During the First Voyage Cook made over 50 observations of magnetic variation: the angle between magnetic north and true north. Those made whilst sailing north along the east coast of Australia gave rise in an earlier paper,1 to further questions that are addressed in this paper.
The introduction of the measurement of magnetic variation into the Royal Navy and the Marine Service of the East India Company can be traced back to the voyages in 1699 and 1700 of the renowned mathematician and astronomer, Edmond Halley, FRS. The young Halley, born in 1656, developed strong interests in astronomy and the Earth’s magnetic field. The first interest led Halley, with the support of the Royal Society, to sail in 1676 to the island of St Helena, where he and his assistant James Clark, observed the constellations in the southern hemisphere, and determined the positions of 341 stars, most of which could not be seen from Europe. His second interest led to an early compilation of the magnetic variation (55 observations from 47 places) that resulted in a paper to the Royal Society in 1692.2 He postulated the idea that the Earth had four magnetic poles, though he recognized the need for more measurements. In turn this led Halley to begin the search for support for his ideas for a scientific voyage devoted to understanding the global changes in magnetic variation.3
Halley’s ideas gained the support of the Royal Society and the Lords of the Treasury. He was encouraged by Queen Mary II, and after initial delays, he embarked on the Pink (a ship with a very narrow stern) Paramore of about 80 tons burthern on 20 October, 1698, carrying 20 persons. The first voyage took Halley down the west coast of north Africa, across to Brazil, returning to the UK via Barbados in July 1699. The second, and longer, voyage left England in September 1699, and sailed south to Rio de Janeiro, and then to 50°S, before turning north to St Helena, Barbados, and Newfoundland, arriving in England on 10 September, 1700.
Halley was equipped with two azimuth compasses: a necessary precaution in case one of them was damaged but also enabling him to investigate the differences between the compasses. The magnetic amplitude is defined as
an arch of the horizon, contained between the sun or a star, at the time of rising or setting, and the magnetic east or west point of the horizon pointed out by the compass; the difference between this and the true amplitude is the variation of the compass.4
With an azimuth compass
Halley could determine magnetic variation at a place by one or, more accurately two observations. A single measurement of the magnetic amplitude at sunset (making allowance for refraction on the horizon) and then compared this figure with the Sun’s computed azimuth from geographical north. The other method used when conditions allow, consisted of taking measurements of the Sun’s magnetic amplitude, i.e. at sunset and sunrise, half the difference between these two amplitudes being the variation, which was applied to the ship’s position at midnight.
In order to calculate true north and longitude from astronomical observations, Halley would have relied on published star catalogues. Particularly useful were the calculations of the Italian astronomer Jean Dominique Cassini (1625-1713), director of the Paris Observatory, who had observed the motions of the planets, and had prepared tables of the revolutions of the satellites of Jupiter, so that they could be used for the determination of geographical longitude. The method (which Halley employed on his voyages) involves observing with a telescope the occultations of the satellites at a certain place and comparing the time of events with the time at another distant point as predicted and recorded in the tables.5
The question arises as to how accurate were Halley’s measurements of magnetic variation, in order to compare his measurements with those of Cook? There were several possible sources of error, and Halley was certainly aware of all these possibilities. In a demonstration in May 1701 to the Royal Society he demonstrated the manner in which he measured variation at sea using two compasses, and noted one compass read 7° 40′ and the other 8°.6
Halley made over 150 measurements of magnetic variation on his two voyages and
Halley’s Atlantic Chart is presumably the first published isogonic (magnetic declaration) chart, and apparently, the earliest printed isoline map.7
Halley commented on the accuracy of his measurements in a number of ways.
Another way to check the accuracy of Halley’s measurements is to compare them with a chart of magnetic variation for 1700 compiled by W. van Bemmelen of the Observatory of the Dutch East India Company in Batavia. Comparison of the two charts in common regions “found that they agreed to within 1° in the declination almost everywhere.”11 This suggests that the true values were measured with an accuracy of plus or minus 0.5°.
Halley does not appear to have recorded any unusual amplitude measurements that might have been caused by a large magnetic storm or naturally-occurring sub-outcrop of a highly magnetic rock.
The hope of Halley and others that measurements of magnetic variation would lead to an accurate method of determining longitude at sea were not realized, but Halley was aware of the importance to navigators approaching an unknown coast in bad weather of having accurate measurements of the magnetic declination. In 1701, he described the advantages to a mariner sailing along a line of parallel towards the Scilly Isles.12 It was this particular conclusion that was picked up by the Navy Board, and subsequently became an instruction to all sea captains, especially Cook.
During the First Voyage, between May 1768 and July 1771, Cook made over 50 observations of the magnetic variation. However, it was only when sailing along the east coast of Australia, and then south of Timor, that he began to record unexpected changes in his measurements. The question arises of why Cook began to observe large changes in magnetic variation between April and September 1770?
