174 
LINES OF MAGNETIC DECLINATION IN THE ATLANTIC. 
the Declination in 1787, published by M. Hansteen in his great work, Magnetismus 
der Erde, and republished by myself in this country in the Reports of the British 
Association for the Advancement of Science for 1834. Of the determinations made 
at sea, all have been corrected for the effect of the ship’s iron when observations on 
which the corrections must depend have been provided. I have discussed those 
corrections in more detail and at greater length than I might otherwise have done, 
on account of the practical importance attaching to this part of the subject since the 
introduction of steam navigation and the increased employment of iron in the con- 
struction and equipment of vessels ; and in the hope and belief that the discussion 
may be found to have a practical as well as a theoretical value. 
The 1480 determinations reduced and corrected for epoch are arranged in a general 
Table in Zones, each zone including 10° of latitude; the determinations comprised 
in each zone are arranged in the order of their longitudes, commencing always with 
the most westerly : they are all likewise inserted in the map, where they are expressed 
in degrees and decimals of a degree. For the purpose of drawing the lines of decli- 
nation in general conformity with the determinations, the latter have been arranged 
in groups, each group having its mean geographical position at or near the point of 
intei’section of every fifth meridian and parallel, (as far as the observations would 
permit,) counting from the parallel of 0°, and the meridian of 280°; and in the moi’e 
frequented parts, and where consequently the number of determinations was greatest, 
at or near the points of intersection of parallels and meridians distant only 2^ degrees 
from each other. Each group contains all the determinations comprised within 
equal distances of latitude and equal distances of longitude on either side of the point 
of intersection ; the distances being so taken that the number of determinations con- 
stituting a group should be generally from ten to twenty. 
If the mean geographical position corresponding to the determinations in a group 
diflfered more than a few minutes from the latitude or longitude of the desired point 
of intersection, one or two determinations adjacent to, but beyond the limit, were 
taken into the group, or one or two pairs of determinations within the limits were 
combined and their mean taken instead of the separate results. This was done for 
the purpose of diminishing the amount of the correction to be applied to the mean 
declination of the group, to reduce it to the corresponding value of the declination 
at the point of intersection itself. The factors employed in making that reduction 
were derived from the map itself. The values of the declination thus obtained at 
the intersection of every fifth degree of latitude and longitude (as far as the determi- 
nations permitted), and at the intersection of every 2^ degrees in the more frequented 
quarters, have been regarded as elements of the declination lines ; and these lines 
have been drawn in accordance with the elements with only such slight deviations 
as were indispensable to preserve an interconformity between the lines, where it was 
evident that the determinations themselves were slightly discordant. The values of 
the Declination at the points of intersection obtained by the process of grouping are 
