16 Prof. Thomson on the Dynamical Theory of Heat. 



that, when an electrical current is continuously produced from a 

 purely thermal source, the quantities of heat evolved electrically 

 in the different homogeneous parts of the circuit are only com- 

 pensations for a loss from the junctions of the different metals, 

 or that, when the effect of the current is entirely thermal, there 

 miist be just as much heat emitted from the parts not affected 

 bv the source as is taken from the source. 



17. Lastly*, when a current produced by thermal agency is 

 made to work an engine and produce mechanical effect, there 

 will be less heat emitted from the parts of the circuit not affected 

 by the source than is taken in from the source, by an amount 

 precisely equivalent to the mechanical effect produced ; since 

 Joule demonstrates experimentally, that a current from any kind 

 of source driving an engine, produces in the engine just as much 

 less heat than it would produce in a fixed wire exercising the 

 same resistance as is equivalent to the mechanical effect pi'oduced 

 by the engine. 



18. The equahty of thermal effects, resulting from equal 

 causes through very different means, is beautifully illustrated by 

 the following statement, drawn from Mr. Joule's paper on mag- 

 neto-electricity f. 



mechanical theory of thermo-electric currents, in various papers communi- 

 cated to the Ro^'al Society of Edinburgh, and to this Magazine, had I been 

 acquainted with'it in time! — W. T. March 20, 1852.] 



* This reasoning was suggested to me by the following passage con- 

 tained in a letter which I received from Mr. Joule on the Stli of July 1847. 

 " In Peltier's cxi)erimeut on cold produced at the bismuth and antimony 

 soldei', we have an instance of the conversion of heat into the mechanical 

 force of the cun-ent," which must have been meant as an answer to a re- 

 mark I had made, that no evidence could be adduced to show tliat heat is 

 ever put out of existence. I now fully admit the force of that answer ; 

 but it woidd require a proof that there is more heat put out of existence at 

 the heated soldering [or in this and other parts of the circuit] than is 

 created at the cold soldering, [and the remainder of the circuit, when a 

 machine is (hiven by the current,] to make the '•' evidence " be experi- 

 mental. That this is the case I think is certain, because the statements of 

 § 16 in the text are demonstrated consequences of the first fundamental 

 proposition ; but it is still to be remarked, that neither in this nor in any 

 other case of the production of mechanical effect from pm-ely thermal 

 agency, has the ceasing to exist of au equivalent quantity of heat been 

 demonstrated otherwise than theoretically. It would be a very great step 

 in the experimental illustration (or verification, for those who consider such 

 to be necessaiy) of the dynamical theory of heat, to actually show in any 

 one case a loss of heat ; and it might be done by operating through a very 

 considerable range of temperatures with a good air-engine or steam-engine, 

 not allowed to waste its work in friction. As will be seen in Part II. of this 

 pajier, no experiment of any kind could show a considerable loss of heat 

 without employing bodies differing considerably in temperature ; for in- 

 stance, a loss of as much as "098, or about one-tenth of the whole heat used, 

 if the temperature of all the bodies used be between 0° and 30° Cent. 



[t In this pajier reference is made to his previous paper " On the Heat 



