Ask Marilyn ® by Marilyn vos Savant is a column in Parade Magazine, published by PARADE, 711 Third Avenue, New York, NY 10017, USA. According to Parade, Marilyn vos Savant is listed in the "Guinness Book of World Records Hall of Fame" for "Highest IQ."
I poked around your Web site looking for one error in physics that I am sure Marilyn made, but so far as I know, was not followed up in print.
This was several years ago at least. Someone wrote in with a question about whether hot or cold water freezes faster. Marilyn replied that it is logical that the cold water freezes faster since it has does not have to cool down as much to hit freezing. There was a considerable debate about this in Scientific American Amateur Scientist an even longer time ago, because the correct answer, that hot water (for equal masses of water) will freeze faster than cold water is certainly unexpected. The debate was not about whether or not this was true, but why. The Boltzman fourth power law for cooling was invoked, but cooling due to evaporation and mixing from a greater thermal differential were also included. Supposedly, folks who live in very cold climates wash their cars with cold water rather than hot (or maybe rinse off the road salt) during winters.
Perhaps hot water provided a partial melt of the frost below the tray (in an older model refrigerator) which ultimately provided a better heat sink to the water in the tray, allowing it to freeze faster.
Ed Foster <erfoster@ziplink.net> wrote to suggest an alternative possibility:
Far more cooling is needed to freeze 32 degree water than cooling the water from even a high temperature down to 32 degrees (heat of fusion and all that.) Perhaps enough of the hot water evaporates off while cooling to 32 degrees that the reduced amount remaining freezes before the water that started out cold. The calculations to show how much of the original hot water needs to evaporate should be easy to do and if I get inspired enough to do it I'll send the results. Doing an experiment and weighing the resultant ice might also prove informative.
Eliot W. Collins <ewcollins@att.com> wrote to suggest yet another possibility:
I believe that boiling water before attempting to freeze it, removes the air from the water and this allows the water to freeze more quickly.
From the alt.folklore.urban FAQ, the answer is that under some conditions (e.g. in wooden buckets, where most of the cooling is caused by evaporation), hot water freezes faster. However, since the cooling is caused by evaporation of a significant amount of hot water, the hot water produces less ice.
Two references are cited:
One of the 350 questions discussed in their archive is Is it true that hot water placed in a freezer freezes faster than cold water? And if so why does this happen?
Bo Briggs <bobriggs@usa.net> recommends How Things Work: Water, Steam, and Ice. (This is a comprehensive web page that covers numerous questions.)