THE COMPASS THAT DIDN'T USE MAGNETIC NORTH
How Ralph Bagnold Used the Sun to Navigate the Desert
Eric Kelley · EMNAV Field Notes · October 7, 2026 · Historical Navigation
In the open desert, Ralph Bagnold encountered a navigation problem.
His compass could be wrong.
Not because the compass was necessarily defective.
The problem was the environment surrounding it.
The vehicle carrying the navigator contained metal. The engine, tools, equipment, weapons, and structure of the vehicle could influence a magnetic compass. In some areas, iron bearing geological formations could introduce another source of magnetic disturbance.
That presented a serious problem for someone attempting to navigate a vehicle across enormous stretches of desert.
Bagnold's solution was elegant.
He did not simply try to make the magnetic compass better.
He changed the reference.
Instead of relying on Earth's magnetic field, he used something his vehicle could not magnetically disturb.
The Sun.
Before the War
Ralph Alger Bagnold was a British Army officer, explorer, engineer, and later a scientist.
Years before the Second World War, Bagnold and other British officers stationed in Egypt began exploring the enormous desert regions west of the Nile by automobile. The Churchill Archives Centre describes those leave expeditions in Model T and Model A Fords across much of the Libyan Desert that had not been mapped. The Royal Society places the traverses in 1929 to 1932, with a last civilian expedition in 1938. The National Army Museum records a first east to west crossing of the Libyan Desert in 1932, using Model A Fords.
The environment exposed weaknesses in both vehicles and conventional navigation methods.
Roads could disappear.
Maps could be limited.
Landmarks might be separated by enormous stretches of terrain.
Soft sand could stop vehicles.
Water and fuel had to be carefully managed.
Distances were large enough that relatively small navigation errors could become significant.
Bagnold and his companions experimented with ways of making long range motor travel across the desert practical. Reduced tyre pressure on loose sand, and methods for crossing dunes, later became part of that practical knowledge. Navigation became one of the central problems they had to solve.
The Magnetic Compass Problem
Bagnold experimented with a magnetic aero compass supplied by the Royal Geographical Society.
The idea was reasonable.
Aircraft depended on magnetic compasses.
But a magnetic compass mounted in a vehicle presented another problem.
A magnetic compass responds to the magnetic field existing at the instrument.
The vehicle itself could alter that local magnetic environment.
The National Army Museum records that the compass was mounted on his vehicle, that he could not compensate it properly because of the magnetic influence of the car's metal, and that he was often forced to stop and take bearings away from the vehicle. Iron ore deposits sometimes found in desert regions affected the same instrument. The Royal Society puts the point in similar terms: outcroppings of iron rich rock, and even the mass of a vehicle engine, might cause a serious deviation in a magnetic compass.
The magnitude of that error depends on the magnetic environment. Metal near a compass is a possible source of error. It does not automatically make every magnetic compass useless.
This reveals a fundamental principle of navigation. A magnetic compass does not possess an independent knowledge of north. It measures a physical phenomenon. It responds to a magnetic field. If that field is disturbed, the indication can be disturbed as well. The same distinction, between the direction a needle indicates and true north, is the subject of Your Compass Doesn't Know True North.
Change the Reference
Bagnold could have treated the problem entirely as a magnetic compass problem.
Improve the compass. Improve the compensation. Move the compass. Shield it. Calibrate it.
Instead, he approached the problem differently.
What other independent directional reference was available?
The answer was above him.
The Sun.
The vehicle might influence the magnetic field surrounding a compass. It could not magnetically deflect the Sun.
That distinction became the basis of Bagnold's solution.
Bagnold Didn't Invent the Sun Compass
Bagnold did not invent the idea of navigating by the Sun, and his was not the only military sun compass. His achievement was adapting the sun compass to the particular demands of fast moving desert navigation.
That claim is narrower than it first sounds, and it is the historically defensible one. Solar direction finding was already an established part of navigation. The Royal Society describes the sun compass as an ancient method of holding a true course, and describes Bagnold as having refined it for the desert. Naval navigation manuals already treated the Sun as a directional body: apparent solar time, the equation of time, longitude, and solar azimuth were standard astronomical problems long before a desert vehicle needed a dashboard instrument. The Admiralty tradition belongs to that larger body of astronomical navigation. It does not, on the evidence used here, describe Bagnold's compass.
