HALL. — THERMAL AND ELECTRICAL EFFECT IN SOFT IRON. 623 



allowance for the change of specific heat with change of temperature. 

 He gets with the unit current (10 amperes), in copper, — 



o- = +(3.01 + 0.00662 t) 10-', 

 and in silver, — 



o- = +(7.363 + 0.00887 t) 10"^ 



t being the temperature on the ordinary Centigrade scale. From 

 these equations we get for copper, — 



. = .H-2-=+(l^ + C6.2)lO-», 

 and for silver, — 



_+(^i5«0 + 88.,),o-... 



Each of these equations indicates a falling value of v with rise of tem- 

 perature. Each indicates for the ordinary thermo-electric diagram a 

 line rising from left to right, with constantly though slowly diminishing 

 steepness. Lecher himself would probably not claim that his observa- 

 tions are sufficiently accurate to establish as a fact the change of cur- 

 vature of these representative lines. They do, however, make it appear 

 probable that the lines for copper and silver should be drawn nearly 

 straight through a great range of temperature on the thermo-electric 

 diagram, and they suggest these metals as good subjects for further 

 and more careful study. For the very difficulty which every one of 

 experience in the matter finds in the measurement of the Thomson 

 effect in any metal impresses upon us the desirability of establishing 

 some one line on the thermo-electric diagram so accurately that others 

 can be satisfactorily determined from it by "thermo-electric height" 

 measurements, without further recourse to direct observations of the 

 Thomson effect. 



To be sure, the lead line, taken as horizontal, has been used in this 

 way, and it may be asked whether it is not still sufficient for the purpose 

 indicated. All observers, so far as we know, agree in declaring the 

 Thomson effect in lead to be very small. Le Roux,^^ who rated this 

 effect in iron as 31 and in copper as 2, found it sensiblement mil in 

 lead. At or near 50° C, Haga^^ found the effect in "pure" lead to 



" Ann. de Cliimie et de Ph\-sique, 4"'"' serie, t. X, 18G7, p. 201. Le Roux does 

 not, we believe, describe the quality of the lead which he used. 



^8 Ann. de TEcole Poly technique de Delft, t. Ill, 1887, p. 51. Analysis showed 

 the lead useil by Ilaga to contain, "beside a trace of silver, 0.09.3 per cent of anti- 

 mony, 0.004 per cent of iron, and 0.013 per cent of zinc." It is easy to get much 

 purer lead now. 



