442 Mr. E. H. Griffiths on 



He adds, " Even if the specific heat of water was accurately 

 known it would not, for any reason, be arithmetically com- 

 mensurable with any other definite physical quantity. 

 Hydrogen is an elementary substance, and its atomic weight 

 is a standard unit in the arithmetic of chemistry. The specific 

 pv of hydrogen is on this account most appropriate for the 

 standard unit of heat and energy in the arithmetic of thermo- 

 dynamics/' 



Although there is much force in Mr. Gray's arguments, 

 there are, however, practical objections. The exact deter- 

 mination of pv presents almost similar difficulties to those 

 encountered in the determination of the capacity for heat of 

 water. We are entirely dependent on thermometric measure- 

 ments, and I doubt if we have yet thoroughly mastered the 

 difficulties of gas thermometry, more especially as regards 

 the influence of the envelope on the enclosed gas. The 

 theoretical advantages are no doubt great, but the practical 

 difficulties would I believe be increased instead of diminished 

 by the adoption of this unit. 



In a letter to l Nature,' May 2, 1895, Dr. Joly suggests the 

 latent heat of steam at the standard pressure as a practicable 

 unit. He remarks as follows : — " One gramme of saturated 

 steam at 760 millim. might be assumed to give up the unit 

 quantity of heat in becoming water without change of tem- 

 perature. This unit might be called a therm, in order to 

 avoid confusion with the existing unit. The specific heat of 

 water would then stand at about 1*8 milli-therms. The larger 

 value of the new unit commends itself as being more applicable 

 to the problems of applied science ; which, indeed, may be 

 inferred from the fact that engineers often understand by the 

 term calorie the kilogramme-degree. 



" I am aware that the change proposed is a radical one ; 

 but an appreciable change is better than a vexatious correction, 

 and we know now that revision and change are inevitable. 



" In the definition of the proposed unit we replace the 

 unreliable thermometer by one of the most trustworthy of 

 instruments — the barometer ; and our quantities of heat may 

 be determined by the chemical balance, and at 760 millim., 

 read directly upon the weights. We are sure of the purity 

 of the material ?' 



This appears to me to be a thoroughly practical suggestion, 

 and, above all, it clears the way of many of the difficulties 

 connected with thermometry. The change, however, is I fear 

 too radical and the magnitude of the proposed unit is dis- 

 tinctly inconvenient. I will not dwell on this point, for I 

 have reason to believe, from a conversation I subsequently 



