A note on the converse Bricard property of

projective planes
arXiv:2311.17209v1 [math.AG] 28 Nov 2023

Zoltán Szilasi

Abstract
We show that the converse Bricard property does not hold in every
Moufang plane.

1 Introduction
An incidence geometry is a projective plane if
(P1) for every pair of distinct points A and B there is a unique line incident
←−→
with A and B (we denote this line by AB);
(P2) for every pair of distinct lines m and n there is a unique point incident
with m and m (we denote this point by m ∩ n);

(P3) there are four points no three of which are collinear.

In a projective plane an ordered triple of noncollinear points is a triangle.
Then the points are called the vertices, and the lines joining the three possible
distinct pairs of vertices are called sides.

We say that two triangles ABC and A′ B ′ C ′ are centrally perspective from
←−→ ←−→ ←−→
a point O if the lines AA′ , BB ′ and CC ′ are incident with O. The triangles are
←→ ←
− −−→ ←→ ← −−→
called axially perspective from a line l if the points AB ∩ A′ B ′ , AC ∩ A′ C ′ and
←→ ←
− −−
→
BC ∩ B ′ C ′ are incident with l.

We consider the following incidence properties of projective planes.

Mathematics Subject Classification: 51A20; 51A30; 51E15; 51A05

Key words and phrases: projective planes; Moufang planes; Bricard property; alternative
division rings; octonions.
Zoltán Szilasi, Institute of Mathematics, University of Debrecen
H-4010, Debrecen, Hungary
E-mail: szilasi.zoltan@science.unideb.hu

1
(D11) If two triangles are perspective from a point, then they are perspective
from a line.
(D10) If the triangles A1 B1 C1 and A2 B2 C2 are perspective from a point O, and
←−−→ ←−−→ ←−−→ ←−−→
O is incident to the line of A1 B1 ∩ A2 B2 and A1 C1 ∩ A2 C2 , then they are
perspective from a line.
(D9) If the triangles A1 B1 C1 and A2 B2 C2 are perspective from a point O,
and the triplets (A1 , B2 , C1 ) and (A2 , B1 , C2 ) are collinear, then the two
triangles are perspective from a line.

(D11) is called the Desargues property and (D10) is called the little
Desargues property. A projective plane is called Desarguesian, if (D11) holds;
and a Moufang plane, if (D10) holds. It is easy to see that (D11) is a stronger
property, than (D10); and (D10) is stronger, than (D9). It can be shown (see
[2]) that if the Fano property holds, i.e., no complete quadrangle has collinear
diagonal points, then (D10) follows from (D9).

In [1] the following properties are investigated:

• The Bricard property: Let ABC and A′ B ′ C ′ be two triangles, and let
← −→ ← −−→ ←→ ← −−→ ←→ ←
− −−→ ←−→ ←−→
P := BC ∩ B ′ C ′ , Q := AC ∩ A′ C ′ and R := AB ∩ A′ B ′ . If A′ P , B ′ Q
←−→ ←−
→ ←−
→ ←→ ←−→
and C ′ R are concurrent, then D := BC ∩ AA′ , E := AC ∩ BB ′ and
←→ ←−→
−
F := AB ∩ CC ′ are collinear.
• The converse Bricard property: Let ABC and A′ B ′ C ′ be two triangles,
←→ ←
− −−
→ ←→ ←
−−→ ←→ ←
− −−→
and let P := BC ∩ B ′ C ′ , Q := AC ∩ A′ C ′ and R := AB ∩ A′ B ′ . If
←−
→ ←−
→ ←→ ←−→ ←
−→ ←−→
D := BC ∩ AA′ , E := AC ∩ BB ′ and F := AB ∩ CC ′ are collinear, then
←−→ ←−→ ←−→
A′ P , B ′ Q and C ′ R are concurrent.

It is shown in [1] that in every Desarguesian projective plane both of these

properties are true. It is natural to ask, if (D11) is necessary for the Bricard
property, or does it follow from one of the weaker conditions. The author of [1]
conjectures that the Bricard property and its converse are equivalent to (D9).
In this paper we show that it is not true: we prove that it does not even
follow from (D10), as we provide a counterexample for the converse Bricard
property in a Moufang plane.
In [1] many other interesting open problems are mentioned about the Bricard
propety. One of these questions is to determine dual of the Bricard property. In
this paper we show that the dual of the Bricard property is its converse.

