The Electrical Conductivity and Fluidity of Strong Solutions

Abstract

In adopting Callendar's association theory of strong solutions (Proc. Roy. Soc., A., Vol. LXXX., p. 466) some difficulty is experienced in getting the strongest solutions of electrolytes to conform to the laws laid down. This is attributed to the inaccuracy of the ionisation data, which are derived from observations on electrical conductivity. It may be supposed that the viscosity of the solution will affect its conductivity, and the author carried out a series of experiments to determine if there were any definite relation between conductivity and fluidity in the case of calcium chloride solutions. The feature of these determinations is the simultaneous observation of viscosity, electrolytic resistance and temperature. Solutions were contained in an unsilvered Dewar cylinder, which permitted a slight adjustment of temperature if necessary. A platinum thermometer records the temperature, and is caused to oscillate in the solution by a stirring mechanism. In so doing it draws a platinum scoop through the liquid and thus acts as an efficient stirrer. The conductivity cell and the viscometer were bound to the thermometer by rubber bands. While the thermometer oscillates the readings of electrical resistance were measured by means of the rotating commutator and bridge. The viscometer was in the form of a capillary pipette, which was immersed in the solution a known depth. The levels of the inflowing liquid were indicated electrically, and hence the rates of inflow could be accurately estimated and compared with that of water. Viscosities eorrect to less than 1 per cent. were obtained. I sothermal Observations. - Perfectly smooth curves for conductivity and fluidity were obtained, even when the supercooled melted crystals were included. No definite connection between conductivity, fluidity and concentration can be derived if the latter is expressed in terms of volume, but if concentration is expressed as a ratio of masses - molecules of solute to 100 molecules of solvent - the ratio conductivity C /fluidity F stands in linear relation to the concentration n when the latter exceeds one-fourth its maximum value-In spite of the enormous variations of the quantities in this ratio C/nF has values differing at most only 2 per cent. from the mean. Examination of similar results by Bousfield and Lowry for sodium hydrate give striking agreement with this conclusion. However, owing to the dependence of C and F on linear dimensions, this relation is difficult to interpret. Variations with Temperature. - One solution of nearly cryohydric strength was examined at temperatures from 40?C. to - 50?C. The same tube and apparatus were employed, and for lower temperatures liquid air was used as a cooling agent. The failure of the fluidity-temperature and conductivity-temperature curves to exhibit the same variations was clearly shown. Conductivities of various solutions were examined from 40?C. to their freezing points and the curves C/n and temperature plotted. The increasing curvature with concentration is shown and the error involved in applying the ratio, molecular conductivity to that at infinite dilution, obtained at one temperature, to indicate ionisation at another temperature, is quite apparent. Moreover, within the limits of accuracy employed no indication is given of the existence of different hydrates. The results obtained above suggest that no reliance can be placed on ionisation data derived from electrical conductivity observations.