Whitehead and Hill — Measurement of Self -Inductance. 151 



no value owing to the electrostatic capacity of the doubly 

 wound coils. Except in the case of a few determinations with 

 Method 25, all the resistances used were made of a special 

 German silver resistance wire, B. and S. gauge, Nos. 30 and 

 27. This wire was not doubled, but wound continuously in 

 one layer on slates or pieces of micanite, each slate containing 

 about 2000 ohms conveniently subdivided. The self-induct- 

 ance and capacity of such resistances is entirely negligible. 

 For fine adjustment a small resistance box was constructed 

 differing from an ordinary box in that its coils were wound 

 flat on thin pieces of micanite. 



Coils and Condenser. — The coils whose inductance were 

 measured were as follows : 



F, external diameter 33*46 cm , internal diameter 23*8 0m , was 

 made of about 1200 turns of "No. 16, B. and S. gauge single 

 cotton-covered copper wire. Self-inductance as determined, 

 *5729 henry. Resistance, 36*3 ohms. 



C. Same dimensions as F except depth. It consisted of 

 1747 turns of No. 22 B. and S. gauge single cotton-covered cop- 

 per wire. Self -inductance found to be 1*3025 henry. Resist- 

 ance, 78 ohms. 



S. External diameter 23*5 cm3 , internal diameter 15 cms , depth 

 about 3*5 cms . It consisted of 2082 turns of No. 22 B. and S. 

 single cotton-covered copper wire. Resistance, about 69 ohms. 

 Self -inductance found to be 1*0331 henrys. 



Condenser. — A ^-microfarad mica condenser. 



Conditions of sensibility. — The deflection of an electrodyna- 

 mometer is proportional to the product of the currents flowing 

 in the fixed and hanging coils multiplied by the cosine of the 

 phase difference of the two currents. In most of these Row- 

 land methods the adjustment consists in so altering non-induc- 

 tive resistances in one or more branches of the bridge that there 

 is a ninety degree phase difference between the currents in the 

 fixed and hanging coils respectively. Evidently in order to 

 secure maximum sensibility the currents through the dynamo- 

 meter coils should be as heavy as possible, heating alone being 

 the limit. 



Sources of Error. — In work with alternating currents great 

 care must be exercised that one part of the network does not 

 exert an inductive action on another, e. g., that the coil whose 

 inductance is being measured does not affect the hanging coil 

 of the electrodynamometer. Induction was carefully guarded 

 against by the arrangement and tested for by means of revers- 

 ing switches. The electrostatic capacity of doubly wound coils 

 and resistances has already been mentioned as well as the pre- 

 cautions taken to avoid it. In the experiments here described 

 all the connections were made as short as possible and no wires 

 were twisted. Heating of the resistances was avoided as much 



