130 
Terrestrial Magnetism. 
angular amount of the deviation of any magnetic needle 
from the true magnetic meridian, the motion of which was 
limited to a horizontal plane,—which is exactly the con¬ 
dition of the compass needle on board of a ship. These 
experiments were at first made with solid balls of from 
300 to 100 lbs. weight: but when similar experiments 
were made afterwards with hollow balls of the same 
diameter, it was not a little surprising to find no difference 
between the results of the former observations with solid 
spheres, and the latter with hollow ones. It was there¬ 
fore clear that the effect of iron upon the magnetic needle 
depended upon the extent of its surface, and was totally 
independent of its mass; excepting that, for the full 
development of the attractive energy of any mass of iron, 
it was necessary that it should possess a certain thickness, 
which some experiments have fixed at the ^th of an 
inch. Further experiments made upon very irregular 
masses of iron have also proved the existence of a plane of 
neutrality in all of them, and completely established the 
fact of there being attractive and repulsive forces in all 
cases, whether the iron of a ship be very generally dis¬ 
tributed, or whether it be collected in isolated masses. 
These last experiments having proved that the effect 
produced depended upon the surface, and not upon the 
mass of iron, Mr. Barlow suggested the use of two 
circular plates, with a thin piece of board interposed 
between them ; by which means a more uniform action is 
obtained, and the compound plate is more powerful 
when the plates of which it is composed are separated 
from each other. The action of this plate upon the 
compass, and the manner of placing it, have already been 
mentioned -when speaking of the ball, and need no 
further description. 
From this time, the subject continued to be carefully 
investigated in various ways and at different places on 