There are four possible explanations for large and sudden changes in magnetic variation.
1.Someone in Cook’s company left something magnetic near the azimuth compass.
One would expect Cook, an experienced observer, to be well aware of such a possibility, and of the influence of the ship’s heading, later studied by Flinders.
2.There were errors in determining true north.
True north and longitude would have been based on observations of the sun, or stars, and comparison in Cook’s words with the “Assistance of the Nautical Almanac and Astronomical Ephemeris”.13 It is my understanding that an error, whether arising from the observations involved or one inherent in the astronomical tables or procedures, could not have led to an error of some three degrees in the magnetic variation.14
3.They were caused by a naturally-occurring sub-outcrop of a highly magnetic rock.
4.Cook’s measurements were affected by magnetic storms.
It has been shown that the regional magnetic surveys around Magnetic Island, offshore Townsville, did not show any large magnetic anomalies to explain Cook’s observation.15 However, if there was a geological explanation it must still be observable today. Recent enquiries amongst Australian yachtsmen has led to the following authoritative statement.
essentially there is no unusual magnetic effect [affecting navigation] from Cape Upstart, Magnetic Island or any other part of the Queensland coast.16
As to magnetic storms, there are six observations to consider. Around New Zealand, between November 1769 and May 1770, Cook mapped the gradual change in the magnetic variation, and made no mention of unexpected deviations.
On 11 April, 1770, between New Zealand and Australia, Cook noted
this 2½°E more than yesterday and expected to have it less for the observations were equally good.
On 19 May, 1770, he observed a magnetic variation 8°36’ East in evening and 8°20’ East in morning. About two or three leagues offshore Sandy Cape, Queensland.
On 5 June, 1770, between sunset and sunrise near Cape Upstart he noted a change in magnetic deviation of 3¼ degrees.
The next day, off Magnetic Island he reported that the
Compass did not traverse well near it.
Between 11 June and 4 August, 1770, Endeavour was grounded for repairs after hitting rocks on the Great Barrier Reef, and no measurements were made of the magnetic variation. However, by observing an Emersion of Jupiter’s first Satellite, Cook made an accurate calculation of longitude.
On 13 September, 1770, offshore Timor, Cook observed
1°10’ West by amplitude and 1°27’ West by azimuth
Three days later, Joseph Banks, offshore Timor, observed an Aurora Australis. This aurora was also observed by the Chinese—the first documented observations of aurora at geomagnetic conjugate points.17
Changes of three degrees in the magnetic variation only occur for short periods of time (five to ten minutes) during magnetic storms. This leaves the possibility that the change of 3¼ degrees overnight in June 1770 was caused by one of Cook’s observations being affected by a large magnetic storm. It is now recognised that, between October 1769 and October 1770, observations in Europe and Asia of large sunspot groups and a series of low-latitude aurorae reflect a period of enhanced solar activity. These widespread observations suggest that in September 1770 the world was affected by one of the largest known magnetic storms, similar to the better known “Carrington event of 1859”.18
When sailing north from Cape Howe, Cook and Banks between 20 and 24 April, 1770, discussed the effect of land on the variation of the Needle. Cook was aware that in several places he had visited
the land had a very remarkable effect upon the [magnetic] variation.
Cook followed up this discussion on 30 May, 1770, whilst at anchor at Thirsty Sound. He made the following comments after his climb to the top of Pier Head, a prominent 110 metre high hill.
having the Azimuth Compass with me… the Needle of which differ’d from its True position something very considerable, even above 30 degrees, in some places more, in other [places] less, for I try’d it in several places. I found it differ in itself above 2 points in the space of about 14 feet. The loose stones which lay upon the Ground had no effect upon the Needle; I therefore concluded that it must be owing to Iron Ore upon the Hill, visible signs of which appeared not only here but in several other places.19
Matthew Flinders visited Pier Head on Sunday, 5 September, 1802, and spent two days confirming Cook’s observations. He found that moving his theodolite just three yards changed bearings by two degrees. The theodolite stood four feet above the ground, whilst Cook probably placed his azimuth compass on the ground, thus explaining the differences. Pier Head is a trachyte plug, and the hill has for aeons been subject to lighting strikes, which also magnetize the surface rocks.20
Flinders also noted changes of up to four degrees in magnetic north with changes in the direction of the ship’s head, and subsequently noted how such changes might be permanently take into account. Flinders’ experiments in 1812 included procedures for magnetic compensation of compass heading errors by the strategic placing of soft iron rods near the binnacle compass, which became known as Flinders bars. This led, in 1812, to the Admiralty instructing all Royal Navy commanders to understand their onboard compass environment, and to standardise and maintain their compass binnacles, free from magnetic interferences.21
Originally published in Cook's Log, page 32, volume 41, number 3 (2018).
your email address will not be published