28 Prof. J. Le Conte on the Discrepancy between the Computed 



action on each other in the transmission of elastic pulses, and 

 that consequently sounds are propagated in it in the same man- 

 ner and with the same velocity as if it were a homogeneous gaseous 

 medium having the known physical properties of the mixture. In 

 fact, so far as the propagation of sound is concerned, some such 

 view is necessary in order to account for the fact that sounds are 

 not duplicated when heard at great distances. The difficulty of 

 conceiving of such a mutual action between the nitrogen and 

 oxygen of the atmosphere, under the assumption that it is a 

 mechanical mixture, has led Mr. Herapath*, and perhaps others, 

 to conclude that the air must be a chemical compound. And it 

 must be confessed that the physical evidence pointing in this 

 direction is much stronger than I had imagined previous to the 

 present investigation. Let us look at this point more closely. 



It is proper to premise that the proofs which chemists have 

 adduced to show that the air is a mechanical mixture are mainly 

 negative. They show that there are no evidences of chemical 

 action between the constituents of the atmosphere ; that there 

 are no indications of chemical forces resisting the separation of 

 the nitrogen and oxygen ; that Graham's law of diffusion, which 

 secures uniformity of constitution, is a physical law depending 

 on the relative density of the gases, and having no relation to 

 their chemical properties, &c. At first sight this seems to con- 

 stitute a strong body of evidence in favour of the Daltonian 

 view. But it must be recollected that in liquid mixtures we have 

 every gradation of manifestation of chemical phenomena, from 

 extreme chemical activity to cases in which it is difficult, if not 

 impossible, to detect any evidence of such action. Moreover it 

 is well known that the more recent investigations of Prof. Graham 

 have demonstrated the existence of analogous physical laws 

 governing the diffusion of saline solutions, in which Dalton's 

 idea, that the diffusibility is referable exclusively to the elasticity 

 or self -repulsion of the particles, is totally inadmissible. Indeed 

 philosophers have for some time past recognized the necessity of 

 modifying Dalton's views, in so far as to admit the existence of 

 a mutual molecular attraction among the particles of the diffusing 

 gases f. In fact the late Mr. Ivory, in speaking of Dalton's 

 hypothesis " that one gas is inelastic to all other gases/ 3 says, 

 "although it it be intelligible in words, is incomprehensible by 

 the understanding." Nevertheless, while physicists have thus 

 been compelled to acknowledge the mutual action of the elastici- 

 ties of intermixed gases, yet, in the absence of evidences of distinct 



* Mathematical Physics, vol. ii. p. 37. London, 1847. Also Mr. T. S. 

 Thomson, ibid. vol. xxv. p. 54 (1844). 



t Mr. Ivorv, in Phil. Mag. S. 3. vol. xx. pp. 81, 197,278 (1842). Also 

 Mr. T. S.Thomson, ibid. vol. xxv. p. 54 (1844). 



