MEASUREMENT OF THE PELTIER EFFECT. 437 



so that the work of the external forces is 



H 2 



As the current is infinitely weak, the heat disengaged, according 

 to Joule's law, may be disregarded; the phenomenon is reversible, 



and we may apply Carnot's principle. Referring to (645), we have 



H 2 



here # = H 2 , b= -- , and equation (5) of (646) gives 



H 2 _ T> /HA _ m 

 T~}T \2 )- W 

 or 



H = AT. 



As experiment does not favour this conclusion, it follows that, in 

 using a condenser, the difference of potential of the plates cannot in 

 general be considered as constant; it depends on the quantity of 

 electricity with which they are charged, and it must be assumed that, 

 no doubt by the action of the surrounding medium, a phenomenon 

 is produced analogous to the polarisation of electrodes. 



1032. MEASUREMENT OF THE PELTIER EFFECT. SOLIDS. The 

 Peltier phenomenon (247) may be considered as furnishing a general 

 method for measuring local variations of potential. Between two 

 points, whose difference of potential H is independent of the exist- 

 ence of a current, the energy furnished or absorbed by a current I in 

 unit time is W=IH. If this energy is simply transformed into a 

 quantity of heat Q, we have 



JQ=IH, 



and the determination of H reduces to a calorimetrical measure- 

 ment. The only difficulty consists in eliminating the heat disengaged 

 between the same points, in accordance with Joule's law. 



If all the measurements have been made in C. G. S. units, the 

 mechanical equivalent J of heat, expressed in grammes and centi- 

 grade degrees, is 4-17 x io 7 (917). Without changing the unit of 

 heat, if the current is in amperes and the difference of potential H 

 in volts, we have simply 



IH = 4-i7, or H-x. 



