hydrometer. Zero is the same in both instruments, but the point marked 30° in Cartier’s is marked 32° in Baume’s, the degrees of the latter being thus diminished in the proportion of 15 or 16. Cartier’s hydrometer is only used for liquids lighter than water.

The alcoholmeter of Tralles is the official instrument for testing alcoholic liquors in the U. S. but the instrument which is most generally used both here and abroad is that of BeaumÉ. There are two instruments bearing BeaumÉ’s name, one for liquids lighter than water, the other for those which are heavier. All hydrometers, alcoholmeters and saccharometers work on the same principle, though they are each differently graduated for the particular work to be done and the details of the measuring process are slightly different. All these instruments are provided with tables whereby their readings may be corrected and the specific gravity of the liquid determined.

The above hydrometric methods can be safely employed only when the spirit tested contains a very small amount of solid matter, since, when such matter is contained in the liquor in quantity, the density alone cannot possibly afford a correct indication of its richness in alcohol. Many methods have been proposed for the estimation of alcohol in liquor, containing saccharine coloring and extractive matters, either in solution or suspension. Undoubtedly the most accurate of these, though at the same time the most tedious, is to subject the liquor to a process of distillation by which a mixture of pure alcohol and water is obtained as the distillate. This mixture is carefully tested with the hydrometer, and the percentage of alcohol in it determined by reference to the tables as above described; from this quantity and the volume of the original liquor employed the percentage by volume of alcohol in that liquor is readily found. The condensing arrangement must be kept perfectly cool, if possible in a refrigerator, as the alcohol in the distillate is very liable to be lost by re-evaporation. When great accuracy is desired, and time is at the operator’s disposal, the above method is preferable to all others.

It is performed in the following manner: Three hundred parts of the liquor to be examined are placed in a small still, or retort, and exactly one-third of this quantity is distilled over. A graduated glass tube is used as the receiver, in order that the correct volume may be drawn over without error. The alcoholic richness of the distillate is then determined by any of the above methods, and the result is divided by three, which gives at once the percentage of alcohol in the original liquor. The strength at proof may be calculated from this in the ordinary way.

If the liquor be acid, it must be neutralized with carbonate of soda before being submitted to distillation. From eight to ten per cent. of common salt must be added, in order to raise the boiling point, so that the whole of the spirit may pass over before it has reached the required measure. In the case of the stronger wines it is advisable to distil over 150 parts and divide by two instead of three. If the liquor be stronger than 25 per cent. by volume of alcohol, or above 52 to 54 per cent. under-proof, an equal volume of water should be added to the liquid in the still, and a quantity distilled over equal to that of the sample tested, when the alcoholic strength of the distillate gives, without calculation, the correct strength required. If the liquor be stronger than 48 to 50 per cent. under-proof, three times its volume of water must be added, and the process must be continued until the volume of the distillate is twice that of the sample originally taken. In each case the proportionate quantity of common salt must be added.

For the estimation of alcohol in wines, liquors, etc., the following method may be employed: A measuring flask is filled up to a mark on its neck with the liquor under examination, which is then transferred to a retort; the flask must be carefully rinsed out with distilled water, and the rinsings added to the liquor in the retort. About two-thirds are then drawn over into the same measuring flask, and made up to its previous bulk with distilled water, at the same temperature as that of the sample before distillation. The strength is then determined by means of Syke’s hydrometer, and this, if under-proof, deducted from 100, gives the true percentage of proof-spirit in the wine.

A quick, if not always very exact, method consists in determining the point at which the liquor boils. The boiling point of absolute alcohol being once determined, it is obvious that the more it is diluted with water the nearer will the boiling point of the mixture approach that of water; moreover, it has been proved that the presence of saccharine and other solid matters has but an almost inappreciable effect upon this point. Field’s alcoholometer, since improved by Ure, is based upon this principle. It is shown in Fig. 53, and consists, roughly speaking, of a cylindrical vessel A, to contain the spirit; this vessel is heated from beneath by a spirit lamp, which fits into the case B. A delicate thermometer C, the bulb of which is introduced into the spirit, is attached to a scale divided into 100 divisions, of which each represents one degree over-or under-proof. This method is liable to several small sources of error, but when a great many determinations have to be made, and speed is an object rather than extreme accuracy, this instrument becomes exceedingly useful. It does not answer well with spirits above proof, because the variation in their boiling points are so slight as not to be easily observed with accuracy. But for liquors under-proof, and especially for wines, beer, and other fermented liquors, it gives results closely approximating to those obtained by distillation, and quite accurate enough for all ordinary purposes. Strong liquors should therefore be tested with twice their bulk, and commercial spirits with an equal bulk, of water, the result obtained being multiplied by two or three, as the case may be.

