220 PROCEEDINGS OP THE AMERICAN ACADEMY 



bar, we may consider / indepeudent of S, and the sum of the heat 

 lost is a miuimum when 



dH , (IH^ _ 

 dS~^ dS —^' 





8=11 



K{t 



-0/ 



\. 



Substituting this value of >S^ in iTand H^, 



H= I\JcK (t^ — <i)]*, and ff, = I[kK (t, — <J]i 



That is, the heat dissipated is a minimum when that generated by 

 the current per unit time is equal to that lost by conduction per unit 

 time. 



Within certain limits the thermo-electric engine follows the ordi- 

 nary law of other heat engines : the work done increases as the differ- 

 ence between the temperatures of the hot and cold junctions increases. 

 Unfortunately, however, the existence of neutral points renders it im- 

 possible to obtain an unlimited electro-motive force from a thermo- 

 electric element. This appears to be one of the most serious objections 

 to the pi-actical use of thermo-electricity. We are at present engaged 

 upon some experiments on the effect of high pressure and high tem- 

 perature upon the position of the neutral point. 



