The behavior of a pure substance on slow cooling is known to most science students. A graph of the temperature of such a system as a function of time is called a cooling curve and exhibits the characteristic plateau which occurs at the freezing point of the substance. The temperature at which the solid first forms from the solution is lower than the freezing point of the pure liquid solvent. Furthermore this freezing point does not remain constant; it slowly falls as more solid forms since in most cases solvent is the first substance to freeze, leaving the remaining solution proportionately richer in solute. Also visible in both graphs on the facing page is the phenomenon known variously as "undercooling" or "supercooling". Sometimes as a liquid or solution is cooled towards its freezing point the temperature will actually drop below the freezing point before any phase change occurs. This behavior is enhanced by very clean or new equipment which lacks scratches or other irregularities that serve as sites for crystallization to begin. Vigorous stirring will eventually overcome any tendency to supercool in a mixture. In an ideal solution (no interactions among particles) there is a nearly linear relationship between the concentration of the solution (expressed in molality) and the drop in the freezing point. Well, freezing point depression is more than a nice way to keep your radiator water from freezing up on a mountain ski trip! By measuring the freezing point depression of an unknown solute in water and knowing Kf, the molality of the unknown solution can be determined. From there it is possible to calculate the molar mass of the unknown solute. In addition to molar mass determination (which actually is not done that much any more in laboratories), some of the first historical evidence for the existence of ions in solution was obtained from freezing point depression data. You will have a chance to compare the freezing points of some solutions of ionic compounds to see for yourself. The thermometer probes were calibrated for reasonable accuracy over a wide range of temperatures but this sometimes means that at the extremes of the range they can be off a little. For this reason the first thing you need to find out is what temperature your probe says water freezes at. All of the other temperatures are then relative to this value. Using about 2.5 mL of distilled water in one of the plastic test tubes will allow you to freeze the sample in 5 minutes or less. As with all of the samples, constant stirring is essential. Be sure to stir with a circular motion and not by banging the probe on the bottom of the test tube. Supercooling may occur, even with pure water. The non-wetting plastic seems to exacerbate this phenomenon. Its opacity also makes it difficult to determine when the solution has frozen. You should notice a marked increase in viscosity when freezing occurs. At that point you can lift the probe out of the mixture and see if it has slush on it. Hard freezing is not required.