4 Professor ApJOH^f upon a new Method of 



long does not hesitate to affirm, " that the results of De la Roche and Berard 

 are still those which should inspire most confidence, and that though they cannot 

 be considered as having attained perfect precision, they are amply sufficient for 

 putting beyond all doubt that the various simple and compound gases have not, 

 under the same volume, an equal capacity for heat." 



Having disposed of these preliminary animadversions upon the labours of 

 some of his predecessors, M. Dulong proceeds to the explanation of the particu- 

 lars of a very ingenious method practised by himself, for determining — not the 

 specific heats — either at a constant volume, or under a constant pressure, but the 

 ratio which subsists between these quantities in the case of the different gases. If 

 a be the caloric necessary to be communicated to a given weight of any gas in 

 order to produce in it, maintained of a constant volume, a given increase of tem- 

 perature, and a-\-b the caloric necessary to produce the same change of tempera- 

 ture, when the gas is permitted to expand so as to retain its primitive elasticity, 



= 1 -|-- expresses the ratio in question, and is the quantity at which Du- 

 long, by his method of research, was enabled in the following manner to arrive. 

 The Newtonian formula for the velocity of sound, viz. 



vii:y^^x(l. +003750, 



is long known to give results appreciably less than the truth, but Laplace was the 

 first who pointed out the cause of the discrepancy, and showed that Newton's ex- 

 pression should be multiplied by the square root of the relation between the specific 

 heat of air under a constant volume and a constant pressure, a correction which is 

 at present found to give results in almost perfect accordance with observation. If, 

 therefore, the velocity of sound in atmospherical air be determined experimentally, 



and that this be divided by the Newtonian expression, the quotient will be / ""'" , 



or the square root of the relation between its specific heat under a constant vo- 

 lume and a constant pressure ; and the same method may obviously be extended 

 to all the gases,* provided we can determine the exact velocity of sound in each. 



• The Newtonian expression, 



i'=y^^(l +,00373 0, 

 is applicable to any gas by substituting for d the s. g. of the gas in relation to mercury. 



