410 Mr. E. H. Griffiths on 



experimental reasons for the conclusion that the ratio 



Mean calorie , . ., , , „ 



tttj ; r: — t-tto is much nearer unity than the result or 



Thermal unit at 15 J 



the above comparison would indicate. I cannot here do more 

 than briefly state the reasons for my conclusions, but full 

 details of the experimental evidence will be found by any 

 who refer to the original paper. I performed some ex- 

 periments on the latent heat of evaporation of water at 

 temperatures 25°, 40°, and 50°, and, having made improve- 

 ments in the apparatus, determined with great care the value 

 of L at 30° and 40°. 



The result of all these experiments is closely represented 

 by the formula 



L = 596-73-O-6OlO0 



in terms of a thermal unit at 15° C. 



XT .p Mean calorie - A11E , P 



JNow it we assume ^ ; -, - F7 r=r011o (mean of 



1 hernial unit at 15 



results given supra), the value of L at 0° as determined by 



Dieterici is 603' 6, whereas if we assume the equality of the 



two units Dieterici's value becomes 596*73, L e., is identical 



with that resulting from my experiments. 



Again, if we deduce the values of L over the range 63° to 

 100° from Regnault's experiments on the Total Heat of Steam, 

 then (assuming as before the equality of the two thermal 

 units) the above formula represents, with extraordinary 

 accuracy, the result of Regnault's experiments. 



This evidence is of course indirect, but, at the same time, 

 its weight cannot be rightly estimated without a study of the 

 Tables xiv. to xvn., as given in the above paper. 



The abstract of that communication came under the notice of 

 Dr. Joly, F.R.S., who thereupon commenced a series of deter- 

 minations by means of his well-known calorimeter. A short 

 account of the experiments and the conclusions to which they 

 led is given by Dr. Joly in pp. 322, 323 of my paper. 

 These experiments are, I believe, the first direct determinations 

 of the ratio of the mean calorie to the thermal unit at 15° 

 which have been performed since the time of Regnault ; and 

 I may be permitted to express my gratification at the extent 

 to which they bear out my predictions of December 1894, 

 regarding the close equality of the two units. The result is : — 

 Mean calorie _ Q'9957* 

 Thermal unit at 15° 1 



* In obtaining this number Dr. Joly assumed the decrease from 0° to 

 15° given by Bartoli and Stracciati ; if we assume Rowland's decrease 

 over that range we get 0'9962. 



