MARINER'S COMPASS. 671 



The masses of steel or hard iron, which have been magnetised 

 during the construction, behave like magnets; this magnetism is 

 sometimes called sub-permanent, for its value, which depends on 

 the position of the ship during building, diminishes slowly at sea, 

 and only becomes stationary after the vessel has made a certain 

 number of voyages. Soft iron, on the other hand, becomes mag- 

 netised by the earth; the temporary magnetisation thus produced 

 varies with the direction of the ship and its geographical position. 

 There are thus two kinds of disturbing actions one set constant, 

 the other varying with the direction of the ship. 



Let us first assume that the ship- is upright. The head of the 

 ship is the plane of symmetry from stern to bow that is to say, 

 from back to front, Let 



f be the azimuth of the ship's head with the magnetic meridian, 

 or the magnetic course, this angle being counted towards the 

 east; 



f the azimuth of the head with the direction of the compass, 

 or the compass course; 



B = f - f the deviation of the compass. 



The angle V which the compass makes with the geographical 

 meridian, or the apparent declination, is sometimes called the 

 variation. If A is the real declination, we have 



We shall assume that both the permanent and sub-permanent 

 magnetism of the ship are independent of the temperature; the 

 variations which it might undergo from this may be neglected in 

 comparison with other sources of error. 



It will be assumed that for masses of soft iron the induced 

 magnetisation is proportional to the magnetising force, from which 

 it will follow that magnetisations produced by different causes will 

 superpose themselves ; these conditions are sufficiently realised with 

 actions of the same order as that of the earth. 



We shall finally suppose that the compass needle is very small 

 compared with its distance from the nearest masses of iron or steel, 

 and therefore that it moves in a sensibly uniform field. It is suf- 

 ficient, therefore, to calculate the disturbance of the terrestrial field 

 at the centre of the compass. 