2 Preliminaries
Let R be a set and +, · be binary operations on R such that
• (R, +) is a commutative group with zero element 0;

2
 Figure 1: The Bricard property

• a · 0 = 0 · a = 0 for all a ∈ R;

• (R\ {0} , ·) is a loop;

• a · (b + c) = a · b + a · c (a, b, c ∈ R);
• (a + b) · c = a · c + b · c (a, b, c ∈ R);
• a · (a · b) = (a · a) · b (a, b ∈ R);

• a · (b · b) = (a · b) · b (a, b ∈ R).
Then (R, +, ·) is called an alternative division ring. In the following we will
write simply ab instead of a · b. We denote the unit of (R\ {0} , ·) by 1. In
every alternative division ring for all a ∈ R\ {0} there are a′ , a′′ ∈ R such that
aa′ = 1, a′′ a = 1, and a′ = a′′ . This element is called the inverse of a and is
denoted by a−1 .
By a difficult theorem of Bruck-Kleinfield and Skornyakov, an alternative
division ring either is associative or is a Cayley-Dickson algebra over some field.
From this it follows that in every alternative division ring we have the inverse
property

a(a−1 b) = (ba−1 )a = b for all a ∈ R\ {0} , b ∈ R,

3
since this holds in every Cayley-Dickson algebra.

The incidence structure (P, L, I), where

• P := {[x, y, 1], [1, x, 0], [0, 1, 0] | x, y ∈ R};

• L := {⟨a, 1, b⟩ , ⟨1, 0, a⟩ , ⟨0, 0, 1⟩ | a, b ∈ R};

• ([x, y, z], ⟨a, b, c⟩) ∈ I if and only if xa + yb + zc = 0
is a projective plane called the projective plane over the alternative division
ring R.

A projective plane is a Moufang plane if and only if it can be coordinatized

by an alternative division ring, i.e., it is isomorphic to a projective plane over
an alternative division ring. (For a proof see [3] or [4].)

A Moufang plane is Desarguesian if and only if the coordinatizing alternative

division ring R is associative, i.e., a(bc) = (ab)c for all a, b, c ∈ R, and hence R
is a skewfield.
The most simple example of an alternative division ring that is not a skew-
field is the alternative division ring of octonions. They can be constructed by the
Cayley-Dickson procedure from the quaternions. An octonion can be written in
form
x = x0 + x1 i + x2 j + x3 k + x4 l + x5 I + x6 J + x7 K,
where xi (i ∈ {0, 1, 2, 3, 4, 5, 6, 7}) are real numbers, and the rule of multiplica-
tion is given by the the following table:

i j k l I J K
i -1 l K -j J -I -k
j -l -1 I i -k K -J
k -K -I -1 J j -l i
l j -i -J -1 K k -I
I -J k -j -K -1 i l
J I -K l -k -i -1 j
K k J -i I -l -j -1

The projective plane over the octonions is called the octonion plane.

3 The dual of the Bricard property

Theorem 3.1 The dual of the Bricard property is its converse.

Proof. The dual of the Bricard property can be formulated as follows:

Let a, b, c and a′ , b′ , c′ be the sides of two triangles. Let p be the line
through b ∩ c and b′ ∩ c′ , q be the line through a ∩ c and a′ ∩ c′ , r be the line

4
 Figure 2: The dual of the Bricard property

through a ∩ b and a′ ∩ b′ . If a′ ∩ p, b′ ∩ q and c′ ∩ r are collinear; then the line

d through b ∩ c and a ∩ a′ , the line e through a ∩ c and b ∩ b′ , and the line f
through a ∩ b and c ∩ c′ are concurrent.

Use the following notations:

A := b′ ∩ c′ , B := a′ ∩ c′ , C := a′ ∩ b′ , A′ := b ∩ c, B ′ := a ∩ c, C ′ := a ∩ b.
Then the converse Bricard property applied to the triangles ABC and
A′ B ′ C ′ is the same as the dual of the Bricard property.

