Tech level is early Rome, China, Ethiopia, etc. ~ first century BC. Unlike history, there's peace and cooperation among all the benign monarchies of the whole Earth. But it's hard to get daily communication from (now) peaceful Rome to (now) peaceful imperial China, for example, and back to Ethiopia.

Horses are not fast enough, heck even the mighty camel is not to the emperor's satisfaction.

What would be a faster and more efficient way to rapidly transfer regular (say "once-a-week") messages across vast land distances, given plenty of existing resources, 1BC tech level, a few decades of research beforehand, humankind cooperation, and our expertise as WorldBuilders?

Note: Keep in mind, there's almost entirely "peace on Earth" among the 300 million or so global population, and since it's land-transport, I can omit the Americas, unless you have a clever solution to that, focusing on the "over land" portion.

  • 4
    $\begingroup$ I think you're looking for low latency, not regularity. I can have multiple messages in transit to permit myself 1 message a week over an arbitrary empire. However, if I need a reply from my message, I am interested in the "ping" time of my empire, which might be 2 weeks to send a message to China and get the reply. $\endgroup$
    – Cort Ammon
    Dec 19, 2015 at 3:39
  • 4
    $\begingroup$ Relevant, but not a duplicate IMO: How quickly and accurately should news travel in a rural medieval setting? (Some of the answers should definitely be useful, but the premise for the question is different.) $\endgroup$
    – user
    Dec 19, 2015 at 15:01
  • 1
    $\begingroup$ @AndreiROM: Your point about encryption is interesting; I think you meant encoding, but it brings up a good point. A rider can be intercepted and their message confiscated. Tower communications might be intercept-able through a literal MiTM attack, but it would require a knowledge of the encoding technique. Both could use data encryption techniques. $\endgroup$
    – Wingman4l7
    Dec 21, 2015 at 2:07
  • 1
    $\begingroup$ @tmh -- the Royal Road was protected by Persian rulers and later used by the Romans. On this road couriers, riding in relays, could travel 1,677 miles (2,699 km) in seven or nine days. This how fast real horse mounted couriers carried the rulers message. -- historyofinformation.com/expanded.php?id=162 $\endgroup$ Dec 21, 2015 at 10:13
  • 1
    $\begingroup$ @GaryWalker That would give us 25-32 days for 6,000 miles. Much better. The heliograph towers still seem to be the fastest solution, though. $\endgroup$
    – tmh
    Dec 21, 2015 at 10:42

6 Answers 6


A string of heliograph towers.

Traditional semaphore lines are unfortunately out; while the construction of the towers is relatively low tech, they rely on the use of optical telescopes, which weren't invented until the early 17th century.

While heliographs weren't invented & put into regular use until the late 19th century, the basic concept is very simple. Wikipedia claims a flash could be seen from 30 miles with the naked eye under ordinary conditions; max range was ~10 miles per inch of mirror diameter -- with a record of 183 miles done by mirrors of only 8 inches square!

Wikipedia tells me that polished copper mirrors appeared 3000-4000 BCE, and bronze mirrors showed up ~2000 BC. Metal-coated mirrors showed up in the 1st century AD, which aligns with your time-frame. Their quality was likely inferior to those used in 19th century heliographs, so their size would have to be correspondingly increased -- but working from a max range of ~30mi visibility by the naked eye, it would be quite feasible to use mirrors even if the quality was an order of magnitude worse (~10mi per 10 inch diameter = > 30 inch mirror).

The biggest issue would probably be building, maintaining, and staffing the towers, which would have to be spaced at regular intervals across all manner of terrain. Regular supply runs for particularly remote towers might prove an expensive challenge. Weather would also be an issue, probably forcing some towers to be built closer together in some regions due to poor visibility. Closer intervals means more towers -- which means more expense.

Actual semaphore lines may have used rotating indicators, symbols made up of straight lines at mostly right angles, articulating arms, or a series of shutters. However, a heliograph is limited to two states -- "on" or "off". Interesting to consider would be how different types of languages (pictographic / logographic / phonographic) would constrain or aid the development of the symbols used to encode messages for transmission. Perhaps there would even be competing standards!

