238 
MESSRS. R. T. GLAZEBROOK AND J. M. DODDS ON THE 
of the telescope, supposing, as in our case, the telescope is placed so as to look just 
over the edge of the scale. 
In making these adjustments the scale and telescope were first fixed so that division 
250 —the middle of the scale—was vertically below the centre of the object-glass of 
the- telescope ; the galvanometer was placed in position and levelled, and the telescope 
adjusted to view the image of the scale in the mirror. 
Then telescope and scale were both moved until the division of the scale which 
coincided with the cross-wire was close to 250. When this was the case the normal 
to the mirror and the axis of the telescope were nearly in the same vertical plane. 
The scale was then turned in a horizontal plane until its two ends, equidistant from 
division 250, were also equidistant from the mirror, taking care at the same time that 
the image of division 250, the centre of the scale, remained close to the cross-wire of 
the telescope. 
The scale was thus put at right angles to the normal to the mirror. 
A long bar magnet resting on a pivot at its centre was then supported close to the 
scale, and it was found that the scale was very nearly parallel to the axis of the 
magnet, the error was certainly not more than 20'. 
Thus the scale has been set very approximately north and south, and since the 
mirror is very nearly parallel to the scale, it is also nearly parallel to the axis of the 
galvanometer needle. Hence, if we set the coils parallel to the scale or mirror they 
will be very nearly north and south, and their plane will be approximately parallel to 
the axis of the needle. 
A piece of plate glass can be screwed on to the galvanometer in a position very 
nearly parallel to the coils. This was done, and the reflection of a lamp placed just 
below the telescope was observed, the galvanometer coils were turned until this 
reflected image was seen in the centre of the field of the telescope. Thus the galva¬ 
nometer coils were placed very nearly north and south. The reading of the vernier 
attached to the galvanometer was noted, and by means of it the coils could readily be 
brought back to the same position, or placed at any required angle to the meridian. 
The adjustments thus described were of course only approximate, but it is easy to 
show that the method of experiment eliminates any small outstanding error. 
Let us suppose the coils are inclined at an angle a to the meridian. The effect of 
this is merely to change G into G cos a in both the equations (1) and (2), and cos a 
disappears from the resulting equation. It is better, however, that the field of force 
produced by the current in the galvanometer coil should be as nearly as possible 
uniform throughout the space through which the needle moves in the throw and 
deflection respectively. This condition is best satisfied if the needle when in equilibrium 
is parallel to the coils. 
Let us now suppose that the scale is inclined at an angle y to the plane of the 
mirror. Let and p. 2 be the scale Amines of the throw to the right and left of the 
resting point. 
