the Coefficient of Expansion of Incandescent Platinum. 43 



Table III. 



Metals. 



Boiling-points. 



Boiling-points. 





Derille and Troost. 



Other Talues. 





Hg 



S 



Cd 



Zn 



360 



440 



860 

 1040 







350 



T448 



l ^} { 



884 



Eegnault. [Journal. 1878). 

 Bennett (Silttman's American 

 Carnelly and Williams (Quart. 



Journ. Cbem. Soc. 1876-78 

 Tiecqu.erel(Comp. Bend. Ira.). 



The substitution of these values in Benoifs formula places 

 it more at variance than before with the measurements of 

 Matthiessen and Siemens — a variation probably clue to the 

 difference of behaviour noticed by the latter physicist in the 

 case of different specimens of platinum. 



The brief discussion of the above-mentioned results suffices ; 

 we think, to show that — 



1st. The formulae in question are based for the most part upon 

 unwarrantable suppositions, such as the constancy of the spe- 

 cific heat of copper and platinum, the constancy of the coeffi- 

 cient of expansion of the latter metal, and upon the accuracy 

 of certain very doubtful values for the boiling -points of zinc, 

 cadmium, <fcc. 



2nd. That, aside from the inaccuracy of those data, the 

 varying resistance of different specimens of platinum renders 

 any formula for the calculation of temperature of that metal 

 from its electric resistance applicable only to the identical wire 

 for which the law of change of resistance with the tempera- 

 ture has been determined. 



3rd. That, from the data at command, we are not in posi- 

 tion to calculate the temperature of an incandescent platinum 

 wire from its change of resistance, nor from its length, nor 

 indeed in any other manner, further than to express the tem- 

 perature in terms of the length or the resistance of the wire. 



4th. That, owing to the great variations shown by different 

 specimens of platinum as regards its resistance, the determi- 

 nation of the expansion of the wire is to be preferred, when- 

 ever practicable, to the measurement of its conductivity. 



