350 JOHN W. SHIVE 



determinations; in no case tested was the difference greater than 

 0.004 atmospheres. The data for the two most concentrated 

 series are given in tables II and III. In these tables the num- 

 bers of the solutions are those employed by Tottingham. The 

 partial, volume-molecular, concentrations and the total salt con- 

 centrations are given for each series. These are followed by 

 the actual depressions of the freezing point (after correction for 

 undercooling of 2.5°C.) and by the diffusion tensions at 25°C. 

 Finally, the last column of each table gives the plus and minus 

 percentage variations of the diffusion tensions here determined, 

 from the corresponding calculated values as given by Tottingham. 



It will be observed that the results obtained by the method of 

 calculation and those here set forth are in very close agreement 

 for the optimal series of solutions. The greatest deviation 

 above 2.50 atmospheres of diffusion tension (calculated for Tot- 

 tingham's series 5) is 6.4%, and the greatest below this calculated 

 value is 5.6%. The average for this whole series is 2.508 at- 

 mospheres of diffusion tension, which represents a deviation, 

 from the calculated value, of only 0.32%. For the more con- 

 centrated series on the other hand, the results obtained by the 

 two methods are not in nearly as complete accord. In every 

 instance the values obtained by the cryoscopic method fall 

 markedly below the calculated value (8.15 atmospheres of dif- 

 fusion tension), the smallest deviation from the calculated value 

 being 6.7%, and the greatest 20.0%. The average diffusion 

 tension for the whole series is 7.22 atmospheres, which represents 

 a minus deviation from the calculated value of 8.9%. It is of 

 interest to note that the whole range of variation between the 

 highest and lowest values of diffusion tension for the optimal 

 series is only 0.30 atmospheres, while the more concentrated 

 series shows a corresponding range of 1.08 atmospheres. 



As has already been remarked, a number of tests made with 

 solutions from Tottingham's least concentrated series showed 

 practically perfect agreement between his calculation and the 

 diffusion tension values derived by cryoscopy; variations were so 

 small as to make it appear unnecessary to test the entire series. 

 From what has just been said it is clear that the optimal series 



