Platonic solid

• 1. Introduction

  ›

  Platonic solids are among the oldest objects of mathematical fascination on earth, and there are exactly five of them. Not four, not six. Five. That absolute limit is itself one of the most startling facts in all of geometry. The tetrahedron, with its four triangular faces. The cube, with its six square faces. The octahedron, with eight triangular faces. The dodecahedron, with twelve pentagonal faces. And the icosahedron, with twenty triangular faces. No other shape can belong to this family. The ancient Greek philosopher Plato, for whom the shapes are named, believed they were not merely geometric curiosities but the literal building blocks of reality. What is it about these five shapes that made them so captivating to philosophers, astronomers, and mathematicians across millennia? And what does it actually mean that only five can exist?

• 2. From Stone Balls to Plato

  ›

  Carved stone balls found in Scotland have sometimes been proposed as early evidence that Neolithic people knew about these shapes. The claim does not quite hold up. Those balls bear rounded knobs rather than polyhedral faces, and the number of knobs frequently does not match the vertices of any Platonic solid. No ball's knobs matched the twenty vertices of the dodecahedron, and their arrangement was not consistently symmetrical.

  Among the ancient Greeks, some sources credit Pythagoras with discovering these forms. Other evidence suggests he may have known only the tetrahedron, the cube, and the dodecahedron. A contemporary of Plato named Theaetetus is credited with mathematical descriptions of all five and may have produced the first known proof that no other convex regular polyhedra exist.

  Plato engaged with the shapes in the dialogue Timaeus, written around 360 B.C. He matched each of the four classical elements to one of the solids: earth to the cube, air to the octahedron, water to the icosahedron, and fire to the tetrahedron. The fifth solid, the dodecahedron, he addressed only obliquely. His words: the god used it for arranging the constellations on the whole heaven. Aristotle proposed a fifth element he called aither but showed no interest in pairing it with Plato's fifth solid, leaving that cosmic assignment perpetually incomplete.

• 3. Euclid's Proof and Kepler's Dream

  ›

  Euclid gave the Platonic solids their most rigorous early mathematical treatment in the Elements. The entire final book, Book XIII, is devoted to their properties. Propositions 13 through 17 cover the construction of each solid in turn, and for each one Euclid calculated the ratio of the circumscribed sphere's diameter to the edge length. Proposition 18 then argues that no further convex regular polyhedra can exist. The mathematician Andreas Speiser has argued that building up to those five constructions was the central organizing goal of the entire deductive system of the Elements.

  Nearly two thousand years later, Johannes Kepler attempted something far more ambitious. In Mysterium Cosmographicum, published in 1596, he proposed that the orbits of the six planets then known could be explained by nesting the five solids inside one another, each pair of adjacent solids separated by inscribed and circumscribed spheres. In his model the octahedron sat innermost, followed by the icosahedron, the dodecahedron, the tetrahedron, and finally the cube, with the outermost sphere representing Saturn's orbit. Kepler eventually had to abandon this planetary architecture. But the research it demanded led him to his three laws of orbital dynamics, including the discovery that planetary orbits are ellipses rather than circles. He also found two nonconvex regular polyhedra that now bear his name, the Kepler solids.

• 4. Why Only Five Exist

  ›

  The proof that exactly five Platonic solids exist and no more is simpler than it might seem. At every vertex of a Platonic solid, at least three faces must meet. The angles of those faces, added together, must sum to less than 360 degrees; otherwise the shape cannot close into a solid and instead tiles flat across a plane. A vertex where the angles add to exactly 360 degrees produces a flat, infinite tiling rather than a polyhedron.

  With triangular faces, each contributing 60 degrees, you can build a valid vertex with three, four, or five triangles meeting. Six triangles would sum to exactly 360 degrees and produce only a flat tiling. That gives you the tetrahedron, the octahedron, and the icosahedron. With square faces, each contributing 90 degrees, only three squares per vertex work; four would again sum to 360. That gives the cube. With pentagonal faces, each contributing 108 degrees, only three pentagons per vertex are possible; four would exceed 360 degrees. That gives the dodecahedron. Hexagonal faces already contribute 120 degrees each, so three of them meet at exactly 360 degrees, producing only a flat tiling and no solid at all. The constraint exhausts itself there.

