II2 
BULLETIN OF THE LABORATORIES 
three, put together in the form of a cross, with the middle of the short 
board about forty inches from one end of the long board. The boards 
should be carefully halved together and be exactly at right angles to 
one another. This cross is mounted on a stand so that it' can be easily 
turned in azimuth, and so that the longer part stands nearly east and 
west. The magnetometer is now placed in the middle of the short 
arm and adjusted so that the middle line of the short arm and the mid- 
dle line of the long arm of the cross, exactly at right angles to the 
former line, intersect directly under the centre of the needle. A scale 
and lamp should now be arranged in front of the magnetometer, with 
the scale supported by the longest arm directly in front of the magnet- 
ometer, parallel with the line on the short arm and distant one hundred 
centimetres from the needle. The middle of the scale should be di- 
rectly over the middle line of its arm. The lamp is placed in front of 
a hole cut in the scale support directly under the centre of the scale. 
The light from the lamp is thus reflected back upon the scale. Now 
turn the cross till the spot of reflected light falls directly on the centre 
mark of the scale. The instrument is now in its proper position 
for use. 
I find that with a plane mirror I can get the best spot of light by 
using two lenses, one of long focal length, placed close to the face of 
the mirror and one of comparatively short focal length placed close to 
the aperture in the scale support. To obtain a dark line across the 
spot of light, I place a very fine wire between the two lenses and move 
it back and forth till the mark is sharply defined across the spot. 
The deflecting bar magnet is now placed, end on, in the east and 
west line, with its centre forty centimetres distant from the needle and 
the deflection of the spot of light noted. Then the magnet is placed 
on the opposite side at same distance and deflection noted. If all the 
adjustments are perfect, these two deflections will be equal ; but the 
mean of them will give nearly enough the true deflection. All the 
measurements being made with the same unit, the number of scale di- 
visions divided by distance of scale from mirror gives tan 2 a; where a 
is the angle of deflection of the mirror. 
If r is the distance of centre of magnet from the mirror, 1 the 
half length of the magnet, m the magnetic moment of the mag- 
m (r2 12 )2 
net, H the horizontal component, then = - — Tana(i) 
The magnet may also be placed in the north and south line, side 
