5
$\begingroup$

I am struggling with following problem. Consider $2A$ class in $Co_1$ having $819*759*75$ elements. Each element $a$ from $2A$ have two representatives in $Co_0$. Element $a$ corresponds to $E_8$ sublattice in Leech lattice defined as $\{v: av=-v\}$ where I call by $a$ also proper preimage in $Co_0$. Now the opposite: having $E_8$ sublattice $L$ in Leech lattice I can find element $a_L$ in $2A$ class.

My goal is to find relation between $Order(ab)$ for $a,b$ in $2A$ and corresponding geometry of two $E_8$ sublattices. The $Order(ab)$ can be $2,3,4,5,6$. (BTW in 2017 I have found this relation already).

Take any other sporadic group $g$ and certain conjugacy class $C$ of involutions. The possible values of $Order(ab)$ for $a,b$ in $C$ can be obtained from character table. Can this help to connect some "lattice" to the group $g$ ?

$\endgroup$
2
  • 2
    $\begingroup$ Regarding your last question: McKay observed that the 2A elements in the Monster, Baby Monster, and Fischer24 yield affine $E_8$, $E_7$ and $E_6$ diagrams of conjugacy classes, where the magic numbers on nodes are the orders of the products. There are some partial explanations in the literature (search for Yamauchi, Lam, and collaborators) using Ising vectors in vertex operator algebras. $\endgroup$
    – S. Carnahan
    Commented Feb 9, 2012 at 5:37
  • $\begingroup$ Thank you for this answer ! Can you explain a little what are "magic numbers on nodes" ? $\endgroup$
    – user21230
    Commented Feb 9, 2012 at 8:48

1 Answer 1

4
$\begingroup$

Here is the answer for $Co_1$ which I obtained in GAP.

For given $2A$ involution $a$ in $Co_1$ the number $n$ of such $2A$ involutions $b$ that product $ab$ is in conjugacy class $C$ is presented in format $[C, n]$. I call hook the order of product.

$[ [ "1a", 1 ], [ "2a", 12870 ], [ "2c", 60480 ], [ "3b", 573440 ], [ "3d", 491520 ], [ "4a", 34560 ], [ "4c", 4838400 ], [ "4d", 2419200 ], [ "5b", 12386304 ], [ "6e", 25804800 ] ]$

Next the relation between two $E_8$ sublattices $L_a$ and $L_b$ of Leech lattice is presented in following format. [hook of a and b, number of vectors belonging to intersection of $L_a$ and $L_b$, dimension of vector space generated by both $L_a$ and $L_b$, $[k,l]$ where $k$ is number of vectors in $L_a$ having exactly $l$ perpendicular vectors in $L_b$.

First two rows correspond to $2A$ hook in $Co_1$. These are two cases in $Co_0$. For given $E_8$ sublattice there are $270$ sublattices perpendicular to it (first row) and $12600$ intersecting in $D_4$ sublattice (second row).

$[ 2, 0, 16, [ [ 240, 240 ] ] ]$

$[ 2, 24, 12, [ [ 192, 84 ], [ 24, 126 ], [ 24, 240 ] ] ]$

$[ 2, 4, 14, [ [ 48, 84 ], [ 132, 126 ], [ 60, 240 ] ] ]$

$[ 3, 0, 16, [ [ 240, 126 ] ] ]$

$[ 3, 6, 14, [ [ 162, 84 ], [ 78, 126 ] ] ]$

$[ 4, 0, 16, [ [ 240, 84 ] ] ]$

$[ 4, 0, 16, [ [ 24, 84 ], [ 192, 126 ], [ 24, 240 ] ] ]$

$[ 4, 2, 15, [ [ 120, 84 ], [ 114, 126 ], [ 6, 240 ] ] ]$

$[ 5, 0, 16, [ [ 120, 84 ], [ 120, 126 ] ] ]$

$[ 6, 0, 16, [ [ 66, 84 ], [ 168, 126 ], [ 6, 240 ] ] ]$

The numbers $84$ and $126$ which appear on second $l$ position correspond to $D_7$ and $E_7$ sublattices of $E_8$. The first case happens when projection of the vector on $E_8$ is collinear with "type 4" vector which is sum of two perpendicular vectors from the lattice. The second case happens when projection of the vector on $E_8$ is collinear with lattice vector.

As we can see one of the 4-hook is obtained by intersection in one line of the two 8-planes. This feature can be used to define the Leech lattice from $Co_0$ in 24-dim space. I don't know whether this is useful for something yet.

$\endgroup$
3
  • $\begingroup$ Probably, researchgate.net/publication/… will be of interest to you. Would you mind sharing your GAP code? $\endgroup$ Commented Apr 24, 2019 at 13:16
  • $\begingroup$ @MichielVanCouwenberghe I can share my GAP code. The main problem is that I need to find it and probably test that it is working. It requires time which I don't have. But I will see... $\endgroup$
    – user21230
    Commented Apr 26, 2019 at 11:23
  • $\begingroup$ @MichielVanCouwenberghe how can I contact you via email ? $\endgroup$
    – user21230
    Commented Apr 28, 2019 at 6:08

You must log in to answer this question.