ii7 
in the direction of the compass needles, &c. 
those of the magnetic meridian. Considering, however, that 
this must depend altogether on the distribution of iron, and 
may be therefore liable to great diversity, he recommends, 
that, in every ship, as soon as a fixed position has been selected 
for a compass, the points of no error should be determined by 
repeated observation. The method that was adopted for this 
purpose in the late voyage appearing both simple and effectual, 
it may be useful to exemplify it by an instance or two. 
The Isabella being at anchor in Brassa Sound, Shetland, her 
head was placed, by means of warps, on each point of the 
compass successively, and the bearing of a pile of stones 
on the summit of a distant hill noted by her compass at each 
point ; at the same time that these observations were made 
on board, her bearing from the hill was also observed by a 
compass placed on the pile of stones ; the agreement in bear- 
ing showed the points of no error, and the differences the 
errors in each point, without the calculation which azimuths 
involve. fSee PL X. fig. i.[J 
The Alexander being along side a floe of ice in Baffin's 
Bay, the true magnetic bearing from the ship, of a very 
distant and well defined object on the main land, was found 
by carrying a compass on the ice in an opposite direction, to 
a distance which insured its being perfectly free from local 
influence. The ship's head being then warped round to each 
point of the compass successively, the errors in each were 
determined by the difference in bearing, as in the last 
instance. [See PL X. fig. a.] 
The regularity in the above results is the best testimony 
that the method is a satisfactory one. Certain precautions 
must be attended to : thus, the object must be sufficiently 
