446 



Prof. XV. 0. Rontoen on Electrical 



were excluded, because with some of the gases the electrometer 

 was not equal to measuring the corresponding differences 

 of potential. I must mention that these experiments are not 

 directly comparable with the preceding ones. The following 

 Table contains the mean values of various satisfactorily ac- 

 cordant determinations. 



Gases. 



M. P. at 205 



millims. 



M.P. at 110 

 millims. 



Hydrogen 



1296 

 2402 

 2634 



2777 

 3188 



3287 



1174 



1975 

 2100 

 2317 

 2543 

 2655 



Oxygen 



Carbonic oxide 



Marsh-gas 



Nitrous oxide 



Carbonic acid 



In this Table the gases are arranged in the order of ascending 

 values of the M. P. If this series be compared with that which 

 is obtained when the gases are arranged in the order of dimin- 

 ishing values of their mean molecular path-lengths, at both 

 205 and 110 millims. pressure perfect agreement will be found. 

 Since the minimum difference of potential is a direct measure 

 of the insulating-power of a gas, the result contained in the 

 above Table can be expressed in the following manner : — The 

 shorter the path of its molecules, the greater is the insulating- 

 power of a gas. Now it is known that the smaller the gas- 

 molecules the greater is the length of their paths ; consequently 

 we can also say : — The larger the molecules of a gas the greater 

 is its power to insulate. 



The connexion between the M. P. and the length of path 

 becomes still more convincingly evident when for each gas we 

 form the product of the path-length and M. P. 



Gases. 



Product of path-length and M. P. 



Pressure 205 

 millims. 



Pressure 110 

 millims. 



Hydrogen 



240 

 254 

 259 

 236 

 217 

 224 



218 

 209 

 207 

 197 

 173 

 181 



Oxygen 



Carbonic oxide 



Marsh-gas 



Nitrous oxide 



Carbonic acid 



The path-lengths are taken from Graham's transpiration 



