of Alloys of Bismuth and Tin. 53d 
exerted on the cylinder, the equilibrium of the balance was 
disturbed. In the space of one or two minutes the balance 
again came approximately to rest, and the new equilibrium 
position was observed. ‘The circuit was then broken and a 
second observation taken of the zero position when no current 
passed in the coil. By subtracting from the second reading 
the mean of the first and third, the deflexion produced by the 
current in the coil was obtained. The current was then sent 
through the coil in the opposite direction and observations 
taken as before, and so on. 
As an example of such observations, the following data 
were obtained with the bismuth cylinder when a current of 6 
amperes was sent through the coil : — 
Seale-reading when current is:— | 
: | Deflexion due to | 
oe = <a oh current of | 
(A) zero. | (B) 6 amperes. praraperes: 
| +65 
| — 59°D 120 
56 
| 64 TS 
53 
| 68 119:5 
50 
70 119 
48 
In the same way the balance was standardized from time 
to time by observing the deflexion of the scale seen through 
the telescope produced by the motion of the rider through a 
known length of the beam of the balance. 
The force acting on the rod due to the current in the coil 
was then calculated on the assumption that this was pro- 
portional to the deflexion produced by it. As far as the 
graduations of the balance-beam would allow—that is, down 
to forces of 4 mgm. weight—this assumption was verified 
experimentally, and as the smallest forces measured were not 
less than 35 mgm. weight, there is no reason for doubting 
the truth of this assumption. 
Hxperiments were performed with currents in the coil 
varying from 1°4 to 10°2 amperes, the sensitiveness of the 
balance being adjusted as far as possible so as to give a 
convenient deflexion for each value of the current. 
Having in this way obtained the absolute value of the 
force acting on the rod, this quantity is substituted in Boltz- 
mann’s expression referred to above (p. 49), and the value 
