428 BELL SYSTEM TECHNICAL JOURNAL 



National Bureau of Standards type,^ calibrated by the Bureau of 

 Standards. 



Experimental Procedure 



The procedure used in the determination of K, was briefly as follows: 

 Separate unit mica condensers were selected for C and Ci. Each 

 was measured on the capacitance bridge by itself to determine its value 

 in terms of the bridge calibration, and in addition its series resistance 

 was measured by comparison with the air condensers of the bridge, the 

 series resistance of the bridge condensers being eliminated by virtue of 

 the construction of the bridge,^ and the method of making the measure- 

 ment. At the same time the resistance of r and R, high quality 

 resistance units, was determined by the customary Wheatstone bridge 

 method, and their phase angles were measured by comparison with 

 standards of which the phase angle was known to .02 minute at 1,000 

 cycles. The series impedance was then placed in one arm of the ca- 

 pacitance bridge which had previously been balanced at the frequency 

 in question, and the parallel impedance in the other arm. The bridge 

 was then rebalanced by varying slightly the small air condenser in the 

 bridge and the frequency. The change in the bridge air condenser 

 represents an algebraic addition to the capacitance C necessary because 

 the quantities C, Ci, R, and r were not perfectly adjusted to their 

 nominal values and because K is not exactly equal to 1. The change 

 in frequency is necessary for the same reason. The true frequency was 

 then determined by comparison with the laboratory standard by means 

 of the cathode ray oscillograph.^ For some of the determinations the 

 bridge condensers were used for C instead of an external unit, and a 

 shielded six dial resistance box, variable from .01 to 10,000 ohms in- 

 stead of a unit resistance for R. In this case the final balance was 

 obtained by varying the bridge condenser and R instead of the bridge 

 condenser and the frequency. The vacuum tube oscillator used as a 

 frequency source was capable of maintaining a frequency constant to 

 better than .001 per cent for the duration of the tests. Sets of tests 

 were made at three different times with an interval of about a month 

 between them. The tests were made at frequencies of 1,000 and 2,000 

 cycles. 



The results of the determination at each frequency are contained 

 in Tables II and III respectively. 



^ E. B. Rosa, "A New Form of Standard Resistance," Bulletin, Bureau of Stand- 

 ards, Vol. 5, p. 413. 



* F. J. Rasmussen, "Frequency Measurements with the Cathode Ray Oscillo- 

 graph," A. I. E. E. Journal, January, 1927. 



