Before radio, armies and governments sent messages across open country with light. A mirror heliograph flashed sunlight in Morse code. Semaphore towers and hand flags spelled out letters by the position of arms or flags, and lamps carried flashed codes after dark. Every one of these systems needed a clear view between stations, so the people who built them put stations on high ground and chained them together, accepting the delay and risk that each relay added. A radio administrator siting a VHF repeater or a microwave hop works under those same constraints. This piece covers the equipment, what the reported ranges actually rest on, the US Army’s 1886 network in Arizona and New Mexico, and what carries over to radio.

Sunlight and a key: how a heliograph sent Morse

The heliograph used a flat mirror to reflect sunlight toward a distant station. The operator broke the beam into long and short flashes, and the receiving operator read them as Morse code. The design most associated with military use is usually credited to Henry Mance, a British telegraph engineer working in the Persian Gulf region, and the date commonly given is 1869. Reflected sunlight had already been used over long distances before that. Surveyors used the heliotrope, an instrument Carl Friedrich Gauss developed in the early nineteenth century, to make distant triangulation points visible, so people knew reflected sunlight could be seen across many miles before anyone used it to carry messages. Mance’s contribution was a portable instrument that a soldier could aim accurately and key quickly in the field.

The field instrument sat on a tripod and had a mirror, a sighting arrangement and a key. The operator used a sighting vane or rod to align the mirror so that the reflected beam fell on the receiving station. In some patterns, pressing the key tilted the mirror slightly and swept the beam off the distant station. In others, a shutter interrupted the beam. When the sun was behind the operator, one mirror could not throw light forward, so a second mirror bounced the sunlight onto the first. The receiving station watched through a telescope, and the observer either wrote the dots and dashes down directly or called them out to someone who recorded them.

Two features of the physics shaped how the instrument was used. The reflected beam spreads at roughly the apparent diameter of the sun, about half a degree, so at 30 miles it is on the order of a quarter mile wide. That is wide enough to aim by hand, and narrow enough that an observer well off the line between the two stations could not see the flashes. The sun also moves across the sky at about a quarter of a degree per minute, which is roughly the beam’s half-width, so the operator had to keep adjusting the mirror throughout a message or the beam would drift off the receiving station. The instrument was useless without direct sun, which is part of why it suited the dry, clear country of Afghanistan, South Africa and the American Southwest far better than places with frequent cloud.

Arms, shutters and flags: semaphore on towers and in the hand

Mechanical semaphore came well before the heliograph. In France in the 1790s, Claude Chappe and his brothers built a chain of towers, each carrying a pivoting beam with movable arms at its ends. An operator in each tower watched the next tower through a telescope and reproduced the position he saw. The first line, between Paris and Lille, is usually dated to 1794. The commonly told story says that news of the French recapture of Conde from Austrian forces reached Paris over that line in about an hour in August 1794. The recapture itself is well attested, but I have not checked the transmission time against the line’s own records, and retellings state it with different degrees of precision. The French network later grew into a national system that operated for decades until the electrical telegraph displaced it.

In the mid-1790s the British Admiralty built its own line from London to the coast at Deal. It used a frame of six shutters, each of which could be opened or closed to form combinations, a design associated with Lord George Murray. Later Admiralty lines switched to semaphore arms. On all of these systems, station spacing depended on terrain, tower height and the telescopes available, and stations were typically several miles to a dozen or more miles apart. Each station needed operators on duty through the daylight hours and a clear view in both directions. As a result, the capacity of the whole line was set by its weakest station on its worst day for visibility.

Hand flags put the same principle into a soldier’s arms. In the United States, Albert J. Myer, an Army surgeon who became the Army’s first signal officer in 1860, devised the wig-wag system. In wig-wag, a single flag waved to the left or right of vertical stood for elements of a code, and torches replaced the flag at night. Wig-wag saw heavy use during the Civil War from hilltops and purpose-built signal towers, and Myer’s manual of signals went through several editions after the war. Two-flag semaphore, in which the angle of each arm marks a letter, became standard in navies and later in scouting. Flags were cheap and needed no sun, but their range was limited by what a telescope could resolve, and an enemy watching from the side could read them as easily as the intended station could.

