A ship carries a gyrocompass, which finds true north by spinning, and a magnetic compass, which finds it by the earth, and they are compared every watch and the difference logged, because a compass error is how ships hit things. Ours agree to a degree everywhere except one stretch, where the magnetic swings four degrees and comes back, and it swings towards a spot in the water.
I am the navigating officer, and the compasses are compared on my watch and written up in my hand.
How the comparison works.
The gyro finds north by the spin of the earth and does not care about iron. The magnetic compass finds it by the earth's field and does care, so it has a deviation card that corrects for the ship's own iron, and once corrected it should agree with the gyro. Every hour the officer reads both and writes the difference, and a difference that changes is a compass that needs looking at.
Ours agree to within a degree from berth to berth, corrected, on every crossing, except on one stretch.
On the stretch the magnetic compass swings. Over about a minute it moves four degrees to port of the gyro, holds for eight minutes, and comes back over a minute. The gyro is steady throughout, and so is the ship, on her track, by the satellite.
A swing that comes and goes on one stretch of water is a magnetic anomaly, which is a real thing, a body of iron on the seabed, and it is on some charts as such.
Not on ours. There is nothing on the chart, and the harbour's survey found sand.
The four degrees is not to port, and that exactness matters, because it took me a year to see it.
It is to port going one way and to starboard going the other. It is towards something. Whatever the magnetic compass swings towards on that stretch, it swings towards the same place on the water on both headings, and I have plotted the two bearings, and they cross, at a position a few hundred metres off the track, over thirty one metres of sand, that the chart marks with nothing.
What has been checked.
The magnetic compass. Swung and adjusted by a compass adjuster, twice. The deviation card is good on every heading. The card does not remove the swing on the stretch, because the swing is not the ship's iron.
The gyro. Serviced and tested. It is right.
A handheld compass, on the bridge wing, away from the ship's iron. It swings. Four degrees, towards the spot, on both headings, in my hand.
A magnetometer. I borrowed one from the survey company for a week and it reads a field on the stretch, from the direction of the spot, of a strength that means a very great deal of iron, close, or something less than that, closer.
What the plot says.
A magnetic anomaly is a fixed thing. It does not move. This one does not move either, which is the reassuring part, and it is a few hundred metres off our track, and the swing is small, and it has no effect on the navigation, because we steer by the gyro.
What I have not been able to put down is the plot of the strength. The swing has grown. Four degrees this year. Three the year I joined. The compass adjuster's records for the ship go back fourteen years and the first note of a swing on the stretch is one degree.
Whatever is at the spot is getting stronger, or nearer, and I can measure that with two compasses and a pencil.
What is done about it.
The stretch is logged as a local magnetic anomaly, in the passage plan and in the compass log, and the deviation is noted hourly as it always was, and the ship steers by the gyro, and nobody has to do anything.
I have told the harbour master, who has passed it to the hydrographic office, who have thanked me and asked for the plot, and I have sent it, and it will be on the next chart, as a note, if they agree it is there.
What I do.
I navigate.
The one thing I have changed is that I do not take the handheld out on the wing on the stretch.
The reason is a night in the summer, with the handheld in my hand on the port wing, watching it swing towards the spot, four degrees, steady, and I turned to look along the bearing at the water, dark, a few hundred metres off, and the compass in my hand swung further. Not four. It went round, the way a compass goes round when you bring a magnet to it, a full circle, slowly, and stopped pointing at the water, and something on the water, at the spot, a few hundred metres off, was pointing back.