3
$\begingroup$

In Hatcher on page 84 there is the following proposition: For a connected graph $X$ with maximal tree $T$, $\pi_1 (X)$ is a free group with basis the classes $[f_\alpha]$ corresponding to the edges $e_\alpha$ of $X - T$.

I tried to apply this to the torus $T^2$ with the two edges $e_a, e_b$ and the maximal tree $T = \{ x_0\}$ where $x_0$ is the point connecting the two edges.

The problem is that then I get the free group $F(a,b)$ instead of $\mathbb{Z} \oplus \mathbb{Z}$.

Where am I making the mistake? Many thanks for your help!

  • 3
    Hmm. How is the torus a *graph* (= 1-dim CW-complex)? In fact, you're not that far off, the relation $[a,b] = aba^{-1}b^{-1}$ from the two-cell gives you the desired $\mathbb{Z}^2$.2011-07-07
  • 0
    well, I thought I could just use the $1$-skeleton to get the fundamental group of the whole complex. Need to work out how to put things together correctly I think...2011-07-07

1 Answers 1

3

In this proposition a graph is a regular, 1-dimensional CW-complex.

In fact you computed the fundamental group of the 1-Skeleton of the Torus, which is a bouquet of $2$ $S^1$ and has therefore $\mathbb Z * \mathbb Z$ as fundamental group.