[Proc. Eot. Soc. Victoria, 32 (N.S ). Ft. II., 1920]. 



Akt. .X. — A New Metliod of Detentiinhig the Mechanical 

 Equivalent of Heat. 



BY 



Professor T. H. LABY, IM.A. 



AND 



J. K. ROBERTS, B Sc. 



(Natural Pliilosopliy Depaitineiit, University of Melbourne). 



(With Plates X. and XI. and 1 Text Figure.) 

 [Read 11th September, 1919.] 



Introduction. 



It appeared to tlie writers that an additional dir<_'(.t determina- 

 tion of the mechanical equivalent of heat would he of valuf, and 

 might, in view of tlie progress of precise measurement, attain 

 higher accuracy than that previously realised, and, incidentally, 

 information might be (djtained as to the accuracy of the electrical 

 units. 



After a long series of preliminary expci-iments, a method, in 

 Avhich a copper cylinder is heated by a rotating magnetic field, has 

 been devised, possessintr the characteristics — (1) of accurately 

 stationary temperatures, and (2) that a small fraction of the heat 

 •developed is lost. This method is being carried out so as to attain, 

 if possible, an accuracy of 1 in 10,000 in J, the mechanical equiva- 

 lent of heat. 



Previous Determinations. 



Theru have been five determinations^ of the mechanical equivalent 

 of heat, to which it is neceissary to refer — namely, those made by 

 Joule, Rowland, Misculescu, Reynolds and iNIoorby and Rispail. 

 Of these exiDeriments only those of Rowland and of l»eynolds and 

 Moorby appear to possess high accuracy, and they are not imme- 

 <iiately comparable, a,s the values of J, which they give, are in terms 

 of different heat units. Several determinations have been made of 

 what may be called the electrical equivalent of heat. If the 

 mechanical equivalent, J, were known to an accuracy of 1 in 

 10,000 (which should not be lieyond attainment), the degree of its 



1. Joule, Phil. Trans., Vol. CXL. p. 61 (1850) ; Rowland. Proc. Amer. 

 Acad., Vol. XV., p. 75 (1879) ; Micule.scu. Journ. de Phys., Vol. I,, p. 104 

 <1892) ; Reynolds and Moorby, Phil. Tran. Vol. CXC, p. 301 (1S97); Ris- 

 j.ail, Aim. Chim. Phys,, Vol. XX., p. 417, fl910). 



