8 Mr. T. R t Edmonds on the Law of Density of 



experimental numbers and the theoretical numbers. On refer- 

 ence, however, to the account given by M. Regnault of the 

 various unsatisfactory experiments which he has made for tem- 

 peratures below 60° C, there will be found no ground for be- 

 lieving that the new law for latent heat (expressed by the equa- 

 tion L = PVp _w ), which is indisputably true for all temperatures 

 above 60° C, is not also true for all lower temperatures. (See 

 Memoires de I'Academie, vol. xxi. pp. 712 and 722.) 



The latent heat of saturated steam adopted by M. Regnault is 

 deduced from a well-known empirical law propounded by him, 

 as affording a near representation of the total heat absorbed by 

 a unit weight of saturated steam during its conversion from 

 water, and its elevation in temperature from 0° C. to t° Centi- 

 grade. This " total heat " is called X, and the formula is 



A,=606-5 + -305*, 



this quantity representing units of heat for temperature measured 

 by the Centigrade thermometer, and a unit of heat representing 

 the quantity of heat absorbed by a unit weight of water on ele- 

 vation of its temperature from 0° C. to 1° Centigrade. The 

 latent heat is obtained by subtracting from the "total heat" (X) 

 the number of units of heat absorbed by a unit weight of water 

 whilst its temperature is being raised from 0° C. to f Centi- 

 grade. For example, if £ = 200° C, then we have 



X=606-5 + -305x 200 = 667-5. 



From this number is to be deducted the heat absorbed by a unit 

 weight of water whilst its temperature is being elevated from 0° 

 to 200° C. This is 200 units of heat, together with 3*2 units 

 addition made for increased specific heat or capacity of water at 

 200° C. We have consequently the resulting latent heat of 

 saturated steam at 200° C. = 667*5 -203*2 = 464-3 units of 

 heat, which is the number adopted by M. Regnault, as stated in 

 the last page of vol. xxi. of the Memoires de V Academie, &c. 



The lowest average temperature at which observations on 

 latent heat were made by M. Regnault was + 10° C. At this 

 point the latent heat observed is stated to have been 5995, 

 whilst the new formula would give 587'4 Centigrade units of 

 heat. The difference is 12*1, which is the maximum difference 

 between theory and a selected series of experiments. M. Reg- 

 nault remarks that the difficulties are great in obtaining results 

 to be relied upon from observations on latent heat of saturated 

 steam made at temperatures below 60° C, or at pressure less 

 than one-fifth part of an atmosphere. He says that he has tried 

 various methods of observation at these low temperatures; he 

 assigns no reason for the preference given to the method of ob- 

 servation finally adopted ; and he expresses no confidence in the 



