ure. The alcohol in the liquid is to guard against its freezing in cold weather.
It has been stated that a single needle suspended by a thread would dip more and more as one proceeded from the equator toward the pole; and that in the dry compass this is prevented and the card always maintained horizontal by an adjustable counterpoise on the needle: no such contrivance is needed in the liquid compass; any downward pull of the earth's magnetism is at once met by such opposite pressure of the liquid on the rim of the card as to neutralize it. Magnetic attraction and liquid pressure counterbalance, and the card remains horizontal.
On the inside of the bowl is traced a fine black line—the lubber's-point, or, as it has of recent years been more appropriately designated in the navy, the keel-line (L, Fig. 4). It is this line toward which the point of the card indicating the ship's course is always directed. The binnacle which holds the compass is screwed down to the deck, so that the keel-line, as its name indicates, is in the vertical plane through the keel of the ship, or in a plane parallel to that one. This plane extends from bow to stern, and divides the ship into two equal and symmetrical parts.
Now, let an observer look at the compass-card and keel-line while the ship's bow swings through a portion of a circle: as each point passes the keel-line, it will seem that the card itself is moving, but this is an illusion; the card is still—ever pointing to the magnetic pole, in obedience to the attraction that there exists for the magnetism in the steel wires it carries. But it must not be understood that this attraction is of a nature to pull the card off its pivot: on the contrary, there is no tendency to motion of translation, but merely of direction—to turn the magnets on their pivot and place them parallel to the earth's lines of magnetic force.
To illustrate this, let us examine Fig. 8: C, C, is a steel arrow free to move upon a pivot P; from the extremities of the arrow light threads t . . . t', extend and pass over revolving wheels at N and S; small, equal weights Q and Q' are attached to the ends of the threads. Under the strain communicated to the arrow by the weights, it will, of course, lie in the straight line joining the points N and S.
Now, with the fingers, turn the arrow into the position C, C, the threads will assume the positions t, t', and both weights will be equally raised. Release the arrow suddenly, both weights will descend, and alternately rise and fall as the arrow makes a series of short and constantly diminishing vibrations, as shown in the positions C', C' and C", C", until it finally comes to rest and all is still. Let us replace the arrow by a magnet, and the threads,