proof , in modern usage, is the amount of pure alcohol in a spirit or other alcoholic beverage, stated as a percentage of the total volume of that beverage; when written, that percentage is usually followed by the abbreviation ABV, for “alcohol by volume.”
Originally, however, proof was a threshold, a strength at or above which a spirit was acceptable for certain uses. From that, it evolved into a benchmark figure, usually 50 percent alcohol, with the strengths of spirits expressed as a percentage of or other relation to that figure, as in “80 proof,” for 80 percent of proof (or 40 percent alcohol), or “10 (degrees) under proof” (also 40 percent alcohol). It was a short step from there to discarding the benchmark strength and simply expressing the strength or “proof” of a spirit as its percentage of alcohol.
Before the invention of the hydrometer, it was impossible to determine the exact strength of a distilled spirit. See hydrometer. The earliest methods of which we have record date back to the seventeenth century, with “proof spirit” being first mentioned in 1639. Some tests were simple, binary ones: when you rubbed the spirit between your hands, either it all evaporated or it didn’t; when you dipped a piece of paper in it, either the moisture spread (a sign of weak spirit) or it didn’t; if you soaked a piece of cloth or a spoonful of gunpowder with it, either it burned or it didn’t. Others required more interpretation: if you ignited the spirit and measured how much was left when the flame went out, you would have a rough measure of strength. Or you could raise bubbles in it, whether by blowing into it with a reed or shaking it in a glass vial. The size and duration of the “bead”—the cap of bubbles—gave a fairly good idea of the strength. This last method, in the hands of an experienced user, can yield surprisingly accurate results, within a fairly narrow band of possible proofs (beyond those the bubbles will not properly form). Unfortunately, it is only reliable with spirits straight off the still, as it is easily tricked by adulterants. Nonetheless, this writer has witnessed a veteran palenquero at Santa Maria Albarradas in Oaxaca come within a degree or two of hydrometer proof with it.
In his 1690 Medicina hydrostatica, Robert Boyle suggested a method of measuring the alcoholic contents of liquids based on their density. This led to a good deal of experimentation with hydrometers, particularly in England, where the excise system demanded a more accurate way of measuring the strength of spirits. In 1762, British Parliament did something revolutionary and established a proof for alcohol: henceforward, proof spirit would have a specific gravity of 0.916, measured at 60° F (the different densities of alcohol and water mean that the ratio between their volumes changes with the temperature of their solution). This translates to 50 percent alcohol, but measured not by volume but by weight (alcohol being lighter than water, that works out to 57 percent ABV). It also adopted the hydrometer John Clarke had been working on since the 1720s to measure where a spirit stood in respect to this proof.
Clarke’s device was crude. We need not go into all the details here, but one aspect has stuck with us to this day. It showed its results as proportions: for overproof spirits, the number was how many gallons of the spirit one gallon of water would reduce to proof (the lower the number the stronger the spirit); for underproof ones, it was how many gallons of the spirit would resolve up to proof if you could extract one gallon of water (here, the higher the number, the stronger the spirit). See overproof. Some of these proportions are still with us, such as the “one in six” underproof, which imported brandy, rum, and arrack could not legally sink below nor domestic gin rise above. Translated into modern proof, that is 47.4 percent ABV, a strength at which a number of London dry gins are still bottled.
Clarke’s system was, of course, perfectly mad, and the British government commissioned Charles Blagden and George Gilpin to find a better way. Their detailed findings, which pointed toward a simple ABV-based system, were published in 1794, but the government did not accept them. Instead, in 1818 it adopted a hydrometer by Bartholomew Sikes (1731–1803), which measured alcohol strictly by weight; at least it had a straight numeric scale. Rather than rejigger all the various proofs enshrined in law, the law simply translated those into the new scale. Thus one in six became “17 degrees under proof,” or 41.5 percent alcohol (83 percent of 50 percent). The Sikes system lasted in the United Kingdom and its Commonwealth until 1980.
Spirits are marketed at a wide range of proofs—indeed, the same spirit can be sold at different proofs in different countries. Some of these are determined strictly by law: for example, straight spirits in the United States cannot be sold below 40 percent ABV, so a great many are bottled at that proof; where the minimum is 37 percent, many of them will be found at that strength. Others follow tradition: some rums and gins are bottled at 57 percent ABV, which is the old British “proof” (50 percent ABW; some of these are labeled as “navy strength,” although in fact the Royal Navy bought its sprits at 5.5 degrees under proof, or 54.5 percent ABV). See navy-strength. Similarly, spirits at 43 percent ABV and 63 percent ABV are at 75 percent and 110 percent of that British proof, respectively. As another artifact of the old Sikes system, spirits are generally not labeled “overproof” unless they are over 57 percent.
Other spirits have their proofs set for reasons of taste or marketing. The same spirit, diluted to different proofs, can taste subtly, and sometimes not so subtly, different at each one, as different esters are emphasized. Conversely, in the modern cocktail world, higher proofs are often preferred, as they better resist the dilution that is part of drink mixing, so many new spirits are launched at 45–55 percent ABV. Spirits labeled at “cask strength,” unless they are also single-barrel, are bottled at an artificial composite proof that suggests the “normal” barrel strength for their category.
Déjean, Antoine. Traité raisoné de la distillation. Paris: 1753.
[De Mayerne, Theodore]. The Distiller of London. London: 1639.
McCulloh, R. S. Report … on Hydrometers. Washington, DC: US Department of the Treasury, 1845.
Smith, George. A Compleat Body of Distilling, 2nd ed. London: 1731.
Tate, Francis G. H. Alcolometry. London: HM Stationery Office, 1930.
By: David Wondrich