76 Determination of Degrees of Heat in Absolute Measure, 

 we shall have for all points of the body 



* = 0, (16) 



since here both the differential equation (14) and the limit- 

 ing conditions in the surfaces a and a^ are satisfied, while the 

 limits for the rest of the surface of the body, where k is nil, are 

 satisfied by any given value of <f>, and therefore as well as by 

 that here taken. From equations (13), (15), and (16) we have 



T a -TJ = (P -P)(P-P 1 ) (17) 



If, then, a constant electrical current be passed for some time 

 through a conductor of any shape which is surrounded by bad 

 conductors, and if the temperature at the two surfaces be kept 

 constant and equal, it will be possible by equation (17) to calcu- 

 late the temperature at each point of the conductor from the two 

 differences of electrical tension at the given point and the con- 

 ducting surfaces. The result found will, conversely, serve for an 

 experimental determination of the Centigrade degree in absolute 

 units. 



The increase of temperature which depends on the electrical 

 current is T — T . Now 



T 2 -T 2 > (T-T ) 2 ; 



and in accordance with the last equation, 



(T-T f<(P -P)(P-P,). 



As the right-hand side acquires its greatest value for 2P= P -f- P p 

 if P is taken as greater than P„ we have also 



P _P 

 T-T < o * (18) 



Prom this we see that the greatest increase of temperature 

 which can result in any point of the circuit is always numerically 

 smaller than half the difference of the electrical tensions in the 

 two conductor-surfaces. It would be equal to just half this 

 difference if the conductor-surfaces could be cooled to the ab- 

 solute zero — that is, for T =0. Thus the differences of elec- 

 trical tension, and the greatest increase of temperature which 

 can be obtained by them, stand in the closest connexion with 

 each other. 



Hence it is not without interest to calculate the difference of 

 electrical tension in the poles of a voltaic element for instance, 

 the electromotive force of the element, in Centigrade degrees. 

 Thustheelectromotive force ofaDanielPselementis about 12 x 10 10 

 absolute units ; half the difference of tension of the elements, or, 

 as it is commonly expressed, the positive tension of the element 

 (the negative taken at the same amount), is therefore 6 x 10 10 



