From Newfoundland to Ireland with Marconi

Winter, 1907. You live in Clifden in western Ireland, and your favorite way to spend the rare clear-skied evening (after a hard day of farming or herding or shopkeeping) is to walk up the hill to the monument.

The sunsets are gorgeous over the Atlantic ocean, and as dusk falls you can look south to see a man-made lightning show. Sparks dance over eight huge wooden masts marking a rectangle 1,000 feet long by 200 wide. The masts support 52 wires running lengthwise across the top, which fan down at one end into a single wire connected to a mysterious building. The entire apparatus is over 200 feet tall, and when the wind is right you might hear the electricity crackling.

Wait, 1907? How did we get here?

I went to Newfoundland, Canada in autumn of 2017, where (as usual) I visited a bunch of lighthouses and historic sites. I became fascinated with the history of long-distance communication: the undersea cables and early transatlantic wireless transmissions that brought the modern world together around the turn of the 20th century. Morse code keys for telegraphy, string oscillographs to make graphs of electrical currents in undersea cables, giant Fresnel lens floating on gallons on mercury, mournful foghorns; such a variety of effort and innovation in service of communication.

In April 2021 I wound up in western Ireland, so naturally I made my way to the remote boglands of Connemara to see the ruins of a Marconi station there. On a sunny day we took a long bike ride to explore, picnic, and read every informational sign on the site. Of course I wanted to know more, and for weeks afterward I was excitedly telling anyone who would listen all about Marconi and the history of radio. Both Newfoundland and Connemara are beautiful, and what better excuse to spend hours reading obscure archival materials and revisiting photos than to write a blog post? And here we are.

The station

On a quiet evening near Clifden in the winter of 1907, you can hear a distant rumble: the din from the enormous transmitter. A steam engine powers a 15,000 volt DC generator by consuming a mix of coal, brought in by railroad, and local peat bricks, shoveled by hand from the surrounding bog and dried in the sun. A 1921 report noted that the station used up to 6,000 tons of peat per year.

Power from the generator charges the capacitor, and oh, this capacitor! In a small radio, a capacitor would look a bit like a battery, though it differs in some crucial ways. It consists of two or more conductive terminals, usually metal plates, with a non-conductive “dielectric” material between them that can hold electrical polarization. The capacitor gathers and stores a polarized electrical charge from a power source, which is then discharged quickly to produce radio waves.

A battery stores energy typically through chemical means and releases it slowly, making it ideal to power things that require a steady supply of electricity. A capacitor stores energy through a charged electrical field and releases it quickly all at once, perfect for situations where you need a large, sudden spike of power (and unlike batteries, capacitors never lose their ability to hold a full charge). Think of a classic camera flash: the camera has a battery, which it uses to charge a capacitor when you prepare the flash. Once the capacitor is fully charged, the flash is ready to use; pressing a button discharges all the energy in the capacitor at once and triggers the flash.

This capacitor (formerly known as a condenser) is… somewhat bigger than what you’d find in a camera. Two conductive metal plates? Try nearly 2,000. The size of an average battery? How about an entire building — 350 feet long and 75 feet wide, three stories tall, mostly windowless, with a single wire stretching out one side and branching up to the aerial array.

Each metal plate is two or three times the height of a person (12′ x 30′ to be exact), and they’re hung vertically throughout the whole building 12 inches apart like laundry hung out to dry in the countryside. The dielectric medium is air, so when the capacitor is charged, the entire atmosphere in the building is polarized. Imagine being inside when it’s running!

An entire building humming with electricity, discharged to produce radio waves. But discharged… how? Huge arcing sparks in quick succession, that’s how. Picture a huge spinning wheel, five feet across, with rivets spaced around the rim. It spins between electrodes, and each time a rivet passes through the gap, it closes the circuit and causes a spark to jump across. This happens so rapidly (~350 times per second) that the sparks appear almost continuous, like a flickering flame.

