Supercooling and freezing of water

This is a strictly preliminary report, rendered necessary by a temPorary suspension of the work. No doubt attaches to the observations reported, but the setting of limits within which they are valid, and the determining of the most satisfactory way in which to explain them, should await further work. Contrary to current opinions, it has been found that the extreme supercooling of water (at least down to -21 0 C) is not dependent upon either extreme quiescence or upon the use of a minute volume, and that motes in the water exert a strong nonspecific effect. For each sp ecimen there is a definite temperature (the spontaneous-freezing-point), characteristic of the specimen , at which icc suddenly appears, at which the water freezes spontaneously . The well-known fact that a preheating of the water may facilitate its s. upercooling has been found to depend upon a lasting change so prodJ.lced in the water-a change that persists at room temperatures for at least 6 months, and that seems to be p ermanent. By sedimentation, the spontaneous-freezing-points of the cxtreme laycrs of a colnmn of water become different, that of the bottom ones becoming higher than that of the top ones. All the observations are consistent with the idea that the spontaneous-freezing-point of a specimen is determined, within undetermined limits, by the size of the largest mote that it contains; the smaller the mote, the lower the spontaneous-freezing-point. The effect of agitation has been considered, and three di stinct types of freezing have been recognized.