The Island

Neutrons run left to right. Protons run back into the dark. Each column is one nucleus anyone has ever measured, and its height is how long it survives, on a log scale: a nanosecond on the floor, flat-topped plateaus for the ones that never decay.

Color is how it dies. Gold throws off a helium nucleus. Blue turns a neutron into a proton. Red goes the other way. Green tears itself in half.

Textbooks call this the valley of stability, because they plot energy, and stable means low. Plot lifetime instead and the valley turns into a ridge, with the 288 stable nuclei along its spine.

Watch it bend. Light nuclei like equal numbers of protons and neutrons. Heavy ones need extra neutrons to dilute all that repulsion, about one and a half per proton by the time you reach lead. Past lead the ridge gives out. Uranium-238 hangs on for 4.5 billion years, which is the only reason there's any left in the ground.

This is the chart filling in, by year of discovery. 1896: Henri Becquerel leaves uranium salts on a photographic plate in a drawer and finds the plate fogged. 1932: the neutron. 1940: neptunium, the first element heavier than uranium.

After that come the accelerator decades, and new nuclei pile up along both edges. The top row of the periodic table was closed out by tennessine in 2010. Nobody has confirmed anything heavier than oganesson, element 118.

Electrons fill shells, and an atom with a full shell is a noble gas: smug, inert, untouchable. Protons and neutrons fill shells too. Hit one of these numbers and the nucleus gets a large bonus in binding energy.

Maria Goeppert Mayer and Hans Jensen explained why, and shared the 1963 Nobel Prize for it. Lead-208 has 82 protons and 126 neutrons, magic twice over, and it is the heaviest nucleus known that never decays.

Everything in amber is the model talking. It is built from the equations physicists were using in the 1960s: the nucleus as a charged liquid drop, corrected for shells, run through the standard formula for each way a nucleus can come apart.

Against the 179 superheavy nuclei that have actually been measured, it's usually within a factor of a couple hundred. On a chart that spans twenty-nine orders of magnitude, that makes it a decent pair of binoculars. Now let the sea in. Anything that lasts less than a second goes under.

In 1966 William Myers and Władysław Świątecki extended their shell correction past the end of the known chart and found the ground rising again: superheavy nuclei near 184 neutrons, held up by a shell closure no lab could reach.

This model finds it too, near 112 protons, with a peak lifetime around 10¹² seconds. Labs have been rowing toward it for sixty years. They can get the protons. They cannot get enough neutrons into the boat.

Here is the fiction, and it is one number. Suppose the proton shell at 126 isn't modest. Suppose it is enormous, worth an extra 24 MeV of binding right at 126 protons and 184 neutrons. Same equations. Watch what they do with it.

That isn't an island anymore. It's a spire, and the nucleus on top doesn't decay. In 2026 a civilian in New York ran exactly this model in a web browser and posted it. Thirty years later Tern came home with a sample. You are looking at the original.