ON TUB SPECIFIC HBAT OF \VATKK. 



217 



as supplied by the makers, which was, of course, ascertained liy bridge measurements 

 with a small current only. 



Curve B, in fig. 5, shows the result of an experiment in which the bath was 

 gradually heated by passing a current of about 5 amperes through the resistance, by 

 \\-hirli means the bath was heated by steps from about 13 C. to 50 C. The current 

 was turned off every five or ten degrees, and the resistance measured on the bridge 

 with a small current. These resistances, plotted on temperatures, gave the curve B. 



10005 



I-OOOO 



3905 



9990 



Fig. 5. 



The bath was then allowed to cool, and the resistances measured from point to point 

 are set out on the curve C, the arrow heads indicating the direction in which the 

 temperature was moving. It will l)e observed that the passage of a current of 

 5 amperes produced a raising of the resistance above the value due to the temperature 

 of the bath with the hysteresis effect which had been previously noted by other 

 observers (see CALLENDAB, " On Continuous Electrical Calorimetry," ' Phil. Trans.,' A, 

 vol. 199, p. 77, 1902). 



In the foregoing experiment the resistance measured is that which jtersists after 

 tin' /ii'ttrtf current has ceased. In curve D is given the result of an experiment to 

 determine the variation of resistance during 

 flic jitis.ttttfc <>/ a licnry current. The man- 

 ganin resistance was arranged for an ordi- 

 nary bridge measurement, as shown in fig. G, 

 in which R is the manganin resistance of 

 In ohms, and M a mercury thermometer- 



To Dynamo 



Fig. 6. 



resistance cit'alMiiit 5'6 ohms, p Ix-ing a fixed resistance of 180,000 ohms, and p an 



adjustable resislam t' alxiut 73,000 ohms. The bridge current was varied step 



\"l.. OCXL A. 2 V 



