Baily's Beads: The Last Sunlight Through Lunar Valleys

What you actually see at second contact

In the final minute before a total solar eclipse, the Sun is reduced to a needle-thin crescent. Then, in the last two seconds, that crescent does something startling: instead of tapering smoothly to nothing, it fragments into a line of separate, brilliant points of light hugging the dark edge of the Moon. They are not uniform. Some are fat, some are pinpricks; the spacing is irregular; a bead can flare, fade, then flare again. This is the moment astronomers call second contact, and the beads are the last light of the photosphere getting through.

Watch carefully and you will see the beads go out one by one, not all at once. As the Moon's limb slides forward, it seals off valley after valley. Usually the very last bead is disproportionately bright — a single blazing point set against a faint, ghostly ring of inner corona. That is the diamond ring effect, the beads' more famous cousin: it is simply the moment when only one bead remains. A heartbeat later even that closes, and a thin crimson arc — the chromosphere — flickers along the limb before it, too, is swallowed and totality begins.

The whole sequence reverses at third contact, when the Sun re-emerges: chromosphere, diamond ring, then a fresh string of beads on the opposite side of the Moon. Because the phenomenon is so brief and so bright, it is also the single most dangerous moment of an eclipse to observe carelessly — even one bead is direct, unfiltered sunlight.

The mechanism: a mountain range silhouetted on the Sun

The beads exist because the edge of the Moon is not a perfect circle. It is a jagged profile of crater rims, mountain peaks, and valley floors. When the Moon's silhouette nearly covers the Sun, the mountains along its limb touch the solar edge first and block the light, while the valleys between them still have open sky behind — and through each of those gaps, a tiny slice of the blindingly bright photosphere shines straight through to your eye. Each surviving gap is one bead.

Think of it as looking at a floodlight through the teeth of a mountain-range skyline: where a peak rises, the light is cut; where a notch dips, a shaft escapes. The lunar limb supplies a real skyline. Elevation differences of just a few kilometers between adjacent limb features are enough, because at the Moon's distance those features subtend only a fraction of an arcsecond — comparable to the razor-thin sliver of Sun still uncovered at second contact. Sunlight passing through a valley a few km deep is precisely the scale of geometry that produces a bead.

Two things make the effect fleeting. First, the Moon's shadow races across the geometry: the Moon moves against the Sun at roughly its own diameter every hour, so the crescent thins to nothing and the last valleys close in seconds. Second, the beads are most dramatic and longest-lasting near the edge of the path of totality, where the alignment is grazing and the lunar limb skims tangentially along the solar edge. Near the path's center the beads flash for barely a second; a grazing observer at the limit of the path can watch beads chase each other along the limb for up to about 90 seconds.

Total versus annular — and the ring of beads

Baily's most celebrated sighting in 1836 was not during a total eclipse at all — it was an annular one. That distinction is worth understanding. The Moon's orbit is elliptical, so the Earth–Moon distance varies by about 11% between perigee and apogee, changing the Moon's apparent diameter by a similar amount. When the Moon is near apogee it looks slightly smaller than the Sun and cannot cover it completely; at maximum eclipse a bright "ring of fire" of photosphere remains all the way around. When the Moon is near perigee it looks slightly larger, blocks the Sun entirely, and totality — with the corona — becomes possible.

In an annular eclipse, the ring is on the verge of closing as it forms and reopening as it breaks; at those instants its thinnest segment is chopped by the limb valleys into beads. In a total eclipse, the beads appear at the very brink of totality (second contact) and its end (third contact). The physics of the beads is identical in both cases: photosphere leaking through lunar valleys. What differs is only whether the Moon is big enough to eventually cover the disk.

This is exactly why Baily saw beads so clearly in 1836: an annular eclipse holds the near-perfect edge alignment for a comparatively long time, giving the limb topography an extended chance to slice the light. Under the right geometry the beads can briefly encircle almost the entire lunar disk — a full necklace rather than a short string.

Baily's Beads as a scientific ruler

Beads are not just a spectacle — they are a precise clock. The moment a given bead appears or disappears is fixed by geometry alone: the positions and angular sizes of the Sun and Moon, and the exact profile of the lunar valley responsible. Crucially, that timing is insensitive to atmospheric seeing — the blurring that plagues most ground-based solar measurements — because you are timing a sharp on/off event, not measuring a fuzzy edge.

