A number of water-droplets, uniform in size at 1.0, 0.6, 0.3 or 0.2mm diameter, were prepared from water holding powder suspended as well as water having no addition, by electrical atomization, on a paraffin-covered glass-plate. This glass-plate was placed horizontally in a flat case tightly made of two brass-plates together with a rubber-frame between, and was cooled at various temperatures down to -30°C in a cryostat. Following this, the plate was quickly taken out from the case, and the droplets which had frozen were counted without delay in a low-temperature chamber, using a phase-contrast microscope and, on occasions, by the naked eye.
The powder-materials suspended were iodides of silver, mercury and lead, and further, oxides of copper, cadmium and zinc. As for lead iodide, it was the most soluble of the materials examined, so that an investigation into the influence of aging of the powder in suspended state upon the supercooling phenomenon was carried out in the following manner; the suspension of lead iodide was, after its production, allowed to stand in a closed vessel with frequent stirring for the periods of 0, 2 and 8 days at 5°C or thereabouts, and these three samplings dispersed each to droplets by the same method as with the other materials.
The results are roughly as follows:
1) The smaller the droplets, the more likely they are to exhibit supercooling, the finding being the same in both the absence and presence of solid particles in water, and there appears to exist a linear relation between the logarithm of droplet-diameter and the freezing-temperature.
2) Of the materials tested, silver iodide is most effective in solidifying water under supercooled condition, and its effectiveness goes up, though slightly, as the amount added to water is increased; lead iodide deteriolates considerably on account of its solubility, it being left to itself in suspended state.
3) The following equation
ln{x/(100-x)} =k•(θ-θ')
is presented as giving the fraction (x%) of droplets frozen at a supercooling (θ°C, expressed in terms of the lowering from 0°C), k being a constant depending on the conditions of experiment such as the droplet-size, the nature and amount of the powder suspended, etc. and θ'the supercooling at which x=50.
View full abstract