180 WORK OF P. B. DAVIS. 



GLYCEROL AS A SOLVENT. 



An examination of the literature bearing on this problem shows that only a little 

 work had been done previous to that of Jones and Schmidt. 



Cattaneo 1 measured the conductivity of several halides of the metals in glycerol, 

 and noted that the values obtained were much smaller than the corresponding values 

 in water or alcohol. 



Schottner 2 measured the viscosity of certain mixtures of glycerol and water. Ar- 

 rhenius 3 studied the viscosity of mixtures of various organic substances with water, 

 among them glycerol, and noted that the temperature coefficients of viscosity were 

 greatest where the viscosity was greatest. Schall and Van Rijn 4 determined the 

 relative viscosities of mixtures of glycerol with water and alcohol. Lemke 5 carried 

 out an investigation on the conductivity and viscosity of water-glycerol mixtures at 

 25, and was led to conclude that ionization, and hence electrical conductivity, is pro- 

 portional to the viscosity of the solvent as well as to the association. He noted the 

 periodic viscosity of sodium chloride in 9.8 per cent glycerol and water and a nega- 

 tive viscosity in water at certain dilutions. Getman 6 studied viscosity of potassium 

 iodide in various organic solvents, including glycerol, and noted negative viscosity 

 only in the case of the latter. This he attributed to the association of the solvent. 



Jones and Schmidt 7 have studied the conductivity of lithium bromide, cobalt 

 chloride, and potassium iodide in glycerol at 25, 35, and 45, and in mixtures of 

 glycerol with water, with ethyl alcohol, and with methyl alcohol at 25 and 35. 

 Measurements of viscosity were also made with the N/10 solution in the various sol- 

 vents. They have shown that glycerol is an excellent solvent and in all probability 

 a comparatively good dissociant, since it has a dielectric constant of 16.5 at 18 and 

 an association factor of 1.8 at that temperature. From these data glycerol, accord- 

 ing to the Thomson 8 -Nernst 9 and Dutoit and Aston 10 hypotheses, should have a 

 dissociating power close to that of ethyl alcohol. The extremely low conductivity 

 values obtained were attributed to the high viscosity of the solvent. 



Schmidt noted that all the salts studied increased the viscosity of glycerol in N/10 

 solutions, but that potassium iodide lowered the viscosity of water, and of 25 and 50 

 per cent glycerol with water at 25 and 35. He also showed that the effect of the 

 several salts studied on the viscosity of glycerol was in inverse ratio to the atomic 

 volumes of the cations, exactly analogous to the observations of Jones and Veazey in 

 aqueous solutions. Schmidt also found glycerol to be an apparent exception to the 

 observations of Walden 11 previously mentioned, that fx x r] x =C. Further exceptions in 

 the case of water and glycerol have been noted by Walden from his own investigations. 



The most extensive investigation of glycerol as a solvent has recently been carried 

 out by Guy and Jones. 12 They took up the behavior of some twenty electrolytes in 

 pure glycerol, and in binary mixtures of glycerol with ethyl alcohol, with methyl 

 alcohol, and with water at intervals of 10 from 25 to 75 for the pure solvent and 

 from 25 to 45 in mixed solvents. Conductivity measurements were made over a 

 range of F=10 to 7=1,600, and viscosity measurements with the N/10 solutions. 



'Rend. R. Accad. Lincei. [5], 2, II, 112 (1893). 'Amer. Chem. Journ., 42, 37 (1909). 



! Wien. Ber., 77, II, 682 (1878). Phil. Mag., 36, 320 (1893). 



Zeit. phys. Chem., I, 285 (1887). Zeit. phvs. Chem., 13, 531 (1894). 



mid., 23, 329 (1897). "Compt. Rend., 125, 240 (1897). 



'Ibid., 52, 479 (1905). "Zeit. phya. Chem., 55, 246 IT. (1906). and 78, 257 (1911). 



Journ. Amer. Chem. Soc, 30, 1077 (1908). I! Amer. Chem. Journ., 46, 131 (1911). 



