EXPERIMENTS AT NORTHFLEET. 
CXXI 
It was evident from these experiments that the common centre of attrac- 
tion of the ship’s iron was forward and very nearly a-midships ; and that conse- 
quently, when the ship’s head was north or south by the compass, the direc- 
tion of the earth’s magnetism and of the local attraction coinciding, the- 
compass indicated the true magnetic bearing of objects. 
The true variation of the Needle, therefore, could be at all times ascertained 
by azimuths observed with the ship’s head on either of those points ; when 
the error with which the result might be affected from local attraction 
might be reasonably expected not to exceed the other incidental errors to 
which such observations are necessarily liable. 
It was also seen that the greatest deviation took place when the ship’s 
head was on the east or west points, or in other words, when the two 
attractions were at right angles to each other; and that it diminished 
regularly as the direction of the local attraction approached that of the 
magnetic meridian. 
It may be remarked indeed of these, as it has been of similar observations 
made in the former voyage, that the errors on the points equidistant from 
north or south (or from the line of no error) do not in all instances exactly 
correspond ; the discordancies are, however, within the limit of inaccuracies 
arising from accidental causes ; such as the motion of the ship ; the difficulty 
of retaining her head steadily and precisely on the required point ; the 
imperfection of the compass itself, and especially of its graduation which 
is rarely carried to a smaller division than 20 minutes. These causes 
are sufficient to defeat an attempt at precision in compass bearings observed 
from ships, without including the parallax on warping the vessel round, 
which ought to be prevented by the distance of the selected object. 
With the knowledge thus acquired of the points of greatest deviation, and 
of those on which none took place, and with occasional azimuth observations 
from time to time, as opportunities presented themselves, the irregularities 
in the direction of the compass proceeding from the ship’s iron, occasioned 
no other practical inconvenience in her navigation, than a little additional 
trouble in computing the day’s works. 
Whenever it could be done, the variation on a particular course steered 
was ascertained by actual observation ; but where the courses were many 
in the 24 hours, one set of azimuths with the ship’s head on the north or 
south points shewing the true variation, and a second set, with her head 
either on the east or west points shewing the maximum of disturbance, were 
