Measurements of Precision in Platinum Thermometry. 553 



potential leads may be used, as these admit of the detection 

 with certainty of secular changes in the resistance of the 

 platinum spiral or "bulb." 



When measurements of the highest precision are to be 

 made we believe that Method 2 will be found most convenient, 

 but for calorimetric work, where rapid balancing is often 

 desired, the dial resistance-bridges (Methods 3 and 4) should 

 be extremely useful. In many such measurements readings 

 accurate within o, 01 C. are ample, and then the accuracy 

 with which the bridge coils need to be adjusted is considerably 

 less than that indicated in the text. 



The resistance of a platinum thermometer is most frequently 

 measured by means of a Callendar-Griffiths bridge *. When 

 the Siemens bridge f (fig. 1) is employed, the thermometer 



Fkr. L 



has three leads L l3 L 2 , and L 3 . Q and 8 are the ratio-coils 

 and R is^ the adjustable resistance which is nominally equal 

 to P. I£ Q = S and L 2 = L 6 , then balance is obtained when 



R=P. 



The leads are then completely compensated for as in a 

 Callendar-Griffiths bridge, but there is an important difference 

 which is often overlooked. In the Callendar-Grriffiths bridge 

 a slide wire joins the thermometer arm with the adjustable 

 resistance, but in the Siemens bridge it is impossible to use 

 a slide wire if the ratio Q/S is to be kept constant. 



A third form of platinum thermometer has two current 

 and two potential leads. Previously the resistance of such 

 a thermometer has been measured in two ways : (1) by the 

 potentiometer, and (2) by the Kohlrausch method of over- 

 lapping shunts, a differential galvanometer being employed. 

 We now describe several bridge methods for measuring the 

 resistance of such a thermometer. 



* Callendar, Phil. Trans. A. clxxviii. p. 184 (1887). 

 t W. Siemens, Proc. Roy. Soc. vol. xix. p. 351 (1871). 



