394 Dr. W. Watson on a Quartz-Thread 
forms of. apparatus suitable for recording the variations of 
the declination and horizontal component. The design of a 
satisfactory form of instrument to record the rapid variations 
of the vertical component presents, however, some very con- 
siderable difficulties. Thus, if the attempt is made to reduce 
to any great extent the moment of inertia, and hence the 
mass, of the balanced magnet in the ordinary form of Lloyd’s 
balance, it is found that irregularities are immediately intro- 
duced owing to imperfections or dirt interfering with the 
free movements of the knife-edge. Further, the slight 
tremors to which the base ef the instrumentis almost always 
subject causes the knife-edge to slip about on the plane, and 
thus the azimuth of the magnet varies, and there is difficulty 
in obtaining a satisfactory record. 
The above considerations led the writer, about two years 
ago, to try various other forms of Lloyd’s balance; and the 
following paper contains a description of the instrument finally 
arrived at, and which has been actually run for some con- 
siderable time. 
In principle the instrument resembles the quartz-thread 
gravity-balance designed by Prot, Threlfall*. In addition to 
the advantages to be expected from this form of balance due 
to the suppression of the knife-edge, it promised to allow of 
a very simple method of compensating for the effects of 
changes of temperature. This is of considerable importance, 
for many vertical-force magnetographs might almost be used 
more efficiently as thermographs than as instruments for 
recording changes in the vertical component. 
The principle of the instrument is to have a magnet NS (fig. 5, 
Pl. XXII.) suspended on a horizontal quartz fibre AB, the 
fibre being kept stretched by means of a spring. The centre 
of gravity of the magnet and the torsion of the fibre are so 
adjusted that the axis of the magnet is horizontal. In order 
to see how the temperature compensation can be effected, let 
us suppose that the axis of the supporting fibre passes through 
the centre of gravity of the magnet, so that in the northern 
hemisphere the fibre has to be twisted, say by turning the 
end B, in the clockwise direction as seen when looking 
along the fibre from Btowards A. If now the temperature 
rises two effects will be produced. In the first place, the 
magnetic moment of the magnet will decrease, and hence the 
couple acting on the magnet, due to the vertical component 
of the earth’s magnetic field, will decrease. The result will 
be that the north end of the magnet will rise. Secondly, 
owing to the fact that the torsional rigidity of fused silica 
* Trans. Royal Society, excili. p. 215 (1899). 