A Wireless Thunderstorm

An interesting description of a visit to the Marconi wireless station at Clifden (county Galway) is given by a correspondent of the “Daily Mail,” who says: “An entire room is given up to strange sheets of steel, which are hung from roof to floor, like washing on a line, until only narrow alleys are left. Queer brown earthenware jars, like old-fashioned receptacles, and all manor of outlandish electrical apparatus now confront the visitor. The plates are for acting as a reservoir to store electrical energy. The jars are transformers. The engineer gave a few directions to his assistant, who, seated before an ordinary Morse telegraph instrument in the operating room, placed a telephone headpiece to his ears, and began to fumble with the key, hastily bidding me to stuff cotton wool in my ears and don a pair of blue-glass spectacles. The engineer beckoned me to the connection room on the floor above, which is equipped with a medley of strange electrical contrivances. The use of the cotton-wool and smoked glasses became at once startlingly apparent.

“From the ‘interrupter’ instrument corresponding exactly in duration to the assistant’s touch of the key below, came three blinding flashes of blue-white flame, followed by a short flash, and then three more short flashes. The two side-mouths of the instrument likewise spout eye-blinding flame of the same color and intensity. Simultaneously, the discharger, a few feet across the room, emitted similar blinding flames, and there came a wearing, tearing boom like the deep bass of some gigantic organ, but immeasurably cruder and louder. The duration of each note again corresponded exactly with the assistant’s dot or dash on the instrument below. This was the electrical discharger, which sends oscillating electrical currents from the building into the aerial wires outside. These at once begin to set up vibrations of the ether, which in loops and waves travel with inconceivable rapidity across the sea.”

Marconi Signal Station, Wagga Wagga Advertiser (New South Wales, Australia), Sat 23 Nov 1907

Calibrating the speed of the wheel to the resonant frequency of the current produces stronger and more consistent sparks and resulting radio waves, but this is the early days of radio and that knowledge hasn’t spread just yet. For now we’re just focused on spinning the wheel fast (more sparks!) and running the whole system at the most power we can generate (more output!) to broadcast low frequency electromagnetic waves.

These long waves (1,500 meters from peak to peak at first, over 6 kilometers for transatlantic communications to Nova Scotia later) propagate by hugging the surface of the earth and refracting off the atmosphere, and can cross an entire ocean and bounce over mountains. This is not a fine-tuned science yet; we’re using a wide portion of the frequency spectrum fairly haphazardly, and the whole aerial setup has had a worrying tendency to collapse in a strong wind at other sites, but we’re learning. By the time we reach Clifden station, we’ve discovered that a long array with a fan shape at one end allows us to aim the waves! This is a vast improvement on the previous method of an upside-down pyramid of wires throwing off waves in every direction.

All this takes an enormous amount of power, and there’s a separate building for that. The power house is a good distance away (about 150 meters/500 feet) to reduce interference from electricity, magnetism and general noise. It’s incredibly loud, with four huge 5,000 volt DC generators, three of which are used to power the array as well as charge the massive standby battery (actually over 6,000 2 volt batteries arranged in sequence) in the condenser house. The fourth generator is kept as a backup to ensure continuous operation if any one generator fails. The generators in turn are powered by steam engines directly coupled to huge alternators (500kw each, weighing 10.5 and 7.5 tons), and there are two mammoth 10-ton flywheels in the building as well.

The buildings and equipment are gone now, and most of the living creatures one encounters in these remote boglands nowadays are sheep. I witnessed a local man, out for a pleasant Sunday stroll with his wife, step off the footpath and casually haul a dripping wet ewe out of a muddy hole in which she had gotten her soggy self well stuck, then continue on his way as though nothing had happened. Later I watched a shepherd and his two working dogs herd a large flock of ewes and new lambs through scattered building stones on a hill. Those particular stones outlined the foundations of the Engineers’ Bungalow, once home to two pioneering electronics engineers.

