OF FERRUGINOUS BODIES. 231 
two inches long, was placed on the top of the tall 
tile, nearly on a level with the upper end of the 
iron tube, and six inches distant on the east side 
of it. 
56. Prior to placing the iron tube in this posi- 
tion, and in the absence of all local magnetic 
action, the meridian line of the compass card was 
adjusted to the magnetic meridian of the place 
where the experiments were carried on. 
57. When the needle had come to rest, the 
iron tube, whilst cold, was placed in its proper 
position (55), and the deflection of the needle, 
due to the magnetic action of the upper end, 
noted. The tube was then inverted, and again 
the deflection noted. By these means the mag- 
netic action exercised on the needle, by the 
different ends of the hollow bar, prior to heating 
it, was ascertained; and the mean action of both 
ends easily determined. 
58. The tube was next heated to whiteness, (not 
to a welding heat), and in that condition replaced 
in its former position (55). The needle was care- 
fully watched, and its indications strictly noted. 
In the first trials the needle was much agitated 