In 1973, David Dunham and Joan Dunham turned this into a technique for measuring the Sun's diameter. The idea: station observers just inside the northern and southern limits of the eclipse path, video-record the beads, and time each one against a known lunar limb profile. Because the beads mark where the photosphere threads between lunar peaks, the pattern and timing of the beads pin down the effective angular size of the Sun. This work has been coordinated for decades by the International Occultation Timing Association (IOTA), including campaigns on the eclipses of 3 October 2005, 29 March 2006, 22 September 2006, and 1 August 2008.

The payoff has been a long-running search for whether the solar radius changes over time. Applying the bead-timing method (Dunham & Dunham) to the eclipses of 1715, 1976 and 1979 yielded a solar radius for 1715 that was about 0.34 ± 0.2 arcseconds larger than modern values — tantalizing but at the edge of the error bars, so the result remains debated rather than settled. The honest summary is that eclipse-bead timings are consistent with a solar radius that is stable to a small fraction of an arcsecond, and any long-term variation has not been robustly confirmed. A key ingredient that made the method far more reliable came later: precise lunar limb profiles from laser altimetry.

The Moon we needed to map first

For most of the beads' history, the biggest source of error was that we did not know the lunar limb well enough. Predicting exactly where beads would appear required a topographic profile of the Moon's edge, and for a century the best data came from painstaking observations of grazing occultations — timing stars as they winked in and out behind mountains at the lunar limb. It was indirect and incomplete.

That changed with spacecraft laser altimetry. Japan's Kaguya (SELENE) orbiter and NASA's Lunar Reconnaissance Orbiter (LRO), launched in 2009, carried instruments that measured the Moon's shape directly. LRO's Lunar Orbiter Laser Altimeter (LOLA) has gathered over 6 billion elevation measurements with a vertical precision of roughly 10 cm and an accuracy near 1 m, producing the best global topography of any body in the Solar System besides Earth. The Moon's full elevation range from these data spans roughly 20 km from its deepest basins to its highest terrain; the limb features that make individual beads are the more modest peaks and valleys of a few km along whatever edge happens to face us.

With modern limb models, the NASA Scientific Visualization Studio could predict the shape and motion of the 2017 total eclipse shadow — and where beads would flash — with unprecedented accuracy. So there is a neat inversion at the heart of this subject: we once used the beads to study the Moon and Sun; now we use a precisely mapped Moon to squeeze more science out of the beads.

History, misconceptions, and how to see them safely

The beads bear Francis Baily's name, but he was not the first to notice them. Edmond Halley — of comet fame — recorded them during the total eclipse of 3 May 1715 over England and correctly attributed them to irregularities of the lunar surface. What Baily did in 1836, observing from Jedburgh in the Scottish Borders, was describe them so vividly and publish so promptly (in the Monthly Notices of the Royal Astronomical Society that December) that the phenomenon captured astronomers' imaginations and helped ignite the great age of eclipse expeditions. The name stuck.

A few persistent misconceptions are worth clearing up:

• "The beads are gaps in the corona." No — they are pure photosphere (the Sun's visible surface), far brighter than the corona, shining through lunar valleys. The corona only becomes visible after the last bead is gone.
• "The diamond ring is a separate thing." It is just the special case of a single remaining bead, made to look like a jewel set in the faint ring of inner corona.
• "The red chromosphere is a bead." No — the chromosphere is a genuine layer of the Sun, glowing crimson from hydrogen's Hα line at 656.3 nm, revealed for a few seconds once the beads close.

Safety is non-negotiable. Baily's Beads and the diamond ring are still direct sunlight — even a single bead can damage your eyes and burn camera sensors. Keep certified solar filters on until the last bead vanishes and the corona is fully out; remove them only during totality; and put them back the instant the first third-contact bead reappears. During an annular eclipse there is no safe filter-off moment at all, because the photosphere is never fully covered.

The fast sequence of second contact: how the eclipse transitions from partial to total in a few seconds

Phenomenon	What you see	Physical source	Rough duration
Thin crescent	Sliver of bright Sun	Full photosphere, unbroken	Minutes, shrinking
Baily's Beads	Broken string of bright dots	Photosphere through lunar valleys	~1–2 s (up to ~90 s at edge)
Diamond ring	One brilliant dot on a faint ring	Last single bead + inner corona	~1–2 s
Chromosphere	Thin crimson arc	Hydrogen Hα emission at 656.3 nm	~2–5 s
Totality	Black disk + pearly corona	Photosphere fully hidden	Seconds to ~7.5 min