Intelligence and Miscellaneous Articles. 327 



Mariotte's law that experiment reveals in the different gases. I 

 do not think that we can accept the explanation that M. Duhrunfaut 

 has lately offered*, an explanation which tends to ascribe these de- 

 viations to small quantities of aqueous vapour existing in even the 

 most perfectly dried gases. When Plucker published his experiments 

 on Geissler's tubes, I succeeded in preparing tubes of nitrogen and 

 of carbonic acid which contained no traces whatsoever of the three 

 brilliant rays which belong to hydrogen and aqueous vapour. To 

 accomplish this, I made use of a good common air-pump ; I exhausted 

 the receiver thirty or forty times, and I dried the gases by the ordi- 

 nary means, except only that the electrodes were of platinum in- 

 stead of aluminium, which is very often employed. 



This is the method, pointed out by Rudberg, that M. Regnault 

 and all experimenters have followed. If, nevertheless, a trace of water 

 does remain, it seems to me impossible that it should produce the 

 great deviations that we observe in the case of imperfect gases. 



I have also proved that for air and carbonic acid the molecular 

 state cannot be considered to result solely from mutual attractions 

 or repulsions, whatever may be the law of these actions ; in short, a 

 cold and expanded gas, being then heated and compressed to the 

 same volume, ought to exhibit the same phenomena with regard to 

 its compressibility, which is contrary to experience. And the re- 

 searches of M. Amagat have lately proved the same thing for am- 

 monia and sulphurous acid. The mechanical theory of heat leads us, 

 as a natural consequence, to regard heat as resulting from the mo- 

 tions of the molecules, and to define a gas as a body whose molecules 

 travel in all directions in space. But MM. Kronig and Clausius 

 have shown that if we suppose these progressive motions in the gas 

 to be rectilinear, we arrive at Mariotte's law ; and M. Clausius has 

 even developed a formula which has enabled him to calculate the 

 mean velocity of these motions for the better-known gases. 



The deviations from Mariotte's law arise consequently from attrac- 

 tions which still exist in the gases, and which are nothing but a par- 

 ticular case of universal attraction : these attractions are more or less 

 feeble according to the mass and the mean distances, greater or 

 less, by which the gaseous molecules are mutually separated. This 

 is the simplest explanation we can offer of the phenomenon ; it is 

 the one which I believe is most generally accepted. 



All this being granted, we may determine the actual velocities of 

 the molecules in imperfect gases. 



Imagine a kilogramme of gas, at temperature zero, and under 



an initial pressure p so slight that the volume v shall be very 



great, so that we may disregard the attractions. Increasing the 



p v 

 pressure to p, the volume will be v', and we shall have -^ = 1 + A p , 



A p being the deviation from Mariotte's law under the pressure p. 



Raising the temperature to t, the pressure being constant, the vo- 



v 

 lume v' Q becomes v, and we have — =l + a p t, qc p being the coeffi- 







* Comptes Rendus, vol. lxviii. p. 1262. 



