Prof. Thomson on the Dynamical Theory of Heat. 15 



from thermal agency*. As yet this has not been done for the 

 electrical method, as far as regards the criterion of a perfect 

 engine implied in the second proposition, and probably cannot 

 be done without certain limitations ; but the application of the 

 first proposition has been very thoroughly investigated, and veri- 

 fied experimentally by Mr. Joule in his researches " On the Ca- 

 lorific Eff'ects of Magneto-Electricity;" and on it is founded one 

 of his ways of determining experimentally the mechanical equi- 

 valent of heat. Thus, from his discovery of the laws of genera- 

 tion of heat in the galvanic circuit t, it follows that when mecha- 

 nical work by means of a magneto-electric machine is the source 

 of the galvanism, the heat generated in any given portion of the 

 fixed part of the circuit is proportional to the whole work spent ; 

 and from his experimental demonstration that heat is developed 

 in any moving part of the circuit at exactly the same rate as if 

 it were at rest, and traversed by a current of the same strength, 

 he is enabled to conclude — 



(1.) That heat may be created by working a magneto-electric 

 machine. 



('<J.) That if the current excited be not allowed to produce any 

 other than thermal eff'ects, the total quantity of heat produced is 

 in all circumstances exactly proportional to the quantity of work 

 spent. 



16. Again, the admirable discovery of Peltier, that cold is 

 produced by an electrical current passing from bismuth to anti- 

 mony, is referred to by Joule J, as showing how it may be proved 



* " There are at present known two, and only two, distinct ways in 

 which mechanical effect can be obtained from heat. One of these is by the 

 alterations of volume which bodies experience through the action of heat ; 

 the other is through the medium of electric agency." — Account of Cai'not's 

 Theor}', § 4. (Transactions, vol. xvi. part .5.) 



f That, in a given fixed jiart of the cu'cuit, the heat evolved in a given 

 time is proportional to the square of the sti-ength of the current, and for 

 difi'erent fixed parts, with the same strength of current, the quantities of 

 heat evolved in equal times are as the resistances. A paper by M. Joule, 

 containing demonstrations of these laws, and of others on the relations of 

 the chemical and thennal agencies concerned, was communicated to the 

 Royal Society on the l/th of December 1840, but was not published in the 

 Transactions. (SCe al)stract containing a statement of the laws quoted 

 above, in the Philosophical Magazine, vol. xviii. p. 308.) It was pubhshed 

 in the Philosophical Magazine in October 1841 (vol. xix. p. 260). 



X [In the introduction to his paper on the Calorific Eft'ects of Magneto- 

 electricity, &c., Phil. Mag. 1843. 



I take this opj)ortunity of mentioning that I have only recently become 

 acquainted with Ilclmholz's admirable treatise on the principle of mecha- 

 nical effect (Ueber die Erhallunij dcr Kraft, von Dr. 11. Ilehnholz. Berlin. 

 G. Ileimer, 1847), having seen it for the fir.st time on the 20th of January 

 of this year ; and that I should have had occasion to refer to it on this, and 

 on numerous otlier points of the dynamical tlicory of heat, tlie mechanical 

 theory of electrolysis, the theory of elcctro-maguetie induction, and the 



