438 Mr. E. H. Griffiths on 



this question of the capacity for heat of water as the thermal 

 standard. His answer was to the effect that the French 

 physicists did not feel that the problem could be profitably 

 attacked until some general agreement was arrived at with 

 regard to the scale of temperature to be adopted. 



The determinations of the capacity for heat of water at 

 higher temperatures are still more uncertain than those below 

 30°. Many attempts have been made to solve the difficulty 

 by the use of Bunsen's calorimeter. Concerning these experi- 

 ments Dieterici writes as follows * : — " The great differences 

 which are shown by observations carried out in this manner 

 render it probable that this instrument is unsuitable for the 

 purpose." Now no one has been more successful in the use 

 of this apparatus than Dieterici himself [vide his determina- 

 tion of L and J by such means), and his opinion therefore 

 carries great weight. 



For example, Veltenf, who performed a very elaborate 

 series of determinations with Bunsen's calorimeter, finds a 

 maximum in the capacity of water between 0° and 7°; from 

 7° to 10° a decrease, from 10° to 20° an increase, then a 

 decrease to 40°, where it reaches a minimum, then rises to 

 70° and remains constant to 100° ! 



Difficulties with regard to the measurement of temperature 

 are, however, much diminished by the use of Bunsen's calori- 

 meter, and the conclusions of different observers as to the 

 amount of mercury drawn into the instrument by unit mass 

 of water when cooling from 100° to 0° are in fair agreement. 

 For example : — 



Bunsen 15*41 



Schiiller and Wartha . . 15*442 

 Velten 15*47 J 



and hence Bunsen, Schiiller and Wartha, Wullner, and 

 Dieterici strongly advocate the adoption of this "mean 

 calorie" as the thermal unit§. 



With regard to this proposal I would remark that the 

 discrepancy between the above standard determinations 

 amounts to 1 in 250, and further, that even if the relation 

 between the heat developed and the weight of mercury is 

 capable of exact determination, the unit thus adopted would 

 be useless unless we knew its relation to the thermal unit " at 

 a temperature £°;" for all thermal determinations cannot 



* Wied. Ann. xxxiii. p. 444 (1888). 

 t Ibid. viii. p. 83 (1879). 

 X Ibid, xxxiii. p. 439 (1888). 

 $ Ibid, xxxiii. p. 431. 



