444 Mr. E. H. Griffiths on 



and more markedly by Professor Schuster and Mr. Gannon 

 when they entitled their recent communication to the Royal 

 Society " The Specific Heat of Water " *. Several objections 

 to this proposal of course immediately present themselves, of 

 which the most important is the want of agreement amongst 

 different observers as to the capacity for heat of water at any 

 given temperature when expressed in terms of such a unit. 

 The divergence is not, however, so great as is usually 

 supposed. 



I will not enter at length into my reasons for especially 

 selecting the work of certain observers ; but the principle by 

 which I have been guided is the same as that which influenced 

 my choice of experiments on the changes in water, viz. the 

 attention given to thermometry. 



I also consider it necessary to reject results obtained by such 

 electrical methods as are dependent on the measurement of 

 resistance, if the observer has neglected the change in resist- 

 ance of the conductor consequent on the rise in temperature 

 due to the passage of an electric current. 



I propose to select the following : — 



(1) The results of the later experiments of Joule, as con- 



sidered in the light of Professor Schuster's recent 

 exhaustive re-standardization of Dr. Joule's thermo- 

 meters, Phil. Mag. June 1895. (Mechanical.) 



(2) The determinations of Prof. Rowland, 1879. (Me- 



chanical.) 



(3) The work of Miculescu. (Mechanical.) 



(4) The work of Dieterici. (Electrical.) 



(5) My own experiments of 1892. (Electrical.) 



(6) Prof. Schuster and Mr. Gannon's results in 1894. 



(Electrical.) 

 Joule. — The results of Joule's later experiments as an- 

 nounced by himself were 772*55 ft.-lbs., i.e. 4*162 x 10 7 ergs at 

 61°* 7 F. Joule's own comparison with Rowland's thermo- 

 meter raised this value, after certain corrections by Rowland, 

 to 4*182 x 10 7 ergs. During this summer Professor Schuster 

 published an account (Phil. Mag. June 1895) of his elaborate 

 comparison of the Joule thermometers with the nitrogen-scale 

 of the Bureau International. His conclusion is as follows : — 

 " Joule's equivalent of heat resulting from his own investi- 

 gation and reduced to the nitrogen-thermometer of the Bureau 

 International, when expressed in ergs, becomes 4*173 x 10 7 ." 

 (Temperature 16°*5 C.) 



* In the discussion following the reading* of this paper, Lord Kelvin 

 remarked that Prof. Rankine had made a similar suggestion. 