The Engineer’s Bungalow at Clifden, then and now plus current residents

Henry Joseph Round lived there in 1910-11. He discovered the LED effect in 1907, developed some of the first vacuum tubes, was the first to try broadcasting on different wavelengths for day vs night, and once used copper mosquito netting to repair a wireless transmitter in the remote Amazon jungle. He filed 117 patents in his lifetime, working on everything from radio transmission, to gramophone recording, to sound for film, to systems for tracking enemy movements during WWI, to the precursor to SONAR in the 1940’s.

Charles Samuel Franklin lived in the same building around the same time, and collaborated with Round. He started his career as a wireless engineer for Marconi in South Africa during the Boer War, later developed an innovative shortwave antenna with unprecedented broadcasting power, designed the antenna used for the world’s first “high-definition” television service launched by the BBC in 1936, and filed a total of 65 patents.

In addition to housing and recreation buildings for the many workers required to run the Clifden station, the logistics required to transport and install all the necessary enormous yet delicate machinery in the middle of a vast bog in the age of steam engines is astonishing (dare I say it BOGgles the mind?!). I mean, look where we are:

So, then: why build a station here, in the middle of a remote Irish bog?

The background

It seems to have been worth the effort, thanks to good old capitalism. Marconi didn’t pursue telegraphy as a hobbyist or a scientist; he was a charismatic entrepreneur who used his improvements on newfangled technology to turn a profit. And his improvements were genuine: he was the first to use an aerial antenna and grounded wire to increase the potential transmission distance of radio waves. This innovation led directly from his teenage experiments to the giant masts and crackling electricity at Clifden and his many other stations.

Marconi was born in Italy in 1874, the middle child in a wealthy family; his father was an Italian landowner and his mother was an Irish heiress. He was homeschooled, spent part of his childhood in England, and was interested in electronics from an early age. He started experimenting with wireless in his late teens and by age 20 he was building his own radio equipment, assisted by the butler, in the attic of the family home.

He did not invent wireless telegraphy, nor was he the only person innovating in the field, but he was certainly one of the first to see the commercial potential. After failing to gain investment from the Italian Ministry of Posts & Telegraphs, he traveled to London and used family connections (his Irish grandfather owned the Jameson distillery) to obtain an introduction to William Preece, Engineer-in-Chief of the General Post Office in Britain. Preece was interested in Marconi’s inventions, and the Post Office invested in his experiments.

Marconi’s successful networking irritated scientists whose work preceded his. They saw him as an opportunist, not a true inventor – as one physicist wrote rather cattily to another in 1897, “Marconi has obtained the ear of the British Post Office officials, some of whom are like him not well versed in Physics”. While Marconi himself readily admitted that he was not a scientific expert, this was perhaps a little unfair to Sir Preece, who may not have been a physicist but was certainly a capable and respected electrical engineer in his time.

Marconi’s ambition to bridge the Atlantic wasn’t an easy sell—a transatlantic telegraph cable had been successfully laid half a century earlier in 1858 (though it failed after just three weeks and wasn’t successfully replaced until 1866; more on that in a future post). With serviceable telegraph communication already well established between Europe and North America, there wasn’t a rush of enthusiastic investors for a brand new, untested system, and many still didn’t believe that radio waves could travel over obstacles and around the earth’s curve. Undeterred, Marconi established The Wireless Telegraph and Signal Company Ltd in London in 1897. His perseverance paid off, leading to a string of successes culminating in the first wireless transmission across the English Channel in 1899.

The station at Clifden was built largely due to the Canadian government, which had already funded a Marconi station at Glace Bay, Nova Scotia to break the monopoly of the telegraph companies who operated an undersea cable. Marconi convinced Canada that he could establish transatlantic wireless telegraphy when he successfully received a message in Newfoundland that was sent from his station at Poldhu in Cornwall, England in December 1901.

