WORK OF P. B. DAVIS. 181 



They found that molecular conductivities in glycerol are all extremely small, but 

 show a regular increase with increased dilution and rise in temperature. Further- 

 more, a study of the temperature coefficients brought out the fact that in the case of 

 those salts that have been shown to have large hydrating power in water, such as salts 

 of barium, strontium, calcium, and cobalt, the relative increase was larger than with 

 salts of sodium potassium and ammonium, which show little or no hydrating power 

 in aqueous solutions. Here we have evidence of solvation in glycerol supporting that 

 obtained by Jones and Strong 1 from spectroscopic methods. 



In mixed solvents, Guy and Jones studied salts of potassium, sodium, ammonium, 

 and strontium in various mixtures of glycerol with water and with ethyl and methyl 

 alcohols. They found that conductivities in such mixtures do not follow the law of 

 averages, but are always lower. This they explained by the facts established by 

 Jones and Murray 2 and Jones and Lindsay, 3 that two highly associated solvents when 

 mixed tend to break down the association each of the other, and hence their combined 

 power of dissociating electrolytes is less than if there were no mutual lowering of 

 their association, i. e., if each solvent acted independently of the other. 



From the viscosity data Guy and Jones have shown that the temperature coeffi- 

 cients of fluidity in pure glycerol are very large and nearly equal to those of con- 

 ductivity. Also that the ternary electrolytes studied increased the viscosity of 

 glycerol to a much greater extent than the binary electrolytes. This is attributed 

 to the smaller atomic volumes of barium, strontium, calcium, and cobalt, and to the 

 solvation of the molecules of the solute. In mixed solvents the curves representing 

 conductivity and fluidity were found to be strikingly analogous. 



Probably the most interesting point brought out by Guy and Jones was the large 

 viscosity lowering observed in the case of certain salts in pure glycerol, e. g., N/10 

 solutions of sodium nitrate, ammonium bromide, ammonium iodide, and rubidium 

 bromide. The explanation of this phenomenon is derived from that of Jones and 

 Veazey 4 for similar salts in water. This fact suggested the closer study of some of 

 these salts over a wider range of concentration, the present investigation being a con- 

 tinuation of the work of Jones and Schmidt and of Guy and Jones. 



Work on this problem was begun in collaboration with Win. A. A. Reinhardt, 

 of Baltimore, a graduate of Johns Hopkins University, whose untimely death proved 

 a serious set-back to the carrying out of the investigation. We wish to pay tribute 

 here to our friend and coworker as an earnest, sincere student, one who was rapidly 

 coming to the front among the research students in this laboratory. 



The conductivity data on ammonium iodide in mixed solvents obtained by Mr. 

 Reinhardt are incorporated in this memoir. 



EXPERIMENTAL. 

 APPARATUS. 



The constant-temperature baths used in this investigation were of the form usually 

 employed for such work in my laboratory. The thermostat for viscosity measure- 

 ments was provided with glass windows, both in the front and rear, to facilitate the 

 reading of the viscosimeters. Both baths were equipped with cooling coils in which 

 the water was maintained at constant pressure. This facilitated temperature regu- 



'Monographs Noa. 130 and 160, Carnegie Institution of Washington. 3 Ibid., 28, 329 (1902). 



2 Amer. Chem. Journ., 30, 193 (1903). 4 Loc. cil. 



