2
$\begingroup$

Let $X=[0,1]\times [0,1]\subset \mathbb{R}$ with the subspace topology. Consider the equivalence relation $\sim$ on $X$ whose non trivial relations are $(0,y)\sim (1,1-y)$ for any $y\in [0,1]$. Let $p:X\rightarrow X^*$ be the quotient map.

I am trying to show that if If $E\subset X$, is defined as $E=[0,1]\times \{ 0,1 \}$, then $P(E)$ is homeomorphic to $S^1$. I am hoping my attempt below makes some sense and is fix-able and if not, I can understand where I am getting confused.

My attempt: $p(E)=[0,1]\times \{ 0,1 \}/\sim$. Then take the mapping \begin{align*} \phi:\;& p(E)\rightarrow S^1\\ &(x,y)\mapsto e^{i\pi(x+y)} \end{align*} We verify that the map is constant on the equivalence class, or that $\phi(0,0)=\phi(1,1)$ and $\phi(0,1)=\phi(1,0)$. The latter is obvious by the symmetric nature of the function. For the former: \begin{align*} &\phi(0,0)=1=e^{2\pi i}=\phi(1,1) \end{align*} But I am a bit hazy about this portion, and the quotient topology in general. I think what I am doing here is defining a map from a space $X$ that factor through $X/\sim$ a la algebra first isomorphism theorem.

In this problem, this would mean I am defining my map $\phi$ which actually goes from $[0,1]\times [0,1]$ and insuring it actually is constant on equivalence classes defined by $\sim$. Is that correct?

Finally, how do I show that this map is actually open and continuous? I am particularly concerned about the point $(1,0)$? If I can actually just work in the product topology on $[0,1]\times [0,1]$, this wouldn't be too bad; the pre-image of any neighborhood around $(1,0)$ will include the end points of the intervals (since both x and y will be near zero) which is fine as those are zero in the subspace topology on $I\times I$ as a subspace of $\mathbb{R}^2$. Is that right?

  • 0
    If you have Munkres, look at Theorem 18.3.2017-01-24
  • 0
    @qbert: Why did you change the title of your question? It was correct before - now it's incorrect (a Mobius strip is **not** homeomorphic to a circle).2017-01-24
  • 0
    @ZevChonoles I'm not sure, I think I got confused, I'll change it back2017-01-25
  • 0
    @N.Owad are you referring to the pasting lemma?2017-01-25
  • 0
    I am. If you look at each side of the interval $[0,1]$ as the two pieces you are mapping into $S^1$, it might be easier to show they are continuous, then the pasting lemma can give you that the whole boundary is continuous. You can probably go in the other direction just as easily, for $f^{-1}$.2017-01-25

0 Answers 0