The Six Regular Polytopes of Four-Dimensional Space

A regular polygon is a shape in two dimensions with all sides equal and all angles equal. There are infinitely many: the equilateral triangle, the square, the regular pentagon, the regular hexagon, and so on for every $n \geq 3$.

A regular polyhedron — a regular shape in three dimensions — is much rarer. There are only five: the tetrahedron, cube, octahedron, dodecahedron, and icosahedron. The Greeks knew these as the Platonic solids, and Euclid devotes the final book of his Elements to proving that the list is complete. Why so few? Because in three dimensions there are tight constraints on which regular polygons can fit around a vertex: the angles must sum to less than $2\pi$, and this excludes most combinations.

What happens in four dimensions? Surely with an extra dimension to work in, there would be more regular shapes. The answer is unexpectedly delicate. There are exactly six regular four-dimensional polytopes. Two of them are the natural analogues of the cube and octahedron. Three more are 4D analogues of the dodecahedron and icosahedron. And one — the 24-cell — has no analogue in any other dimension at all. Above dimension four, the number drops back down to just three regular polytopes per dimension, forever. Dimensions $3$ and $4$ are uniquely rich, and the 24-cell is the most uniquely rich object inside that uniqueness.

This article is about Schläfli’s astonishing 1852 classification, the six regular 4-polytopes, and why one of them stands alone.

The pattern across dimensions

Begin with the broad shape of the answer. Let $f(d)$ denote the number of regular polytopes in dimension $d$. The values:

• $d = 1$: $f(1) = 1$ (just the line segment).
• $d = 2$: $f(2) = \infty$ (regular polygons for every $n \geq 3$).
• $d = 3$: $f(3) = 5$ (the Platonic solids).
• $d = 4$: $f(4) = 6$ (the four-polytopes listed below).
• $d \geq 5$: $f(d) = 3$ for every $d$.

The dimensions $3$ and $4$ are the only places where the count exceeds the universal three. Why? Because the constraint that lower-dimensional regular shapes fit around an edge or vertex with positive angular defect (think of the deficit in Descartes’s angle-deficit formula) only allows a few combinations, and these combinations are most numerous in dimension $4$.

The Schläfli symbol

The systematic notation for regular polytopes is the Schläfli symbol, introduced by Schläfli around 1850.

• A regular polygon with $n$ sides has symbol $\{n\}$.
• A regular polyhedron with faces $\{n\}$ meeting $q$ at each vertex has symbol $\{n, q\}$. For example, the cube is $\{4, 3\}$ (square faces, three meeting at each vertex), and the icosahedron is $\{3, 5\}$ (triangular faces, five meeting at each vertex).
• A regular four-polytope with cells $\{n, q\}$ meeting $r$ at each edge has symbol $\{n, q, r\}$.

The pattern continues to higher dimensions. The symbol $\{n_1, n_2, \dots, n_{d-1}\}$ encodes the complete combinatorial type of a regular polytope in $d$ dimensions.

For a regular polytope $\{n_1, \dots, n_{d-1}\}$ to exist, the symbol must satisfy certain compatibility constraints. Schläfli’s central observation was that these constraints can be translated into an inequality (the angular-defect condition), and that the solutions can be enumerated dimension by dimension.

The six regular polytopes of dimension four

The six regular polytopes of 4D, with their Schläfli symbols, vertex counts, and 3D cell types:

Polytope	Symbol	Vertices	Cells	Cell type
5-cell (simplex)	$\{3, 3, 3\}$	$5$	$5$	Tetrahedron
Tesseract (8-cell)	$\{4, 3, 3\}$	$16$	$8$	Cube
16-cell (orthoplex)	$\{3, 3, 4\}$	$8$	$16$	Tetrahedron
24-cell	$\{3, 4, 3\}$	$24$	$24$	Octahedron
120-cell	$\{5, 3, 3\}$	$600$	$120$	Dodecahedron
600-cell	$\{3, 3, 5\}$	$120$	$600$	Tetrahedron

Each row corresponds to a unique 4-polytope. They satisfy the four-dimensional analogue of Euler’s formula:

$V - E + F - C \;=\; 0,$

where $V, E, F, C$ are the numbers of vertices, edges, 2-faces, and 3-cells. (Compare with the polyhedron formula $V - E + F = 2$.) The right-hand side is zero because the Euler characteristic of the 3-sphere (the topological boundary of a 4-ball) is zero.

Visualising in three dimensions

Four-dimensional shapes cannot be drawn directly. The standard visualisation is a Schlegel diagram, a kind of stereographic projection that flattens a 4-polytope into 3D so that one cell becomes the outer “container” and the others nest inside. The picture below shows the Schlegel diagram of the tesseract (the 4D analogue of the cube): an outer cube containing an inner cube, with corresponding vertices connected.

