92 ~Mr. W. Duddell: Jnstruments for the Measurement 
through one coil reacts on the same current flowing through 
anotuer coil, no magnetic material of any kind entering into 
the instrument. (b) Instruments containing iron either as a 
core to the two coils, or as the part reacted on by the mag- 
netic force of onecoil. In all these instruments the essential 
part is a coil or coils carrying the current to be measured 
which produce magnetic fields; and it is in the design of 
these coils that the main difficulties are encountered. If it is 
required to measure a small current, then the coils must 
be wound with fine wire and contain many turns, so that 
both their resistance and self-induction are high, which in 
many experiments is very undesirable, if not absolutely 
prohibitive. 
I came across a very good example of the importance of the 
self-induction of dynamometers for measuring small currents 
in connexion with some experiments in which I wanted to 
measure Q'l amp. at a frequency of 100,000 ~ per second, 
A small reflecting dynamometer was available having 8 turns 
in fixed and 40 turns in moving coil, which gave about 100 
scale-divisions deflexion for 0-1 ampere. On attempting to 
use it no appreciable deflexion could be obtained, although 
everything appeared to be working satisfactorily ; the trouble 
was finally traced to the dynamometer behaving like an 
insulator owing to the great importance of its self-induction 
at this high frequency. Another effect which may occur at 
high frequencies with coils of many turns and layers, is that 
the capacity effect from wire to wire and layer to layer may 
shunt some of the current which should pass round the 
instrument, and so destroy any accuracy in the reading. In 
fact, given a good instrument, the evaluation of high- 
frequency alternating currents is of great difficulty owing to 
the errors introduced by the self-induction and capacity of the 
leads, resistances, &c. which may be used in the measuring 
circuit. 
if the resistance and self-induction of the instrument are of 
no consequence and the frequency is not high, then about 
0-001 ampere can be accurately measured by means of a 
dynamometer without much difficulty. 
The measurement of large alternating currents by electro- 
magnetic means offers quite another set of problems. To 
carry large currents large conductors are necessary. which 
have to be carefully stranded so as to avoid skin-effects. 
This stranding presents many serious practical difficulties, as 
in order to make the current divide equally between the 
strands it is necessary to make the self-induction and resist- 
ance of each strand the same, and also the mutual induction 