British and Commonwealth forces already had more than one sun compass by the Second World War. They were not the same instrument, and they were not used by the same procedure.
The Cole pattern is the best documented of them. The Science Museum Group holds a Cole pattern sun compass with a vertical gnomon, used for directing armoured vehicles in the Libyan desert. Royal Museums Greenwich holds a Cole Universal Sun Compass Mk 3 associated with J. H. Cole, the British survey of Egypt, and the North African campaign, calibrated from the equator to 39 degrees north. A related Evans Lombe instrument is also recorded. A regimental obituary credits T. B. A. Evans Lombe with further work on a sun compass that higher authority then named the Universal Sun Compass. A British Sundial Society account describes an Evans Lombe card instrument, later superseded by the metal Cole pattern. Those accounts do not make the two instruments identical.
The National Army Museum also publishes a November 1942 photograph captioned only as officers at sun compass practice. The caption does not identify the unit, and it does not identify the instrument as a Compass, Sun No. 3 Mk I. This article does not add either identification.
The classic Bagnold instrument was deliberately simpler. Bagnold's own 1945 description, quoted by the National Army Museum, is a knitting needle set vertically in the centre of a horizontal disc, three inches in diameter, graduated in 360 degrees, and rotated against a card of the Sun's azimuth. The vertical gnomon cast a shadow. The navigator compared that shadow with the calculated azimuth of the Sun. The shadow did not itself announce north. It became useful because the Sun's true azimuth was already known.
The operational difference was what the navigator could read after the vehicle turned. Some military sun compasses functioned primarily as course keeping instruments. If the shadow fell on the expected time graduation, the vehicle was still on the course that had been set. A change of direction could require a halt and a reset. Desert driving made that expensive. Dunes, rock, soft sand, obstacles, and corrections meant the vehicle changed direction constantly. The navigator needed the vehicle's actual true azimuth now, not only confirmation that an earlier course was still being held.
David Lloyd Owen, a veteran of the Long Range Desert Group, put the distinction in 1958, in the words quoted by the National Army Museum: Bagnold's compass showed the true bearing of the course followed at any moment, while the other types only made certain that if the Sun's shadow fell on the correct time graduation the truck was following a set course. A change of course meant halting and setting the compass again. That is a veteran's operational comparison, not a claim that every other sun compass was mechanically identical.
The design point follows from that. The clever part was not producing a shadow. The underlying astronomy was already known. The achievement was identifying the information the desert navigator actually needed, and making a simple instrument that supplied it while the vehicle was maneuvering. The question was not only, am I still on the course I set earlier? It was, what is my true azimuth now?
Indian personnel met this instrument in a different setting from India. The National Army Museum holds a Bagnold sun compass used by the Indian Squadron of the Long Range Desert Group, accession 1960-10-23-1. The museum dates the squadron to December 1941, raised from volunteers of the 2nd Royal Lancers (Gardner's Horse), Prince Albert Victor's Own Cavalry (11th Frontier Force), and 18th King Edward VII's Own Cavalry. That is a North African LRDG thread. It is not the same thing as sun compass practice in India, and this article does not merge the two.
What Bagnold Built
Bagnold later described the instrument in his own words, quoted by the National Army Museum from 1945:
"Simply, it was a knitting needle set vertically in the centre of a horizontal disc, three inches in diameter. The face of the disc was graduated in 360 degrees of bearing, and the disc could be rotated in its fixed mounting to follow the sun through the day from east through south to west, according to a card giving the sun's azimuth every 10 minutes of the day."
The gnomon cast a shadow across the instrument.
Mechanically, the idea was remarkably simple.
But the simplicity of the hardware can hide the sophistication of the navigation principle.
The shadow alone did not tell the navigator where north was.
The navigator also needed to know where the Sun should be.
That is the heart of the instrument.
The Sun Has an Azimuth
At any given location and moment, the Sun occupies a particular direction in the sky.
Its horizontal direction from true north can be expressed as a true azimuth.
