279 



gative, in terms of the reciprofals of tlio powers of the function of 

 the heat before mentioned, I'epresenting the effect of the actions of 

 the nuclei or centres in modifying the superficial-atomic elasticity. 

 The numerators of the terms of this series are functions of the 

 density, diminishing along with it, and requiring to be determined 

 by experiment. 



The last term of the expression represents the effect of the 

 mutual action of separate atoms, and is a function of the density, to 

 be determined by experiment. 



The third section treats of Temperature and of Real Specific Heat. 

 Bodies are defined to be at the same temperature, when the powers 

 of their atoms to communicate heat are equal ; and the proper mea- 

 sure of temperature is defined to be the elasticity of a perfect gas at 

 constant volume, or its volume under constant pressure. Those 

 quantities are, in all perfect gases, proportional to the temperature, as 

 measured from a point 274-6 centigrade degrees, or 49428 de- 

 grees of Fahrenheit's scale, below the temperature of melting ice. 

 This point is called the absolute zero, and temperatures, as mea- 

 sured from it, absolute temperatures. 



It is shewn from the equations in the preceding section, that ab- 

 solute temperature, as thus defined, is simply proportional to the 

 quantity of heat in one atom, plus a constant, multiplied by a con- 

 stant coefficient. The constants depend on the nature of the sub- 

 stance, and the coefficient especially on its chemical constitution. 



The reciprocal of this coefficient is, of course, the real specific 

 heat of one atom, which, being divided by the atomic weight, gives 

 the real specific heat of unity of weight. 



The following laws, whicii have been to a great extent established 

 experimentally by Dulong, are inferred from the theory — 



That the specific heats of all simple atoms are either the same, or 

 vary only in certain simple 7iumerical ratios. 



That the specific heats of atoms of sindlar chemical constitution 

 are either the same, or vary only in simple numerical 7'atios. 



The fourth section relates to the actual coefficients of elasticity 

 and expansion of gases. The coefficient of increase of elasticity with 

 temperature at constant volume, and the coefficient of expansion 

 under constant pressure, are the same, and equal to each other, for 

 every substance in the state of perfect gas, being the reciprocal of 

 tho absolute temperature of melting ice, (or -00364106 per centi- 



