204 
LINES OF MAGNETIC DECLINATION IN THE ATLANTIC. 
Comparison with M. Gauss's Theory . — As many persons may be desirous of seeing 
the result of a comparison of the Declinations computed by M. Gauss’s general 
theory, with those derived from observation over a field of considerable extent, I 
have subjoined in Table XI. in degrees and decimals of a degree, the differences of 
Declination at the intersections of every five degrees of latitude and longitude. 
The sign — signifies that the theoretical Declinations are mot e easterly (i. e. greater 
east or less west Declination) than the observed ; and the sign + that the theo- 
retical are more westerly (i. e. greater west or less east Declination) than the ob- 
served. 
It must be remembered that the coefficients of M. Gauss’s theory in their present 
numerical values do not profess to be more than a first approximation ; that they 
rest on maps of the phenomena drawn indeed from observation, but in which care 
was not always taken to use only observations of a definite epoch ; that the points 
of the globe in which the elements of the theory rest upon the observed pheno- 
mena are only eighty-four points on the whole surface of the globe, viz. twelve 
points on seven parallels ; no point being taken in a more southern parallel than 20° 
south latitude ; and lastly, that the coefficients are limited to terms of the fourth 
order. 
On the other hand, it will be remembered that the Declination is the easiest, and 
has been by far the most frequently observed, of the three magnetic elements ; that 
from a very early period maps of the Declination, particularly in the Atlantic, and 
professing to be adapted to definite epochs, have been in much request on account 
of their use in navigation ; and that consequently it might naturally be expected 
that the differences between the theory and observation should be less in the com- 
parison here instituted, than might be the case in parts of the globe where the ele- 
ments of the theory have had a less direct, or a less satisfactory derivation. 
The differences of greatest amount which appear in this comparison are those 
over the north-west portion of the Atlantic, amounting to from five to nearly eight 
degrees generally between the meridian of Newfoundland and the United States, 
(meridians of 287° to 307°, or 53° to 73° west longitude), in the well-traversed 
parallels from 45° to 50° N. ; and increasing to 1 8° and upwards in the latitude of 60° 
in the vicinity of Hudson’s Strait, where the correct value of the declination has 
been well known for several years by the observations of the different British expedi- 
tions of discovery. In the southern Atlantic, where, as already remarked, the data 
of observation on which the theoretical coefficients are based have not been taken 
from a higher latitude than 20° S,, there appears a tendency to systematic differences; 
in excess (or too small easterly declinations) on the west side of the Atlantic, and in 
defect (or too small westerly declinations) on the east side of the Atlantic. 
A discrepancy deserving of notice in the theoretical lines (i. e in the lines of mag- 
netic Declination as they may be drawn from calculation with the present numerical 
elements of M. Gauss’s theoi'y), is in the value of the remarkable isogonic line, the 
