the Velocity of Sound. 95 



sudden condensation or rarefaction of air contained in space 

 limited by partitions which prevent the dispersion of the gene- 

 rated heat and cold into the surrounding space. The tendency 

 to such dispersion is proved in the common experiment by which 

 German tinder is ignited. It may, however, be admitted that 

 the effects of the development of heat in the condensed part of 

 a wave, and of cold in the rarefied part, do not completely destroy 

 each other, and that there is a residual gradation of temperature 

 and elasticity from point to point of the wave, sufficient to 

 account for a slight augmentation of the velocity of propagation. 

 If my mathematical argument be good, this is all that we have to 

 account for. 



But if an elastic fluid does not possess the property of trans- 

 mitting heat-undulations without their undergoing partial 

 degradation and dispersion, the foregoing conclusion is no 

 longer true. The Theory of Molecular Forces on which the 

 conclusion rests, assumes that the medium is perfectly trans- 

 mittent with respect to heat. But Dr. TyndalFs experiments 

 show that while this is the case very approximately in simple 

 gases and atmospheric air, other gases, as especially olefiant gas, 

 do not possess the same property. In these, accession of heat 

 was observed to be accompanied by radiation, and abstraction of 

 heat by absorption, under circumstances in which such effects 

 do not take place in atmospheric air. Consequently in such 

 gases the changes of caloric action produced by forcible altera- 

 tion of the density of the fluid have not their full dynamic effect 

 on the atoms in producing alteration of pressure, being partly 

 expended in the observed radiations and absorptions. This con- 

 version of the undulations is analogous to that which light- 

 undulations undergo when transmitted through a partially 

 transparent medium. Hence, so far as this cause of disturbance 

 operates, the fluid is imperfectly elastic, and transmits sound- 

 waves with a velocity less than that which accords with the rela- 

 tion between the pressure and the density in the quiescent state. 

 It is particularly to be noticed that the alteration of that relation 

 here considered altogether depends on a transient disturbance of 

 the statical condition of heat, produced by a sudden alteration of 

 density, and that the law of Mariotte holds good as soon as the 

 equilibrium is restored. 



But it may be urged that the property of radiation and ab- 

 sorption of heat in olefiant gas has itself the effect of altering the 

 velocity of propagation. While this may be admitted, the fol- 

 lowing reasons may be given for concluding that the effect is 

 very small. According to the views explained above, the radia- 

 tion of heat and cold by olefiant gas when in a state of rapid 

 vibration, is an effect of the same kind as that which is pro- 



