LOCAL ATTRACTION OR THE BAFFIN. 91 
at least parallel, 1 now observed, by the compass 
in the crow’s-nest, the course on which the ship 
lay, merely by taking the bearing of the fore-top¬ 
gallant-mast, and noted at the same time the ex¬ 
act course by the binnacle-compass, as read off by 
the mate and surgeon. Then, putting the ship 
successively on every point, as far round as the 
wind would permit, I continued to observe the 
bearings of the fore-topgallant-mast, and to com¬ 
pare them with the course by the binnacle-com¬ 
pass. These observations afforded data for the 
deviation on twenty points out of thirty-two. 
Fortunately the wind shifted during the opera¬ 
tions, so that with the aid of a few observations 
taken “ in stays,” the remaining twelve points 
were likewise determined. On many courses 
two, three, or more observations were taken. The 
mean results of the whole are contained in the 
following Table *. 
* In this Table, the signs + and — are used in the se¬ 
cond column, to indicate that the errors (or corrections) are 
to be added to, or subtracted from, the westerly variation; 
the sum or difference obtained on any course being the cor¬ 
rection for reducing the magnetic bearings taken on that 
course to the True North and South. In the third column, 
the letters N, S, E, and W, are used to point out the direc¬ 
tion (whether towards the north, or south, or east, or west) 
in which the correction of “ approximate errors” is to be ap¬ 
plied to the course by the binnacle-compass, for giving the 
true course. 
