188 
LINES OF MAGNETIC DECLINATION IN THE ATLANTIC. 
quitted England ; nor could it have existed when the ship was swung in the River 
Thames, at Port Praya, or at the Cape of Good Hope in 1 840, or it would have ap- 
peared to its full amount in the value derived for the coefficient A. The evidence, 
however, afforded by the consistent results of the several trials that were made both 
at the Falkland Islands and at St. Martin’s Cove in 1842, and both on shore and on 
board, leaves no doubt of the existence of index error at that period. 
I cannot find that any trial was made of the index error of this compass on the 
arrival at the Cape in April 1843 ; but in the neighbourhood of the Cape the Erebus 
crossed the track of the Pagoda, the observations of which ship are published in the 
Philosophical Transactions for 1846, Part III., and the declinations observed in the 
two ships agree when the index error observed at the Falkland Islands is employed 
for the compass of the Erebus, but would disagree to an amount of nearly two de- 
grees if that correction were not employed. 
Mr. Tucker’s memorandum is as follows : — 
“The compass error of the standard compass of the Erebus was ascertained at 
Port Louis in the Falkland Islands, on the 13th of August 1842, in the following 
manner. The compass with its card CCH was taken on shore, and its tripod was 
fixed over the spot on which the transit had been established ; the bearings of the N. 
and S. meridian marks were then taken with the compass as follows : 
North mark N. 15° 36''8W. 
South mark S. 15° 40' E. 
The true declination was 17° 33' E. in the month of August by the declinometers of 
the observatory; the error of the standard compass was therefore 17° 33' —15° 38''5 
= 1° 54''5, or the north end pointed 1° 54''5 to the west of the true magnetic north. 
The compass was taken again on shore at the same place in September and December, 
and at St. Martin’s Cove in November, and the compass error was tried in the same 
way, and was found on all those occasions to Le within a few minutes the same as 
that above stated. Also when the ship was swung at Port Louis on the 19th of 
August 1842, the sum of the declinations observed with the standard compass in its 
usual place on board, on the thirty-two points, divided by thirty-two, made a mean 
declination of —15° 39'’3 ; the true magnetic declination in the same month by the 
declinometer was —17° 33', whence the compass error equals 1° 53''7 to the westward 
of north.” 
In conformity with this memorandum I have employed —1° 54' as an index cor- 
rection from August 1842 to April 1843. 
Determinations in H.M.S. Terror, in 1842 and 1843, between Cape Horn and the 
Cape of Good Hope. 
The disturbance of the compass of the Terror was examined at the Falkland Islands 
on the 17th of August 1842, and at the Cape of Good Hope on the 20th of April 
1843, by azimuths observed with the ship’s head successively on the thirty-two 
points. Assuming the mean of the azimuths on the thirty-two points to give the true 
