In February 1869, a chemistry professor in St. Petersburg was late with a textbook, short on sleep, and out of ideas. He wrote the elements on cards and started shuffling them the way a man shuffles a deck of patience. What came out of that afternoon was not a filing system. It was the discovery that matter has a shape.

Dmitri Mendeleev arranged the elements by the only quantity he could measure, atomic weight, and then did the thing that made him great rather than merely industrious: when the pattern disagreed with the measurements, he sided with the pattern. He left holes in his table for elements nobody had found. He corrected the accepted weights of beryllium, indium, and uranium because they spoiled the order. He described a metal he had never seen so precisely that a Frenchman who did find it had to remeasure his own sample and admit the Russian was right.

And then, forty years later, it turned out that Mendeleev had ordered the elements by the wrong number after all. The true index was not weight but nuclear charge. The periodic law had never been a law about chemistry. It had been a law about the nucleus the entire time, and chemistry was only its outermost shadow.

The Periodic Universe follows that discovery all the way down. Beneath the table on the classroom wall lies a second table nobody hangs up: the chart of nuclides, with its own periodicity and its own magic numbers, 2, 8, 20, 28, 50, 82, 126. This book argues that those numbers are not shell closures in a fitted potential. They are polyhedra. Two poles at opposite ends of an axis. Eight at the corners of a cube. Twenty at the vertices of a dodecahedron. The alpha particle is a cube, and it is the most durable object in the light half of the chart because a cube has no direction to be pulled in. The deuteron is a torus with one small blemish in its symmetry, and that blemish is exactly the quadrupole moment experimenters measure.