1
$\begingroup$

By means of the Hopf fibration $\pi: S^3 \rightarrow S^2$, U. Pinkall showed that every compact Riemannian surface of genus one can be conformally embedded in $S^3$.

More precisely: Let $p$ be a closed curve on $S^2$ of length $L$. Lifting it to $S^3$ yields a torus isometric to $R^2 / \Gamma$, with $\Gamma$ generated by $(2\pi, 0)$ and $(A/2, L/2)$, where $A$ is the area enclosed by $p$. (This is Proposition 1 in Pinkall's paper.)

Now it's claimed that if you lift a great-circle you should get the Clifford-torus. The above proposition then yields that the Clifford-torus is isometric to $R^2 / \Gamma _1$, with $\Gamma _1$ generated by $(2\pi, 0)$ and $(\pi, \pi)$. The usual definition of the Clifford torus is $R^2 / \Gamma _c$ with $\Gamma _c$ generated by $(2\pi, 0 )$ and $(0, 2 \pi)$.

Who know's how this fits together?

$\endgroup$
1
  • $\begingroup$ I see you used `\\` to try to get a new line. This does not work here, and you should not do that when writing a (La)TeX either. The proper way to start a new paragraph is to leave an empty line in your source file. If what you are trying to achieve is to avoid the indentation, there is a correct way to do that. $\endgroup$ Jan 16, 2012 at 19:18

2 Answers 2

5
$\begingroup$

Your definition of the Clifford torus is off. The usual definition of the Clifford torus is the set $(z_1,z_2)\in\mathbb C^2$ in the unit sphere $|z_1|^2+|z_2|^2=1$ with $|z_1|^2=|z_2|^2=\frac 1 2$. This is a square torus isometric to $\mathbb R^2/\Gamma_c$ with $\Gamma_c$ generated by $(2\pi/\sqrt 2, 0), (0, 2\pi/\sqrt 2)$ (not $(2\pi, 0), (0, 2\pi)$) which is isometric to $\mathbb R^2/\Gamma_1$ by a $\pi/4$ rotation.

$\endgroup$
2
  • $\begingroup$ Thanks for your answer. You are right about the factor $1/\sqrt{2}$, of course. I forgot about it since I'm actually just interested in the conformal type. Still, I don't see how a rotation by $\pi/4$ yields a isometry between $\R^2/ \Gamma _c$ and $\R^2 / \Gamma _1$, since $\Gamma _c$ is a square and $\Gamma _1$ is a parallelogram, hence the corresponding tori can't have the same conformal type. Where am I wrong? $\endgroup$
    – user18589
    Jan 16, 2012 at 15:30
  • 3
    $\begingroup$ no, it's a square, you just have to look at it the right way. let's skip the $2\pi$ factors. look at the integral lattice $\mathbb Z^2$ in $\mathbb R^2$. Now look at the sublattice generated by $a=(1,1)$ and $b=(-1,1)$. It produces the square torus with side $\sqrt 2$. But it also has a different basis given by $a=(1,1)$ and $a+b=(0,2)$. That's the basis we are looking at here. $\endgroup$ Jan 16, 2012 at 16:02
0
$\begingroup$

If $T_1,T_2$ are any 1-parameter subgroups of $S^3=SU(2)$ viewed as the unit sphere of quaternions, Clifford tori are realized as $T_1.q.T_2$, $q\in S^3$.

But this is isometric to the quotient of $T_1\times T_2$ by $(-1,-1)$, hence the $(\pi,\pi)$ in the lattice $\Gamma_1$, since $T_i\simeq\mathbb{R}/2\pi\mathbb{Z}$.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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