HALL AND CAMPBELL. — MAGNETIC EFFECTS IN SOFT IRON. 



649 



introducing the electron theory, which we do not propose to discuss in 

 this paper. 



The Voigt Foj-mula. — Before the appearance of Moreau's equation 

 Voigt^2 1jj^(^ ijegij }gj \yy theoretical considerations, not involving con- 

 sideration of electrons, to propose a formula very like that of Moreau, 

 which, using our own symbols in part, we shall write thus : ^^ 



Tc -^ p = hTe -^ ®'- 



(13) 



0', replacing the s of the ^loreau formula, is a quantity the definition 

 of which we shall discuss at considerable length, as a number of refer- 



s — > </s 



Figure 11. 



ences to it which we have seen do not make its meaning clear. ^"^ We 

 cannot well show the significance of 0' without the use of a diagram. 



The ordinary thermo-electric diagram has two rectangular axes, the 

 horizontal one representing temperature, the vertical one a variable not 

 usually named here, but which, as one of us pointed out several years 

 ago,^^ is really entropy. Accordingly we make our diagram (Figure 

 11) as an ordinary temperature-entropy diagram is made, with tem- 

 perature, absolute C, vertical and entropy (ergs -r- T) horizontal. In 

 accordance with common practice the T axis is assumed to be identical 



" Wied. Annalen, 67 (1899). 



*^ Voigt has a — sign in his equation, which we avoid by making our hTe 

 equal his — Q. 



" For example, Zahn appears to confuse 0', which = d@'dT, with itself. 



" Proceedings, A. A. A. S., 54 (1904). The paper contains much that I 

 regret pubhshing, but in its early pages there is this one suggestion which still 

 seems to me valuable. — E. H. H. 