That azimuth changes continuously as the Earth rotates.
It also depends on date and location.
Bagnold therefore needed more than a gnomon and a shadow.
He needed solar information appropriate to the journey.
His 1945 description refers to a card giving the Sun's azimuth every 10 minutes of the day. The National Army Museum also describes tables of different dates and latitudes, used to match the disc to the approximate azimuth of the Sun at regular times. The Churchill Archives Centre holds period instructions and correction material for the Bagnold sun compass, including file BGND C.24, but that instruction text is not published online. This article does not invent details from it.
The navigator therefore possessed two pieces of information: where the Sun should be, and where the shadow actually appeared.
The sun compass connected the two.
How the Principle Works
Imagine that the true azimuth of the Sun is known for a particular location and time.
The Sun illuminates the vertical gnomon.
The gnomon casts a shadow across the horizontal compass plate.
The shadow extends in the opposite horizontal direction from the Sun.
If the direction of the Sun is known, the relationship between the Sun, gnomon, and shadow can establish true direction.
The instrument therefore does not ask a magnetic needle where north is.
It asks the sky.
Two Different Physical References
This is the important distinction.
A magnetic compass observes Earth's magnetic field.
A sun compass observes an astronomical relationship.
Both systems attempt to answer a similar navigation question: what direction am I traveling?
But they derive the answer from different physical phenomena.
That means they do not necessarily fail for the same reasons.
What Made Bagnold's Compass Important
The practical advantage was not simply that the instrument could indicate direction from the Sun.
Bagnold developed a system useful aboard a moving vehicle.
That distinction mattered.
The National Army Museum notes that the Long Range Desert Group also experimented with other sun compasses. Those instruments were unaffected by magnetism, but they had to be reset in line with the Sun whenever the vehicles changed course, whether to cross rough terrain or to avoid enemy troops. That delayed the convoys.
Bagnold's instrument could be used while still driving. The disc was turned to match the approximate azimuth of the Sun, taken from tables of different dates and latitudes, at regular times. It was fitted to the dashboards of LRDG vehicles.
David Lloyd Owen, a veteran of the LRDG, later described the operational difference. The National Army Museum quotes him from 1958:
"The compass we used was invented by Bagnold, and the advantage that it gave us over the sun-compasses used by the rest of the Army lay in the fact that it showed the true bearing of the course followed at any moment, whereas the other types only made certain that if the sun's shadow fell on the correct time-graduation the truck was following a set course. This meant that if one had to change course for any reason (and this happened all the time in rough country or sand dunes), the truck had to be halted and the compass set again."
Desert travel rarely meant driving along one uninterrupted straight line.
A driver might maneuver around dunes.
Soft sand might force a detour.
Broken terrain might require another change of direction.
Military operations could impose still more changes.
The navigator needed to know the direction actually being traveled, not merely the direction originally intended.
That made the sun compass particularly useful for vehicle dead reckoning.
The Compass Does Not Know Where You Are
This distinction is fundamental.
A compass provides directional information.
It does not automatically provide position.
Even a perfectly accurate directional observation cannot tell a navigator where he is unless it is combined with other information.
For vehicle dead reckoning, the navigator needed a known starting position, direction, distance, and a record of significant course changes.
Conceptually: starting position, plus direction, plus distance, plus course changes, equals a running position estimate.
The final word matters.
Estimate.
Dead Reckoning Across the Desert
The National Army Museum records that LRDG navigators could continuously log bearing and mileage regardless of course changes. That record permitted a reliable plot of the preceding route at every halt.
This becomes particularly important when the vehicle cannot follow the intended route directly.
Suppose a planned course runs west.
A dune field forces the vehicle southwest.
Later the driver turns northwest to regain the intended line.
The sun compass may correctly indicate both directions.
But the compass alone does not calculate the resulting position.
The navigator has to account for the movement.
A Correct Azimuth Can Still Produce a Wrong Position
This is an important advanced navigation principle.
Direction is only one variable.
Distance is another.
Course changes introduce additional geometry.
Errors in any of them can propagate into the dead reckoning position.
