the Thermal Unit. 443 



had with Dr. Joly, that he is now in favonr of some such 

 proposal as that which 1 have to make at the close of this 

 paper, and I hope to be able to add a promised communica- 

 tion from him as an appendix. 



In < Nature/ May 23, 1895, Mr. Spencer Pickering 

 advocates the adoption of the latent heat of fusion of ice as 

 the unit. He writes as follows : — " A unit of convenient 

 magnitude would be one equivalent to about 100 small 

 calories, and 100 calories has, indeed, been adopted as a unit 

 by more than one writer on thermo- chemistry. There is, 

 however, what may be termed a natural quantity which is 

 nearly equivalent to such a unit, namely, the heat of fusion of 

 one gram, of water at 0° C, which is nearly eighty calories. 

 This appears to be just as suitable from other points of view 

 as the heat of vaporization of one gram of water at constant 

 temperature and 760 millim. pressure ; and if this latter can 

 be recommended on the ground that in defining it we replace 

 the thermometer by the barometer, the former will possess 

 the superior claim of (for all practical purposes) not depending 

 even on the barometer. 



" Neither of these proposed units, however, possess what 

 should be the chief characteristic of a physical unit, namely, 

 a simple relation to other units ; and before adopting either 

 of them, it would be well to consider whether some convenient 

 unit related to, say, the electrical units, could not be adopted. 

 A committee of the British Association would be a body most 

 suited to investigate this matter." 



I doubt, however, if there is any prospect of ascertaining 

 the value of this " fusion-unit " with sufficient accuracy. 

 Mr. Pickering says " it should be determinable with quite as 

 much accuracy as the latent heat of vaporization ; " but here 

 I join issue with him. The attempts which have so far been 

 made to determine- this quantity with precision cannot be 

 considered as successful, and I believe that there are great 

 experimental difficulties in the way. When we have come to 

 a general agreement as to the magnitude of this constant, the 

 time may have arrived for a consideration of Mr. Pickering's 

 suggestion. Both the " fusion-unit " and the " vaporization- 

 unit " would, however, be purely arbitrary, and, to repeat the 

 words of Mr. Gray, " would not be arithmetically commen- 

 surable with any other physical quantity." 



I now pass to the consideration of what appears to me to be 

 the only solution of the difficulty, viz. the adoption of the 

 thermo-dynamic unit. Such a unit is by no means a novelty, 

 for it was in reality made use of by Rowland when tabulating 

 the results of his experiments on the mechanical equivalent, 