  • $\begingroup$ Why did you not link directly to Wikipedia? $\endgroup$
    – celtschk
    Dec 19, 2015 at 10:27
  • $\begingroup$ @celtschk: Wasn't paying attention, keep forgetting that that browser extension modifies the URL. Thanks for the heads-up; fixed! $\endgroup$
    – Wingman4l7
    Dec 19, 2015 at 17:57
  • 1
    $\begingroup$ @PyRulez: Yeah, I left out the finer points on the creation of a signalling code as per "necessity is the mother of invention". Those curious can simply look at the historical record on what semaphore lines came up with. I would find it interesting to speculate on how differing languages (pictographic / logographic / phonographic) might constrain or aid in coming up with workable symbol systems; I might add a note to that effect into my answer. $\endgroup$
    – Wingman4l7
    Dec 19, 2015 at 22:05
  • 3
    $\begingroup$ @PyRulez: A simple Morse code equivalent is sufficient; I don't think you need computer science to develop that (did Samuel Morse know anything about computers or binary?) $\endgroup$
    – celtschk
    Dec 20, 2015 at 11:51
  • 1
    $\begingroup$ @tmh, the Chappe semaphore reportedly could cover 10 miles per minute on the Paris-Lille run. At that rate, Rome-Chang'an would take between eight and ten hours, depending on how straight the route was. $\endgroup$
    – Mark
    Dec 20, 2015 at 20:20

The Chinese actually used a system of beacon towers that predate your scenario. These were installed along the great wall, using fire by night and smoke by day. Actually the beacon tower system predates the great wall and were in use around 200 BC.

The message communicated was very simple though. Basically the signal meant that invaders were approaching. Modulation and coding of the signal would be necessary to carry arbitrary messages.

Any of the signal tower methods have some weaknesses, mainly expensive to build and man, the man on duty can fall asleep, etc. and fail in duty. Men can be attacked or bribed to render the system inoperative.

For comparison the book of Exodus records that the children of Israel were guided by a pillar of fire / smoke for night / day usage. This is often dated 15th century BC (though there is disagreement). Even if the Exodus story is a fable, the account of using fire/smoke to guide was written centuries earlier than even the Chinese beacon towers.


Pigeon Express

Messenger birds have been around for a long time, and can travel much faster than a man on horse.
The Romans were good at roads, and with peace comes trading, so it would be in everyone's best interest to establish a good route with inns every 30-40 miles. These would also be great places to house messenger birds, so a message could be written in Rome and sent down the line from bird to bird until it got to the emperor of China.

While kind of a joke, there is a protocol for IP over Avian Carrier.
In a race between a pigeon, a car, and Australia's Telestra internet provider, the pigeon transfered a 700 mb video file in 1 hour 5 minutes. The car took 2 hours 10 minutes, and the file upload was estimated to take 4 hours, but kept losing connection.

While slower than light, their carrying capacity is pretty high, and things like maps, diagrams, mathematical calculations, pictures and other non alphanumeric items are much easier to transmit.

With training, pigeons can carry up to 75 g (2.5 oz) on their backs. The German apothecary Julius Neubronner used carrier pigeons to deliver urgent medication. In 1977 a similar carrier pigeon service was set up for the transport of laboratory specimens between two English hospitals.

Pigeons can transmit data on overcast days, and even during the night.
They can also be used in parallel, so that larger messages can be carried between multiple birds, and important messages can be sent via more than one bird at a time to reduce the chance total packet loss.

Pigeons can travel between 600 and 700 miles in a single day, meaning the 5000 mile distance between Rome and Beijing could be traveled in as few as 9 days with only 9 birds.
Pigeons also have an advantage to ground based transportation in that they don't have to worry about terrain.

This is much better than a horse, which as an average galloping speed of 25 miles per hour, and will be slowed by things like swamps, rivers, oceans, mountains, etc.

