1 865 . ] Character of Armour-Plated Ships. 117 



the diminution of the directive force, is the constant and regular diminu- 

 tion of both effects with the lapse of time, showing apparently a change 

 in the molecular structure of the iron, by which it becomes less susceptible 

 of induced magnetism. This change has not yet been connected with any 

 observations on the strength of the iron. 



The amount of heeling error in these ships is very considerable, averag- 

 ing about 1 for every degree of heel. In those which have been built 

 head north it is greater. Thus in the ' Warrior,' which was built head 

 north, it is at the standard compass aft 1 49' for every degree of heel. 

 This error maybe corrected by means of a vertical magnet. In the wood- 

 built armour-plated ships, from the armour-plating causing the vertical 

 force to act upwards, the heeling error is very small, and generally to 

 leeward. Thus it is, for each degree of heel for the standard compass, in the 

 following ships- 

 Royal Oak 7' to windward. 



Prince Consort ....:. 8' to leeward. 

 Ocean 15' to leeward. 



The method of obtaining the heeling error by observations of horizontal 

 and vertical force, in addition to observations of deviation, is practically 

 used in this paper for the first time. The formulae for the purpose were 

 given for the first time in the ' Admiralty Manual.' 



Among the practical conclusions drawn by the authors, the most import- 

 ant are, that the best position for the ship to be built in is head south ; that 

 armour-plated ships should be plated with the head in the opposite direc- 

 tion to that of building ; that there should be as little iron as possible 

 within a cone traced out by a line passing through the compass, and mak- 

 ing an angle of 54 45' (cos- 1 = \ with the vertical ; and that in the con- 

 struction of iron-built and iron-plated ships, regard should be had to pro- 

 viding a suitable place for the standard compass. 



The separation of the constituent parts of the various coefficients is not 

 only of great scientific interest, as giving the causes of the different 

 amounts of these parts in different ships, but some of these quantities are 

 so nearly the same in ships of the same class, that when a sufficient num- 

 ber have been observed, we are able, by means of observations of deviation 

 and horizontal and vertical force, made without swinging a ship, and even 

 when she is on the stocks or in dock, to construct by anticipation a table 

 of deviations and of the heeling error. This method was applied in certain 

 of the cases given in the Tables when there was not an opportunity of 

 swinging the ship. This method may be expected to be of much use 

 when the selection of a place for the standard compass comes to be consi- 

 dered part of the duty of the naval constructor. 



The second part of the paper treats of the effect on the compass of 

 masses of iron of various shapes, bearing some analogy to shapes for 