Another very expeditious, but somewhat rough, method was invented by Geisler. It consists in measuring the tension of the vapor of the spirit, by causing it to raise a column of mercury in a closed tube. The very simple apparatus is shown in Fig. 54. A is a small glass bulb, fitted with a narrow tube and stop-cock. This vessel is completely filled with the spirit, and is then screwed upon a long, narrow tube B, bent at one end and containing mercury. This tube is attached to a graduated scale showing the percentage of absolute alcohol above or below proof. To make the test the cock is opened, and the bulb, together with the lower part of the tube, is immersed in boiling water, which gradually raises the spirit to its boiling-point. When this is reached, the vapor forces the mercury up the tube, and, when stationary, the degree on the scale to which it has ascended gives directly the percentage of alcohol.

Another method, which is not to be relied on for very weak liquors, but which answers well for cordials, wines, and strong ales, is that known as Brande’s method. The liquor is poured into a long, narrow glass tube, graduated centesimally, until it is half-filled. About 12 or 15 per cent. of subacetate of lead, or finely powdered litharge, is then added, and the whole is shaken until all the color is destroyed. Powdered anhydrous carbonate of potash is next added until it sinks undissolved in the tube, even after prolonged agitation. The tube is then allowed to rest, when the alcohol is observed to float upon the surface of the water in a well-defined layer. The quantity read off on the scale of the tube and doubled, gives the percentage by volume of alcohol in the original liquid. The whole operation may be performed in about five minutes, and furnishes reliable approximative results. In many cases it is necessary to add the lead salt for the purpose of decolorizing the liquid.

For the investigation of the amount of sugar in, or the concentration of the mash, or beer, a specially scaled hydrometer is used which is termed a saccharometer. Sugar possesses a higher degree of specific gravity than water, and hence it follows that the greater the amount of sugar in the mash the higher will be the specific gravity. The less the hydrometer sinks into the fluid the greater the amount of sugar present. Saccharometers are provided with thermometers whereby the reading may be corrected to a standard temperature, usually 59° F. The saccharometer is correct for solutions containing sugar alone but it is only approximately correct for mash liquor which contains a variety of other matters in variable quantities.

It is a prime necessity that the distiller should be able to determine if the mash has been completely saccharified by the malt. For this purpose a solution of iodine is used. Iodine gives to starch a blue color. If the starch however, has been completely changed into sugar there will either be no discoloration or the filtered mash liquid which is at first a yellowish red becomes blue, then violet, and at last red.

Determination of the Purity of Alcohols. While the knowledge of the amount of alcohol contained in a liquid is of great practical utility, this does not give any idea of the impurities present.

An alcohol of 100 degrees or an absolute alcohol, may contain numerous impurities which may greatly affect its quality. It is therefore necessary in addition to analyze the purity of the alcohol.

In commercial practice there are certain simple processes which will give a basis by which to determine the impurities left after distillation and rectification. These processes are largely empirical. They are based on the perception of the senses and are consequently of an entirely relative degree of precision. Nevertheless, when made by a practical expert, the operation may give very useful preliminary indications.

This test is made in a glass of special shape broad at the bottom and narrowing at the top in order to concentrate the aroma of the product. Ordinary brandies are tested undiluted. Commercial alcohols, of about 95 degrees must be diluted with water to a maximum of 30 degrees. Otherwise the burning tang of the alcohol would preclude any delicacy of perception and allow impurities to pass unnoticed.

The operation is begun by examination by sense of smell. The glass is half filled with the liquid diluted with one half of pure water. The glass is covered with one hand and shaken violently for a few seconds. Immediately upon uncovering it, the quality of the alcoholic vapors may be ascertained by their odor.

For the examination by sense of taste, the operator rinses his mouth for a moment with the liquid itself. The taste of ethyl alcohol is fairly transient;—it disappears quickly allowing the taste of the accompanying foreign matter to be perceived almost immediately afterward. With a little practice this test enables one to distinguish by their flavor the primal origin of alcohols and to judge of their purity. Some professionals succeed by training in arriving at high degree of skill in the art of tasting alcohol as it should be done.

In order to determine the purity of alcohol there are besides chemical tests used by the trade. These tests, which consist in characterizing and measuring separately the impurities which alcohol may contain, such as acids, ethers, aldehydes, bases, etc., belong exclusively to analytical chemistry; they are extremely delicate and complicated. We will not venture to touch upon them here.

One of the simplest tests for purity is that of Barbet. This is based upon the time taken to discolor a solution of permanganate of potash under the action of the tested alcohol. It is not only very rapid but in general more practical than other tests. It allows the aggregate of the impurities contained in an alcohol to be ascertained in a single operation.

The permanganate solution used is very weak (0. gr. 200 of salt), and of a violet-red color. The technique of the proceeding is as follows: 50 cubic centimeters of the alcohol to be tested are placed in a glass vessel the temperature of which is maintained at 64.40°F. 2 cubic centimeters of the permanganate solution are abruptly added and the time noted to within a second. Discoloration is awaited and as soon as it takes place the time is again noted. The total discoloration of the permanganate is not very marked and passes through intermediate stages; therefore it is preferable not to await complete discoloration but to stop at a pale salmon tint, which tint may be comparatively fixed by a sample of colored liquid (say a solution of fuchsine and chromate of potash).

The comparative times of discoloration obtained by M. Barbet with various commercial alcohols, are as follows: