Wandering Compass

What older compasses assumed

In China a lodestone that turned to point south was described from about the first century AD, though how early it was used for finding the way is debated. Around 1088 the scholar Shen Kuo described a magnetised needle and noted that it points a little east of due south: the oldest known record of declination.

In Europe Alexander Neckam described the sailor's needle about 1190, and for centuries it was taken to point north. On Columbus's first crossing in 1492, his log records the needles turning away from the Pole Star.

Some north European compass makers of the 1500s set the needle a little east of the card's north, so that the card read true in home waters. Carried far, or kept for decades, such a card was wrong again. Qibla compasses, which give the direction of Mecca for prayer, carry a table of bearings for each city; read off a plain needle, they are out by the local declination unless the table allows for it.

The present-day numbers come from IGRF-14 for the chosen place and year. The history is from standard accounts of the compass (Needham's Science and Civilisation in China; the Royal Museums Greenwich and NOAA on declination); early dates are uncertain.

A compass does not point to the North Pole. It lines up with the Earth's magnetic field, which is made by molten iron churning in the outer core, 2,900 km down. The angle between true north and the needle is the declination.

The core never stops moving, so neither does the needle. Henry Gellibrand showed it in London in 1635, and navigators' charts have printed a yearly correction ever since.

East of true north is drawn above the line, west below. The last five years are a forecast, hatched. Tap the chart to pick a year.

The dip pole is where a freely hung needle stands straight up and down. The grey circles are every five degrees of latitude.