290 Messrs. J. T. Bottomley and A. Tanakadate. [June 20, 



in C.Gr.S. units, and must be divided by 10 8 if it be desired to reduce 

 them to volts. The differences of temperatures are given in centi- 

 grade degrees. The direction of the current in each of the cases 

 represented, is from platinoid to the second metal of the pair through 

 the hot junction. 



Table I shows, in the way now well known,* the multiplier, at any 

 temperature centigra'de, which must be used, as factor with the 

 difference of temperatures between the hot and cold junctions, in 

 order to calculate the electromotive force in C.G.S. units. The 

 algebraic sign corresponds with that used by Tait, and now adopted 

 by Everett (' Units and Constants,' 2nd Edition, 1886). 



Table I. 



Platinoid -platinum. — 925 — V16xt. 



Platinoid- aluminium — 985— 4'3lx£. 



Platinoid-iron —2916 + 0-86 Xt. 



Platinoid-copper (A) — 1246 - 5'44 x t. 



Platinoid-copper (B) —1294 -4"88 X t. 



Combining the results of Table I with those of Tait, reduced by 

 Everett, we obtain the thermo-electric distance of platinoid from lead, 

 taken as zero, at various temperatures centigrade. If any one of the 

 wires platinum, aluminium, iron, or copper used by us, were identical 

 with the wire of the same name used by Professor Tait, we should be 

 able to deduce with exactness the distance of our platinoid wire from 

 his lead wire. That, however, was not the case ; and each of the 

 secondary wires used by us gives us, as it were, a different result. 

 Thus we have : — 



Table II. 



Platinoid to lead. 



From experiment with platinum ...... —986 — 2*26 X t 



., „ aluminium — 1062 — 3*92 x £ 



„ iron -1182-401 



copper (A) — 1110— 4'49x£ 



copper (B) ... . -1158-3"93x£ 



Taking the mean of all of these, with the exception of the result 

 for platinum, which we omit because different specimens of platinum 

 are well known to differ thermo- electrically enormously among them- 

 selves, we obtain for the thermo-electric distance of platinoid from 

 Professor Tait's lead wire —1128 — 4*1 x t. 



This result enables us to place platinoid in Tait's thermo- electric 

 diagram. Its line is nearly parallel to those of palladium and German 



* Tait, ' Edinburgh Roy. Soe. Trans.,' vol. 27, 1873, and Everett's ' Units and 

 Constants,' 2nd Edition, 1886,—" Thermoelectricity." 



