3
$\begingroup$

The following statement seems true, but I don't know a proof or a reference for it (and I would like one).

Let $\Gamma< \operatorname{PSL}(2,\mathbb R)$ be a nonuniform lattice with one cusp. We may conjugate $\Gamma$ so that an element $ \begin{pmatrix} 1 & s\\ 0 & 1 \end{pmatrix} $ generates the cusp of $\Gamma$. Let $\Gamma'$ be such that $\Gamma<\Gamma'$ be an index $2$ subgroup. Then, the element $ \begin{pmatrix} 1 & s/2\\ 0 & 1 \end{pmatrix} $ generates the cusp of $\Gamma'$.

I imagine a similar statement would be true if both $\Gamma$ and $\Gamma'$ had two cusps.

$\endgroup$
2
  • $\begingroup$ I think you should have -1 on the diagonal. $\endgroup$ – Moishe Kohan Sep 9 '20 at 22:18
  • $\begingroup$ Sorry I meant that $\Gamma< \operatorname{PSL}(2,\mathbb{R})$, and I edited the question to reflect this. So I don't think such things matter. $\endgroup$ – Chris Z Sep 11 '20 at 15:04
4
$\begingroup$

Assume just that $\Gamma$ has index $k$ in $\Gamma'$. Let $C \subset \mathbb R \cup \{\infty\}$ be the set of parabolic points for the action of $\Gamma$. Then $C$ is also the set of parabolic points for the action of $\Gamma'$, because if $\gamma \in \Gamma'$ is parabolic with fixed point $x$ then for some integer $i \ge 1$, $\gamma^i \in \Gamma$ is parabolic with the same fixed point $x$.

In general every $\Gamma'$ orbit of $C$ decomposes as a union of $\Gamma$ orbits, so $$\#\text{cusps}(\Gamma') \le \#\text{cusps}(\Gamma) $$ with equality if and only if the every $\Gamma'$ orbit of $C$ is just a single $\Gamma$-orbit. So assuming that $$\#\text{cusps}(\Gamma') = \#\text{cusps}(\Gamma) $$ it follows for each $x \in C$ the group $\text{Stab}(x;\Gamma)$ has index $k$ in $\text{Stab}(x;\Gamma')$.

For the case $x=\infty \in C$ it follows that if $\begin{pmatrix} 1 & s\\ 0 & 1 \end{pmatrix}$ generates $\text{Stab}(\infty;\Gamma)$ then $\begin{pmatrix} 1 & s/k\\ 0 & 1 \end{pmatrix}$ generates $\text{Stab}(\infty;\Gamma')$.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.