

Law of the Diffusion of Gases. 325 



It remains only to be shown that during the completion of 

 the experiment the same equality of moving force exists ;, and 

 that consequently the final volumes exchanged ought to be, 

 a Mr G-aham^uds them, proportional to the mitia 1 velo- 

 cities. For this purpose we shall refer to the case before 

 mentioned of hydrogen escaping from the diffusion mstru- 

 men Z 'the atmosphere. At any period of the experiment 

 Ster diffusion has commenced, let us call h the quan > y of 

 hydrogen remaining in the instrument, and a the quantity of 

 return air which has entered. The gaseous mixture being 

 maintained at atmospheric pressure, by keeping the , level of 

 the water outside the instrument the same as that of the in, 

 side, its volume will be a + k, and calling atmospl ^"C pres- 

 sure unity, the tension of the hydrogen in the bulb will be 



-A-, and that of the air in the bulb will be equal to ~^ f 

 The diffusive tension or elasticity of the hydrogen is conse- 

 quently proportional to ^ and the impelling force of the 

 atmosphere, which is the excess of its external pressureover 

 the internal tension of the air, is proportional to ^« 

 h Hence, since a and h are indeterminate, it follows 



5,"at A at every intermediate period between the commencement 

 and the conclusion of the phenomenon, the impelling force 

 of the air will be equal to the expelling force of the hydrogen ; 

 "he quantity of motion lost on each side will be the same; t he 

 resulting momenta of the two currents will be equal and by 

 easing similar to that already used, it follows that ^ final 

 volumes exchanged mill necessarily to inversely proportional to 

 the square roots of the densities. This progressive decrease of 

 elasticity or diffusive tension likewise explains why he ra- 

 pidity of diffusion is so great at the commencement of the ex- 



eriment, and gradually diminishes towards its conclusion. 

 P Mr Graham's observations, in short, extend to gases n 

 motion; Mr. Dalton's theory, to their relations when in a state 

 of equilibrium. One is the statical, the other the dynamical 

 exhibition of the same law. M 



It is not easy to assign a reason why the results of Mr. 



Graham's experiments °on the velocities of gases flowing nto 



„n, should differ so considerably Iron, those ind.cah,! 



v theory. Unfortunately, in the few experiment^ ...vest.ua- 



„ 1 i.l.erto made on "this subject, sufficient attention has 



„en paid to the difference between a gas under pressure 



