90 
GREENLAND VOYAGE. 
which is fixed on the very summit of the main- 
topgallant-mast, is just such a situation. No iron 
whatever was employed in its construction, and 
not any of this metal was to be found within fif¬ 
teen feet of it; the iron that then occurs about the 
mast, being of little bulk, and in a position directly 
belowthe crow’s-nest, was considered to be incapable 
of producing any sensible influence on a compass 
placed directly above it. Hence, it was presumed, 
that, in a compass so situated, there could be no 
deviation. This being the case, all that was ne¬ 
cessary for determining the Baffin’s deviation, on 
any one course, was merely to compare the direc¬ 
tion of the ship’s head by the mast-head compass, 
with that pointed out by the binnacle compass. 
The difference of the two was the deviation on 
that course. But for determining the deviation 
on each point, I proceeded as follows:—The bin¬ 
nacle-compass was removed, and an azimuth-com¬ 
pass substituted in its place: an azimuth-compass 
on Captain Kater’s principle, furnished to me 
by the Board of Longitude, was also substituted 
for the little compass I usually carried in a box 
fixed on the topgallant-mast-head. With these 
instruments I could make the observations more 
accurately than with the ordinary compasses. 
Having previously ascertained that the main¬ 
mast and fore-mast of the Baffin were upright, or 
