318 



ME. W. DUDDELL ON THE RESISTANCE AND 



To compare R with the wire, they were adjusted to have practically the same 

 resistance. A current, either direct or alternating, having a R.M.S. value of 

 O'l ampere as indicated by T, was sent through R and the wire in the place of the 

 arc in series, fig. 3. The P.D. between the terminals of R and between the terminals 

 of the arc lamp (i.e., wire) was measured by means of M, in the same way as in the 

 experiments on the arc, the condenser G being short-circuited. The amount by 

 which the impedance of R exceeded the impedance of the arc lamp and wire is 

 tabulated below for the strips 1 to 7 having a resistance of 3 '50 ohms. Each of the 

 results is the mean of at least 1 2 comparisons. The self-induction of the loop formed 

 by the frame of the arc lamp has been found to be about 2'4 X 10~ 7 henry, and the 

 self-induction of the wire itself was calculated to be about 3 X 10~ 8 henry ; so that 

 the total self-induction of the lamp and wire is 27 X 10~ 7 henry. Assuming that 

 the alternator gives a sine-wave form, which was approximately the case, and 

 allowing for this self-induction and for the difference in resistance 0'03 per cent. 

 between the lamp and R, the true power-factor, cos 77 of R, has been calculated, 

 from which TJ the lag of the current in R behind the P.D. and the time constant 

 have been deduced. 



TABLE III. --Test of R, Resistance 3'50 ohms. 



Frequency -f- per second. 



(direct current) 

 32,400 

 41,000 

 50,700 

 60,000 

 80,000 



100,000 

 120,000 



Method o/" Experiment. 



The arc length, the direct current, and the frequency of the alternating testing 

 current having been decided upon for any experiment, the test was carried out in 

 the following manner. The carbons having burnt into shape corresponding with the 

 required current and length, a rough experiment on the impedance of the arc was 

 made, and the value of R adjusted by moving 3 (fig. 3) to that contact on R which 

 made V A and V K most nearly equal, as this gives the greatest accuracy in the power- 

 factor by the 3 -voltmeter method. 



The main direct current was now interrupted and the arc short-circuited by pressing 