A navigator can therefore make a perfectly reasonable compass observation and still eventually be wrong about position. How a small angular error grows with distance is the subject of How Far Off Course Are You?
This is why advanced navigation cannot be reduced to simply following an azimuth.
Navigation is also the management of uncertainty.
The Long Range Desert Group
When the Second World War reached North Africa, Bagnold's years of desert exploration suddenly had military value.
The National Army Museum dates the founding of the Long Range Desert Group to July 1940, under Major Ralph Bagnold of the Royal Engineers. The Churchill Archives Centre records that the proposal was approved in 1940, and that Bagnold was given six weeks to prepare the unit. The Royal Society notes that a troop ship collision took him away from an East Africa posting and back to Cairo, where he established the group.
The LRDG operated across enormous areas of desert, conducting reconnaissance and raids deep behind enemy lines.
Its capability depended on much more than a compass.
Vehicles had to survive.
Fuel and water had to be managed.
Communications mattered.
Mechanical knowledge mattered.
Maps mattered.
Distance mattered.
Desert experience mattered.
Navigation connected all of them.
The sun compass was one component of a larger system that allowed experienced navigators to move vehicles across terrain where conventional road based navigation offered little help.
The LRDG and the SAS
The LRDG's desert mobility also supported Special Air Service operations in North Africa.
The National Army Museum states that the LRDG drove Special Air Service raiders behind enemy lines.
The historical significance is not that the sun compass somehow navigated an SAS patrol by itself.
It did not.
Its importance was that it provided experienced desert navigators with a practical directional reference that could be integrated with maps, distance measurement, route logging, and dead reckoning.
The instrument was part of the navigation system.
It was not the navigation system.
True North Instead of Magnetic North
There is another important consequence of changing the reference.
A magnetic compass fundamentally observes Earth's magnetic field.
The direction it provides therefore has a relationship with magnetic north.
A solar navigation system can establish direction relative to true north through astronomical geometry.
Lloyd Owen's point was exactly that: Bagnold's instrument showed the true bearing of the course being followed.
That removes magnetic declination from the fundamental astronomical reference.
But it does not eliminate error.
It changes the sources of error.
The Error Budget Changes
Consider some of the uncertainties affecting a magnetic compass: magnetic declination, local magnetic interference, instrument error, reading error.
Now consider a solar directional system: time error, incorrect date, incorrect location information, incorrect solar data, instrument leveling, gnomon alignment, shadow definition, reading error.
The second system does not magically eliminate uncertainty.
It substitutes a different collection of uncertainties.
This leads to one of the most important principles in advanced navigation:
Every navigation system has an error budget. Changing the reference changes the error budget.
Time Becomes a Navigation Input
A magnetic compass normally does not care whether it is 0800 or 1000.
A sun compass does.
The Sun moves continuously across the sky.
Its azimuth changes with time.
That means the solar information used by the navigator must also change.
A solar direction calculated earlier in the day eventually becomes an outdated reference. Bagnold's own description used a card of solar azimuth at ten minute intervals.
The navigator has traded magnetic dependency for astronomical dependency.
That is not necessarily better or worse.
It is different.
The Instrument Did Not Eliminate Navigation
The Bagnold sun compass did not calculate the vehicle's position.
It did not measure distance.
It did not eliminate dead reckoning error.
It did not remove the need for maps.
It did not work without a useful view of the Sun.
It did not make an inexperienced navigator into an experienced one.
Its contribution was narrower and more important.
It provided a practical directional reference that did not depend on the magnetic environment surrounding the vehicle.
Bagnold Was Not the Only Sun Compass
Bagnold's instrument was not the only military sun compass.
Other designs approached the same astronomical problem differently.
A useful period comparison appears in U.S. War Department TM 5-9422, Maintenance Manual and Operating Instructions for the Universal Sun Compass, Abrams Model SC-1, dated 1 November 1943.
The instrument documented in TM 5-9422 is not Bagnold's compass. It is an American Abrams instrument. Do not read the manual as a Bagnold manual.
Its value here is comparative.
