Effusion of Argon, Helium, and some other Gases. 435 



This deviation being in the direction indicated by the 

 adiabatic theory, the next thing was to measure some other 

 gas with a diatomic molecule and specific-heat ratio 1*4 : for 

 in that case agreement might be expected between the ob- 

 served and calculated values. Experiments with CO gave the 

 following results : — 



m. s. in. s. 



Oxygen ....... 11 17-2 



CO 10 29-5 10 30-6 



Oxygen 11 16 



Mean time for oxygen . = 11 16 "6 



Mean time for CO . . = 10 30 



The calculated value for CO is 10 m 32 s '9. The observed 

 value is only *4 per cent, smaller. Probably this deviation is 

 due to experimental error (compare the two results with 

 argon). 



The next gas investigated was carbon dioxide ; for as this 

 possesses a lower specific -heat ratio than 1*4, its time of 

 effusion ought to be greater than that calculated. Experiment 

 proved the contrary : — 



m. s. m. s. 



Oxygen 11 16*3 



CO, 12 59 12 58-9 



Oxygen 11 15-3 



Mean value for oxygen . 11 15*8 



The value calculated for C0 2 is 



675-8 x ~= 13 m 12". 

 4 



The observed time is 1*7 per cent. less. 

 This was confirmed by a second measurement with a fresh 

 preparation of C0 2 : — 



m. s. 



Oxygen 11 18-7 



C0 2 13 2 



Oxygen 11 18 



Calculated for C0 2 from oxygen 13 15*4 



Here, again, the observed time per cent, is 1*7 per cent. less. 



This deviation * being in the opposite direction to that 



indicated by the adiabatic theory, it was considered advisable 



* Graham found a similar deviation in the case of C0 2 . He used 

 brine, however, instead of mercury as enclosing fluid, and ascribed the 

 deviation to the solution of the C0 2 in the brine (a single experiment 

 lasted about an hour). 



2H2 



