1865.] Character of Armour-Plated Ships. 119 



that when 1=^-$, the above proportion was as 3:2, which gives for K a 

 value =35, agreeing closely with the values found by Weber and Thalen. 

 In the case of iron of the quality experimented on by the authors, this ratio 

 would be 



t+ m :t - 



Hence in the case of a tank y^th of an inch thick and 4 feet in diameter, 

 t would be about ^4-jj, and the effect about jpg that of a solid mass of iron 

 of the same size. These results, which, however, are not new, as they are 

 involved in Poisson's paper of 1824, explain the mistakes into which various 

 magneticians have fallen as to the magnetism residing entirely on the sur- 

 face, and as to the effect of a body such as a tank depending on its sur- 

 face, not on its mass. 



The same formulae show that to correct a quadrantal deviation of 6 10 f 

 by two cannon-balls placed one on each side of the compass, the distance 

 of the centre of each ball from the centre of the compass should be three 

 radii of the balls. If the distance is greater or less, the quadrantal devi- 

 ation corrected will vary inversely as the cube of the distance. 



The investigation also shows that the effect of a sphere of iron, if its 

 centre be within the cone of 54 45', will be prejudicial both bv diminishing 

 the directive force, and by increasing the heeling error to windward. 

 "When without that cone it will be beneficial in both respects. Hence, as 

 far as possible, no iron should be within this cone. 



Masses of iron which may be compared to a sphere, and near the level 

 of the compass, but in the fore or aft quadrant, are beneficial in increasing 

 the directive force, but prejudicial in increasing the quadrantal deviation. 

 If they are on the port or starboard quadrant, they are doubly beneficial, 

 by increasing the directive force and diminishing the deviation. 



Bodies which may be compared to infinite vertical cylinders, such as 

 iron masts placed before or abaft the compass, are prejudicial by increasing 

 the quadrantal deviation, and they do not increase the directive force. 



March 23, 1865. 

 Major- General SABINE, President, in the Chair. 



Pursuant to notice given at the last Meeting, Count Strzelecki proposed, 

 and the Master of the Mint seconded, the Right Honourable the Earl of 

 Dououghmore for election and immediate ballot. 



The ballot having been taken, the Earl of Donoughmore was declared 

 duly elected. 



The following communications were read : 



