294 PHENbMENA DEPENDENT ON MOLECULAR PATHS 108 



From these numbers, which exhibit a very delightful 

 agreement between theory and experiment, we may also 

 answer the question whether the conductivity f really in- 

 creases as strongly with the temperature as the product of 

 the coefficient of viscosity 77 and the specific heat c. For this 

 we need only find the ratio of the values for 100 and 0, 

 both of the calculated values and of the observed. In this 

 manner Wiillner obtained the annexed table, which shows 

 as good an agreement as can be desired. 



Values of f 100 /f ' 



Wiillner has given in his text-book 1 a still more com- 

 plete table of results. I first extract from this the values of 

 the ratio in which the numbers obtained for 100 by different 

 observers stand to those for 0. The second column, marked 

 ' calculated/ gives, as in the last table, the ratio of incre- 

 ment of the product rjc for the change of temperature from 

 to 100 ; the last column gives the ratio of increment of 

 the observed values of the conductivity for the same interval 



of temperature. 



Values of 



According to this table the differences between the values 

 theoretically determined and experimentally measured are 

 not greater that the errors necessitated by the uncertainty of 

 the observations. 



1 Wiillner, Lehrb. d. Experimentalphysik, ii. ' Warme ' 5. Aufl. 1896, 

 p. 574. 



