WORK OF J. SAM GUY. 177 



Table 120 gives the corresponding viscosity data at 55, 65, and 75. The same 

 general relations seem to hold at the higher as at the lower temperatures. The 

 temperature coefficients of fluidity at these higher temperatures are very similar to 

 those of conductivity at the same temperatures. 



From the data obtained, we are justified in concluding that curves representing 

 change in conductivity and change in fluidity with rise in temperature are very 

 similar to one another. In a word, conductivity seems to follow fluidity quite 

 closely over the range of temperature from 25 to 75. 



The fact that glycerol has such a very large temperature coefficient of viscosity 

 presents the possibility of throwing some light upon the relation between viscosity 

 and reaction velocity. It has long been felt that the viscosity of the medium in 

 which the reaction is taking place must be taken into consideration, and if the velocity 

 of some reaction could be followed, using glycerol as a solvent, it is highly probable 

 that interesting results would be obtained. Glycerol, being such an excellent sol- 

 vent, seems well adapted to such work. 



The viscosities and fluidities of solutions in the various mixtures of glycerol with 

 the alcohols and with water are given in tables 121 to 123, inclusive. Measure- 

 ments were made only with the tenth-normal solutions, since the viscosities of the 

 more dilute solutions differ very slightly from that of the solvent in each case. 

 Curves representing the change in fluidity with concentration of glycerol are given 

 in fig. 80. These curves are, in general, strikingly analogous to the curves repre- 

 senting the conductivities in the same mixtures, though it is seen that the increase in 

 fluidity is more rapid than the increase in conductivity. The viscosities of the solu- 

 tions are in nearly every case greater than that of the pure solvent. 



NEGATIVE VISCOSITY COEFFICIENTS. 



One of the most interesting points brought out in this investigation is the fact that 

 certain salts have been found to lower the viscosity of glycerol. The fact that certain 

 electrolytes have the power to lower the viscosity of water has been known for some 

 time. Jones and Veazey were the first to offer an apparently satisfactory explana- 

 tion, the large atomic volumes of the metals whose salts produced such a change 

 being the key to the phenomenon. The presence of elements with large atomic vol- 

 umes, as has been stated, would decrease the amount of skin friction in a given 

 volume of solution, and thus, in terms of the theory of Thorpe and Rodger, would 

 decrease the viscosity. Jones and Veazey pointed out that only salts of potassium, 

 rubidium, and caesium produce a decrease in the viscosity of water, and that these 

 salts do so in a direct ratio to their respective atomic volumes. Schmidt had noted 

 that the increase in viscosity of solutions in glycerol over that of the pure solvent was 

 in an inverse ratio to the atomic volumes of the metals whose salts he studied; but in 

 no case did he find a negative viscosity coefficient in pure glycerol. 



The results showing negative viscosities in glycerol are given in table 124. From 

 this table it can be seen that one-tenth gram-molecule of rubidium bromide lowers 

 the viscosity of glycerol about 2 per cent, while one-half gram-molecule lowers the 

 viscosity of the solvent over 8 per cent. 



