19.S Proceedinc/s of the Royal Society of Victovifx. 



The same Daniell, when connected in series with the gal- 

 vanometer, gave a deflection g for resistance R. If the 



current through the galvanometer be kg, then -^ = /c g, 

 or /.• = ^^^. Now the mean value of Rg is 1729000, 

 therefore Z; = ivoguoo- ^^ ^ ^^® ^^^® electromotive force of the 

 thermal circuit, and i' its resistance, then e = ksr, or the 

 electromotive foice corresponding to one scale division is 

 ^^1729000 '" '^''^^t' or 92-3 absolute units of electromotive force. 



Now s = 2S6t — 0051 2f-, or if we measure the tem- 

 perature from the neutral point of the two metals, s = 

 •00512/^. If the temperatures be the neutral point, and 

 100° above or below it, and if m be the relative Thomson 

 effect, then 512 x 92-3 = 5000 ?>i, and m = -94. 



To find the neutral point— ^Jf = 0, i.e., 2SQ - 2 x 

 •00512^ = 0, t = 230, i.e., 249° C. ; the height at 0' C. is 

 therefore 249 X '94 = 234. The thermo electric height of 

 thallium above lead is thus— 234 — 94 1, t here being 

 temperature centigrade. 



Another similar but independent set of observations 

 gave as the height 198 — -65 t, the mean of these being 

 217 — -79 /. Of the four metals, each pair was taken 

 together, and the following results obtained. In each case 

 the higher metal is the first : — 



- 217- -79^ 



- 43 + 1-79^ 



- 252 + i-oe^; 



- 12 + 10/ 



- 40 + 147 / 



- 200 + 1-35/ 



If we add the Thallium-Lead to the Copper-Thallium, we 

 get 200 + 100 t, which agrees fliirly well with the directly 



