MOTIVE POWER OF HEAT, 73 



any direct experiment, and doubtless such an ex- 

 periment would be very difficult, but there exists a 

 datum which is very nearly its equivalent. This 

 has been furnished by the theory of sound. It de- 

 serves much confidence because of the exactness of 

 the conditions which have led to its establishment. 

 It consists in this : 



Atmospheric air should rise one degree Centi- 

 grade when by sudden compression it experiences 

 a reduction of volume of T fg-.* 



Experiments on the velocity of sound having 

 been made in air under the pressure of 760 milli- 

 metres of mercury and at the temperature of 6, 

 it is only to these two circumstances that our 

 datum has reference. We will, however, for greater 

 facility, refer it to the temperature 0, which is 

 nearly the same. 



Air compressed T fg-, and thus heated one degree, 

 differs from air heated directly one degree only in 

 its density. The primitive volume being supposed 



* M. Poisson, to whom this figure is due, has shown 

 that it accords very well with the result of an experiment 

 of MM. Clement and Desormes on the return of air into a 

 vacuum, or rather, into air slightly rarefied. It also ac- 

 cords very nearly with results found by MM. Gay-Lussaq 

 and Welter. (See note, p. 87.) 