Or at least, he said he received it. Marconi’s team in England had been instructed to transmit three dots in Morse code (the letter S) continuously during certain hours for the test, but only Marconi and his longtime assistant George Kemp heard the dots through the crackling static in Newfoundland. Many experts today doubt whether the message could have gotten through given the limitations of technology at the time (for context, he was holding a receiver to his ear that was wired to a giant kite flying over 400 feet in the air). The transmission happened during full daylight, the worst possible time for long-distance radio communication due to the effect of sunlight on the ionosphere, and the test was never reliably repeated.

Regardless of the actual success of that first transatlantic transmission, Marconi’s claim won him a contract with the Canadian government. That meant he needed a larger, more powerful transmitter for reliable signals; he had a better design in mind, but there wasn’t enough space for it on his rented land in Cornwall.

Marconi selected the land near Clifden for his new station for several reasons: its location (as close as he could get to his targets on the opposite side of the ocean), its abundant peat (free fuel for his generators) and water (for his boilers), and its open outlook to the west. Knowing Marconi’s priorities, I figured he also got a good price on the 300 acre plot, and further research confirmed that guess: the English landowners accepted his offer of £6/acre (£1,800 total or about £233,221 in 2022 value), less than half the average going rate of about £13/acre for land in Ireland at the time.

Construction began late in 1905, and it wasn’t easy – early transmission tests started toward the end of 1906, but the station wasn’t fully operational until October 1907. Photographs from the opening celebrations show equipment and building materials still scattered about:

Understandable given the working conditions, though at least the site had an adorable little railway (initially to facilitate moving building materials, later primarily used to transport fuel):

While the terrain certainly made construction a challenge, the wet ground of the peat blanket bog was also useful. The aerial was grounded into the earth with wires running under the array, and the superior conductive quality of the damp, acidic ground helped amplify the radio waves. Whether or not Marconi was aware of this effect ahead of time is unclear, but it certainly helped the performance of the new station.

And perform it did: by 1908 Clifden station was able to transmit 14,000 words per day. For monetary context, the Marconi company charged $0.10 per word, or around $3 today (half price for the press) at stations on the other side of the Atlantic. At those rates, Clifden station could have brought in up to $42,000 per day in today’s dollars (though it was likely closer to half of that, since the press were the primary users at that time).

One of the main ways Marconi’s company made money was by transmitting nightly news from his stations scattered along the coasts of Britain, Ireland, and North America to ships who subscribed to his service at sea, which would then publish it in their onboard newspapers. Another paid service was transmitting updates from ships back to shipping companies on land, increasing profits by, for example, notifying them of a ship’s arrival far enough in advance to prepare crews to unload it, without having to pay laborers to wait around for hours or even days based on an estimated shipping schedule.

You may wonder (as I did) exactly why these shipping magnates were paying Marconi. Couldn’t they just send and receive their own messages? The answer, for much longer than I expected, was no; partly due to the complexity of the equipment and the training required, and partly due to Marconi’s stellar commercial instincts. You see, Marconi didn’t sell a ship captain a radio apparatus and a manual to tell him how to use it; oh no. Using a Marconi system to transmit and receive radio signals from a ship required the ship owner to lease the apparatus itself, and a skilled operator to run and maintain it, from Marconi’s company.

Can you imagine if you had to not only rent your phone, but also rent a person along with it, and that person was the only one allowed to touch it? Oh, and aside from life or death emergencies, up until 1912 all Marconi stations (both ship and shore) only acknowledged and relayed messages sent from Marconi operators, giving him a near monopoly. Now the person you pay to operate your phone refuses to accept calls unless the caller also pays them a fee. It’s a fascinating setup (and obviously quite profitable for Marconi).

Marconi’s company established regular ship to shore (and ship to ship) communication in 1899. And long after Marconi’s leasing system ended, having at least one skilled radio operator onboard was an absolute necessity for every ship, strongly recommended by the International Radiotelegraph Convention in 1912 and subsequently required by the SOLAS international treaty from 1914 all the way up until 1979.

But why 1912? Did something happen to trigger that international treaty (which is still in force today)?

