horizontal needle when under the influence of magnets. 355 
nearer to the needle, their nearest ends being at the distance 
of 9.6 inches from the centre ; and they were carefully ad- 
justed in the magnetic axis, or line of the dip, so that the north 
end of the needle pointed very accurately to the south, or 
180 0 of the instrument. In this position I vibrated the needle, 
as before, and found that it made ten vibrations in one hun- 
dred and five seconds. From this it appears that the force 
by which the needle was now directed towards the south, 
was to the undiminished terrestrial force towards the north, 
as .0625 to 1 ; and was therefore rather less than the force 
towards the north in the observations from the 6th to the 
12th of April. The instrument was placed near the outer 
wall in a room having a single window facing N 40°E; the 
only iron in the room being a large lock to the door and the 
weights to the windows, which were always in the same po- 
sition when the observations were made. 
The observations were begun on the 21st of April, and 
continued to the 27th, when they were unavoidably inter- 
rupted. Whilst making them, after the first two days, I per- 
ceived a gradual increase in the deviation easterly, but as I 
could only attribute this to a small change taking place by 
degrees in some part of the instrument, any correction of it 
must have been in a great measure arbitrary. Had I reduced 
the observations each day, by considering the morning and 
evening nearly stationary points as zero, it would have had 
the effect of separating the evening observations of one day 
from the morning ones of the next : I therefore preferred 
giving the observations as they were made, although on the 
last day the accumulated increase amounted to 2°. 