After dark and under cloud: signal lamps and their limits

Lamps extended visual signalling into the night and into conditions that blocked the sun. A signal lamp concentrated its light with a lens or reflector and used a shutter or key to make Morse flashes, so the same operators and codes could work the heliograph by day and the lamp at night. Armies used oil and later acetylene lamps in the late nineteenth century, and in the twentieth century navies adopted electric shuttered lamps, of which the Aldis lamp is the best known by name. Navies kept light signalling long after radio arrived because a narrow light beam could pass a message between ships in company without producing a radio emission that an enemy could detect or take a bearing on.

A lamp did not depend on the sun, so it worked through the night, and a link that went dark under cloud could sometimes carry on with lamps. Its light source, however, was far weaker than direct sunlight, so over the same path a lamp generally reached a shorter distance than a heliograph did on a clear day, and signalling instructions of the period treated lamp ranges as shorter. Fuel, wicks, mantles or batteries all had to be carried up to a hilltop station and kept dry, which added to the supply burden that every visual station already carried.

Every kind of visual signalling failed in the same weather. Haze, dust, smoke, rain and fog shortened the range or stopped a link entirely. Heat shimmer near the ground could make a flash hard to read even when it was visible. A route that worked at mid-morning might stop working in afternoon dust, and the operators had no way to tell the distant station why the link had gone quiet. For that reason, commanders who relied on visual links kept mounted messengers in service alongside them for the hours when the signal stations could not see each other.

Reported ranges: what governed distance and why the figures disagree

Popular accounts of the heliograph give ranges that differ by a factor of several, and most of the disagreement comes from mixing different kinds of figure. A working range is the distance a unit could expect to signal reliably with a standard instrument in normal conditions. A record is the longest link anyone achieved under ideal conditions with carefully chosen sites. An anecdote is a distance someone remembered. A retelling that says the heliograph could reach 50 miles, or 100, or more than 180 may be quoting any of these without saying which.

The physics explains why the figures spread so widely. The brightness of the flash at the receiving station depends mainly on the mirror’s area and falls off with the square of the distance, so doubling the distance requires four times the mirror area for the same apparent brightness, all else being equal. Clear dry air, a high sun, a path elevated above the dusty layer near the ground, a good telescope and a skilled observer all extend range. Haze shortens it more than any practical increase in mirror size can recover. Field instruments were built around small mirrors sized to be carried by a man or a pack animal, and larger mirrors suited fixed stations or deliberately long links.

The longest link usually cited was a test in the 1890s between mountain peaks in Utah and Colorado, at a distance of more than 180 miles, which encyclopedia entries and popular histories attribute to the US Army Signal Corps. I have not seen the original report of that test, so I treat the figure as a reported record rather than an established one. Even if the figure is taken at face value, it describes the best case between two very high peaks in exceptionally clear air and says little about what a field station could expect. The British and American signal manuals of the period generally gave working ranges for their standard instruments under stated conditions, and those figures were far shorter than the records. I have not reproduced specific numbers here because the manuals went through many editions and the instruments changed between them. Anyone who needs a figure should take it from a named edition of the British Army’s signalling instructions or the US Signal Corps manual and quote it along with the edition and the conditions it assumes.

Arizona and New Mexico, 1886: the heliograph network against Geronimo

The best-known American use of the heliograph came in 1886, during the final campaign against Geronimo and the Chiricahua Apache who were with him. In March of that year, after talks at Canyon de los Embudos in northern Mexico, Geronimo and a group of followers left before the surrender agreed with General George Crook was completed. Crook asked to be relieved and was replaced in April by Brigadier General Nelson A. Miles, who took command of the Department of Arizona. One of Miles’s early measures was a network of heliograph stations on peaks across southern Arizona and southwestern New Mexico.

The stations were sited to watch the valleys and passes the Apache used, to report sightings, and to pass messages between posts and field columns faster than couriers could ride. Each station needed a crew who could work the instrument and defend themselves, and in that country the crew also needed water and rations carried up to the peak. Accounts of the network give different station counts and different totals for messages handled. I have not reconciled them against Miles’s own report, so I give no single figure here. The department reports for that year were published with the annual report of the Secretary of War, and anyone who needs the station list and message counts should look there rather than rely on a secondary summary.