4 A counterexample for the converse Bricard

property in the octonion plane
Theorem 4.1 The converse Bricard property does not hold in every Moufang
plane.

Proof. Consider the following ABC and A′ B ′ C ′ triangles in the octonion

plane:
A[1, 0, 0], B[0, 1, 0], C[0, 0, 1]; A′ [i, −1, 1], B ′ [−1, j, 1], C ′ [k, −k, 1].
←→ ←−→
− ←→ ←−→
First, we show that the points D := BC ∩ AA′ , E := AC ∩ BB ′ and

5
 ←→ ←−→
−
F := AB ∩ CC ′ are collinear.
← −
→ ←−→
The equation of BC is x1 = 0 and the equation of AA′ is x2 + x3 = 0.
Therefore D = [0, −1, 1].
←→ ←−→
The equation of AC is x2 = 0 and the equation of BB ′ is x1 + x3 = 0.
Therefore E = [−1, 0, 1].
←−→ ←−→
The equation of AB is x3 = 0 and the equation of CC ′ is x1 + x2 = 0.
Therefore F = [1, −1, 0].
Here, D, E and F are incident to the line x1 + x2 + x3 = 0, therefore they
are collinear.
←−→ ←−→ ←−→
We are going to show that A′ P , B ′ Q and C ′ R are not concurrent.
←−→ ← −−
→
First, we calculate the coordinates of P := BC ∩ B ′ C ′ .
←−−→
The coordinates of B ′ C ′ are ⟨e1 , 1, e2 ⟩ such that
)
−e1 + j + e2 = 0
.
ke1 − k + e2 = 0

From these equations

−e1 + j = k(e1 − 1),
therefore
 
−1 1 1 1 1 1 1
e1 = (k + 1) (k + j) = − k (k + j) = + j + k + I,
2 2 2 2 2 2
1 1 1 1
e2 = e1 − j =
− j + k + I.
2 2 2 2
P is the intersection of this line and x1 = 0, so the coordinates are
 
1 1 1 1
P 0, − + j − k − I, 1 .
2 2 2 2
←→ ← −−
→
Next, we calculate the coordinates of Q := AC ∩ A′ C ′ .
←−′−
→′
The coordinates of A C are ⟨e1 , 1, e2 ⟩ such that
)
ie1 − 1 + e2 = 0
.
ke1 − k + e2 = 0

From these equations

ie1 − 1 = k(e1 − 1),
therefore
 
−1 1 1 1 1 1 1
e1 = (i − k) (1 − k) = − i − k (1 − k) = − i − k + K − ,
2 2 2 2 2 2

6
 1 1 1 1
e2 = 1 − ie1 =
+ i + k + K.
2 2 2 2
Q is the intersection of this line and x2 = 0. For Q[x1 , 0, 1],
 
1 1 1 1 1 1 1 1
x1 − i − k + K − + + i + k + K = 0,
2 2 2 2 2 2 2 2
so   −1
1 1 1 1 1 1 1 1
x1 = − + i+ k+ K − i− k+ K − =
2 2 2 2 2 2 2 2
  
1 1 1 1 1 1 1 1 1 1 1 1
=− + i+ k+ K − + i + k − K = + i − k + K.
2 2 2 2 2 2 2 2 2 2 2 2
Therefore the coordinates are
 
1 1 1 1
Q + i − k + K, 0, 1 .
2 2 2 2
←→ ←
− −−
→
Now we calculate the
← coordinates of R := AB ∩ A′ B ′ .
−′−
→
The coordinates of A B ′ are ⟨e1 , 1, e2 ⟩ such that
)
ie1 − 1 + e2 = 0
.
−e1 + j + e2 = 0

From these equations

ie1 − 1 = −e1 + j,
therefore
 
−1 1 1 1 1 1 1
e1 = (i + 1) (1 + j) = − i (1 + j) = − i + j − l,
2 2 2 2 2 2
1 1 1 1
e2 = e1 − j =
− i − j − l.
2 2 2 2
R is the intersection of this line and x3 = 0, so the coordinates are
 
