On 16 July 1962, a 23-year-old French speleologist named Michel Siffre climbed down into the Scarasson cavern, an underground glacial cave in the Ligurian Alps, and stayed there for about two months. He had no clock, no calendar, and no sunlight. He ate when he felt hungry and slept when he felt tired.
Siffre had a one-way telephone line to a support team at the cave entrance. He could call up to them; they could not call down to him, and they were under instruction never to tell him the time or date. Each time he woke, ate, or prepared to sleep, he phoned the surface so the team could record it. The record they built was a log of his life kept on his own internal schedule, with no reference to the world above.
When the team finally told him, in mid-September, that the planned period was over, Siffre was not ready for it. By his own running estimate, far less time had passed. Accounts of the experiment differ on the exact size of the error, with some putting it at around a month and others at about two weeks, but the direction is consistent and not in dispute. Siffre believed he still had a substantial stretch of the experiment left to go. He had lost weeks without noticing.
What the error actually showed
The misjudged date is the memorable part of the story. It points to two separate effects, and they are worth keeping apart.
The first is in the daily log the surface team kept. When Siffre’s sleep and waking times were laid out against real, external time, they did not hold to a 24-hour pattern. His sleep-wake cycle ran slightly long, settling at about 24 and a half hours rather than exactly 24. That mattered, because it showed his body was keeping time on its own, with nothing in the cave to reset it.
This is what later came to be called a free-running circadian rhythm. The term describes a body clock operating without its usual external corrections. In ordinary life, daylight, clocks, meals, and the schedules of other people constantly reset the human body clock to the 24-hour day. Those resetting cues are known as zeitgebers, from the German for “time-givers.” Strip them all away, as the cave did, and the clock keeps running, but on its own slightly different period.
The second effect is the larger error in his date estimate, and it was not a simple accumulation of that half-hour daily drift. Half an hour a day over two months adds up to only a little more than a day. The stranger finding was psychological. In the cave, with no changing light, no calendar, and little variation from one period of wakefulness to the next, Siffre’s memory of elapsed time compressed. He did not merely run late. He lost track of duration itself.
Why this counted as a discovery
Before Siffre’s cave stay, it was not settled whether humans had a genuine internal clock at all. One reasonable view held that the human sense of a day was essentially learned and externally driven, a response to sunrise, hunger, and habit, with nothing autonomous underneath.
The Scarasson experiment was strong evidence against that. Siffre had been sealed away from every external time cue for two months, and a rhythm had not vanished. A rhythm had persisted. It had simply persisted at its own pace rather than the calendar’s. That points to a clock that is internal, that runs whether or not the outside world is feeding it information.
Siffre, by training, was a geologist. He had gone to Scarasson partly to study its underground glacier, and the decision to turn the trip into a study of his own experience of time was, as he later described it, the central idea of his life. He was not, in 1962, a chronobiologist. The field of human chronobiology did not really exist yet. His experiment is one of the things that brought it into being.
What followed
The 1962 results drew serious attention, including from researchers interested in how humans would cope with long-duration spaceflight, where ordinary day and night also disappear. NASA, Siffre later said, funded mathematical analysis of his first experiment.
Siffre went on to run many further isolation studies, and in 1972 spent six months alone in a cave in Texas. In those longer experiments the free-running rhythm sometimes drifted much further, with sleep-wake cycles in some cases stretching toward 48 hours. He was candid, in later interviews, that long stretches of solitary time underground were difficult and damaging, and that the kind of experiment he had run on himself would be hard to justify now.
The core result has held up. The human body keeps time on its own, with an internal clock that runs close to the 24-hour day but not exactly on it, and that needs daily correction from light and routine to stay aligned with the world. That correction is happening, quietly, every ordinary morning. Siffre’s two months in the dark are part of how we know it is there, by showing what the clock does when nothing is left to correct it.
