Consider the representation of the Klein bottle $K$ as a square $[0, 1]$ $X$ $[0, 1]$, with the top and bottom edges identified by $(x, 1)$~$(x, 0)$, and the left and right edges identified by $(0, x)$ ~ $(1, 1-x)$. We have $\mathbb{Z}_2$ homology groups given by $\mathbb{Z}_2$, $\mathbb{Z}_2 \bigoplus \mathbb{Z}_2$, and $\mathbb{Z}_2$ in dimensions $0$, $1$ and $2$ respectively.

I want to find the map induced on homology by the map that rotates $K$ $180$ degrees about its midpoint. I calculated the homology by using a CW-complex, and so I am confused as to how to interpret the rotation. I have a hunch that the map on homology will be the identity map but I have little to back it up.

We have a class $\alpha$ that corresponds to a simplex that maps to the top edge of the square. When rotated, it maps to the bottom edge of the square in the opposite direction. This induces a map $1 \rightarrow -1 = 1 \in \mathbb{Z}_2$ in homology, since the composition of $\alpha$ with the reversed chain of $\alpha$ is clearly a boundary. How can we interpret the action of the rotation on the single $2$-cell of $K$? I have no idea how to deal with this.


You can give this a CW-complex structure with one $0$-cell, two $1$-cells and a $2$-cell. Over $\Bbb Z_2$ the boundary maps are all zero, so the homology groups over $\Bbb Z_2$ have dimensions $1$, $2$ and $1$.

You needn't worry about the effect of your map $\rho$ on your two-cell. It's an automorphism, so can only act trivially on $H_2(K,\Bbb Z_2)\cong\Bbb Z_2$.

The only thing you care about is the action on $H_1$. But that is generated by the homology classes of the upper and left edges of your square. But each of these is taken by $\rho$ either to itself or its negative. Over $\Bbb Z_2$ that is the same homology class. So $\rho$ acts trivially on $H_1(K,\Bbb Z_2) \cong\Bbb Z_2^2$.

  • $\begingroup$ Thanks! If the H_2 group had been the integers rather than the integers mod 2, how would one identify where the generator of H_2 is sent? If it had been a reflection rather than rotation would the generator be sent to -1? Is it always possible to classify such maps of CW complexes by examining the degree of the induced map on each n-sphere? (When we quotient out by lower dimensional cells the n-cell becomes the n-sphere. This is the n-sphere I refer to) $\endgroup$ – Daven Jun 12 '19 at 18:20
  • 1
    $\begingroup$ In that case, the map $\rho$ is orientation preserving on the $2$-cell. But again, that doesn't matter, since $H_2(K,\Bbb Z)=0$. @Daven $\endgroup$ – Lord Shark the Unknown Jun 12 '19 at 18:27

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.