1 1 1 1
R 1, − + i − j − l, 0 .
2 2 2 2
←−→
Following similar method we calculate the coordinates of the line A′ P
through A′ [i, −1, 1] and P 0, − 12 + 12 j − 12 k − 12 I, 1 , and we obtain
 

←−→ 1
 
1 1 1 1 1 1 1
A′ P i + l − J − K, 1, − j + k + I .
2 2 2 2 2 2 2 2
←−→
The
 1 coordinates of  the line B′Q through B ′ [−1, j, 1] and
Q 2 + 12 i − 12 k + 12 K, 0, 1 are
←− → 1
 
′ 1 1 1 1 1 1 1
BQ j − l − I − J, 1, − j − l+ I− J ;
2 6 6 6 2 6 6 6

7
 ←−→
and the the line C ′ R through C ′ [k, −k, 1] and R 1, − 21 + 12 i − 12 j − 12 l, 0 is
 

←−→ 1 1
 
′ 1 1 1 1 1 1
CR − i + j + l, 1, k + I − J − K .
2 2 2 2 2 2 2 2
←−→ ←−→ ←−→
We have to check if A′ P , B ′ Q and C ′ R are concurrent. We calculate the
←− → ←−→
intersection of A′ P and C ′ R. This point has coordinates [x1 , x2 , 1] such that
 
1 1 1 1 1 1 1 1

x1 i + l − J − K + x2 + − j + k + I = 0 
2 2 2 2 2 2 2 2 
  .
1 1 1 1 1 1 1 1 
x1 − i + j + l + x2 + k + I − J − K = 0 
2 2 2 2 2 2 2 2
From these equations
 
1 1 1 1 1 1 1 1
x1 i + j + J + K − + j − J − K = 0,
2 2 2 2 2 2 2 2
  −1
1 1 1 1 1 1 1 1
x1 = − j+ J+ K i+ j+ J + K =
2 2 2 2 2 2 2 2
  
1 1 1 1 1 1 1 1 1
= − j+ J+ K + i− j− J − K =
2 2 2 2 4 2 4 4 4
1 1 1 1 1 1 1
= + i − j + k + l + I + K;
4 4 2 4 4 4 4 
1 1 1 1 1 1 1 1
x2 = − + j − k − I − x1 i+ l− J − K =
2 2 2 2 2 2 2 2
  
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
= − + j− k− I− + i− j+ k+ l+ I + K i+ l− J − K =
2 2 2 2 4 4 2 4 4 4 4 2 2 2 2
5 1 1 5 1 3 1 1
= − + i + j − k − l − I + J + K.
8 4 2 8 2 4 2 8
So
←−→ ←− →
 
′ ′ 1 1 1 1 1 1 1 5 1 1 5 1 3 1 1
A P ∩C R = + i − j + k + l + I + K, − + i + j − k − l − I + J + K, 1 .
4 4 2 4 4 4 4 8 4 2 8 2 4 2 8
←−→ ← −→ ←−
→
Therefore A′ P , B ′ Q and C ′ R are concurrent if and only if this point lies on
←−→
B ′ Q, i.e.,
  
1 1 1 1 1 1 1 1 1 1 1 5
+ i− j+ k+ l+ I + K j− l− I− J − +
4 4 2 4 4 4 4 2 6 6 6 8
1 1 5 1 3 1 1 1 1 1 1
+ i + j − k − l − I + J + K − j − l + I − J = 0.
4 2 8 2 4 2 8 2 6 6 6
It is enough to calculate the real part of the left side. It is
        
1 1 1 1 1 1 5 1 1 1 5
− j j − l l − I I− = + + − ̸= 0.
2 2 4 6 4 6 8 4 24 24 8
Therefore the right side cannot be zero, so the three lines are not concurrent.

8
References
[1] Ann Linehan: Relationships between geometric propositions which charac-
terise projective planes, Thesis, University of Western Australia, 2021.
[2] A. A. Heyting: Axiomatic projective geometry, Bibliotheca mathematica,
1980.
[3] D. R. Hughes, F. C. Piper: Projective Planes, Springer, New York, 1973.

[4] F. W. Stevenson: Projective Planes, W. H. Freeman and Company, 1972.