I haven't directly tried this, but I think it is easy. First, you want a way to measure the salt concentration. For a first try, not very quantitative, taste isn't bad. To get a little more quantitative, you could use a cheap volt-ohm meter to measure the electrical resistivity, which goes down as more salt is dissolved. Youíd want to use two electrical contacts with fixed geometry (maybe glued to the side of a cup) and just measure the electrical resistance across the water. This also isnít extremely accurate, unless youíre very careful with cleaning the electrodes and perhaps switching their connections to the ohm meter, taking the average of the two readings, but itís a good start. You can also evaporate the water (boil it if youíre in a hurry) and weigh the salty residue, but you may need a very sensitive scale or balance for the salt you might find in an ice cube after the de-salting process. There are also commercially-available water salinity meters. A quick web search indicates that you may be able to find these where fancy fish and aquariums are sold. Some fish need just the right amount of salt in their water and so a salinity meter is just right to have around. They mostly rely on the electrical conductivity principle described above, but a few take advantage of the different refractive index saltwater. Others can measure the density very precisely and compare that against unsalty water.
For the de-salting, try putting the water in an ice-cube tray and let most of it freeze. If itís very salty, and your freezer wonít freeze any of it, you may need to use a styrofoam cooler with some dry ice in it instead of the freezer. Try to arrange it so that the freezing is slow, to avoid getting saltwater trapped in little pockets in the ice. Then pour off the liquid. Melt the ice in a clean cup. You should find that the salt concentration is very much lower. You can repeat the procedure with the new liquid to get even less salty water, in case too much salt got trapped in those pockets the first time.
mike w (and Tom J.)
(published on 10/22/2007)