LINES OF MAGNETIC DECLINATION IN THE ATLANTIC. 
187 
land Islands in Aug-ust 1842, and at the Cape of Good Hope in April 1843; from 
these we obtain, — 
At the Falkland Islands .... B = — -0377; C=+-0009. 
At the Cape of Good Hope . . . B = — ‘0517; C=— *0040. 
It has been already noticed that the term x/B^-j-C^, derived from these values, 
was on both occasions numerically greater than would have been assigned from the 
dips at the respective stations, and the values of the same coefficients at other stations 
where the ship had remained sufficiently long for the full development of the changes 
in its induced magnetism corresponding to changes of geographical position. The 
experiments at the Falkland Islands and at the Cape of Good Hope, afford however 
the best indication which we possess of the magnetic state of the ship in the interval 
comprised between their respective dates, and must betaken as the foundation of the 
corrections during that interval. In September 1842 the Erebus quitted Port Louis 
in the Falkland Islands for Cape Florn, and after remaining some weeks at St. Mar 
tin’s Cove, where the dip was between —58"^ and —59°, returned in November to the 
Falkland Islands; from whence she sailed immediately afterwards to resume the 
magnetic survey of the higher latitudes, arriving at the Cape of Good Hope in April 
1843. In the whole interval between August 1842 and April 1843, the ship was at 
no time in a lower dip than that at the Falkland Islands, and we may presume there- 
fore, with much probability, that the ^B^-l-C^ was in no part of the interval less 
than its value at Port Louis. We have thus a minimum value for this term. During 
January 1843 the dips observed in the Erebus ranged, in different localities, from 
— 60° to —63°. In February from —58° to —62°; in the first week of March a 
favourable opportunity presenting itself for pressing to the southward, the dip in- 
creased to between —65° and —66°; but from the 8th of March it progressively di- 
minished until the arrival at the Cape on the 6th of April. The experiments made 
at the Cape on the 20th of that month gave a value of \/B^-}-C^, corresponding to 
—63°, or thereabouts. It is not probable from this review that ^/B'^+C^ was at any 
time much higher than it was found at the Cape ; it may possibly have been a little 
higher for some days in March, but I have thought it safer to keep within the limits 
which were actually observed than to assume a conjectural maximum ; and in cor- 
recting the declinations of this period I have accordingly taken the value of ^B^-j-C^ 
observed at Port Louis as applicable until the Erebus sailed for the higher latitudes 
in December 1842, and have then increased it uniformly and progressively with the 
time until the first week in January 1843, when the dip was —63°, corresponding to 
the highest observed value of -^/B^-j-C^; and I have used that value thenceforward 
until the arrival at the Cape, where it accorded with the experiments. 
By the following memorandum with which I have been furnished by Mr. Tucker, 
Master of the Erebus, it appears that the standard compass of that ship had an index 
error of considerable amount during this portion of the voyage. I have not been 
able to learn anything satisfactory in regard to the cause of this error, which did not 
exist when the compass was tried by Captain Johnson, R.N., before the expedition 