Most people assume the Pony Express riders galloped their entire route. In fact, the speed of a pony express rider averages out to 10 miles per hour- meaning they spent most of their time alternating between a trot (about 8-9 mph) and a canter (12-13mph). The Pony Express riders switched to fresh horses every 10-15 miles.

By using a route through northern UK, Iceland, Greenland and northern Canada, it is possible to use carrier birds to send messages into the Americas too.

A China, Siberia, Alaska route would also work with a lot less ocean to cross.

  • $\begingroup$ Would they be reliable for vast distances or would you require many many pigeons? $\endgroup$
    – Mikey
    Dec 20, 2015 at 21:27
  • 1
    $\begingroup$ Quoting Wikipedia, "Their average flying speed over moderate distances is around 80 km/h (50 miles per hour)" -- that wouldn't make them so much faster than horses in the context of this question. A long-distance message over several thousand miles would still take a couple of months. $\endgroup$
    – tmh
    Dec 21, 2015 at 7:41
  • $\begingroup$ @tmh The average speed of a galloping horse is 25-30 miles per hour, and only for couple miles at a time. Also terrain would slow them down. At 600-700 miles a day, a series of pigeons could travel the 5000 miles between Rome and Beijing in 9 days. Other birds could also be trained to carry messages in areas where pigeons are less suited. $\endgroup$
    – AndyD273
    Dec 21, 2015 at 15:24
  • 1
    $\begingroup$ @Mikey They are pretty reliable. A single pigeon traveled 7000 miles in 55 days. This sounds slow, but it had to do things like forage for food and spend time resting/recovering. A more realistic distance is 600-700 miles in a day, so a series of pigeons could do that distance much faster. I added some more details. It would be much much faster than a horse. $\endgroup$
    – AndyD273
    Dec 21, 2015 at 15:29

One option that was not covered is using musical instruments. Historically this has only been used in situations where the line of sight is obstructed, market places, forests, and mountains, which is not directly applicable to your case. But since humans are already capable of audio communication it could feasibly support larger bandwidth, something close to speaking slow and careful.

Instruments used historically have been horns, drums, and gongs. But I think for an infrastructure project like this you could go for something more expensive. Hydraulis or water organ was known early enough and while it was large and expensive, it needs a water reservoir, that might not be an issue for a permanent communications line. The main benefit is that it produces actual music which can be used to to encode messages fairly efficiently and user friendly.

The main difficulty is obviously the resources needed to build an entire infrastructure of water reservoirs and water organs. But realistically all attempts to build a a communications network this wide would have similar issues. And the ancients had good enough practical understanding of acoustics to focus the sound produced in the right direction and a hydraulis could in theory scale to produce lots of sound, so the distance between stations could fairly large in comparison to other methods. And the need for water is somewhat balanced by not needing line of sight unlike the visual methods.

  • $\begingroup$ Signal drums were an excellent idea, but the actual implementation had problems because the signalling method was complicated and sometimes the wrong message was communicated. Kinda like Microsoft Windows. +1 $\endgroup$ Dec 19, 2015 at 14:45
  • $\begingroup$ @GaryWalker Or any other method of communication used by humans. LOL. But yeah I am not enamoured by drums myself. The water organ should allow more natural signalling less prone to operator error. But I don't think anyone has tried that in practice, so it is just a baseless assumption by me. A vast network of huge water powered organs would be cool though... (<- my real motive exposed) $\endgroup$ Dec 19, 2015 at 15:06
  • 2
    $\begingroup$ It's worth noting that [whistled languages(en.wikipedia.org/wiki/Whistled_language) are actually a thing, and can be used for communication in mountainous regions or across a long, flat area. Doesn't directly relate to using water organs, but the concepts behind it can be useful to base an instrumental form of communication on. $\endgroup$
    – Bobson
    Dec 20, 2015 at 1:00
  • $\begingroup$ Also note that yodeling originally was a means of communication at a distance. $\endgroup$
    – celtschk
    Dec 20, 2015 at 11:56

Tins can and strings.

