550 



PROCEEDINGS OF THE AMERICAN ACADEMY. 



When either of these methods is used it is an advantage to have the 

 leads of low resistance ; this means large wire and the possibility of a 

 great cooling effect of these large l^ads on the fine platinum wire. 



It is the first object of this paper to show a method of bridge connec- 

 tion whereby the resistance of the leads may be entirely and directly 

 eliminated. 



A difficulty peculiar to resistance thermometers results from the con- 

 dition that in general we have to measure a resistance varying over a 

 very great range. The regular form of resistance box presents condi- 

 tions which make it insufiicient for this purpose, involving, as it does, the 

 calibration of resistances which are high multiples and sub-multiples of 

 a chosen standard. 



The potentiometer methods of measurement suggested a modification 

 of the Carey-Foster form of the Wheatstone bridge, which was finally 

 ^ adopted. A resistance may 



be measured by connecting 

 it as part of one side of 

 a wire bridge and tapping 

 the galvanometer lead suc- 

 cessively to its two ends. 

 Its value is given by the 

 resistance between the bal- 

 ancing points of the other 

 galvanometer terminal, 

 multiplied by the ratio of 

 the resistances of the two 

 sides of the bridge. This 

 may be modified in the fol- 

 lowing way, so that only a 

 difference has to be meas- 

 ured : — Let T (Figure 1) 

 be the resistance to be measured, with D and F the leads and external 

 resistance in the bridge construction, S the known resistance, M and 

 N the ratio coils, with a low resistance slide wire between them. The 

 resistance of the two portions of this wire we shall denote by w and w'. 

 The wire between the points 3 and 4 can be neglected for the present. 

 The shaded portions of the figure are thus heavy copper connectors, the 

 resistance of a single one being equal to^. 



By shifting the galvanometer lead from El to C a large change in the 

 positions of the other galvanometer terminal on the wire is called for. 



Figure 1. 



