IN WOOD-BUILT AND IRON-BUILT SHIPS. 
77 
H and S have small positive values, but so small that their combination would not 
produce, in the Thames, an error of half a degree. N has a well-marked positive 
value. In this ship the magnetism would be sensibly corrected, by placing (by trial) 
a mass of soft iron abaft the compass and at a lower level, in such a position as to 
correct the deviation with head east and head west, and then placing a mass of soft 
iron at the level of the compass, starboard or larboard, so as to correct the com- 
bination of the original quadrantal deviation and the new quadrantal deviation pro- 
duced by the first corrector ; and this would be sensibly effective, without change, 
in all localities. Or, permanent magnets might be applied to neutralize the errors 
at the cardinal points, and soft iron at the level of the compass, starboard or larboard, 
to correct the quadrantal deviation ; the soft iron would then be effective in all 
latitudes, but one of the magnets would require alteration in different latitudes, and 
would require reversion in opposite hemispheres. 
The Pandora. 
14. H-fNX 9-67=+0T66 S = -f0-009 
15. H-Nx 11-32 = — 0-272 S=+0-082 
S has changed suflSciently to produce at Auckland (No. 15) a disturbance whose 
maximum is 40'. If this be neglected, the compass may be sensibly corrected in the 
same manner as for the Erebus. H is small. 
The Mseander. 
16. H-f-Nx 9-67= +0-070 S= +0-034 
17. H — Nx6-19= — 0-195 S=+0-008 
H would appear to have a value of —0-091, which at Sheerness would produce a 
deviation of 1° 20'. S is practically insensible. This ship would require for perfect 
correction a weak magnet with its marked end towards the stern, in addition to the 
soft iron as in the Erebus and Pandora. It will probably be better to use, for the 
polar-magnet-correction, a magnet alone, adjusting or reversing it as may be 
necessary. 
The Spy. 
22. H+Nx 9-67= +0-097 S=— 0-113 
23. H — Nx3-09= + 0-431 S=— 0-006 
If the headward subpermanent magnetism has not changed, II is positive and N is 
negative. These forces would requii-e correctors in positions opposite to those of the 
Mseander. S seems to have changed, to an amount which would produce at Sheer- 
ness a deviation of 1° 30'. It will probably be best to correct the polar-magnet- 
deviation by adjustible magnets. 
