WORK OF P. B. DAVIS. 



Prior to Jones and Veazey's investigation, Euler 1 had attempted to explain the 

 lowering of the viscosity of a solvent by a dissolved substance, on the basis of the 

 "electrostriction" of the solvent caused by the charge upon the ions of the solute as 

 proposed by Drude and Nernst. 2 But Wagner and Mublenbein 3 showed that Euler 's 

 explanation was not valid, since certain non-electrolytes in organic solvents also show 

 negative viscosity, e. g., cyanobenzol in ethyl alcohol. 



The explanation offered by Jones and Veazey is based on their own observations 

 and on the classical work of Thorpe and Rodger. 4 The latter have indicated that 

 viscosity phenomena are in all probability dependent upon the frictional surfaces 

 of the ultimate particles present in any liquid or solution. A review of the data 

 obtained by Wagner 5 from his study of the viscosity of a number of inorganic salts in 

 water will show that he found negative viscosity only in the case of caesium, rubidium, 

 and potassium salts, and in some instances thallous salts. The first three metals 

 occupy the maxima on the atomic volume curve of Lothar Meyer, 6 i. e., they have 

 the largest atomic volumes of all the elements. That some salts of potassium give 

 positive viscosity is to be expected, since it has been shown that viscosity is an addi- 

 tive function of both the cation and the anion of the dissolved salt. In the case in 

 question the one might tend to lower the viscosity of the solvent, the other to 

 increase it, the final results depending upon whether the sum of these two opposing 

 forces was positive or negative. Potassium has the smallest atomic volume of the 

 three, and in many instances the positive viscosity effect of the anion would entirely 

 overcome the negative effect of the cation. 



In view of these facts, Jones and Veasey offered the apparently satisfactory explan- 

 ation that "negative viscosity" is due to a lessening of the skin friction between the 

 solvent and the molecules or ions of a solute, because of the large atomic volume of 

 the cations as compared with the molecular aggregates of the solvent. This explana- 

 tion follows directly from the work of Thorpe and Rodger. 



W. Taylor 7 measured the viscosity of one, two, and three molar solutions of potas- 

 sium chloride, bromide, and iodide in water at different temperatures, and noted that 

 negative viscosity may pass over into positive viscosity with rise in temperature, and 

 confirmed the view that viscosity depends upon the nature of both cation and anion. 

 He noted also the negative viscosity effect of ammonium iodide when dissolved in water. 



Reference should be made to the extensive work of Walden, 8 on the relations 

 that exist between viscosity and conductivity at infinite dilutions. He finds that 

 Moo??*, = C for more than thirty organic solvents. Exceptions have been noted to this 

 relationship and these will be taken up later. 



J Zeit. phys. Chem., 25, 536 (1898). 6 Zeit.. phya. Chem.. 15, 31 (1890). 



mid., 15, 79 (1894). 'Ann. Chem. (Liebig), Suppl., 7, 354 (1870). 



"Ibid., 46, 867 (1903). TEdin. Proc. 25, 228 (1904) and Edin. Trans.. 45 ,397 (1906). 



Phil. Trans., 185-4, 307 (1894). 8 Zeit. phya. Chem., 78, 257 (1911). 



179 



