312 STAFF COMMANDER EVANS AND MB. A. SMITH ON THE MAGNETIC 
Transverse longitudinal masses of Iron extending from side to side as Iron beams. 
Let m be the length of the beam, or in general the breadth of the vessel, r the 
distance of either end of the beam from the compass, S the area of the section of the 
beam. It is easily seen that such a beam will give no coefficient except 
xS m 
Every such beam therefore diminishes the directive force and produces a + qua- 
drantal deviation, the effect being directly proportional to the mass of the beam, 
inversely proportional to the cube of the distance of its ends. 
If we have a rectangle of four beams, two fore-and-aft and two transverse, the compass 
being in or directly above or below the centre of the rectangle, l being the length of 
the two fore-and-aft beams, m of the two transverse beams, we shall have 
a=-2zSL 
/ 
/ 
whence 
X=1-^S 
xs l — m 
T ~r*-' 
Such beams may be compared to the armour-plating of a ship, and we thus see that 
for a compass near the centre of the ship, l being greater than m , the effect of such 
plating will be to diminish the quadrantal deviation. 
In accordance with this result, we find that in the wood-built iron-plated ships, when 
the compasses are inside the rectangle of the armour-plating, the quadrantal deviation is 
very small. 
When, as in the case of the Warrior and Black Prince, the plating does not extend 
from end to end, and the compasses are near or even outside one end, the case is 
different. 
Thus if the fore-and-aft coordinates of the ends be ad and x, and the distances from 
the compass r' and r, we shall have 
a=2*s{-^+*}, 
e=2*s{-^+^}, 
_ -As(x+y z'-y ] 
r 3 “ r 13 f 
When the plating extends abaft the compass x is negative, and when this is the case, 
ad being of course greater than y, so long as x is greater than y, or so long as the plating 