The dramatic version of this story, in which the heliograph ran Geronimo down, is a myth, or at best a large overstatement. Lieutenant Charles Gatewood, accompanied by two Apache scouts, Martine and Kayitah, located Geronimo’s band in Mexico in late August and negotiated with him. Geronimo surrendered to Miles at Skeleton Canyon, Arizona, in early September 1886. Captain Henry Lawton’s column had spent months pursuing the band across rough country in Mexico, and Mexican forces were also in the field. Miles’s own accounts gave the heliograph network prominent credit, while later historians have generally placed more weight on the scouts and the negotiation.

What the network did show was that a visual relay chain could cover a large stretch of mountain and desert and keep carrying traffic over many hops for months, in country whose weather suited it. It also showed the costs. Every peak had to be occupied and supplied, and a dust storm or cloud bank over a single station broke the chain at that point, so the speed advantage over a courier depended on short messages and on every station in the route being able to see its neighbors at the moment the message was sent.

Relay stations: siting, staffing and the delay each hop added

A visual link was a chain of point-to-point hops, and planning one was nearly the same problem a microwave engineer faces today. The planner needed sites that could each see the next, with some margin so that ordinary haze did not close the path, and the route had to bend around terrain that blocked the view. Every added hop meant another site to occupy and supply. Where no direct path existed, an intermediate peak station existed only to relay, the visual equivalent of a microwave repeater site built for no purpose except getting around a ridge.

Relaying cost time. Under the usual practice, an intermediate station received the whole message and wrote it down before sending it on, rather than passing each flash through as it arrived. A message crossing five hops therefore took at least five times as long to transmit as it would have over a single link, plus the time each station spent getting the next one’s attention. Heliograph keying was slow compared with a telegraph line, because each flash had to be distinct enough to read through a telescope at a distance, so long messages were expensive and signallers were trained to keep them short.

Relaying also multiplied the chances of error and of outright failure. Every station that copied and resent a message could misread a letter or drop a word, and nothing in the chain corrected an error once it had been passed along. Every station was also a place where weather or an absent operator could stop traffic. The same arithmetic applies to a modern backhaul chain. If each of five hops in series is available 99.9 percent of the time and their failures are independent, the chain as a whole is available about 99.5 percent of the time, which is roughly five times the downtime of a single hop. Those figures are an arithmetic illustration rather than a specification for any real system, but they show why the number of hops in series matters as much as the quality of each one.

Horizon and Fresnel zone: where light and radio share rules and where they part

Visual signalling and VHF or microwave radio share a basic constraint, which is that the curve of the earth and the terrain between two stations can block the path. A station on low ground sees a nearby horizon, and height is what pushes that horizon out, which is why the heliograph planners put their stations on mountaintops. For radio under standard atmospheric conditions, a common rule of thumb gives the distance to the radio horizon in miles as the square root of twice the antenna height in feet. Over smooth ground, the usable distance between two antennas is roughly the sum of their two horizon distances. A repeater antenna 100 feet up and a portable held at about 6 feet give roughly 14 plus 3.5 miles, or about 17 to 18 miles over flat terrain, and buildings, trees or hills shorten that further.

Light and radio part ways over how much clear space the path needs. Radio energy does not travel only along the thin line between two antennas. A significant share of it travels within an ellipse around that line called the first Fresnel zone, and an obstruction that intrudes far enough into the zone weakens the signal even when the two antennas can see each other. The width of the zone depends on wavelength and path length. The standard formula, applied at the midpoint of a 20-mile path, gives a first Fresnel zone radius of roughly 400 feet at 150 MHz and roughly 65 feet at 6 GHz, while for visible light the radius is a few inches. A heliograph beam could graze a ridgeline and still be read, whereas a microwave hop over the same ridge needs tens of feet of clearance, and a commonly cited design rule calls for at least 60 percent of the first Fresnel zone to be clear.

VHF also differs from light in the opposite direction. Its longer wavelength bends over and around obstacles to a degree light cannot, which is why a VHF portable often works some distance into a terrain shadow that no visual signal could reach, and why VHF coverage is better predicted with a propagation model than by asking whether two points can see each other. Under standard conditions, atmospheric refraction bends radio somewhat more than it bends light. Unusual layering in the atmosphere can occasionally carry VHF or microwave signals well beyond their normal range, or bend a microwave beam off its target, and heavy rain absorbs microwave energy at the higher frequencies, generally becoming a design concern above roughly 10 GHz. A visual signaller could at least see haze rolling in, but radio’s weather effects are invisible from the station and have to be found in path studies and system logs.

