96 



AUSTIN: HIGH FEEQUENCY RESISTANCES 



from 20° to 30° above the temperature of the room. The uncer- 

 tainty of the individual high frequency current readings is approx- 

 imately one part in fifty. The mean of thirty or more readings 

 is taken as a basis for each calculation. The calorimeters pro- 

 vided with motor driven stirrers are of glass 15 cm. high and 10.8 

 cm. in diameter and contain sufficient petroleum to cover the 

 coils under experiment. The coils of 0.40 mm. double silk-covered 



copper wire are wound on glass, and the principal constants are 

 given in Table I and their resistances in Table II. 



After the determination of the resistance of the six pairs of 

 standard coils a roller inductance of the Fessenden type was cali- 

 brated by comparison with the standard coils. The method 

 used is as follows: A buzzer-driven wave meter is used to excite 

 an oscillatory circuit containing a sensitive thermoelement and 

 variable air condenser. By means of switches either one of the 

 standard coils or the variable inductance can be inserted in this 

 circuit, the deflections in each case being observed on the galva- 

 nometer of the thermoelement. Sufficient fine wire resistance is 



