Measurements of Precision in Platinum Thermometry. 557 



If, however, the Coils are badly adjusted so that instead of a 

 resistance bein£ 9*9 ohms it is 9*95 ohms, the change 

 when this coil was used as a shunt resistance would be 

 5 millionths of an ohm less than O001, but an error of this 

 amount would not necessarily be made, as the bridge is very 

 easily calibrated; Badly adjusted shunt coils are equivalent 

 to a non-uniform slide wire. It will be seen that only 

 ordinary care is required in the adjustment of the coils. In 

 the application of the first method, there is, in addition to the 

 three dials and in series with them, a coil of about 0*4 ohm 

 resistance ; this, together with the resistance in the three dials 

 (when reading 0.0.0), makes a total resistance of about 

 2 ohms. 



With such a resistance system for R the bridge may be 

 standardized by inserting a standard coil of 2 ohms resistance 

 in the place of P and obtaining a balance by shunting this coil, 

 or by the use of the three dials. P and R are then reversed 

 and a second balance obtained. This measurement enables 

 the resistance of the arm R to be calculated when the dials 

 read 0.0.0 and the other coils are shunted through the 

 mercury contacts. To standardize the main resistance-coils 

 a 10-ohm coil may be placed in P and R adjusted for a 

 balance. 



A complete calibration of R is most readily made in a 

 manner similar to that Used for the Callendar-Griniths 

 bridge, but the latter process need not be described here as 

 it is so well known. 



The resistances Q and S are conveniently of 200 ohms 

 resistance. 



■Second Method* 

 Fig-. 6. 



Principle. — At first the connexions are as shown in fig. §•, 

 in which X is the adjustable resistance and is of a form 

 similar to that described for R in Method I. Q, S 5 and H 