In the picture, the outer large cube is one of the eight cubical cells of the tesseract. The inner small cube is another one, and the six “frustum-like” shells between them — bounded by a face of the outer cube on the outside and a corresponding face of the inner cube on the inside — are the remaining six cubical cells. Eight cells in total, each one a cube. The vertices on the diagram are eight of the sixteen total tesseract vertices; the other eight project to the same locations on the page (each of the eight “visible” vertices is really a pair).

The astonishing 24-cell

Of the six regular 4-polytopes, four have direct analogues in three dimensions. The 5-cell is the 4D analogue of the tetrahedron — five tetrahedral cells meeting at every vertex. The tesseract (or 8-cell) is the 4D analogue of the cube. The 16-cell is the 4D analogue of the octahedron — eight tetrahedral cells. The 120-cell is the 4D analogue of the dodecahedron, with $120$ dodecahedral cells, and the 600-cell is the 4D analogue of the icosahedron, with $600$ tetrahedral cells. These five fit a recognisable pattern.

The sixth — the 24-cell $\{3, 4, 3\}$ — does not.

It has $24$ vertices, $96$ edges, $96$ triangular faces, and $24$ octahedral cells. Its dual is itself (it is self-dual, like the tetrahedron). It has no analogue in three dimensions and no analogue in any dimension above four. It is unique to the fourth dimension, and it has been one of the most curious geometric facts in mathematics since Schläfli identified it in 1852.

The 24-cell is also intimately connected to the lattice of integer quaternions (specifically the Hurwitz quaternions), to the densest known sphere-packing in four dimensions, and to certain exceptional Lie algebras. Its appearance in dimension four is a “sporadic” phenomenon similar in spirit to the exceptional structures that occur sporadically in algebra — the Mathieu groups, the monster group, the $E_8$ lattice, the Leech lattice. The 24-cell sits in the same intellectual neighbourhood as those: a beautiful object that exists only because dimension four happens to have just enough symmetry to support it.

Above dimension four: the great simplification

In dimension $5$ and every dimension above, the number of regular polytopes collapses to exactly three:

• Simplex $\{3, 3, \dots, 3\}$ — the generalisation of the tetrahedron. In $d$ dimensions, it has $d+1$ vertices.
• Hypercube $\{4, 3, 3, \dots, 3\}$ — the generalisation of the cube. In $d$ dimensions, it has $2^d$ vertices.
• Cross-polytope or orthoplex $\{3, 3, \dots, 3, 4\}$ — the generalisation of the octahedron. In $d$ dimensions, it has $2d$ vertices.

These three families exist in every dimension, including $1, 2, 3, 4, \dots$. The 5-cell, tesseract, and 16-cell from the 4D list are the four-dimensional members of the simplex, hypercube, and cross-polytope families. The other three regular 4-polytopes — the 24-cell, 120-cell, and 600-cell — exist only in dimension four.

The dodecahedron $\{5, 3\}$ and icosahedron $\{3, 5\}$ exist only in dimension three; they are the analogous “exceptional” Platonic solids unique to that dimension. The pattern across dimensions is:

• $d = 1$: $1$ regular polytope.
• $d = 2$: infinitely many.
• $d = 3$: $5$ regular polyhedra ($3$ universal: tetra, cube, octa; plus $2$ exceptional: dodeca, icosa).
• $d = 4$: $6$ regular 4-polytopes ($3$ universal + $3$ exceptional: 24-cell, 120-cell, 600-cell).
• $d = 5$ and above: only $3$ universal regular polytopes (simplex, hypercube, cross-polytope).

Dimensions $3$ and $4$ are the only places in the hierarchy with more than three regular polytopes. And the 24-cell is unique even among the exceptional ones in being the only regular polytope ever to be self-dual in this way.

Why so few?

The reason regular polytopes thin out so quickly in high dimensions is rooted in the angular defect condition. For a regular polytope $\{n_1, \dots, n_{d-1}\}$ to exist, the lower-dimensional regular polytopes that bound it must fit around their meeting points with strictly positive angular defect. Translating this condition into inequalities on the Schläfli symbol gives a system of integer inequalities that has only finitely many solutions in every dimension above $2$.

A careful count of solutions in each dimension:

• In dimension $3$: $\{n, q\}$ with $\tfrac{1}{n} + \tfrac{1}{q} > \tfrac{1}{2}$, giving $5$ Platonic solids.
• In dimension $4$: $\{n, q, r\}$ with both the 3D constraint on $\{n, q\}$ and a similar constraint for $\{q, r\}$, giving $6$ regular 4-polytopes.
• In dimension $5$: the analogous constraint admits only $3$ solutions.

The drop from $6$ at $d = 4$ to $3$ at $d = 5$ is sharp and complete. There is no in-between case; either dimension $4$ is special, or it isn’t. The reason it is special — beyond the bare counting — lies in the structure of the symmetry groups, especially the Coxeter group $F_4$ associated with the 24-cell, which has properties not present in any other Coxeter group.

