I 
294 Mr Christie’s Observations 
(exclusive of four observations which I had reason to suspect 
were incorrect,) to be 6° 46' ; the greatest value 7° 6' ; and the 
least 6° S3'. This coincidence was, as near as I could possibly 
expect, from the nature of the experiments. 
In three other sets of experiments, the compass was adjusted 
somewhat differently ; and in these the deviations of the hori- 
zontal needle were observed at every 1 0° of longitude, for the 
latitudes 30°, 45°, 60°, having the ball at the distance of 18 
inches from the centre of the needle. These gave the following 
results : 
Distance of centre of ball from centre of needle — 18 inches, 
diameter of ball nr 12.78, weight = 288 lbs. 
Latitude 
of the 
ball. 
Values of s <7, computed from the horizontal deviations. 
Mean. 
Greatest. 
Least, 
cs 
o 
o 
3° 36' 
3° 51' 
3° 25' 
45 
3 55 
4 00 
3 47 
60 
3 18 
3 25 
3 15 
The remarkably near agreement of the several values of s <r 
for the respective latitudes, fully confirmed the views with which 
the experiments were undertaken, and proved clearly, that as 
the ball was carried round the magnetic axis iCw, the horizon- 
tal deviations were such as would arise from the deviations of 
the particles in the line s C n. 
As a further test of the correctness of my hypothesis, I ap- 
plied it to the deviations of the dipping-needle. According to 
the foregoing principles, s <r being the deviation of the particles 
in the line s Cn 9 which is the direction of the 
dipping-needle, if «r <? be drawn perpendicular 
to the meridian, s <S would be the deviation of 
the dipping needle when it is placed in the plane of the meridian, 
since it can then only move in that plane, and <r </ would be the 
deviation of the same needle, when placed in a plane at right an- 
gles to the plane of the meridian : <r *■' being observed, the angle 
r Z <r, or the horizontal deviation, might be computed from it. 
The dipping-needle was placed due west, at the distance of 
18 inches from the centre of the table, and the mean of the de- 
