8-2 
MR. AIRY ON THE OBSERVED DEVIATIONS OF THE COMPASS. 
after making half the voyage, the errors perhaps would have been only half as great ; 
but these errors would have been equally gratuitous. 
In other instances, such as that of the Simoom, in which the change of subper- 
manent magnetism is real and unusually great, the tabular method (supposing, for 
illustration, that there had been no opportunity of sufficiently investigating the 
errors during the whole voyage) would have united the gratuitous errors with the 
errors produced by the real change, and would have produced at the Cape of Good 
Hope an error of 11°; whereas, if the correction by magnets had been used, the 
error would have been under 6°. At intermediate places, as the neighbourhood of 
St. Helena, where the earth’s directive force differs still more from that in England, 
the gratuitous error would have been much greater, and the error really depending 
on change in the ship would probably have been less, as occurring in an earlier part 
of the voyage. 
The mere comparison of magnitudes of errors, however, in this way, does not 
sufficiently exhibit the disadvantage of the method of ‘^Tables of Deviations.” It is 
an important defect that no good new table can be formed, without observations for 
the error on numerous points of azimuth ; whereas the operations for readjustment 
of magnets require observations on only two points of azimuth. And, I apprehend, 
that the necessity of using a table at all (that is, of steering by one nominal course 
when another course is intended) is, especially in difficult channels, a very serious 
evil, from which the method of steering by a corrected compass is entirely free. 
11. I have alluded above to the possible changes in the energy of the correcting 
magnets ; but I am bound to state that these changes (when ordinary care is taken 
for the conservation of the magnets) are, to the best of ray knowledge, extremely 
minute. It is known, as a matter of experience, that the diminution of the subper- 
manent magnetism of a new iron ship, though small, is usually greater than that of 
the magnets ; inasmuch as it usually becomes necessary to increase the distance of 
the magnets from the compass. 
I subjoin the Table of Polar Magnet Deviations, which has been used in the pre- 
ceding investigations, and which may perhaps be useful, in future, for similar investi- 
gations. 
