106 
GREENLAND VOYAGE. 
and altitudes, I obtained the mean longitude of 
17° 54' 30" W., for the place of the ship on the 
14th of June, and found the error of the chrono¬ 
meter to be nearly four minutes of time *. 
These satisfactory observations for the longi¬ 
tude (established by many subsequent proofs), ena¬ 
bled me to ascertain the exact effect, in a particu¬ 
lar case, of the extraordinary refractive property of 
the atmosphere in the Arctic Seas, which, with¬ 
out such proofs, would scarcely have been credi¬ 
ble. The coast that has just been described, is 
in general so bold, as to be distinctly visible in 
the ordinary state of the atmosphere, at the dis¬ 
tance of sixty miles; but on my last voyage into 
these regions, one part of this coast was seen, 
when at more than double this distance. The 
particulars w r erc these:—Towards the end of July 
1821, being among the ice in latitude 74° 10', 
and longitude, by lunar observation and chrono¬ 
meter, (which agreed to twenty-two minutes of 
longitude, or within six geographical miles), 12° 
* This lunar observation was afterwards proved, by com¬ 
paring my chronometer with one of Captain Bennet’s, and 
by correcting its rate by subsequent observations. All the 
longitudes mentioned, therefore, in this narrative, are cor¬ 
rected longitudes, and not exactly those given by the chro¬ 
nometer, at its original rate, wliich proved to be nearly two 
seconds per day wrong. 
