116 * 
LIEUT.- CO LON EL SABINE ON TERRESTRIAL MAGNETISM. 
east, will find that the westerly variation, which increases the whole way from the 
Brazils to about the meridian of the Cape, begins there to diminish, and continues to 
diminish, passing- into easterly variation increasing-, for above 100 degrees of longi- 
tude east of the Cape. The separating- line which now passes through the Cape 
divides spaces distinguished by opposite magnetic characteristics ; on the west side 
of the Cape the north end of the needle moves to the west, and on the east side to 
the east, as east longitude increases. 
The maps which exhibit the results of the observations in the two ships, of the 
Declination, Inclination and Intensity, in the voyage of 1841-1842, and the isogonic, 
isoclinal, and isodynamic lines traced approximately in conformity with them, are a 
continuation of the maps published in No. V., which embodied in a similar manner the 
results of the preceding voyage. The results in the Erebus are distinguished from 
those in the Terror by a different character, for the purpose of permitting the degree 
of accordance in the two series of independent determinations to be readily judged 
of by the eye. These maps afford the best reply to those who have expressed doubts 
of the success of observations of the inclination and intensity made at sea. 
Magnetic lines, drawn from observations made in parts of the globe to which ob- 
servation had not previously extended, are the proper test by which we may judge 
of the degree of approximation with which the values of the numerical elements have 
been obtained in a general mathematical theory of terrestrial magnetism, such as 
M. Gauss’s. The portion of the observations of the Antarctic Expedition which has 
been placed before the Royal Society in No. V. and in the present number of these 
Contributions, permits us already to form some conclusion on this point. Plate XIII. 
exhibits the lines of one of the magnetic elements, i. e. the intensity, computed by 
M. Gauss’s theory, and drawn in Plates XVIII. and XIX. of the Atlas des Erdmagne- 
tismus, compared with the lines which are the direct results of observation. 
The very imperfect resemblance between the two systems of lines is of course no 
impeachment of the sufficiency of the theory, with corrected numerical elements, to 
represent the natural phenomena in parts of the globe which observation may not 
have reached. The degree of approximation to which it will do this must depend 
upon the extent and correctness of the observation-basis from whence the numerical 
elements are derived, and upon the order of the magnitudes comprehended in the cal- 
culation. 
The evidence which the plate affords, that the calculations in the elaborate work 
referred to differ so widely from the facts in the southern latitudes, shows how much 
observations were wanting in those latitudes for the purpose of perfecting the theory ; 
and is an ample justification (if indeed any justification were necessary) of the 
exertions which the last few years have witnessed to obtain them. 
Since these pages were written I have received from Mr. Archibald Smith the fol- 
lowing note. Regarding it as a continuation of the memorandum with which he 
