Alternating Currents to Calibration of Capacitii-hoxes. Ill 



The following results were obtained by this modification of 

 Anderson's method for a copper coil wdiich we have been 

 using throughout these experiments. 



Table V. 





0. 



r. 



L. 





•1x10-6 farads 



17000 



•442 henries 



S=Q=1000, and P = 130 



•2x10-6 , 



8300 



•441 „ 



B.A. ohms, throughout. 



•3x10-6 „ 



5340 



•440 „ 



C is the nominal value of the 



•4x10-6 „ 



3900 



•440 „ 



capacity. 



•5x10-6 „ 



2960 



•439 „ 



The value of L is given cor- 



•6x10-6 „ 



2400 



•440 „ 



rected for the true value of 



•7x10-6 „ 



1990 



•441 „ 



C and for S, Q, r, and P 



■8x10-6 „ 



1680 



•441 „ 



being in B.A. ohms. 



•9x10-6 „ 



1440 



•440 „ 





1-0x10-6 „ 



1240 



•439 „ 



The sensitiveness of adjustment is about 1 in 250, or 

 •4 per cent., in these experiments. 



In what follows it is necessary to remember that we are no 

 longer using a sensitive high-resistance galvanometer, but a 

 moving coil instrument of from 30 to 500 ohms' resistance. 

 On the other hand, we are using a much higher pressure, 

 viz. 100 volts., of which from 10 to 50 per cent, is applied to 

 the high-resistance coils forming the bridge. This pressure 

 is only applied for a very small fraction of a second at a time 

 during the act of testing. It is of course necessary at first 

 to balance the bridge for direct currents as before, and for 

 this purpose the magnetic field in the galvanometer must be 

 generated by a direct current. Thus it is necessary to supply 

 both direct and alternating currents to both the bridge and 

 the field-magnet coils. Arrangements were made as shown 

 diagrammatically in fig. 2. 



A series of experiments were performed with instrument A, 

 the sensitiveness of adjustment being approximately 1 in 4000, 

 or '025 per cent. 



A second series of experiments were performed, using the 

 coil B on the new instrument, the results of which are given 

 below in Table VI. (p. 718). The sensitiveness here was 

 about 1 in 8000, or •012 per cent. 



Measurement of very small Inductances. — The foregoing 

 experiments show that it is possible to measure an inductance 

 with certainty to something well under ^^ of a millihenry. 



Phil, Ma.g, S. 6. Vol. 6. No. 36. Dec, 1903. 3 B 



