150 Whitehead and Hill — Measurement of Self -Inductance. 



motor. As long as the load on the dynamo was not varied the 

 speed remained very constant. Greater constancy of speed 

 was secured by operating the motor from storage batteries. 

 The voltage furnished by the alternator could be controlled by 

 slightly altering the strength of the held. For the determina- 

 tion of the frequency a chronograph was used. A kind of 

 speed counter was connected to the shaft of the dynamo and a 

 contact so arranged that in every ninety-seven revolutions of the 

 dynamo armature a circuit was closed and a record made on 

 the chronograph sheet. The chronograph was set up close to 

 the rest of the apparatus and by means of a thread the pen 

 could be raised for an instant from the sheet and so the exact 

 time of making an adjustment could be readily noted. By 

 interchanging the gear wheels the cylinder of the chronograph 

 was made to run at about three and one-half times the custo- 

 mary speed. The distance between the checks indicating the 

 seconds was so increased to over five and one-half centimeters. 

 A scale similar to that shown in fig. 1 was then constructed, 

 consisting of fine lines scratched on 

 transparent celluloid and blackened 

 with India ink. By means of this scale 

 it was possible to read the chronograph 

 sheet directly to fiftieths of a second 

 and to estimate to five hundredths. 

 However, in reading a chronograph 

 sheet as exactly as this, certain precau- 

 tions must be taken. First, the checks 

 indicating the seconds should be sharp 

 and distinct. Secondly, an error may 

 arise if the chronograph does not run 

 quite regularly. This can generally be 

 remedied by properly adjusting the 

 governor. By this method the fre- 

 quency could be quite accurately determined, to 1/10 per cent 

 or closer, if the sheet was read with sufficient care. It is to 

 be noted that if sidereal seconds are recorded on the chrono- 

 graph sheet they should be reduced to solar seconds, otherwise 

 an error of 3/10 per cent is introduced in the absolute value 

 of the frequency and, therefore, of 6/10 per cent in the value 

 of self -inductance. 



Apparatus. 



Hlectrodynamometer. — A Rowland electrodynamometer was 

 used. The self-inductance of the fixed coils was -0165 henry, 

 and of the hanging coil *0007 henry. 



Resistances. — In the measurements of self -inductance w T ith 

 alternating currents, ordinary resistance boxes are of little or 



