Ice colder than itself
Many people believe the ice must have been below freezing to make a drink colder than 0 degrees C, but ice-chilled cocktails are routinely as cold as minus 6 degrees C. Arnold proves it with a martini and a thermometer. First stir a pitcher of ice and water until it reads 0 degrees, to see that the ice is at 0 degrees and nothing is stored up in it. Put 2 ounces of gin and a little dry vermouth into a metal tin at room temperature, add about 120 grams of ice patted dry, and stir with the thermometer. After about 10 seconds the drink is near 5 degrees, after about 30 seconds near 0 degrees, where most bartenders stop. If you keep stirring it keeps falling: perhaps minus 4 degrees after a minute and about minus 6.75 degrees after two, when it levels off at something close to equilibrium. He would never stir a drink for drinking that long, because it would be too diluted.
The reason is a tug of war between heat and entropy: heat energy favours freezing, entropy favours melting, and which one wins depends on temperature. In a cocktail the contest always ends in a tie, but where it ends, the freezing temperature of the drink, can change.
The chilling power of ice
Every gram of ice that melts supplies 80 calories of chilling power. A 3.5-ounce daiquiri shaken for 10 seconds melts 55 to 65 grams of ice, which averages about 2,000 watts of chilling power for that one drink. All ice has the same power per gram; what differs is how fast it is delivered, and that depends on size and shape.
Surface area, surface water and holdback
Smaller pieces have more surface for their weight, so they chill faster, but the same surface causes two problems.
- Chilling rate. Ice melts at its surface, and the drink must also keep moving over it. Stirring or shaking brings fresh liquid to the ice and carries the cold meltwater away, which is the main way a drink is chilled quickly. Ice sealed in a plastic bag or a gel pack chills much more slowly, and cannot take a drink below 0 degrees, because its meltwater never mixes with the drink. Even a block of steel from liquid nitrogen does not chill as fast as melting ice cubes.
- Surface water. Ice near its melting point is wet, and small ice carries more water, which has already given up its chilling power but still goes into the drink. This is why bartenders think thin or small ice overdilutes. Arnold made crushed ice, spun the extra water off in a centrifuge and found that the drinks ended up as diluted as with big cubes once they reached the same temperature. The remedy is to put a strainer over the tin or mixing glass and snap the water off the ice before using it, or to crack large pieces into smaller ones just before use.
- Holdback. Some of the drink stays stuck to the ice. Stirring with crushed ice and sloppy pouring held back 12 to 25 percent of the drink, small machine ice 7 to 9 percent, and large cut block ice only 1 to 4 percent. Draining, giving the vessel a sharp snap, setting it upright and straining once more brings every kind of ice down to about 1 to 4 percent. Smooth block ice holds back less than its surface area alone would predict, because the shape of the surface matters as well as its size.
Ice that has been used is fine for a second use, for example to serve a stirred drink over the ice it was stirred with, as long as it still looks good: it is colder than fresh ice and wet with cocktail instead of water.
Very cold ice and equilibrium
Ice colder than 0 degrees actually chills more slowly at first, because it has to warm to freezing before it can melt. The effect is small: ice at minus 1 degree stores only as much extra chilling energy as one more gram of melted ice per 160 grams. Ice is also a fairly good conductor of heat, about three and a half times better than still water, so ice that has been out of the freezer a while is at about 0 degrees and its starting temperature can be ignored.
As a drink chills, the rate falls, and after a couple of minutes it hardly changes, because it is approaching the freezing point of the cocktail. You never quite reach that point, but blended drinks come closest, since their chilling is both extremely fast and extremely efficient. A drink stops getting colder when you remove the ice, or when its chilling rate falls below the rate at which heat enters from the room. The different methods, stirring, shaking, building and blending, each have their own chilling limits, and these shape each style of drink.
The Fundamental Law of Traditional Cocktails
Arnold defines a traditional cocktail as one made of liquor, mixers and ice at 0 degrees, with ice as the only chilling and no heat gained or lost to the surroundings. From that he states his law.
There is no chilling without dilution, and there is no dilution without chilling.
A drink is only diluted because ice melts, and ice only melts by chilling the drink. This explains why stirred drinks are warmer and less diluted than shaken drinks, and why traditional cocktail ratios work.
Proving it with Manhattans
His consequence for stirred drinks is that two drinks with the same recipe that reach the same temperature are diluted by the same amount, whatever ice, stirring speed or time got them there. In Experiment 2, a Manhattan of 4 ounces of rye (50 percent), a fat three-quarter ounce of Carpano Antica Formula vermouth and four dashes of Angostura is divided in two and stirred with a thermometer, one half over large ice and one over the same ice broken in half. When both reach minus 2 degrees C, the drinks taste the same, look the same and have the same wash line in the glass. Each picks up between 1.25 and 1.5 ounces of water and finishes at 26 to 27 percent alcohol. His home trick follows: stir with a thermometer and stop at the temperature you like, and the drink is the same every time.
Experiment 3 covers the habit of leaving ice in the mixing tin while making a round of drinks. Three identical Manhattans are made: ice sits in one for 105 seconds before it is stirred for 15 seconds, ice sits in another for 105 seconds after it is stirred, and the third is stirred and poured at once. The three drinks come out almost identical. Standing ice dilutes only marginally, and any differences were random and not repeatable, so a short wait of a couple of minutes with large ice does no harm. Even so, Arnold admits that he still feels uncomfortable seeing ice sit unused in a stirred drink.