IN THE IRON AND ARMOUR-PLATED SHIP NORTHUMBERLAND. 
489 
At the end of December 1866, the ship was completed in her equipment for tem- 
porary service and steamed to Sheerness, where she remained swinging to the wind 
and tide till the early part of March 1867; when she steamed to Devonport, at which 
place, with the exception of two days’ trial at sea to test the machinery in the middle 
of May, she has remained in a dry dock with her head directed S. 84° E. magnetic till 
the present time. 
The positions of the compasses at which observations were made were the following : — 
Standard Compass . — The ‘ Northumberland ’ having been built with her head nearly 
N.E., the magnetism principally developed was in the upper part of the stern and star- 
board quarter, and it was therefore desirable that this compass should be as far forward 
as possible. It was accordingly placed 172 feet from the stern, and 8f feet above the 
iron deck. 
Steering Compasses . — The upper deck steering-wheel is 52 feet from the stern, under 
the fore part of the poop wooden deck ; two compasses were placed close in front of it 
6 feet apart, each 4 feet above the iron deck, and 3 feet 8 inches below the iron beams 
supporting the poop-deck. 
Poop Compass . — This compass was placed on the fore extreme of the poop-deck, and 
9 inches before the line joining the steering compasses, and 4 feet above the poop-deck. 
In the selection of a place for these latter compasses there was no room for choice ; the 
arrangements of the architect and the requirements of the seaman could be alone con- 
sulted. 
The results of the observations will be found in the Table appended to this paper, and 
will include a few made at temporary positions not necessary to describe in detail. 
For a complete explanation of the meaning of the quantities tabulated, and the method 
of obtaining them by observation, I must refer to the last of the two papers mentioned 
above. Here it may suffice to say, that if 
£ represents the magnetic azimuth of the ship’s head, 
the azimuth by disturbed compass, }>=%—%' the deviation, then 
sin($=9t cos &-+-9B sin £'-|-(5 cos sin (2^ + ^) + @ cos (2£'-|-d) exactly, 
or 
d=A-|-B sin £'+C cos £' + D sin 2£'+E cos 2£' approximately. 
Of these coefficients 91, 25, (S (or A, D, E) depend solely on the transient magnetism 
induced in the soft iron, and therefore cannot be affected by any artificial magnetization, 
or demagnetization. 
9$ (or B) depends partly on the magnetism induced in soft iron by the earth’s vertical 
force, partly on the permanent or subpermanent magnetism of the hard iron. (5 (or C) 
depends on the last. It is therefore to the changes in SB and (5 only that we are to look 
for the effects of polarization or depolarization. 
X is a factor almost always less than unity, representing the mean force, to north, as 
affected by the soft iron in the ship. 
