In 1859 the telegraph broke, and two operators kept working on the current from the sky
On the night of 2 September 1859, the telegraph line between Boston and Portland was failing. Pylons were throwing sparks. Operators up and down the eastern seaboard were getting shocks off their own equipment. The Boston operator sent this to Portland:
Please cut off your battery entirely for fifteen minutes.
Portland: Will do so. It is now disconnected.
Boston: Mine is disconnected, and we are working with the auroral current. How do you receive my writing?
Portland: Better than with our batteries on. Current comes and goes gradually.
Boston: My current is very strong at times, and we can work better without the batteries, as the aurora seems to neutralize and augment our batteries alternately, making current too strong at times for our relay magnets. Suppose we work without batteries while we are affected by this trouble.
Portland: Very well. Shall I go ahead with business?
Boston: Yes. Go ahead.
They ran the line for about two hours on no power supply at all — on current induced in the wire by the sky. It is the first time on record that more than a word or two was sent that way. The exchange survives because the Boston Evening Traveler printed it.
What was hitting them
Just before noon the previous day, two English astronomers looking at the Sun through separate instruments — Richard Carrington and Richard Hodgson — each saw a patch of the surface flare white. Neither knew what he was looking at. Those are the first recorded observations of a solar flare, and they were published side by side in the Monthly Notices of the Royal Astronomical Society because neither man could corroborate the other in advance.
What followed the flare was a coronal mass ejection aimed at Earth. A typical one takes several days to cross the 150 million kilometres. This one took 17.6 hours — the standing explanation, and it is offered as a belief rather than a measurement, is that another ejection a few days earlier had swept the ambient solar wind out of its path. It arrived on 1–2 September and produced the most intense geomagnetic storm in recorded history.
The aurora went everywhere. Gold miners in the Rocky Mountains got up and started cooking breakfast because the light woke them and they thought it was morning. In the north-eastern United States people read newspapers by it. It was seen from Cuba, Hawaii, southern Japan, Queensland, New Zealand, south-central Mexico, and Colombia — which is nearly on the equator. A miner at Rokewood, in Victoria, remembered it fifty years later well enough to write to a Perth paper about the colours “always curling round at the zenith”, and to note who saw what in it: “the superstitious and the fanatical had dire forebodings, and thought it a foreshadowing of Armageddon.”
The sentence I keep coming back to
Carrington had, by any reasonable standard, cracked it. He had seen a flare on the Sun at noon. A magnetometer at Kew, watched by Balfour Stewart, had jumped at the same moment. The next day the Earth’s magnetic field convulsed and every telegraph line on two continents went strange.
He wrote that he was not sure the two were related. “One swallow does not make a summer.”
He was right to be unsure, and being right about that cost him the credit. What turned his suspicion into knowledge was not a better insight but a larger sample — the American mathematician Elias Loomis collected reports of the storm’s effects from around the world and published them, and it is the compilation, not the observation, that made the solar–terrestrial connection stand up. A single spectacular coincidence, watched by the only two people on Earth in a position to see it, is exactly the kind of evidence that is most tempting and least sufficient.
The part that has aged badly
The 1859 storm broke the only long-distance network that existed, and two men renegotiated their protocol mid-failure and kept sending traffic. It is easy to read that as sturdiness. I don’t think it is. A telegraph circuit is a wire, a battery and a person paying attention, and a person can decide to work with the fault. There is nothing in it fast enough or automatic enough to save itself in ninety seconds — which is the number from the next time this happened to somebody.
On 13 March 1989, a weaker storm induced currents in the ground across eastern Canada. Most of Quebec sits on a rock shield, so the current found a less resistant path: Hydro-Québec’s 735 kV transmission lines. The variations tripped the circuit breakers. The James Bay network went offline in under ninety seconds and Quebec was dark for nine hours. The aurora that night reached Texas and Florida, and because it was 1989, some people watching it wondered whether they were seeing a first strike.
Everything that made 1989 worse than 1859 is a thing we would call an improvement: longer lines, higher voltages, tighter coupling, protective relays quick enough to disconnect a continent before anyone can be consulted. The operators at Boston and Portland had two hours to negotiate. Hydro-Québec’s breakers had already acted before a human knew there was weather.
A 2013 study by Lloyd’s of London and Atmospheric and Environmental Research put the cost of a Carrington-scale event today at 600 billion to 2.6 trillion US dollars in the United States alone. Other work traces a second-order effect most people would not think of: not the blackout, but the loss of industrial production of fertiliser and pesticide, and through that a modelled global crop yield reduction of 38–48%.
And on 23 July 2012, a Carrington-class ejection crossed Earth’s orbit and missed, by roughly nine days of orbital travel. Nobody had to negotiate anything. Nobody was even asked.
Sources: English Wikipedia, “Carrington Event” (page id 19325218) and “March 1989 geomagnetic storm” (page id 22318042), both fetched 2026-08-08. The Boston–Portland exchange is quoted there from the Boston Evening Traveler, 3 September 1859; the Rokewood recollection from C. F. Herbert’s 1909 letter to the Perth Daily News.