The unsinkable ship

Marconi had difficulty convincing people that he could send messages across an entire ocean, plus the lawyers at the Anglo-American Telegraph Company (operators of the undersea cable) threatened to sue him following his Newfoundland experiment in 1901 based on being the only ones licensed to operate telegraph communications in North America. He backed down briefly until their legal monopoly agreement expired in 1906, and focused instead on communicating with ships at sea, something a submarine cable was certainly incapable of doing.

But while ship to shore (and ship to ship) communication wasn’t possible (unless you count semaphore flags and Morse lanterns) before wireless telegraphy, it took quite some time for shipping companies to see value in the service. One leading German shipping magnate once scoffed that he wouldn’t pay for a system to enable his captains to say “Good morning” to each other, but Marconi’s marketing skill worked on others. By late 1901 his commercial, political, and military rhetoric had won him 54 land-based stations on both sides of the Atlantic, plus systems in use on 44 ships from major companies.

Though his motivations were primarily commercial, Marconi also realized early on that his system had great potential to save lives, calling it “the most potent safeguard that has yet been devised to reduce the peril of the world’s sea-going population” in 1901.

Just over a decade later the famous wireless transmission from the sinking Titanic reached Cape Race lighthouse in Newfoundland 370 miles away, as well as multiple ships that raced to help. That tragedy led to the international conference which established the first Safety of Life at Sea (SOLAS) treaty in 1914, part of which required every ship over a certain capacity to have a wireless operator on board.

Jack Phillips was the head Marconi operator on the Titanic. He broadcast the distress calls and continued transmitting calls with its estimated location right up to the end, going down with the ship. The Titanic was equipped with a 5Kw rotary spark generator; 10 times more powerful than the 1/2Kw wireless used by most Marconi operators at sea. As a result, the ship was able to broadcast further than normal, and when the Titanic hit the iceberg, Phillips was already in communication with Cape Race, busily sending a large number of personal telegrams from ship passengers that had accumulated when the wireless was temporarily out of order the previous day.

Jack Phillips joined Marconi’s company in 1906, and was stationed at Clifden from 1908 to 1911. You can just see the white house where he lived in the back right of this photo:

It’s a bit hard to imagine how small the world of wireless communication was back then. Many Marconi operators knew each other and became friends: training together, serving together, and talking to one another in Morse code over long shifts. It wasn’t uncommon for them to message each other informally in between official business, to check in and say hello (perhaps that German shipping magnate wasn’t entirely wrong about paying for his ships to say “Good morning” to one another).

Walter Gray (27) was the head operator at Cape Race, the closest land-based station to the Titanic and the main relay point for news about the disaster. He was friends with Jack Phillips, having attended Marconi training at the same time. Harold Cottam, the 21 year old wireless operator on the Carpathia and the man responsible for bringing the first rescue ship to the Titanic, was friends with both Jack (who turned 25 the day after the Titanic set sail) and Phillips’ assistant Harold Bride (22). This friendship comes across in the informal quality of some of the messages sent back and forth during the disaster, perhaps most stirringly in Phillips’ last complete message to Cottam, which Cottam remembered at the official inquiry:

Come as quickly as possible, old man, the engine room is filling up to the boilers.

Jack Phillips, Harold Bride, Harold Cottam, and (later) Walter Gray

Phillips’ and Cottam’s communication allowed the Carpathia to reach the Titanic before any other ship, but sadly still too late for many of the passengers who succumbed to hypothermia or injuries in the hour and a half or so between when the Titanic sank and when the Carpathia arrived at the scene. Harold Bride was among those rescued, and despite his injuries he helped relieve his friend Cottam, who was exhausted after working non-stop for hours as the only radio operator on the Carpathia.

