JOVANA
Explore Library Glossary Getting Started Three Levels Fields How it works Mission
Join the mission
All guides

The Fourteen Bravais Lattices

Seven crystal systems times four centerings does not make twenty-eight — it makes exactly fourteen. Here is why the arrangements of lattice points close into one short, complete, proven list, and how the Wigner-Seitz cell shows each point its own patch of space.

Where the rung has brought us

We are at the top of the rung, and every piece is now in hand. Guide 1 split a crystal into a lattice of identical points and a motif hung on each point. Guide 2 packed one repeat into a unit cell and measured it with six numbers, the lattice parameters a, b, c and the angles between them. Guide 3 let a cell carry extra points at its body centre or on its faces, so a centered cell trades a bigger box for a shape that shows off the true symmetry. Guide 4 sorted all lattices into seven crystal systems by the relations among those six numbers. One question is left, and it is the crown of the whole rung: exactly how many genuinely different lattices are there?

The naive guess multiplies. Seven systems, and in each you might place a point at the centre (P for primitive, I for body-centered, F for all faces, C for one pair of faces): seven times four is twenty-eight. The real answer is fourteen, the Bravais lattices, and the gap between 28 and 14 is not sloppiness or a shortcut. It is a hard geometric fact: most of those 28 combinations are either impossible in that system or turn out to be a lattice we have already counted, just drawn in a clumsier box. Understanding why is what makes the number stick.

Why twenty-eight collapses to fourteen

Two scissors trim the list. The first cuts away centerings that would destroy the very symmetry that defines a system. Take a cube and centre just one pair of opposite faces (base-centered cubic). That single choice singles out one axis over the other two, so the three cube axes are no longer equivalent — and equal, interchangeable axes are exactly what makes a lattice cubic. The would-be base-centered cubic is really just a tetragonal lattice wearing a cubic costume, so it is not allowed to join the cubic family. This is the deeper reason a cube gets only P, I, and F: base-centering is forbidden because it breaks cubic symmetry.

The second scissors cuts away centerings that are real but redundant — they describe a lattice we can draw with a smaller, already-listed cell. Consider a base-centered tetragonal cell. Rotate your view 45 degrees about the tall axis and a smaller simple (P) tetragonal cell falls out, with half the base area and the very same set of points. Since the two boxes hold the identical lattice, we keep only the simpler one and drop C-tetragonal as a duplicate. The same trick shows face-centered tetragonal is just body-centered tetragonal in a bigger box. Run both scissors over all 28 candidates and exactly 14 survivors remain — no more, no fewer.

The fourteen, system by system

Here is the full census. Read across: each system keeps only the centerings that survive both scissors, and the running total climbs to fourteen. The last column reminds you how many lattice points each cell actually owns, counted the way guide 3 taught — a corner point is shared among eight cells so it contributes 1/8, a face point is split between two cells so it gives 1/2, and a point sitting fully inside counts as a whole one.

CRYSTAL SYSTEM   ALLOWED LATTICES     # NEW   RUNNING TOTAL
--------------   -----------------    -----   -------------
triclinic        P                      1           1
monoclinic       P  C                    2           3
orthorhombic     P  C  I  F              4           7
tetragonal       P  I                    2           9
hexagonal        P                       1          10
rhombohedral     R                       1          11
cubic            P  I  F                  3          14
--------------   -----------------    -----   -------------
                                   TOTAL  ->        14

  points per conventional cell:
    P  primitive     = 1     (8 corners x 1/8)
    C  base-centered = 2     (corners + 2 faces x 1/2)
    I  body-centered = 2     (corners + 1 inside)
    F  face-centered = 4     (corners + 6 faces x 1/2)
The fourteen Bravais lattices. Only cubic and orthorhombic are rich enough to carry the full P/I/F set; triclinic, hexagonal, and rhombohedral get one lattice each.

Look at the cubic row, because you already know its residents from earlier rungs by another name. Primitive cubic (P) is the bare simple-cubic lattice; body-centered cubic (I, two points per cell) is the skeleton of iron, tungsten, and chromium; face-centered cubic (F, four points per cell) underlies copper, aluminium, gold, and nickel. Here lattice points per cell pays off directly: an FCC conventional cell genuinely contains four lattice points, which is why its cube edge is chosen larger than the shortest point-to-point distance — a bigger box bought in exchange for showing the four-fold cubic symmetry plainly, exactly the trade guide 3 described.

The Wigner-Seitz cell: each point's own territory

A conventional cell is a convenient box, but it draws the boundary in an arbitrary place and often centres it on empty space rather than on a lattice point. There is a more natural way to hand out space, and it gives us the last idea of this rung. The Wigner-Seitz cell of a lattice point is simply all the space that is closer to that point than to any other. Think of one fire station per lattice point and colour every spot of the city by which station is nearest: each station's coloured patch is its Wigner-Seitz cell, the region it alone serves.

  1. Pick one lattice point and draw a straight line from it to each of its near neighbours.
  2. At the midpoint of every such line, slice space with a plane perpendicular to the line (the perpendicular bisector).
  3. The smallest volume around your chosen point that all these planes enclose is the Wigner-Seitz cell.

This cell has three qualities that the conventional box lacks. It is primitive — it wraps exactly one lattice point, so it holds one point's worth of volume, no more. It is unique — there is only one such region, no arbitrary choice of axes. And it wears the full point symmetry of the lattice, so its very shape advertises the lattice type: the Wigner-Seitz cell of a body-centered cubic lattice is a truncated octahedron, and that of face-centered cubic is a rhombic dodecahedron. The idea returns with force in the next rung: build the same closest-point cell in reciprocal space and you get the Brillouin zone, the arena in which a solid's electrons and waves are described.

What the fourteen do — and do not — tell you

Guard the beginner's trap one last time, because it is easiest to slip on right here. The fourteen Bravais lattices classify the lattice — the bare array of points — and nothing else. They say nothing about the motif. This is why 'FCC' can be dangerously ambiguous: copper is an FCC lattice with a single atom on each point, but rock salt is also an FCC lattice, carrying a two-atom motif (a sodium and a chlorine), and diamond is an FCC lattice with a two-carbon motif. Same lattice, wildly different crystals. Remember the master equation, crystal = lattice + motif: the Bravais lattice fixes only the first term.

To capture a real crystal's full symmetry you need the motif's own symmetry too, and that larger accounting closes into its own exact lists further up the ladder: 32 point groups for the symmetry about a point, and 230 space groups for the symmetry of the whole periodic pattern. Fourteen, thirty-two, two-hundred-and-thirty — every one is a complete, proven enumeration, never a 'so far'. They are the reason crystallography can be a finite, closed subject rather than an endless catalogue.