IN WOOD-BUILT AND IRON-BUILT SHIPS. 
57 
paper in the Philosophical Transactions, 1839), I was guided by experiments on hori- 
zontal intensity as determined by vibrations ; but even then I found the computa- 
tion of polar-magnet-deviation so troublesome, that I executed the calculations by 
graphical construction. But in other cases, where there is no determination of hori- 
zontal intensity, the computation of polar-magnet-deviation would be very much 
more troublesome. This consideration, together with the paucity of instances in 
which a comparison of the ship’s magnetism in different localities was possible, pre- 
vented me from entering further into the numerical calculations of ships’ magnetism. 
But having lately received from Captain Washington, R.N., Hydrographer to the 
Admiralty, the records of observations in several ships, which I desired to treat 
numerically; I remarked that the trouble of calculation might be much diminished, 
and the process might be made perfectly direct and definitive, by the previous prepa- 
ration of a Table of Polar-Magnet-Deviation ; and I proceeded therefore at once to 
compute the Table which is appended to this paper. Of this Table I will now give a 
short description. 
The table is a Table of double-entry. One of the arguments is the “ Modulus,” 
which is the same as the proportion of AB to AC in fig. 1. It is given to every 
*01 from *00 to *80. The other argument is the “Apparent Azimuth of the Ship’s 
Head from the Neutral Position,” which is the same as the “apparent azimuth of the 
polar magnet” or “azimuth of the polar magnet as measured from the disturbed 
position of the compass-needle,” or the angle EBF in fig. ] . This is used as the 
argument of the Table, because, in the examination of the disturbance of ships’ com- 
passes, it is usually most convenient to fix the ship in position by means of its own 
compass ; and in fact all the observations supplied to me have been made in positions 
of the ship so determined. As the observations of deviation of ships’ compasses are 
usually made from “point” to “point” of azimuth, the division of the circle here 
employed is that by points and decimals of a point. The Table is carried to 8 points 
only, as the polar-magnet-deviations from 8 to 16 points are the same in reversed 
order; and those from 16 to 32 points are the same as those from 0 to 16 points with 
change of sign. At the bottom of each column is the “Mean of all tlie Polar-Magnet- 
Deviations for each value of the Modulus,” M*hich is necessary for enabling us to deter- 
mine the value of the modulus in any given case. 
In ascertaining, from a given series of observed compass-deviations, the neutral 
position and modulus to be used in the application of this Table, it will be necessary 
to recognize the existence of a deviation following very nearly the law of quadrantal 
deviation, and the given numbers must therefore be so combined that quadrantal 
deviation will be ipso facto eliminated. This will be done by so arranging the pro- 
cess that the numbers for a whole semicircle of apparent azimuth will be added 
together algebraically. This being understood, we may now proceed with advan- 
tage to investigate the nature of the terms produced by combining the effects of the 
polar-magnet-force and the quadrantal force. 
MDCCCLVI. 1 