Procedure on a relayed link: word counts, repeats and what radio inherited

Signallers developed procedure to protect messages that had to pass through several hands. The general practice, taken from telegraph work, gave each message a number and a header and counted its words, so a receiving station could confirm that the count matched and ask for a repeat of the missing word rather than the whole message. Receiving stations acknowledged each message so the sender knew it had arrived. Operators used short procedural signals for calling, waiting, repeating and ending, so the chain could be managed without plain-language conversation that would take more flashes than the message itself.

Much of that survives in radio. Amateur radio’s formal message format, used in emergency traffic nets, still carries a message number, a precedence, the station of origin and a check that is the word count of the text. An operator relaying a formal message passes the header along unchanged so the final receiver can compare the count. Military radio procedure draws on prowords and acknowledgment rules from the visual and telegraph era. Public safety voice radio mostly dropped this formality, which is reasonable for short dispatch traffic between two parties who can hear each other directly. It is less reasonable for a message that has to pass through a relay operator on a large incident or through a patched gateway, because each handoff can lose a word in the same way a heliograph relay station could.

The deeper inheritance lies in site planning. Signal officers chose stations from maps and reconnaissance and then confirmed each link by trying it. A radio system is planned the same way, with terrain databases and propagation software, and then verified with drive tests and path measurements. The heliograph era also established that a relay site is only as useful as its worst operating day allows, and that a chain of hops needs a planned fallback for the hour when one station goes dark. Both points apply directly to the microwave and fiber backhaul behind a multisite system and to the temporary relays a COML sets up on a large incident, where a single mountaintop portable repeater can carry every tactical channel on the fire.

What to do at your agency

• The radio system administrator should pull the path profile for each microwave hop in the backhaul, confirm that it shows Fresnel zone clearance at the design criterion and the date of the last vegetation or construction survey along the path, and list any hop that has no profile on file.
• The radio system administrator should write a one-page list of every site that sits in series between the dispatch consoles and each transmitter site, mark which ones are relay-only, and give the list to the communications center manager with a line for each site stating what coverage is lost if that site fails.
• The emergency management director, or whoever maintains the communications continuity plan, should add one paragraph to that plan naming the fallback path for traffic if the busiest relay site goes off the air and naming the role that decides to switch to it.
• The radio shop lead should confirm, for each remote or mountaintop relay site, the date of the last generator test and fuel check and whether winter or storm access is documented, and should record any gaps in the site maintenance log.
• The COML or communications training officer should add one exercise to the next scheduled communications or ARES drill in which a written message of thirty to fifty words is relayed through two intermediate stations, then compare the received text and word count against the original and record every error.
• The training officer who maintains history or orientation material should review any slide that credits the heliograph with capturing Geronimo or gives a single heliograph range without a source, and should correct it to name the source and describe the figure as a reported range.

Takeaways

• The heliograph sent Morse code by flashing reflected sunlight. The field design is usually credited to Henry Mance around 1869, and the instrument depended entirely on direct sun and constant re-aiming as the sun moved.
• Chappe’s tower semaphore in France in the 1790s, the British Admiralty’s shutter and semaphore lines, Myer’s wig-wag and two-flag semaphore all relied on clear sight lines between stations, with station spacing set by terrain and tower height.
• Signal lamps carried the same codes at night and under cloud, but they generally reached shorter distances than a heliograph in full sun, and navies kept them because light leaves no radio emission for an enemy to detect.
• Heliograph range figures in popular accounts mix working ranges, records and anecdotes, so any figure should be quoted with its source, its edition or date, and the conditions it assumes.
• Miles built a heliograph network across southern Arizona and New Mexico in 1886, but the claim that it ran Geronimo down is a myth, because the surrender came through Gatewood’s negotiation with the help of two Apache scouts after months of pursuit.
• Each relay hop added transmission time, a site to supply and a fresh chance of error, and the same arithmetic means a chain of microwave hops in series has more downtime than any single hop in it.
• Light and radio share the horizon problem, but microwave also needs Fresnel zone clearance of tens of feet that a heliograph never needed, while VHF bends into terrain shadow where light cannot reach.
• Word counts, numbered messages and acknowledgments developed to protect relayed visual traffic, and they survive in the formal amateur radio message format that agencies can use whenever traffic must pass through a relay operator.