184 
LINES OF MAGNETIC DECLINATION IN THE ATLANTIC. 
be ascribed either to errors of observation or to accidents ; and the prominency which 
lias been given to them on this occasion will not be misplaced, if it should serve to 
impress upon those who have the power of carrying out practical suggestions, the 
importance of giving a sufficient trial to the method proposed by Mr. Archibald 
Smith in the eighth number of the Magnetic Contributions, whereby the variable 
term in the correction formula may be at all times determined experimentally at sea, 
by deflections of the compass needle obtained with the ship’s head on two opposite 
points of the compass. The observations needed are extremely simple, require no 
unusual circumstances of weather and no reference to celestial objects, and need 
occupy but a very few minutes. 
Mr. Smith has shown that the variable term may also be determined at sea by ob- 
servations of azimuths with the ship’s head placed on the points of greatest disturbance; 
but the deflection method promises to be even more simple than that by azimuths. 
By the addition of a brass bar attached at right angles to the prism and sight vane 
of the azimuth ring of the standard compass, deflecting magnets may be temporarily 
fixed at a convenient distance from the compass needle, and the deflections measured 
with the ship’s head on two opposite points ; as was first practised by Captain (then 
Lieut.) Henry Clerk, R.A., F.R.S., in his Antarctic voyage*. 
If this rnetliod of determining the variable coefficient in the correction formulae be 
found to answer its purpose on a further and sufficient trial, the correction of the dis- 
turbances occasioned by the ship’s iron might be still further simplified by the formation 
of tables of each term for every probable value of the coefficients, when the only calcu- 
lation remaining to be made would be the addition of the quantities to be taken out 
from the tables. In wooden ships, two terms, and consequently a single addition, 
would probably, in most cases, be sufficient for the whole amount of the correction -f'. 
With reference to the corrections which we have now occasion to employ for the 
declinations observed in the Erebus in her passage from England to the Cape of Good 
Hope, we have the following values of the constant coefficients A, D and E in the 
formula (6.):|:, derived from the observations on the thirty-two points of the com- 
pass, at the several stations at which these observations were repeated, by the equa- 
tions (16.), (19.) and (20.) §. 
A. 
D. 
E. 
Gillingham 
-f 16 
- 3 
Port Praya 
. . -23 
-1-24 
— 8 
St. Helena 
+27' 
+ 3 
Cape of Good Hope, 1840 . . . 
. . -1-23 
-1-23 
+ 15 
Mean . . . 
. . -H 5 
-1-22 
+ 2 
* Philosophical Transactions, 1846, p. 347. 
t Whilst these pages were in the press, tables such as are here referred to have been drawn up and printed 
under the direction of the Admiralty in a tract entitled “ Directions for ascertaining at any time, whether at 
sea or in harbour, the changing part of the Deviation in the Compass occasioned by the Ship’s Iron.” 
I Philosophical Transactions, 1846, p. 348. § Ibid. pp. 350 and 351. 
