The compass has been used as a primary navigational aid for at least a thousand years. Unfortunately compasses do not always indicate true (or grid) North. Compasses show a general North-South alignment but some adjustment is generally required to account for the error between True North and Magnetic North. The errors become very great in high latitudes due to the dislocation between the magnetic and the geographic poles. An extreme example was noted on the western journey of Scott's Discovery expedition. Setting out in mid October 1903, Scott and his party followed the trail blazed by Bert Armitage in 1902 through the Admiralty Range and beyond onto the polar plateau. As they sledged across the ice they would have encountered a position directly south of the South magnetic pole and north of the geographic south pole.The error of their compass passed from East to West and was nearly at its maximum of 180 degrees. The North end of their compass was pointing almost exactly due South. The images below are from the Maritime Museum in Hobart.
A navigator's compass error is due to two main influences. "Deviation" is the error caused by the effects of magnetic materials in the ship. The value of deviation changes according to the direction in which the ship is heading. Account of this is made by "swinging the ship" and noting the difference between compass and grid headings. These differences are recorded to make a "Compass Card" which serves as a reference for the navigator or physicist. Scott's Discovery was purpose built with a magnetic observatory. The fabric of the ship was totally free of any metals that could influence the magnetic observations within 30 feet of the observatory. Nevertheless when in the high latitudes of the Ross Sea the process of swinging the ship was undertaken often. This would have been an arduous task using rowing boats to haul the stern of the ship full circle. Most of the food supplies were tinned food. As they were consumed or relocated to trim the ship (as coal was used) the ship's deviation would have changed causing a critical influence on the physicists magnetometer observations. Von Drygalski's Gauss had a similar metal-free arrangement around the magnetic observatory on board. The Carnegie Institution commissioned a ship for precise magnetic observations to be built entirely of materials free of magnetic influence. It was a schooner powered by a bronze motor. The Carnegie cruised the world for over 20 years until fire destroyed it in 1927.
"Variation" is the difference between the compass' indicated bearing and the true (grid) bearing due to anomolies in geomagnetism. These are related to the location of magnetic poles and to local disturbances. In his journal Joseph Banks recorded a discussion with Cook aboard Endeavour about compass variation as they passed Pigeon House mountain on the New South Wales coast on April 23rd, 1770.
The Master today in conversation made a remark on the Variation of the Needle which struck me much, as to me it was new and appeared to throw much light on the Theory of that Phaenomenon. The Variation here is very small, he says: he has three times crossed the line of no variation and that all those times as well as at this he has observed the Needle to be very unsteady, moving very easily and scarce at all fixing: this he shewd me: he also told me that in several places he has been in the land had a very remarkable effect upon the variation, as in the place we were now in: at 1 or 2 Leagues distant from the shore the variation was 2 degrees less than at 8 Lgs distance.
The "line of no variation" is a theoretical line on the Earths surface connecting points of zero magnetic declination (an agonic line). At those points the indicated magnetic North and grid North are the same. Maps showing lines of equal variation across the oceans were created as aids to navigators.
Antarctic expeditions used sophisticated instruments to determine variation and dip. The magnetic Pole is defined as the area where the needle's dip is 90 Degrees. The magnetic poles are constantly shifting by around 15 kilometres each year. James Clark Ross located the North Magnetic pole in 1831 and came close to also finding the South Magnetic pole on his epic expedition of 1839-1843. Accurate location of the South Magnetic Pole was achieved by Louis Bernacchi after a year of painstaking observations at Cape Adare during the Southern Cross expedition of 1989-1900. Mawson (with Edgeworth David and Mackay) sledged to the area of the magnetic pole during Shackleton's Nimrod expedition of 1907. In 1912 he sent out a sledge party (Bage, Webb and Hurley) that approached within 50 miles of the magnetic pole from the north during his own "Aurora" (Australian Antarctic Expedition) expedition.