We think of digital communications as a modern invention. But the reality is that semaphores, smoke signals, and Aldis lamps are all types of digital communication. While telegraphs are not as old as smoke signals, they, too, are a digital mode. The problem with all of these is that they require the operator to learn some kind of code. People don’t like to learn code because it is difficult, and employers don’t like to pay high wages to trained operators.
In the late 1830s, a man named William Cooke proposed a complex telegraph to a railway company. The company didn’t care for it and asked for something simpler. The railway didn’t like that either, so Cooke joined up with Charles Wheatstone and patented something that was a cross between a telegraph and a Ouija board.
Text by Needles
The Cooke and Wheatstone telegraph used multiple needles–the patent used five needles as an example. The needles were magnetic and could be made to turn to the right or the left by way of an electromagnet. Above and below the needles were a pyramid of letters (see figure to the right). When the needles moved, they would both point to the letter (the figure shows the two yellow needles pointing to the letter G).
You can learn to read this code very quickly, but the system required one wire for each needle. The really cool feature is that since only two needles need to turn, and they turn in opposite diretions, current flowing through one needle’s wire and returning via another’s handles the problem with just one active current loop. Some systems had an extra wire to serve as a return so you could move any needle separate of the others.
Great Western Railway was the first to use one of these unique telegraphs, with a span of 13 miles in 1838. You might notice that because the number of letters available is only 20, some letters aren’t available (in particular, C, J, Q, U, X, and Z). This required creative spelling in some cases.
Hung by Telegraph Wire
For example, a story that caught the public’s attention in 1845 was when a murder suspect, [John Tawell], was arrested thanks to the telegraph. Tawell boarded a train at Slough to avoid authorities. The telegraph carried the message:
A MURDER HAS GUST BEEN COMMITED AT SALT HILL AND THE SUSPECTED MURDERER WAS SEEN TO TAKE A FIRST CLASS TICKET TO LONDON BY THE TRAIN WHICH LEFT SLOUGH AT SEVEN FORTY TWO PM HE IS IN THE GARB OF A KWAKER WITH A GREAT COAT ON WHICH REACHES NEARLY DOWN TO HIS FEET HE IS IN THE LAST COMPARTMENT OF THE SECOND CLASS COMPARTMENT
Since there was no J, JUST became GUST and the lack of Q required the spelling of KWAKER. The message and the arrest became quite famous, and the instruments that carried the message wound up in the Science Museum in London. The video below from ARM shows a display from the museum about the telegraph.
Next time you send some text over a TCP socket using an ESP8266 between two Arduinos, just remember that people have been hacking digital messages for centuries. We tend to think of hacking as having to do with sophisticated technology, but the hacker spirit will work with whatever presents itself–from fire and smoke signals to magnets and needles, to silicon and square waves.
If you long for the old days but still want to read e-mail, you might try this project. Then again, there’s always Twitter.
Cooke-Wheatstone Needle Diagram By Wrrglla [CC BY-SA 3.0]
Photo of Telegraph By Geni [CC BY-SA 4.0-3.0-2.5-2.0-1.0]
“Ouiji” -> Ouija. Interesting read though.
Fixed by a friendly spirit.
Also fix the word “directions”.or where you using the Cooke machine and it had no letter C? :-)
Great article, weak video (I was hoping for a lot more info from it).
Very interesting technology, especially considering the fact that this happened a good 60-ish years before electricity would become something that was available via a distribution system. Am I correct in thinking that each station would have its own dynamo/generator/magneto?
Early telegraphs and telephones used battery power, this probably did too.
For sure this system did use batteries. There was one earlier attempt at electric telegraphy in France that was powered by a static electric generator that signaled by making pith balls jump in a glass tube, It was not developed due needing a separate conductor for each letter.
Thanks for clearing that up!
Cool. And almost certainly way faster for a message like that than land semaphore networks, which started being used seriously about 40 years earlier. (Semaphore systems generally had code books, so if there was a combination of flags/positions for something you could get a message across a country in 15 minutes or so, but if there wasn’t, good luck.)
It is strange that the idea of encoding with simple signals like Morse code took so long to be invented. In hindsight it seems so obvious.
Morse code isn’t just abstract dots and dashes. They had to develop the hardware as well, which involves coming up with audio oscillators etc. to generate the signal tones, and then filter the tones back to solenoid action at the recieving end.
The idea may have well existed, but the implementation lacked behind because it wasn’t a trivial thing to early 19th century level of understanding of electricity. Nobody had any clue about how signals would travel down wires or why they did travel, so the breakthrough was really about making an oscillator and coupling it to a transmission line in an efficient manner so the result could be picked up miles away at the other end. Seems easy to us now, but we forget they burned the first transatlantic cable by cranking up the power source, because they didn’t understand that the problem was the impedance of the cable rather than its resistance.
The Cooke and Wheatstone system operates on simply switching the polarity of pairs of wires, which didn’t need any fancy hardware. Just some switches and galvanometers.
There were optical telegraph systems before this as well as a few other long distance signaling methods that could have supported a morse-like code, and have been the better for it. I suspect it was more a case of no one having given it much thought until faced with the limitations of signaling down a wire.
I always thought the Heliograph was the first VLC digital optical communicator. In 1820 Carl Friedrich Gauss dabbled with one. This gadget was instrumental in our Civil War here in America. The Native Americans used to notice the US Army troops flashing enigmatic signals between mountains during the Indian Wars. I think they thought it was some sort of magical thing as the flashes would usually lead to a quick interdiction by troops when they emerged from a canyon or other hiding place. Akin to their smoke signals but much more message content.
The British polymath Robert Hooke, gave a vivid and comprehensive outline of visual telegraphy to the Royal Society in 1684, but the first practical system was built by the French engineer Claude Chappe and his brothers in 1792, who succeeded in covering France with a network of 556 stations.
DV82XL – Yes you are right about the SEMAPHORE systems. However, I was referring to the ability of the Heliograph to reflect sunlight to a remote location without aid of telescope using Morse Code. It was a man portable system with just a small pocket-sized gadget on top of a tripod.It had aiming holes to line up the sun and its reflection to the intended target.Then the sender would tap out a message via a aperture-shutter mechanism.
Dax I’m not disputing you, however, are you saying early telegraphs used “audio oscillators” versus just a Leyden Jar (battery), hand switch tapper, and a electromagnetic relay armature on the other end to receive the tapping? Also your comment about impedance vs. resistance and transatlantic cable… Impedance would suggest that they used AC for the early transatlantic cables. I could be wrong but I always thought they used DC. And the receiving end was a mirror-galvanometer that would deflect very slightly from the medium-voltage on one end being dropped by resistance of 3,000 miles of copper. Just say’n…
It appears that Dr. Edward Orange Wildman Whitehouse MD screwed up the first cable by inducing 2,000 THOUSAND volts from his side in Ireland. The cable’s insulation just could not take it. He was a self-taught electrician and was summarily dismissed by Atlantic Telegraph Company after this complete fail in 1858.
Just a slight correction a ‘Leyden Jar” was an early type of capacitor not a battery. I think the term you are looking for is ‘Galvanic cell’.
Actually I was conflating Benjamin Franklin’s use of several Leyden Jars to form a battery source. He coined the phrase “battery”. I was conflating it with the “Daniell cell” which was used for 19th century telegraph systems.
There isn’t a ‘C’, yet several words in the message are spelled correctly with a ‘c’: ‘committed’, ‘class’, ‘coat’, ‘second’, and ‘compartment’
Probably transmitted as “K” and transcribed as “C”.
Omitted in the Hackday writeup is clarification that the quoted telegram was transmitted with a later refinement of the design which used only two needles total and was able to encode all letters except for J, Q and Z. The linked article at http://www.cntr.salford.ac.uk/comms/johntawell.php provides a good description of how the encoding worked in that version.
Thanks for the article, it is very interesting!. An alternative headline could have been “Playing Carmen Sandiego, this time for real”, I mean that message that led to the capture of the murder looks like it has been sent by ACME headquarters (https://en.wikipedia.org/wiki/Carmen_Sandiego).
Reminds me of the 25 letter alphabet used by POWs in North Vietnam. It consisted of a 5×5 grid of the alphabet (I think Q was omitted). The first set of taps indicated which row the letter resided in and the second set of taps indicated how deep in the row. E.g. “tap-tap”, “tap” was the letter “F”
“K” was omitted as “C” could easily be used in its stead without loss of understand as both of them could be pronounced the same for most words.
Thanks, so my example, “tap-tap” “tap” would be the letter “G”
“HE IS IN THE GARB OF A KWAKER”, well a duck costume would be a dead giveaway.
B^)
Go to your food cabinet and look for your QUAKER OAKS box and look at the man on the front. He too is in the “garb of a Quaker”.
https://www.quaker.co.uk/images/default-source/products/quaker-traditional-oats/quaker-oats-1kg.png?sfvrsn=8
*woosh*
You list ‘U’ as one of the missing letters, but it’s in the diagram (which is missing ‘V’ instead).
So opposite to the old Romans in Latin.
Yes Cooke’s machine was quite amazing. I would like to add this one in circa 1840: Royal Earl House’s printing telegraph (http://www.google.com/patents/US4464). Some how he took 28-keys on a piano keyboard (56 characters when shift key used) and transmitted a digital message to a distant receiver over only 2-wire pair! It could transmit 33~40 words per minute or 2,000 per hour. It works on the Daisey-wheel principle. It’s a mystery to me. I would like for someone to analyze and explain how this thing worked with 56 characters over a 2-wire pair in 1840’s. Unlike Cooke’s machine it was heavily used by Great Western Telegraph Company.
http://www.telegraph-history.org/george-m-phelps/house5.jpg
According to wikipedia (https://en.wikipedia.org/wiki/Printing_telegraph) it used synchronous transmission (probably the first to do so). very cool stuff!
Yes I assumed that but HOW? What I can analyze is this: The receiver’s print wheel with all 56 characters spins in synchronicity with the transmitter’s same wheel. When the keyboard sends a character SOME HOW both wheels STOP and and an impact hammer makes the ink impression on paper that is also feed line by line. So far my brain is hurting!!! This was 1840 darn it! How the heck did he figure out how to synchronize two wheels and stop/start them so they kept synchronicity? And all over TWO DARN WIRES! They said it was difficult to manufacture but they made loads of them for deployment for financial institutions. This guy was thinking outside the box 176 years ago!
By the looks of it a continual clock pulse was sent by the transmitter to the receiver that incremented the print wheel. Which character sent would then be selected in the time domain.
Wow! Now that does make sense! So the transmitter pulses one clock pulse over the telegraph line pair, the local wheel and the remote wheel ratchet pawls one alphanumeric unit. Probably did this very fast. Then when a character needs to impact the ink and paper tape the transmitter halts on that character **, then reverses polarity to fire the print head relay; the ratchet pawl prevents the wheel from turning backwards during this pulse. Thanks! That idea can be expanded on in some modern day invention too maybe. If you see the Smithsonian’s model you do see a clock lead weight under the unit. This is probably what keeps the clock going. This means the telegraph line was active all the time with clock pulses. Must have been dedicated lines and the return pair must have been separate (technically 3~4-wires).
** this part is unclear – how would the machine KNOW which character was up that moment?
It’s amazing Mr. House’s mind could work like this. He was a recluse and did not want his picture taken (only one exists). His nephew was smart like him too. House was sued by Morse and Bell but won each patent infringement law suit. If you look at his mental evolution since his Vermont birth, he was not initially an electrical genius. He had to learn electromagnetical sciences after giving up law school. As a kid he caught a frog, skinned it, and put in springs in its legs and made it hop even though it was dead. He designed a much better telephone than Bell but did not market it well.
His printer was hooked up between NYC to Boston initially. Then to other places around east coast to Ohio. I think it was only good for 2 stations each pair. Telegraphs could be several attached and only needed 1-pair for 2-way communications. But 2,000 instantly recognized words per hour must have been better than training someone who to learn Morse Code. Even though House’s printer was successful, it was only short-lived. He did make other inventions before he died at 81 years old. He even added steam-power to his telegraph printer. STEAMPUNK?!
.
>”** this part is unclear – how would the machine KNOW which character was up that moment?”
There was a null indexing position on the wheels that would sync up every revolution.
Learning Morse is no big deal, and sounder type telegraphs are less expensive, far less complicated, and far more robust and therefore far more reliable than indexing types. They could also work on single wire circuits with multiple stations which was important for railway signaling, telegraphy’s killer app.