Note: for this question a topological manifold is defined to be a locally Euclidean, second countable, Hausdorff space. Also, I am using the subscript $x$ and $y$ just to keep track of which copy of $\mathbb{R}$ I am calling the $x$ and $y$ axis, respectively.

Intuitively I can see why the wedge sum of the $x$ and $y$ axes are not a manifold, since they do not have a well defined dimension. I am having trouble making this idea more rigorous.

Let $\mathbb{R}_x$ denote the $x$-axis and $\mathbb{R}_y$ denote the $y$-axis. Also, let $0_x$ and $0_y$ be the zero on the $x$ and $y$ axis, respectively. The wedge sum of the $x$ and $y$ axes is defined as

$$\mathbb{R}_x\vee\mathbb{R}_y=\mathbb{R}_x\amalg\mathbb{R}_y/(0_x \sim 0_y).$$

Any open neighborhood of $0$ in $\mathbb{R}_x\vee\mathbb{R}_y$ would have to be open when intersected with both $\mathbb{R}_x$ and $\mathbb{R}_y$. This leads me to believe that the open neighborhoods of $0$ in $\mathbb{R}_x\vee\mathbb{R}_y$ are wedge sums of intervals containing $0_x$ and $0_y$ in the $x$ and $y$ axes. Then is $\mathbb{R}_x\vee\mathbb{R}_y$ not a topological manifold because it is not locally Euclidean around $0$?


Yes, that's correct. To finish the proof, you need to show that a wedge sum of two open intervals is not homeomorphic to an open disc in $\mathbb{R}^n$ for any $n$. This can be done, for example, by considering what happens if you delete the origin from the wedge sum: the resulting space will have four connected components. Deleting a point from an open disc in $\mathbb{R}^n$ will leave nothing (if $n=0$), two components (if $n=1$), or one component (if $n\ge 2$), but never four components.


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.