A small list, an enormous shadow

The complete classification of regular polytopes was, in 1852, an obscure piece of mathematics in an obscure manuscript by an obscure Swiss professor. Schläfli’s Theorie der vielfachen Kontinuität was difficult to read, partly because it required readers to think in dimensions above three. The manuscript circulated privately during his lifetime; most of it was published only posthumously in 1901. Schläfli received almost no recognition during his career.

A century and a half later, the list of six regular 4-polytopes appears in nearly every introductory text on higher-dimensional geometry. The 24-cell is the central exceptional object of the subject. The 120-cell and 600-cell — vast and beautiful structures with hundreds of cells — are the four-dimensional analogues of the dodecahedron and icosahedron, and they have been visualised in countless ways through computer graphics, animations, and physical models (including the famous Roman dodecahedron and Klein’s icosahedron book).

Beyond their intrinsic beauty, the 4-polytopes have unexpected applications. The 24-cell is the kissing configuration in 4D — the optimal arrangement for $24$ unit spheres all touching a central one, the maximum possible in dimension four. The 600-cell is the basis for quasicrystalline structures found in certain alloys. The lattice generated by the 24-cell is the densest known sphere-packing in four dimensions. And the symmetries of the 600-cell appear in the action of the binary icosahedral group, which is central to certain constructions in algebraic geometry and the theory of singularities.

What Schläfli saw in 1852 was that geometry does not end at the third dimension. Dimensions $4, 5, 6, \dots$ have their own geometric inhabitants, and the systematic study of those inhabitants — eventually called higher-dimensional geometry — has become one of the major themes of 20th-century mathematics. The six regular 4-polytopes are the gateway into that subject. They are the rare exceptional objects whose existence in dimension four foreshadows the entire richness of higher-dimensional geometric thinking — and the 24-cell is the most rare and exceptional of them, an object whose unexpected presence is one of the small wonders of dimensional analysis.

The Greeks counted five regular polyhedra and stopped. Schläfli, working in nineteenth-century Bern with no apparent audience, counted regular polytopes in every dimension and found the pattern. The list of six in dimension four is, in a precise sense, the most beautiful single result of that count.

Frequently asked

Who discovered the four-dimensional regular polytopes?

Ludwig Schläfli, a Swiss mathematician working in nearly complete obscurity at the University of Bern, classified all six regular four-dimensional polytopes between 1850 and 1852. His manuscript Theorie der vielfachen Kontinuität (Theory of Multiple Continuity) anticipated by half a century the systematic study of higher-dimensional geometry. Most of it was published only posthumously in 1901. Independently and slightly later, Stringham (1880), Forchhammer (1879), and a handful of others rediscovered some of the polytopes. Schläfli's priority is now universally acknowledged.

What is a Schläfli symbol?

A compact notation for regular polytopes, introduced by Schläfli. A polygon with n sides is {n}; a regular polyhedron with faces {n} meeting q at each vertex is {n, q}; a regular four-polytope with cells {n, q} meeting r at each edge is {n, q, r}. So the cube is {4, 3} (square faces, three meeting at each vertex), the tesseract is {4, 3, 3} (cube cells, three meeting at each edge), and the 120-cell is {5, 3, 3} (dodecahedron cells). The symbol encodes the complete combinatorial type of the polytope in a few digits.

Why is the 24-cell so special?

Because it has no analogue in any other dimension. The 24-cell is self-dual (its dual is itself), has 24 vertices, 96 edges, 96 faces, and 24 octahedral cells. It is the only regular polytope in any dimension that does not fall into one of the three families of regular polytopes that exist in every dimension (the simplex, the hypercube, and the cross-polytope/orthoplex). Its appearance in dimension four is a genuine 'sporadic' phenomenon, similar in spirit to the exceptional Lie groups E6, E7, E8 in algebra. Mathematicians have spent considerable effort understanding why it exists at all.

How many regular polytopes are there in higher dimensions?

Exactly three, for every dimension n ≥ 5: the simplex (the n-dimensional analogue of the tetrahedron), the hypercube (the analogue of the cube), and the cross-polytope or orthoplex (the analogue of the octahedron). These three exist in every dimension above 2. Dimensions 3 and 4 are exceptional, with 5 and 6 regular polytopes respectively. The drop from 6 to 3 between dimensions 4 and 5 is one of the most striking patterns in geometry — and the 24-cell is the singular non-classical polytope that vanishes in the transition.

Continue reading

Blog

Knot Polynomials: How to Tell Two Knots Apart with Algebra

The same knot can look completely different from one drawing to another. Knot polynomials assign an algebraic fingerprint to each knot, usually distinguishing two knots in one calculation.

Read →Blog

The Gauss–Bonnet Theorem: When Curvature Becomes a Topological Constant

Bend a surface, stretch it, dent it, smooth it. The local curvature changes everywhere — but its total integral does not. It equals 2π times the Euler characteristic, forever.

Read →Blog

Penrose Tilings: Patterns That Cover the Plane but Never Repeat

The crystallographic restriction theorem forbids fivefold symmetry in any repeating wallpaper. Penrose tilings have fivefold symmetry anyway — by refusing to repeat at all.

Read →