Ball Lightning: The Phenomenon Science Took Too Long to Accept

The story is always more or less the same, and people have been telling it for centuries. A thunderstorm is overhead. A glowing ball of light, somewhere between the size of an orange and a beach ball, appears in the air. It hovers, or drifts slowly, sometimes at head height, sometimes along the ground. It may pass through an open window or come down a chimney and move across a room, apparently indifferent to walls and furniture. It glows white, or red, or a strange blue, and it lasts only seconds. Then it either fades away silently or bursts with a loud crack, sometimes leaving a sharp smell of sulphur or ozone behind. The witnesses are farmers, sailors, housewives, soldiers, pilots and, occasionally, physicists, and for the better part of two hundred years the official scientific response to all of them was the same: you are mistaken.

Ball lightning is a rare and genuinely strange case in the history of belief, because here the ordinary roles are reversed. The people reporting the uncanny thing were, on the whole, telling the truth, and the experts insisting it could not be real were the ones clinging to a comforting story. This is a piece about a phenomenon that took an embarrassingly long time to be accepted, and about what its long exile teaches, which is that the line between careful scepticism and stubborn denial is thinner and more human than science likes to admit.

The thing in the room

Begin with how convincing the reports are, because their sheer quality is the heart of the matter. Ball lightning has been described consistently, across centuries and continents, by observers who had never met and often had no shared vocabulary for it. One of the most famous early cases is the Great Thunderstorm of 1638 at the church of Widecombe-in-the-Moor, on Dartmoor in Devon, where during a violent storm a great ball of fire was said to have burst into the church, killing four people and injuring around sixty, scorching the building and filling it with a sulphurous stench. Contemporaries attributed it to the Devil, which was the available explanation, and it is now regarded as one of the best early records of ball lightning.

The reports kept coming, and some of them came from people impossible to dismiss as ignorant. In 1753 the physicist Georg Wilhelm Richmann, working in St Petersburg on the electrical nature of storms in the wake of Franklin’s kite experiments, was killed when a detached globe of pale fire reportedly leapt from his apparatus to his head during a thunderstorm, leaving him dead on the spot and his engraver stunned beside him. He may be the first person recorded to have died while scientifically investigating atmospheric electricity, and a form of ball lightning is the leading candidate for what killed him. Over the following two centuries the accounts accumulated into the thousands, from every kind of witness in every kind of place, and their consistency is exactly what a real, rare, physical phenomenon would produce.

Why the experts said no

And yet, deep into the twentieth century, a substantial part of the physics community held that ball lightning did not exist. The reasons are worth understanding sympathetically, because they were not stupid, and because they are the fork where the scientific consensus departed from the evidence in front of it.

The first problem was reproducibility. Science trusts what it can summon on demand in a laboratory, and ball lightning stubbornly refused to be summoned. It appeared at random during storms, lasted seconds, and left almost nothing behind to measure. A phenomenon that cannot be reproduced sits uneasily with a discipline built on the controlled experiment, and many physicists concluded that a thing which could not be produced and measured probably was not a thing at all. The second problem was theoretical: no one could explain how a ball of glowing energy could persist for several seconds, drift against the wind, and pass through glass, without a mechanism to hold it together. The known physics said such a structure should dissipate almost instantly. If theory said it was impossible and the lab could not make one, the anecdotes could be waved aside.

So they were, and here the sociology gets uncomfortable. The witnesses were often ordinary people, and their testimony was discounted precisely because it was anecdotal and because they were not experts. A convenient alternative was offered: ball lightning was probably an optical illusion, an afterimage burned onto the retina by a nearby ordinary lightning flash, which the startled observer then “saw” floating in the room. This explanation had the great advantage of requiring no new physics and no respect for the witness’s account of what happened. It explained the phenomenon away by explaining the witness away. For a long time it was the respectable position, and to report having seen ball lightning was to risk being gently classed with those who saw ghosts.

The witness who could not be dismissed

The wall began to crack when the witnesses became harder to patronise. A frequently cited turning point came in 1963, when R.C. Jennison, a respected radio astronomer and academic, observed ball lightning at close quarters aboard an aircraft: a glowing sphere emerged from the cockpit area after the plane was struck, floated down the aisle past his seat at a steady, unhurried pace, and continued toward the rear of the cabin. Jennison was a trained scientific observer with no incentive to invent, describing a slow-moving object in good light at arm’s length, and the retinal-afterimage explanation simply would not stretch to cover him.

As more accounts arrived from pilots, scientists and engineers, the “everyone is mistaken” position grew harder to hold with a straight face. The accumulation of credible testimony did to ball lightning something similar to what happened with other long-doubted natural phenomena that turned out to be real, and it is a familiar arc: the reports that a discipline finds inconvenient are dismissed until they become too numerous and too well-sourced to dismiss, and then, quietly, the consensus turns. The recurring luminous plasmas studied for decades at the Hessdalen Lights in Norway sit in a similar zone, real enough to instrument and too strange to fully explain, waiting on the physics to catch up.

