INTRODUCTION, XVil 
this can only be corrected practically. The magnet being removed, and the brass 
bar inserted (without any directive magnet), the extremities of the arc of vibration 
were observed; the marble slab beneath having radii drawn for every 5°, the de- 
viation of the position of detorsion from the magnetic meridian is obtained, and the 
arms of the torsion-circle turned an equal number of degrees in the opposite direc- 
tion. Much care and time was bestowed on these observations, so as to eliminate 
the torsion as completely as possible. Considerable annoyance was experienced 
from the breaking of the fibres of the first suspension-threads. They were formed 
of 16 fibres, as recommended in the Royal Society’s Report. The necessity of 
removing the declination magnet during the observations of inclination, the diffi- 
culty of holding the thread with a force exactly equal to the weight of the magnet, 
and the consequent liability to strain, or, by loosening, to alter the disposition of 
the fibres, were frequent sources of torsion. 
18. The principal facts relating to the suspension-threads are as follow :— 
May 18. 1841. The suspension-thread was formed of 16 fibres of untwisted 
silk ; the torsion was removed approximately for the observations of adjustment. 
June 29. 1841. One of the fibres found broken; it was removed, and the tor- 
sion eliminated. The brass bar was suspended for several days, and was sometimes 
on one, sometimes on the other side of the magnetic meridian; attributed to mois+ 
ture. 
July 11. 1841. Before this date the torsion had been removed with much care. 
July 20. 1841. Another fibre found broken ; the loose part was removed with- 
out disturbing the thread. No mention is made of the torsion being removed; it 
is, therefore, probable that a considerable torsion existed during the term observa~ 
tions of July 22. 1841. 
July 26. 1841. A new suspension-thread attached. Great care was taken to 
render the tension of each fibre as equal as possible: several threads were rejected 
in succession, until a satisfactory one was obtained. The brass bar was suspended 
for several days, and the torsion occasionally removed. 
Dec. 4—8. 1841. The declination magnet was found resting on the copper 
ring, the suspension-roller being untwisted, and the thread cut at its contact with 
the ring. A new thread was prepared, and the screws of the roller tightened. As 
much care was taken in forming the thread, and removing the torsion, as on July 26. 
Dec. 20, 1841. On Mr RussEL1’s return from Edinburgh, where he had been 
for a short time, the thread was found much stretched, bringing the stirrup into con- 
tact with the copper ring; it was wound up a little, and the plane of detorsion exa- 
mined preparatory to the term-day, Dec. 22, 23. 1841. 
Dec. 23. Noon, Gott. From the uniformity of the declination readings during 
several hours of the term-day, the box was lifted, and the lower part of the stirrup 
found grazing on the copper ring ; the suspension-roller was quite tight; it must, 
therefore, have been due to the stretching of the thread. The observations of decli- 
MAG. AND MET. OBS. VOL. I. é 