  A purely algebraic version of the same argument runs through Euler's formula for polyhedra, V minus E plus F equals 2. Combining that relation with the constraints on how many edges meet at each face and vertex narrows the field to exactly the same five pairs of integers. Descartes' theorem adds a geometric gloss: the total angular deficiency across all vertices of any convex polyhedron is always 720 degrees, and dividing that by the deficiency at a single vertex gives the vertex count.

• 5. Symmetry and Dual Pairs

  ›

  Each Platonic solid has a dual, formed by placing a new vertex at the center of each face and connecting adjacent ones. Swapping faces and vertices in this way produces another Platonic solid. The tetrahedron is its own dual. The cube and the octahedron are duals of each other; a cube has eight vertices and six faces, while the octahedron reverses those counts. The dodecahedron and the icosahedron form the third pair.

  Because a polyhedron and its dual share the same symmetry group, the five solids reduce to only three distinct symmetry groups. The tetrahedral group has an order of 24 including reflections. The octahedral group, which also governs the cube, has an order of 48. The icosahedral group, which also governs the dodecahedron, has an order of 120. The proper rotation subgroups have orders of 12, 24, and 60 respectively, each equal to exactly twice the number of edges in the corresponding polyhedra.

  All five solids also possess a remarkable property called the Rupert property: a hole can be cut through each solid large enough for a copy of itself to pass through. All five share this feature.

• 6. In the Natural World

  ›

  The tetrahedron, the cube, and the octahedron all appear in crystal structures. The icosahedron and the dodecahedron do not, though one mineral form called the pyritohedron has twelve pentagonal faces arranged in the same pattern as a regular dodecahedron. The pyritohedron's faces are not perfectly regular, so it falls short of true membership in the family.

  Boron is more obliging. Allotropes of boron and many boron compounds, including boron carbide, incorporate discrete clusters of twelve boron atoms arranged as icosahedra within their crystal structures. Carborane acids have molecular shapes that closely approximate regular icosahedra.

  In the early twentieth century, the zoologist Ernst Haeckel documented species of Radiolaria whose skeletons took the forms of regular polyhedra. Among them: Circoporus octahedrus, Circogonia icosahedra, Lithocubus geometricus, and Circorrhegma dodecahedra. Many viruses, including the herpes virus, also adopt the icosahedral form. The icosahedron suits viruses because it can be assembled from many identical protein subunits, reducing the genetic coding required to specify the entire shell.

  For liquid crystals, icosahedral symmetry was proposed theoretically by H. Kleinert and K. Maki in 1981. Three years later, Dan Shechtman discovered icosahedral structure in aluminum, work that earned him the Nobel Prize in Chemistry in 2011.

• 7. Higher Dimensions

  ›

  Three dimensions turn out to be unusually rich for convex regular polytopes. In two dimensions there are infinitely many regular polygons. In four dimensions there are exactly six convex regular polytopes. In every dimension higher than four there are only three. The Swiss mathematician Ludwig Schläfli discovered the four-dimensional analogues in the mid-nineteenth century. Five of his six four-dimensional polytopes correspond directly to the Platonic solids, and a sixth, the self-dual 24-cell, has no three-dimensional counterpart at all.

  In any dimension greater than four, the only convex regular polytopes are the simplex, the hypercube, and the cross-polytope. In three dimensions those three coincide with the tetrahedron, the cube, and the octahedron. The dodecahedron and the icosahedron are, in a precise sense, a three-dimensional exception that higher dimensions do not repeat. Their icosahedral symmetry group, with its order of 120, is the most complex of the three polyhedral groups and has no analogue among the polytope families that persist into higher dimensions.