The spectrograph in Tibet

The decisive moment, the equivalent of finally reading the code, arrived by accident in 2012 on the Qinghai plateau in China. A team of researchers led by Jianyong Cen at Northwest Normal University in Lanzhou had set up spectrographs and video cameras on the Tibetan plateau to study ordinary cloud-to-ground lightning. During a storm, a bolt struck the ground perhaps a kilometre from their instruments, and out of the strike point rose a glowing ball that drifted for a second or so before vanishing. By pure luck, their equipment was pointed the right way, and for the first time in history the spectrum of a natural ball lightning event was recorded, its light split into the fingerprint of the elements producing it.

What the spectrum showed was decisive. The glowing ball’s light contained the emission lines of silicon, iron and calcium, the very elements of soil. The ball was, in effect, vaporised ground, burning in the air. This was powerful confirmation of a theory proposed in 2000 by John Abrahamson and James Dinniss at the University of Canterbury in New Zealand, who had suggested that when lightning strikes soil, the intense heat vaporises silicon compounds and the shockwave ejects them into the air as a fluffy network of silicon nanoparticles, which then slowly oxidise, glowing as they burn, held loosely together in a floating ball for a few seconds until the fuel is spent. The Tibetan spectrum matched the prediction. The peasants and pilots had been describing a real object all along, and it was, remarkably, a cloud of burning earth.

The laboratory catches up

The Tibetan spectrum did not stand alone; by then the laboratory had begun, haltingly, to make things that behaved like ball lightning. Working from the burning-silicon idea, researchers passed powerful discharges across silicon wafers and produced small glowing balls that floated briefly above the surface before dying, crude cousins of the real thing but persuasive proof of the mechanism. In a separate line of work, the Israeli engineer Eli Jerby and colleagues at Tel Aviv University found in the mid-2000s that firing a microwave drill into certain solid materials could eject a floating, glowing plasmoid a few centimetres across, a fireball that hovered and drifted for a fraction of a second. None of these laboratory objects reproduces natural ball lightning exactly, and the phenomenon still cannot be conjured reliably on demand, which is why physics continues to regard it as not fully solved. But the gap between “impossible” and “not yet fully modelled” is an enormous one, and crossing it was the work of the past two decades.

Surveys have also quietly demolished the idea that ball lightning is vanishingly rare or confined to the credulous. When researchers have polled large populations, a small but consistent fraction, on the order of a few per cent, report having seen something matching the description at some point in their lives, a rate broadly comparable to the odds of being caught near an ordinary lightning strike. That is a great many witnesses, spread across every class and country and level of education, all describing the same short-lived glowing sphere. Treating that body of testimony as mass illusion was always a heavier assumption than treating it as evidence, once the theoretical objection had been answered.

What the long denial was really about

The obvious moral is that science should have listened to the witnesses sooner, and there is truth in it, but the honest lesson is more uncomfortable and more interesting. The physicists who doubted ball lightning were doing something reasonable most of the time. Demanding reproducibility, distrusting isolated anecdotes, and preferring explanations that require no new physics are exactly the habits that make science reliable and protect it from the endless tide of things people sincerely believe they have seen. The very same scepticism that wrongly rejected ball lightning correctly rejects a thousand other claims that really are illusions and misperceptions. The retinal-afterimage explanation was the discipline’s immune system working as designed.

The failure was one of degree, the point at which a healthy caution curdled into a refusal to look. When the credible reports numbered in the thousands, when trained observers were among the witnesses, the balance of evidence had shifted, and the continued insistence that it must be illusion stopped being scepticism and became its own article of faith, a story told to preserve the comfort of a theory that had no room for the phenomenon. Expertise had quietly become a reason not to believe the evidence rather than a tool for weighing it, and that inversion is the real cautionary tale, because it can happen to any authority that mistakes the limits of its current models for the limits of the world.

There is a particular dignity in the ball lightning story that the usual myths lack. Most of the phenomena that draw a crowd, the sounds that become monsters like the Bloop, the desert glows that become spacecraft as at the Marfa Lights, turn out on inspection to be ordinary things magnified by longing. Ball lightning went the other way. The witnesses were right, the experts were wrong, and the truth, when it finally came, was stranger than either side had quite imagined: the ground itself, struck into a burning sphere and set adrift through a farmhouse for a few impossible seconds. The people who saw it and were disbelieved deserved better, and what they were owed, in the end, was the simplest and hardest thing to give a witness whose story does not fit the model. They deserved to be believed, and it took science two centuries to manage it.