184 WORK OF P. B. DAVIS. 



70 to 80, drained as quickly as possible into the receiving flask, then washed out at 

 least three times with the warmed solvent. The flask containing the more dilute 

 solution was then filled up nearly to the neck with the solvent, warmed and shaken 

 thoroughly without wetting the neck, then filled to a point slightly above the mark, 

 and finally diluted after cooling to 20 in a thermostat. Final warming in the air- 

 bath and thorough shaking completed the operation. The flasks used were of Jena 

 normal glass, eliminating as far as possible the danger of the action of the solvent on 

 the glass. Some time was allowed before the final dilution to the mark to permit the 

 glass to contract properly after the unavoidable heating during the process of dilution. 



On account of the extremely high resistance offered by glycerol solutions con- 

 ductivity measurements were not made at dilutions greater than N/1600. Even at 

 that dilution considerable difficulty was experienced in obtaining concordant read- 

 ings. As an example it might be well to note that N/1600 solutions in glycerol in a 

 cell with a constant as low as 2.3 required a balancing resistance of from 8,000 to 

 10,000 ohms, while pure glycerol in the same cell required 20,000 to 22,000 ohms. 



Conductivity and viscosity measurements in pure glycerol were made at intervals 

 of ten degrees from 20 to 75, and in glycerol-water mixtures at 25, 35, and 45. 

 Viscosity measurements were not made at a greater dilution than N/10, since at 

 lower concentration 17 approaches r} x too closely to be accurately differentiated. 



SOLVENTS. 



Glycerol. The glycerol used was from a new lot of Kahlbaum's "Doppelt-Dist. 

 1.26" and had a mean specific conductivity of about 0.6X10 -7 at 25, and a mean 

 specific gravity of 1.257 at the same temperature. No attempt was made to redistil 

 it, since Schmidt had already shown that redistillation did not appreciably lower the 

 conductivity. 



Water. The water was purified by the method of Jones and Mackay 1 with the 

 improvement as mentioned by Schmidt and had a mean specific conductivity of 

 1.5XlO _7 at 25. 



SALTS. 



The rubidium and ammonium salts used in this work were from Kahlbaum's best 

 products. These were recrystallized two or three times from conductivity water and 

 carefully dried each time at 130 to 135 before weighing. In addition the rubidium 

 salts were examined spectroscopically and showed the presence of only traces of 

 sodium and no potassium. The ammonium iodide was pure white after drying, and 

 solutions of it in glycerol were only slightly tinted after standing for some time. 



PROCEDURE. 



Conductivity data were calculated in the usual way. 

 Viscosity measurements were calculated from the formula 



1 .$L 



Vo &o to 

 in which 77 is the viscosity coefficient for the liquid in question, 770 the absolute vis- 

 cosity of water, S the specific gravity of the liquid at the given temperature, t the 

 time of flow of the same, *S and t the density and time of flow of water at the same 

 temperature. 



Amer. Chem. Journ., 17, 83 (1895). 



