Laiv of the Diffusion of Gases. 323 



doubled. Mr. Graham concludes with remarking that the 

 law which he has discovered is not provided for or explained 

 by any of the existing theories of corpuscular philosophy. 



The object of the following remarks is to show that Mr. 

 Graham's facts, so far from being inconsistent with every pro- 

 posed theory of the mechanical relations of mixed gases, af- 

 ford an elegant and striking confirmation of the truth of Mr. 

 Dalton's hypothesis on this subject, which is, that the parti- 

 cles of one gas are not elastic or repulsive in regard to the 

 particles of another gas, but only to those of their own kind. 

 The most obvious and remarkable feature of Mr. Graham's 

 law is, that the mutual diffusive velocities of gases are exactly 

 proportional to those which theory assigns for their relative 

 velocities of escape into a vacuum. 



That portion of Mr. Graham's experiments last alluded to 

 on this point are, I admit, at variance with the acknowledged 

 law of gaseous mechanics, which pronounces that the velo- 

 cities of different gases rushing into a vacuum are inversely 

 proportional to the square roots of their densities. The de- 

 monstration of this law is, however, so rigorous and unexcep- 

 tionable, as naturally to inspire a suspicion that there is in 

 Mr. Graham's facts either some inaccuracy of observation, or, 

 what is more probable, some defect in the principle of the 

 mode of operating, which has led to an erroneous conclusion. 

 That this is really the case, is rendered still more probable 

 by the fact that there is a certain degree of accordance be- 

 tween Mr. Graham's observations and the proportions as- 

 signed by the theoretical law. For instance, he finds that the 

 velocity of hydrogen flowing into a vacuum is considerably 

 greater than that of common air under similar circumstances, 

 but not quite so much so as theory would indicate. But a 

 still more suspicious circumstance affecting the accuracy of 

 this Table is, that the density of the gases, which in the case 

 of their mutual diffusion is, according to Mr. Graham's own 

 law, an element of the first importance, should in the case of 

 their escape into a vacuum have little or no agency. Ac- 

 cording to Mr. Graham, the four gases, whose specific gravities 



are as follow : Nitrogen 0*972 



Common air 1*000 



Oxygen 1*111, 



Carbonic acid 1*527 



all flow into a vacuum with the same velocity ! So singular 

 an anomaly unavoidably creates a distrust of tlie principle of 

 operating, or of the accuracy of the observation. 



Relying, therefore, on the validity of tliis law of gaseous 

 mechanics, wc shall proceed to examine how, in cooperation 

 2 T 2 



