332 
GREENLAND VOYAGE. 
determining the bcaringsof the land,orthe azimutli 
of the sun, were taken at the mast-head, because 
this was the only part of the ship where compass- 
bearings could be relied on. In every other part 
of the ship, indeed, that could be conveniently 
resorted to, there was so much “ local-attraction,” 
or “ deviation,” that observations taken therein, 
with the magnetic needle, were useless. During 
boisterous weather, when the mast was so un¬ 
steady as to prevent the employment of the azi¬ 
muth-compass in the crow’s T nest, all required an¬ 
gles were taken with a sextant, and the bearings 
derived from them, by connecting one of the se¬ 
ries with the angular distance of the sun. The 
longitudes were chiefly determined by the chro¬ 
nometer, its rate being corrected by subsequent 
observations made on known coasts,—by positions 
formerly determined, — by lunar observation,— 
and by a comparison with another chronometer. 
Only two opportunities for obtaining lunar obser¬ 
vations occurred during the voyage. One of 
these was taken advantage of; but, on the other 
occasion, I was unfortunately on shore at the time 
the moon appeared, which set before I returned 
on board, and never rose again during the luna¬ 
tion. Sometimes, when the sun was invisible, 
both latitude and longitude were determined by 
the bearings of two objects, the correct positions 