Ok, this is not really a serious recommendation in the sense that I would recommend it, but Tin Can Telephones were actually a commercial product in the US at one time. After a number of patented improvments, they were used at a range up to about 1/2 of a mile.

Unlike semaphore towers, they could be used in heavy fog, etc. A half mile range is clearly a problem for a serious long-distance communication network though.

Tin cans are in short supply in 1 BC too.

The Hydraulic Telegraph I don't remember where I read about this years ago, but I remember thinking it was pretty inventive. Limited to pre-defined messages and fairly expensive to build, but it was actually used in military situations in 4th century BC.

I was foolish for not mentioning this originally, but the historical example for given for sending a message is between Sicily and Carthage. These cities (or just the island and the city) are hundreds of km apart and includes a considerable distance of the Mediterranean Sea. So, I don't know whether this reference is mythical, confused, or very impressive ancient engineering.

  • $\begingroup$ Thanks. I was just writing a post and basically re-inventing the Hydraulic Telegraph--you saved me a lot of effort. I gave up because I didn't think we could have made good enough pipes back then--the only good material was wood and it would compress too much. $\endgroup$
    – Bill K
    Dec 20, 2015 at 7:12
  • $\begingroup$ Um... It's called the Iron age for a reason. Roman aquaducts were concrete with metal parts where needed. Before that, you had the bronze age, so why wood? Also, why must the network be homogenous? Where sending message by bottle down the mountain is faster, just do that. Each node only needs to know how to share a message with all those who connect directly. $\endgroup$
    – The Nate
    Dec 22, 2015 at 16:44

Okay, we didn't solve this problem until the telegraph, so it's hard.

Gary Walker mentioned the Water Telegraph (which I was trying to invent in my answer) but I don't think the materials could convey pressure reliably over huge distances. While commenting on that I came up with another way...

How about a small pipe full of standing water with a long wood "Stick" running down the middle? You couldn't move it fast, but you could move it quite a distance, say 10 feet just to pick a random number.

If every 2 inches you calibrated a mark on the side of the pipe, each mark would mean a message, that would give you a vocabulary of 120 messages.

In order to send a message you have some person apply steady pressure to the end of the stick until it reaches the correct position. The water would make pushing the stick slow but it shouldn't take too much pressure---perhaps you could use a ratcheted lever to push the stick so it could push fairly hard and keep the pressure on for a long time.

It might even take days, but as long as the stick didn't buckle in the middle it should move--Very Slowly. A waterproof rope that floats might work as well, you would pull instead of push.

Anything but a liquid seemed like it would have too much friction--with water it's more about overcoming inertia than friction so a steady pressure could work.

If necessary it could be relayed at certain distances. Relays would be absolutely necessary to change elevation--this thing would have to be completely flat.

To relay messages up and down mountains, a rope and pulley system should be able to go a good vertical distance and could be calibrated with the same message marks.

  • $\begingroup$ This was solved repeatedly prior to the invention of the electrical telegraph. The telegraph merely made distance entirely irrelevant, rather than mostly irrelevant. $\endgroup$
    – Mark
    Dec 20, 2015 at 20:23
  • $\begingroup$ @Mark Really? a communication mechanism significantly quicker than horses crossing a distance significantly longer than LA to NY? Train might do it--but I don't think they were used for that kind of distance until post telegraph and I've never even heard of smoke signals being used at those distances, but I know I'm often surprised about what past human accomplishments I've missed--. Could you elaborate? $\endgroup$
    – Bill K
    Dec 20, 2015 at 20:55
  • 1
    $\begingroup$ Although now that I think of it, any mechanism would work with enough relay stations, even shouting through rolled up paper. $\endgroup$
    – Bill K
    Dec 20, 2015 at 20:58
  • 1
    $\begingroup$ Signal drums. Heliograph. Semaphore tower. True, nobody did a Rome-to-China run prior to the electrical telegraph, but if you've got the manpower, there's no practical difference between a signal line crossing France and one crossing Eurasia. $\endgroup$
    – Mark
    Dec 20, 2015 at 21:17

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .