How the Morse Telegraph Worked: Key, Register, Sounder and Relay

There was no single "Morse code machine." A working line was a set of instruments, each with its own name:

Between the offices ran the line wire, carried on poles. In the station wiring Pope shows, the line came in through a lightning arrester, passed through the relay and the key, then through the main battery to a connection with the ground.

Morse's first transmitter was not a key. Britannica describes a "portarule," which ran molded type with built-in dots and dashes through a mechanism that made and broke the contact. Alfred Vail replaced it with a simple make-and-break key.

Pope's handbook describes the key as a brass lever four or five inches long, hung on a steel pivot so it could move up and down. One wire of the line was connected to the key's frame and the other to an insulated anvil underneath. Pressing the knob brought the lever down onto the anvil and closed the circuit, "precisely as if the wires themselves had been brought together." The contact points were platinum ("platina"), because ordinary metal would be melted by the spark when the circuit broke. A spring lifted the lever again when the operator let go.

The key also had a small switch called a circuit closer. On American lines the battery power came only from the offices at one or both ends of the line, so every idle key along the way had to be closed with this switch. That kept the circuit complete, so the next station's sounder could still answer to the line, as Wikipedia's article on the telegraph key explains.

The register was the receiving instrument of the first Morse lines, and the reason the code is written in dots and dashes. Its electromagnet pulled an iron armature fixed to a lever. The other end of the lever carried a steel point, or style, which pressed up against a narrow strip of paper. Clockwork, driven by a weight the operator wound up, pulled the paper steadily between grooved rollers.

While current flowed, the style embossed a raised line on the moving paper. When the key opened, a spring pulled the style away. The length of each mark matched how long the key was held down: a short press left a dot, a longer one a dash. The National MagLab notes that Morse's early version wrote the code with a marker and that by 1836 he had changed it to embossing.

The strip above draws the word TELEGRAPH the way a register would have marked it, in the American Morse used on US lines. The single long mark is L, and the narrow gap inside R is one of American Morse's spaced letters.

Operators learned to read the clicks of the receiving magnet, and writing the message down straight from the sound proved quicker than reading the tape. From 1848, starting on the New York to Boston line, some telegraph companies hired "sound operators" who took messages by ear and wrote them out by hand. By 1881, Pope writes, many larger offices had dropped the register for a sounder: just the electromagnet, armature and lever on a base. A "box sounder" let the lever strike a hollow wooden box to make the sound louder.

A sounder doesn't beep. It makes two sounds for every dot or dash: a heavy click when the current starts and the lever is pulled down, and a lighter one when the current stops and the lever springs back. In Pope's words, "A dot makes as much noise as a dash, the only difference being in the length of time between the two sounds." Without the lighter stroke, he adds, the operator could not tell E, T and L apart. Each of them is a single press in American Morse.

What that means for one word, counted with this site's Morse engine at 20 words per minute:

At 20 WPM one unit lasts 60 milliseconds. A dot is 60 ms between its two clicks and a dash is 180 ms. The operator heard 42 clicks and had to judge every gap.

Current weakens over a long wire, and leaks away where the insulation is poor. Pope says a register could be worked by the line current alone only for "a distance not exceeding 20 or 30 miles," and only on a well-insulated line. Most telegraph lines were longer than that.

The relay solved it. It was a second electromagnet, wound with very fine wire and built "with great delicacy," so even a weak line current could move its armature. That armature did not mark anything. It closed a pair of contacts that connected a local battery in the receiving office through the register or sounder. A feeble current from a distant office could then drive a loud sounder at full strength.

The relay needed constant care. In rainy weather, when current leaked from the line, the operator had to move the relay's magnet farther from its armature so it would not stick. Pope calls this "adjusting" the instrument, "one of the most important of an operator's duties."

You can try the telegraph online. The telegraph simulator gives you a straight key and a telegraph sounder, and decodes what you send as you key it. It reads International Morse, so key TELEGRAPH as - . .-.. . --. .-. .- .--. .... and listen for the click and clack of each press.

At the end of a line, a terminal station was wired as described under the parts table: line, lightning arrester, relay, key, main battery, ground. Stations along the way, called way stations, sat in the middle of the same wire. The line came in, ran through their relay and key, and went out again to the next office. Several offices could share one wire. That is why Pope warns that the key "should never be opened" without first adjusting the relay, "to be sure that no other offices are using the line." When an operator left for the night, or a thunderstorm came, a switch called a cut-out let the current pass straight through without going through the instruments.

The first news dispatch went out the day after the line opened. On 25 May 1844 a telegram from the Capitol told Baltimore's Patriot newspaper how the House had just voted, according to the House historian. The lines spread quickly. Western Union opened for business in 1851, and the first transcontinental line joined the coasts on 24 October 1861. In the Civil War the US Military Telegraph Corps laid about 15,000 miles of line and handled about 6.5 million messages.

"Telegraph code" is the general name for any code used to send text by telegraph. Morse code is the best-known one. But the telegraph used two versions of it:

• American Morse on US landlines. Pope teaches it with the dot as the unit of time, a short dash of three dots and a long dash, plus "spaced letters" with a gap inside them.
• International Morse, based on Friedrich Gerke's 1848 revision. The German-Austrian Telegraph Union adopted the Morse telegraph in 1851, and a conference in Paris made Gerke's code the international standard in 1865. US companies kept American Morse at home, so messages to and from abroad had to be resent in the other code.

Every letter that differs, with its sound, is on the American Morse code page.

The key outlasted the wire. The same straight key moved to radio, where it keyed a transmitter instead of a line. The J-38 straight key was used by the US military in World War II and later by many radio amateurs. Faster operators used a "bug," a semi-automatic key whose vibrating arm made the dots by itself, and which a skilled operator could run at more than 40 words per minute. Every operator's rhythm, called their "fist," was distinctive enough that in both world wars listeners used it to track individual ships and submarines.

Radio took Morse code out to sea, where the wireless operators of ships such as the Titanic used it. You can still practice with a straight key, paddles or a cootie on the Morse code simulator.

This guide is written and maintained by Translate Morse Code (independent site). Every code, click count and duration on the page comes from the same Morse engine and code tables the translator uses, the instruments are described from Pope's 1881 handbook, and every other fact is checked against the source listed below.