Loading [MathJax]/jax/output/CommonHTML/jax.js
\usepackagecolor\usepackage[matrix]xy\usepackageskak Skip to content →

Tag: Sloan

A projective plain (plane) of order ten

A projective plane of order n is a collection of n2+n+1 lines and n2+n+1 points satisfying:

  • every line contains exactly n+1 points
  • every point lies on exactly n+1 lines
  • any two distinct lines meet at exactly one point
  • any two distinct points lie on exactly one line

Clearly, if q=pk is a pure prime power, then the projective plane over Fq, P2(Fq) (that is, all nonzero triples of elements from the finite field Fq up to simultaneous multiplication with a non-zero element from Fq) is a projective plane of order q.

The easiest example being P2(F2) consisting of seven points and lines

But, there are others. A triangle is a projective plane of order 1, which is not of the above form, unless you believe in the field with one element F1

And, apart from P2(F32), there are three other, non-isomorphic, projective planes of order 9.

It is clear then that for all n<10, except perhaps n=6, a projective plane of order n exists.

In 1938, Raj Chandra Bose showed that there is no plane of order 6 as there cannot be 5 mutually orthogonal Latin squares of order 6, when even the problem of two orthogonal squares of order 6 (see Euler’s problem of the 36 officers) is impossible.

Yeah yeah Bob, I know it has a quantum solution.

Anyway by May 1977, when Lenstra’s Festschrift ‘Een pak met een korte broek’ (a suit with shorts) was published, the existence of a projective plane of order 10 was still wide open.

That’s when Andrew Odlyzko (probably known best for his numerical work on the Riemann zeta function) and Neil Sloane (probably best known as the creator of the On-Line Encyclopedia of Integer Sequences) joined forces to publish in Lenstra’s festschrift a note claiming (jokingly) the existence of a projective plane of order ten, as they were able to find a finite field of ten elements.



Here’s a transcript:

A PROJECTIVE PLAIN OF ORDER TEN

A. M. Odlyzko and N.J.A. Sloane

This note settles in the affirmative the notorious question of the existence of a projective plain of order ten.

It is well-known that if a finite field F is given containing n elements, then the projective plain of order n can be immediately constructed (see M. Hall Jr., Combinatorial Theory, Blaisdell, Waltham, Mass. 1967 and D.R. Hughes and F.C. Piper, Projective Planes, Springer-Verlag, N.Y., 1970).

For example, the points of the plane are represented by the nonzero triples (α,β,γ) of elements of F, with the convention that (α,β,γ) and (rα,rβ,rγ) represent the same point, for all nonzero rF.

Furthermore this plain even has the desirable property that Desargues’ theorem holds there.

What makes this note possible is our recent discovery of a field containing exactly ten elements: we call it the digital field.

We first show that this field exists, and then give a childishly simple construction which the reader can easily verify.

The Existence Proof

Since every real number can be written in the decimal system we conclude that

R=GF(10ω)

Now ω=1.ω, so 1 divides ω. Therefore by a standard theorem from field theory (e.g. B. L. van der Waerden, Modern Algebra, Ungar, N.Y., 1953, 2nd edition, Volume 1, p. 117) R contains a subfield GF(10). This completes the proof.

The Construction

The elements of this digital field are shown in Fig. 1.

They are labelled Left1,Left2,,Left5,Right1,,Right5 in the natural ordering (reading from left to right).



Addition is performed by counting, again in the natural way. An example is shown in Fig. 2, and for further details the reader can consult any kindergarten student.

In all digital systems the rules for multiplication can be written down immediately once addition has been defined; for example 2×n=n+n. The reader will easily verify the rest of the details.

Since this field plainly contains ten elements (see Fig. 1) we conclude that there is a projective plain of order ten.

So far, the transcript.

More seriously now, the non-existence of a projective plane of order ten was only established in 1988, heavily depending on computer-calculations. A nice account is given in

C. M. H. Lam, “The Search for a Finite Projective Plane of Order 10”.

Now that recent iPhones nearly have the computing powers of former Cray’s, one might hope for easier proofs.

Fortunately, such a proof now exists, see A SAT-based Resolution of Lam’s Problem by Curtis Bright, Kevin K. H. Cheung, Brett Stevens, Ilias Kotsireas, Vijay Ganesh

David Roberts, aka the HigherGeometer, did a nice post on this
No order-10 projective planes via SAT
.

Leave a Comment

The Leech lattice neighbour

Here’s the upper part of Kneser‘s neighbourhood graph of the Niemeier lattices:



The Leech lattice has a unique neighbour, that is, among the 23 remaining Niemeier lattices there is a unique one, (A241)+, sharing an index two sub-lattice with the Leech.

How would you try to construct (A241)+, an even unimodular lattice having the same roots as A241?

The root lattice A1 is 2Z. It has two roots ±2, determinant 2, its dual lattice is A1=12Z and we have A1/A1C2F2.

Thus, A241=2Z24 has 48 roots, determinant 224, its dual lattice is (A241)=12Z24 and the quotient group (A241)/A241 is C242 isomorphic to the additive subgroup of F242.

A larger lattice A241L of index k gives for the dual lattices an extension L(A241), also of index k. If L were unimodular, then the index has to be 212 because we have the situation
A241L=L(A241)
So, Kneser’s glue vectors form a 12-dimensional subspace C in F242, that is,
L=C×F2(A241)={12v | vZ24, v=v mod 2C}
Because L=L, the linear code C must be self-dual meaning that v.w=0 (in F2) for all v,wC. Further, we want that the roots of A241 and L are the same, so the minimal number of non-zero coordinates in vC must be 8.

That is, C must be a self-dual binary code of length 24 with Hamming distance 8.



Marcel Golay (1902-1989) – Photo Credit

We now know that there is a unique such code, the (extended) binary Golay code, C24, which has

  • one vector of weight 0
  • 759 vectors of weight 8 (called ‘octads’)
  • 2576 vectors of weight 12 (called ‘dodecads’)
  • 759 vectors of weight 16
  • one vector of weight 24

The 759 octads form a Steiner system S(5,8,24) (that is, for any 5-subset S of the 24-coordinates there is a unique octad having its non-zero coordinates containing S).

Witt constructed a Steiner system S(5,8,24) in his 1938 paper “Die 5-fach transitiven Gruppen von Mathieu”, so it is not unthinkable that he checked the subspace of F242 spanned by his 759 octads to be 12-dimensional and self-dual, thereby constructing the Niemeier-lattice (A241)+ on that sunday in 1940.

John Conway classified all nine self-dual codes of length 24 in which the weight
of every codeword is a multiple of 4. Each one of these codes C gives a Niemeier lattice C×F2(A241), all but one of them having more roots than A241.

Vera Pless and Neil Sloan classified all 26 binary self-dual codes of length 24.

Leave a Comment