432 Dr. F. G. Don nan on the Relative Rates 



of 



had a form which was somewhat different to that described 

 previously and used in the subsequent experiments. In other 

 respects the apparatus was the same. All the experiments 

 were carried out at the temperature of 25° C. It is of 

 course necessary to keep the temperature very constant during 

 an experiment, as the apparatus acts like a gas-thermometer. 

 The following times of effusion were observed : — 



m. s. m. s. 



Hydrogen ... 3 47\% 3 47*8 



Oxygen .... 12 21*3 



Carbon monoxide . 11 l(r4, 11 18 



Nitrogen .... 11 18, 11 17 -8. 



It is evident from the results for hydrogen and oxygen 

 that there is a large viscosity- effect. Thus for pure effusion 

 the time for hydrogen should be very nearly one-quarter of 

 that for oxygen, i. e. 3 m 5 S, 3. It was found, however, that the 

 effect of viscosity could be allowed for by employing, instead 

 of the equation t = tc*>/d, where d = relative density and /c = 

 a constant depending on the apparatus, the equation 



t = K x V ' tl + K 2 fJi. 



where //,= relative viscosity, and k u «:2 ==a PP ara tus-constants. 

 Thus if we employ the table of relative viscosities given by 

 Graham *, viz. : — 



Oxygen =1 



Hydrogen . . . . = '4375 



Nitrogen or CO . . = '875.0 



C0 2 = -7272 



we can use the results obtained with oxygen and hydrogen 

 to calculate k x and k 2 . We obtain the equations 



^YT008 + -4375^ = 227-9, 

 4/ti -f- k 2 =74 1 *3, 

 from which 



*!= 129-2, * 2 =224-4. 



Hence the time of effusion for carbon monoxide should be 

 given by 



i= V 14. 129-2 + -875x 2242, 

 whence 



^=ll m 19 s -6. 



The difference between this and the observed time is about 

 3 per cent. In the case of nitrogen the agreement is a 



* " On the Motion of Gases," Part II., Phil. Trans. 1849, pp. 349-392. 



