404 MR C. RANKEN AND DR W. W. TAYLOR 



Discussion of the Results. 



1. The relation between concentration and relative viscosity. — The previous 

 determinations of the relative viscosity of potassium chloride solutions were insufficient 

 to show conclusively the nature of the curves below concentrations of 1 mol per 

 litre. It was obvious, however, that below that concentration the curves which were 

 practically straight lines must change their character, since the line did not pass 

 through the origin at zero concentration. Whether the viscosity of the solution at 

 some finite concentration became the same as that of water below which concentration 

 it was less than that of water (so-called " negative viscosity "), finally at infinite dilution 

 to again become the same as that of the solvent, could not be settled. Our present 

 results indicate that the course of the curve is somewhat complicated. At 25° C. at 

 concentrations between 1 mol and 2 mols per litre, the value of *]/% is greater than 1, 

 at 1 mol per litre it is less than 1, but steadily increases with increasing dilution until 

 at O'l mol per litre the ratio is again = 1 ; after which it is probably a little less than 1, 

 the observed values being "9990 and '9995. 



2. The results obtained for potassium chloride were recalculated and curves 

 drawn to show the relation of the ratio n\% to the concentration of the solution 

 expressed in mols per kilogram of solution, to the number of mols of solvent to one 

 mol of solute, and to the number of mols of solution to one mol of solute, but none of 

 these methods of representation introduced any simplification of the curve. 



3. In all the solutions examined, with the one exception of mercuric cyanide, increase 

 of temperature increases the ratio njn^. For electrolytes the rate of increase is very 

 much greater than for non-electrolytes ; mercuric chloride, which is very little ionised, 

 ranges itself with the non-electrolytes in this respect also. 



At high temperatures solutions of mercuric cyanide might be expected to exhibit 

 " negative viscosity," but at 45° C. the ratio nj% is still distinctly greater than unity, 

 the temperature-coefficient being very nearly the same as for water. 



4. Our measurements give further support to the view that the relative viscosity of 

 all solutions passes through at least one minimum, the position of the minimum being 

 determined by the nature of the substance and the temperature. The minimum value 

 may be smaller than the corresponding value for the solvent. Whether this occurs or 

 not depends on some factor yet unrecognised. 



5. Carbamide. — Rudorf* found that, at 25° C, the relative viscosity of aqueous 

 solutions of carbamide decreased with dilution, and at concentrations below *234 mol 

 per litre became less than that of water ; the values are — 







* Zeit, 



m tj/ij 



937 1010 



409 1-002 



234 -996 



117 -993 



058 -995 

 /. Phys. Ohem., 43, p. 257, 1903. 