The American instrument also depended on true direction, date, latitude, time, a gnomon shadow, and the known movement of the Sun. Its operating instructions required magnetic or grid azimuths to be converted to true azimuth before use. The operator accounted for date, latitude, and local apparent time. The shadow relationship then provided the astronomical directional reference. The manual also required the instrument to be leveled, and it told the operator to reset the shadow reference at regular intervals, about every 15 minutes, because the Sun moves.
The mechanical implementation differed.
The physical reference was still the Sun.
Local Apparent Time
Period military sun compass instructions reveal how much information can exist behind a simple analog instrument.
Ordinary clock time is not automatically local solar time.
Standard time covers an entire time zone.
Longitude changes the relationship between clock time and local solar time.
The equation of time introduces another correction between mean solar time and apparent solar time.
TM 5-9422 therefore included procedures for converting ordinary time into local apparent time. Its method included a longitude correction of about four minutes per degree of longitude, added if the position was east of the time meridian and subtracted if west, and then a separate equation of time correction taken from the date.
Those two corrections are not the same thing. Longitude places the observer inside the time zone. The equation of time accounts for the difference between mean solar time and the apparent Sun.
The details belong to a different sun compass, but the larger lesson applies broadly.
Analog celestial navigation often trades electronic dependency for knowledge.
The hardware may be simple.
The information behind it is not.
The Sun Is a Moving Reference
A sun compass cannot treat the Sun's direction as fixed throughout the day.
The Sun's azimuth continually changes.
Bagnold's navigators therefore needed updated solar directional information. His card gave azimuth every 10 minutes. Other military sun compass designs incorporated their own methods for dealing with the same problem. The Abrams instrument in TM 5-9422, again a different compass, reset its shadow reference about every 15 minutes.
The mechanical implementations differed.
The astronomical requirement did not.
The sky is moving.
The navigator must account for that movement.
Why Not Just Move the Magnetic Compass?
That remained an option.
Bagnold had taken magnetic observations away from his vehicle.
But consider the operational difference.
Stop. Move sufficiently far from the vehicle to reduce its magnetic influence. Take the observation. Return to the vehicle. Continue. Repeat whenever another clean observation is required.
A practical vehicle mounted sun compass changed that process.
The directional reference could remain with the vehicle.
The Simplest Part Was the Hardware
This may be the most interesting feature of the entire system.
The physical principle could be implemented with remarkably simple components: a horizontal angular scale, a vertical gnomon, and a shadow.
But the navigator still needed to understand true direction, solar movement, time, distance, course changes, dead reckoning, and accumulated error.
The instrument did not replace navigation knowledge.
It converted that knowledge into something useful in the field.
From Bagnold to EMNAV
More than eighty years later, the design lesson remains relevant.
Modern navigators possess extraordinary technology.
GPS can calculate position almost instantly.
Digital maps can continuously display that position.
Electronic compasses can provide heading.
Astronomical software can calculate the Sun's position in seconds.
But the underlying navigation problem has not changed.
A navigator needs trustworthy references.
No instrument should be confused with the thing it measures.
A magnetic compass does not create north.
GPS does not create position.
A map does not create terrain.
Each is a system for observing, calculating, or representing something that exists independently of the instrument.
The EMNAV Navigator
The EMNAV NAVIGATOR includes a Bagnold style sun compass capability based on the same fundamental principle.
The Sun provides an independent astronomical directional reference.
The NAVIGATOR is not a reproduction of Bagnold's original instrument.
It integrates the underlying solar navigation principle into a modern analog navigation tool using contemporary land navigation conventions.
That distinction matters.
The historical Bagnold sun compass demonstrates the principle.
The NAVIGATOR applies that principle within a different instrument designed for modern field navigation.
When One Reference Fails, Change the Reference
Bagnold's most important lesson may not be the physical compass at all.
It may be his approach to the problem.
The magnetic reference was unreliable in the environment where he needed it.
Instead of assuming that magnetic north had to remain the solution, he asked what other independent reference was available.
The Sun provided one.
That principle extends far beyond desert navigation.
If one source of navigation information becomes unreliable, do not automatically abandon navigation.
Ask what other independent reference is available.
Terrain. The Sun. Stars. A known point. A linear feature. Elevation. Time. Distance. Another observation.
A different physical reference can provide information the failed reference cannot.
