188 
Mr Searle, An experiment with a pair 
Let AG = r-y, BC = n, AD = r 3 and BD = r 4 and let p 1} p 2 , jh 
and p 4 be the lengths of perpendiculars from 0 upon r u r 2 , r 3 and 
i\. Also let the perpendiculars from A and B upon A 0 R 0 be h 4 
and h 2 . Let the pole strength of CD be m and let that of AB 
be m'. 
The force which G exerts on A is mm jr-c dynes along GA away 
from G, and its moment about 0 in the clockwise direction is 
— mm'pjr* dyne-cm. in the case illustrated in Fig. 4. Treating 
the other three forces in the same way and taking account of the 
signs of their moments about 0, we find that, in the case of Fig. 4, 
the resultant clockwise moment is 
If H be the horizontal component of the earth’s magnetic force, 
the counter-clockwise moment due to the action of the earth is 
m'H (hj + h 2 ). Equating these moments, we have 
From this expression m can be found in terms of H when the four 
distances r x ... and the six perpendiculars h 1} h 2 , p 1 ... have been 
measured. 
The value of m deduced from (1) may be compared with that 
obtained from the readings of an ordinary magnetometer, having 
a short magnet and provided with a pointer moving over a 
graduated circle. The instrument should be placed in approxi- 
mately the same position in the laboratory as the Robison 
magnetometer so as to avoid errors due to local variations of H. 
If the Robison magnet GD produce a deflexion 6, when placed 
in the “ end on ” position, and if R 4 and R 2 be the distances of the 
centres of its spheres from the centre of the magnetometer, 
The usual precautions should be observed. Thus the magnet 
should be placed both East and West of the magnetometer and 
should be turned end for end in each of these positions, while both 
ends of the pointer should be observed on each occasion. The 
mean value of the angle deduced from the eight readings is 
used in (2). 
The axis of the Robison magnet should pass through the axis 
of suspension of the magnetometer magnet. If this adjustment 
be not secured, the deflexions will differ considerably among 
themselves. 
( 1 ). 
