516 



ELECTRICAL MEASUREMENTS 



course, each indicator has its characteristic rate of response to the 

 current. 



Another point may be noted : after the device has cooled down, 

 owing to shutting off the current, there will be no increase of 

 reading when a current slightly larger than that previously reg- 

 istered, is turned on until the larger current has been maintained 

 for a time longer than the normal time lag of the indicator; for 

 that time must elapse before there has been sufficient expansion 

 of the air in bi to cause the liquid again to begin to flow into the 



I'm 



510152025303540455055 

 Time in Minutei 



FIG. 302. Illustrating the influ- 

 ence of the size of a Wright-demand 

 indicator on the rate at which the 

 final reading corresponding to a 

 given current is attained. 



Time in Minutes 



FIG. 303. Showing the delay of 

 a Wright-demand indicator in be- 

 ginning registration after the indi- 

 cator has cooled down. 



index tube. This is illustrated in Fig. 303. Curve A shows the 

 normal rise of the indication, when, after the device has been set, 

 the current is maintained at 10 amp. After the current had been 

 cut off and the indicator allowed to cool thoroughly, a run at 

 10.5 amp. gave curve B, the indicator not being reset. 



Owing to differences in the tubes, it is impracticable to print 

 the scales, for each must be graduated by experiment to fit the 

 particular tube to which it is applied. It is usual to determine 

 either four or five points by passing measured currents through 

 the indicator for a sufficient time, and marking on the scale the 

 corresponding heights of the liquid in the index tube i^', the 



