266 STAFF COMMANDER EVANS AND MR. A. SMITH ON THE MAGNETIC 
indicating the place which should be selected for the compass, and also in enabling 
us to anticipate or account for the subsequent changes which take place in the 
deviation. 
Another and perhaps even more important result is that we are enabled by observa- 
tions made with the ship’s head in one direction, and therefore when she is in dock or 
even on the stocks, to determine the coefficients and construct a table of deviations, 
including the heeling error, without swinging the ship. To explain this, we may observe 
that for the complete determination of the deviations of the compass when the ship is 
upright and in one geographical position, six coefficients are required. But of these 
two vanish when the iron is symmetrically arranged, two more are so nearly the same 
in ships of the same class that they can be estimated with a near approximation to the 
truth ; we have therefore only two coefficients left, and these can be determined by an 
observation of deviation, and an observation of horizontal force made without altering 
the direction of the ship’s head. 
So as regards the heeling error, to determine this three additional quantities are 
generally necessary, but of these one is zero when the iron is symmetrically arranged ; 
another may be estimated, and the third may then be determined by a single observation 
of vertical force. 
The quantities so estimated change little after the ship is completed, so that any 
assumption made as to their value may he checked by subsequent observations. 
These considerations will show the importance of not only making the observations 
we have mentioned, but of reducing the observations made, and of tabulating, discussing, 
and publishing the results of the observations. In the Tables it will be seen that the 
original observations are not given ; they, as well as the curves and computations by 
which the coefficients are derived, are carefully preserved among the records of the 
Admiralty Hydrographic Office, and may at any time be referred to ; but the coefficients, 
at least so far as regards the deviation of the horizontal needle, represent so exactly 
the observations made, that to give them here at length would he an unnecessary waste 
of space. 
The observations, the results of which are tabulated, were made in the following 
manner. The deviations of the Standard Compass were observed by reciprocal simul- 
taneous bearings of the Standard Compass and an azimuth compass on shore, in the 
manner described in the ‘Admiralty Manual.’ The admirable construction of the Admi- 
ralty Standard Compass, as regards design and workmanship, accuracy of adjustment 
and magnetic power, leaves nothing further to be desired for such observations. The 
arrangement of its four needles obviates, as we have shown in a former paper*, the 
sextantal error caused by the length of the needle when acted on by iron placed near it. 
The deviations of the steering and maindeck compasses were obtained by observations 
of the direction of the ship’s head by those compasses, made simultaneously with the 
observations of the Standard Compass. These compasses in the Royal Navy are of 
* Philosophical Transactions, Part II. 1862. 