The Desert as a Laboratory
The desert exposed navigation weaknesses that could remain hidden elsewhere.
Small directional errors became significant across long distances.
Distance errors accumulated.
Detours mattered.
Sparse landmarks reduced opportunities to confirm position.
The vehicle itself complicated magnetic navigation.
The environment forced navigators to understand what their instruments were actually measuring.
That is why Bagnold's solution remains interesting.
The sun compass was not simply an ingenious wartime gadget.
It was a response to a failure of assumptions.
The conventional directional reference was unreliable.
So Bagnold changed the reference.
The Bigger Lesson
The Bagnold sun compass can look primitive beside a modern navigation system.
That is deceptive.
Its elegance lies in how little hardware was required to connect the navigator to an independent external reference.
A vertical object casts a shadow.
The Sun has a calculable direction.
An angular scale measures the relationship.
The navigator supplies the knowledge connecting them.
No magnetic needle is required.
No satellite is required.
No battery is required.
The system works because the navigator understands what the instrument is actually measuring.
Bagnold did not make north.
He found another way to observe it.
From Field Notes to the Field
Advanced land navigation, coming soon
EMNAV founder Eric Kelley is currently developing Advanced Land Navigation, a book for experienced navigators who want to go beyond basic map and compass instruction.
The book explores advanced field techniques, navigation error, dead reckoning, solar and celestial navigation, historical and forgotten methods, terrain analysis, field expedient techniques, and the underlying geometry that makes navigation work.
Historical instruments such as the Bagnold sun compass are particularly valuable because they reveal how navigators solved difficult problems with limited technology.
Advanced Land Navigation will examine not only how techniques such as these worked, but why they worked, where they could fail, and which underlying principles remain useful to modern navigators.
Many of the techniques explored in EMNAV Field Notes will be investigated in greater depth, tested, refined, and expanded for Advanced Land Navigation.
Follow EMNAV Field Notes for updates as the book develops.
Sources and Further Reading
National Army Museum, "Desert innovator: Bagnold's sun-compass," nam.ac.uk/explore/sun-compass. Prewar exploration, the Royal Geographical Society aero compass, vehicle and iron ore interference, Bagnold's 1945 description of the instrument, solar azimuth cards and tables, vehicle use, bearing and mileage logs, the LRDG, the SAS relationship, and David Lloyd Owen's 1958 description.
Churchill Archives Centre, University of Cambridge, Bagnold Papers research guide, archives.chu.cam.ac.uk/collections/research-guides/bgnd. Prewar desert exploration, specialist equipment including the sun compass, formation of the LRDG, and later scientific work. Instructions for the Bagnold sun compass are held as BGND C.24 and are not quoted here because the full text is not online.
The Royal Society, "Sea of Sand," royalsociety.org/blog/2019/08/sea-of-sand. Exploration, desert travel, magnetic interference from iron rich rock and the vehicle, and the sun compass as a refined ancient method.
U.S. War Department, TM 5-9422, Compass, Sun, Universal Type, Abrams Model SC-1, 1 November 1943. Comparative technical material only. This is not Bagnold's instrument and not a Bagnold manual. Used for true azimuth conversion, local apparent time, the four minute longitude correction, the equation of time, leveling, and the need to update a moving solar reference.
Science Museum Group Collection, Cole pattern sun compass, 1939 to 1945, vertical gnomon, used for directing armoured vehicles in the Libyan desert. Royal Museums Greenwich, Cole Universal Sun Compass Mk 3, object NAV0398, associated with J. H. Cole, the survey of Egypt, and the North African campaign. Queen's Royal Hussars Museum obituary of Col T. B. A. Evans Lombe, on further work that became known as the Universal Sun Compass. Malcolm Barnfield, "The Sundial Goes to War," British Sundial Society, on the Evans Lombe card instrument and the Cole pattern. National Army Museum accession 1960-10-23-1, Bagnold sun compass of the Indian Long Range Desert Squadron. The November 1942 National Army Museum photograph is cited only by its published caption, officers at sun compass practice. No Compass, Sun No. 3 Mk I manual, and no 10/16th Punjab identification, was verified for this revision.