So water will transition from liquid to solid and back through a phase change which consumes or produces energy. So any melting will cool the drink assuming it's warmer than 0 degrees C.

So let's take as our starting point a mass of ice uniformly 0 degrees C. Regardless of geometry, any cooling of whisky will inevitably produce a phase change to liquid water in the exact amount required by the energy absorbed by the ice. So you get 0 degree ice, 0 degree water, and cooler whiskey. Assuming a 0 degree ice ball geometry is irrelevant.

Of course water can be cooler than 0! So let's assume some really cool ice and do some thought experiment geometries. Take super thin sheets of ice for example. The total amount of energy available to cool the drink without melting is the specific heat of the ice times the temperature difference from zero (there might be some non-linearity here but I doubt it's relevant) times the volume of the ice. So if you have enough volume and cold enough ice you should be able to achieve a 0 degree drink at which point you have no more cooling and therefore no melting and no water.

Unless the heat is transferred out of the water unevenly, say a large mass of it with a central core inaccessible from the outside liquid. Then it would matter how fast heat travels from the interface to the core relative to how fast heat travels from the drink to the interface. This probably has something to do with the outside liquid being turbulent and the inside solid being a lattice but I'm not sure.

So, pure speculation here, I'd guess a large ball would have smaller surface to volume and so would cool more slowly for equivalent volume, but you get a large volume so you might be okay assuming the cold core of the large ice ball does the job of cooling the surface faster than the drink can heat it up.

It seems to be it would be more effective to get very small ice chunks, a ton of them, make them really cold, then you have lots of surface area and if you do the math you could approximately guarantee no melting and you'd cool fast. Then strain out the chunks and put whatever you want in there since it's already 0 (maybe something cool looking you can condescend about?)

But whatever you do don't put it in a glass you grab with your whole hand, the air is a really good insulator, glass much less so and palms work like a circulating hot-fluid heat exchanger.

Starting off with the opposite, I'll compare two situations. One with whiskey and snow, the other with whiskey and an ice ball. (mass of snow = mass of ice ball)

Immediately when you drop the snow in, the temperature of the solid water in the snow raises from some negative temperature to 0 C, the temperature of the whiskey lowers to 0 C, and the surface of the snow particles melts. The slurry stays in equilibrium at 0 C until external heat inputs melt all of the solid water.

Also because it is at 0 C, heat transfers from the environment quickly (heat transfer rate depends on the difference in temperature, the bigger the difference the more energy is transferred in the same amount of time)

Downsides:

* Cold whiskey

* Immediately diluted whiskey

* Faster heat transfer.

Now for a solid ball of ice. As opposed to the snow, there isn't an instant change to a constant temperature. The ice (which in my freezer is at -19 C) initially only warms towards 0, there isn't a significant amount of melt until the whole ball is at 0 C. The whiskey likewise don't instantly, or perhaps ever cool to 0 C. This warmer whiskey accepts energy from the environment less quickly, which warms the ice ball less quickly, which results delayed and slower ice melt.

This all happens because the surface area in contact with the fluid is much much smaller than with the snow.

So you get

* warmer whiskey (desirable for more flavor)

* less water immediately in the whiskey

* slower introduction of water into the whiskey

This is all because the large mass, low surface ice ball slows down the thermodynamic process which has all desirable effects. * The key is not cooling the whiskey as quickly & not keeping the whiskey at 0 C. If it _did_ keep the whiskey at 0 C (for example with lots of stirring) there wouldn't be a difference between snow and ice ball.

Normal ice cubes are a middle ground with undesirable effects to a lesser degree.

Room temperature on a cool day is what many will say is the 'right' temperature, but there is no right or wrong, despite what many questionable experts will tell you.

Nobody is going to agree that X is the best temperature for every whiskey. Whiskies and drinkers are all different. Room temperature whiskey and ice give the most control with the least complication.

The ice press sells for $700 to $1,100, depending on the size. http://www.williams-sonoma.com/products/cirrus-ice-ball-pres...

> My experience using these has resulted in more than a few cracked ice balls, whose deep fissures essentially increase the surface area of the ice ball, negating any benefits in that department.

I had pretty much the same experience with my molds (I was using the Tovolo brand, rather than the one you linked to, though).

Or you could go for a pricier option such as http://www.wintersmiths.com/ (which uses directional freezing).

Williams Sonoma is likely adding a huge markup.

Although I wonder if my posting about plastic ice cubes really misses the point of the article.

Yes.

The ice melts at a rate governed by how fast it absorbs the heat.

No.