The tragedy of the Titanic (though about 700 people were rescued, over 1,500 died) and the subsequent international treaty led to a number of safety changes, including:

  • Messages regarding ice or shipwrecks must be transmitted free of charge
  • SOS is recognized as the official international radio-telegraph distress signal
  • Shipbuilding safety requirements put in place (fireproofing, separate watertight compartments, exits, emergency lighting, etc)
  • Ships must carry sufficient lifeboats and life jackets for all passengers, and
  • A variety of requirements for radio installations and use like this one:

All merchant ships belonging to any of the Contracting States, whether they are propelled by machinery or by sails, and whether they carry passengers or not, shall, when engaged on the voyages specified in Article 2, be fitted with a radiotelegraph installation, if they have on board fifty or more persons in all.

Those wireless systems had to be “capable of transmitting clearly perceptible signals from ship to ship over a range of at least 100 sea miles by day under normal conditions and circumstances”, required a backup power source that could run for at least six hours, and had to be placed as high as possible in the safest possible position within the ship to ensure their ability to continue transmitting for as long as possible in an emergency.

Distress signals, and plans to listen for them, were not new in 1912, but they were codified more and more specifically over time. Even in the early days operators were required to give priority to distress signals, and there were suggestions early on that operators should observe radio silence for at least two minutes every fifteen minutes beginning at the top of each hour. By 1932, best practices were regulated with a treaty; 500Khz was established as the international band for distress signals, and all maritime radio operators were required to observe radio silence to listen for distress calls on this band twice every hour during their shift, from :15-:18 and again from :45-:48.

Since the Merchant Shipping Act of 1919, British ships above a certain size had been required to maintain a continuous watch, monitoring frequencies for distress calls around the clock. Even though there were separate certifications for “watchers” vs fully licensed operators (watchers only had to be qualified to recognize distress signals and summon the operator, they didn’t receive messages nor send any), this added employee costs to shipping companies’ bottom lines.

Marconi continued to innovate, and in 1920 he conducted the first public test of an automated distress alarm. He refined this over the next few years, and by 1927 these alarms were mandatory on all British ships carrying between 50-200 people on board. The alarms responded to a series of 12 four-second steady signals (hence the dashed markings around the edge of the clock in the photo above), interspersed with one second of silence between each. When a receiver detected this sequence, it triggered alarm bells to ring in the radio room and on the bridge of the receiving ship which could only be silenced by an operator in the radio room. This ensured that all ships in range would receive distress alerts even if their wireless operators were off duty or asleep⁠—as had been true for the ship closest to the Titanic when it sank.

To see a Marconi shipboard auto-alarm system circa 1920, check out the History of Science Museum at Oxford.

He also developed portable emergency radios for use on lifeboats like the one shown above. It could transmit the automatic emergency signal, or transmit manually, and could use both of the multiple emergency wavelengths for both Morse code and voice. The use of 2182KHz in this example places it after 1947 when that band was designated, but the exact date is unclear as this model was still in use into the 1980s. It had a telescoping aerial that could extend to 15′, and both an external power supply and a backup hand-crank. I especially like the instructions to throw the earth into the sea.

Later years

Marconi won international acclaim and built a successful (despite regular financial crises) company that morphed through varying forms but retained his name until 2005. He drove wireless radio innovation for decades and was directly involved with its commercial, safety, and military development. His company even developed directional tools, from systems that used radio waves to locate and track military targets in WWI, to loop antennas that helped ships determine their exact location at sea nearly a century before satellites and GPS.

Clifden station was damaged in 1922 during the Irish Civil War and was never repaired. By that time, wireless technology had advanced enough for Marconi’s stations in Wales and England to take over for Clifden.

From trench warfare at the Somme and ships in distress at sea to daily news bulletins and personal messages, so many people relied on Marconi’s inventions to communicate. It’s hard to imagine a world without his influence, and I feel very lucky to have found myself semi-accidentally tracking his history. Cheers, Marconi! from the banks of Newfoundland to a blanket bog in Connemara, I salute you.

Sources & further reading

Historical accounts


Articles and research papers

Additional pieces of interest

Historical documents

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s