The jerky motion of a small craft will often cause an ordinary compass card to oscillate and even to completely revolve, so that the direction in which one is steering cannot be ascertained. It is therefore necessary to use a heavy compass card whose movement will be more sluggish. A spirit compass—such as Dent’s spirit compass—in which the card is surrounded by fluid—is the best adapted for small vessels.

But when a little craft is pitching quickly the very best compass will oscillate to some extent. In this case, one must observe the extreme points on either side reached by the swinging card and take the mean between them as the ship’s direction.

The needle points to the magnetic north, which coincides with the true north in certain portions of the earth only. The difference between the magnetic and real north is called the variation of the compass, it differs in different places, and is constantly but very slowly changing.

The variation at the mouth of the Thames, for instance, is now roughly 20° W., that is, the needle is dragged 20° to the west of north, so that if a vessel’s head be pointing true north, her compass will indicate a direction of about N.N.E.

The compasses designed on a chart generally show, as we have already explained, the magnetic and not the true bearings. Consequently, in order to ascertain the course that should be steered from one place to another, we lay one edge of the parallel rules on the two places on the chart, and by opening the rule we slide the other edge along—always preserving the direction—until it is on the centre of one of the compass designs on the chart. The direction thus indicated represents the course to be steered by compass.

But if the compasses on the chart are, as sometimes happens, true and not magnetic, the course shown by the rule is not the one to be steered by compass. We must, in this case, discover the variation existing in the locality—generally given on the chart—and convert the true course as indicated by the edge of the rule into a compass or magnetic course. To do this, if the variation is so many degrees west, the course to be steered will be that number of degrees east of the true course; if the variation is so many degrees east, the course to be steered will be that number of degrees west of the true course.

Where there is local attraction, as in an iron ship, the compass bearings will not be correct magnetic, but the needle will be dragged on one side or the other of the magnetic meridian. This error is called the deviation of the compass.

On board a wooden yacht, if certain precautions are observed, there should be no perceptible deviation of the compass. As iron ballast is generally used on a yacht, the compass should be as high as possible above it. If the boat have a well, the binnacle can at least be on a level with the raised hatch on the deck.

Care must be taken that no elongated iron mass, especially if vertical, such as a bolt, be near the compass.

Deviation is not uniform, like variation. When the vessel’s head is pointing about E. or W. the deviation is generally greatest, and it is different with every direction the vessel is turned to.