Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

When we introduce a smooth structure on $m$-manifold we want to have an atlas which covers all the points of the manifold such that both of its charts are smoothly related. Does it mean that nay space $C$ homeomorphic to $D\subseteq \mathbb R^m$, $D$ - an open set, is a smooth $m$-manifold?

My reasoning: we can take an atlas $\mathcal A = \{(C,h)\}$ where $h:C\to D$ is a homeomorphism.

Edited: Since the first version of the question is unclear, I'd like to clarify it. If the space $M = \{(y,x)\mid y=|x|,x\in (a,b)\}$ with an induced topology is a smooth manifold? If yes, then what is the tangent space $T_{(0,0)}$? Finally, if it is a smooth manifold, what is the advantage of raising assumptions on the smooth relations between charts if a non-smooth curve $M$ is a smooth manifold?

The atlas was asked: $(M,\pi)$ for $\pi(x,y) =x$.

share|cite|improve this question
Yes. Smoothness is not relevant to the question, since you are only asking for homeomorphisms. – Zhen Lin Aug 4 '11 at 14:21
@Zhen: I've edited a question. I wonder if it is a smooth $m$-manifold. – Ilya Aug 4 '11 at 14:26
That's absolutely true (and the first example in any reasonable expositions of manifold theory). A large part of the construction is then to extend to manifolds all the differential calculus you know on open sets in $\mathbb R^n$... – PseudoNeo Aug 4 '11 at 14:30
@PseudoNeo: I thought that once we have a smooth manifold, we have also a differential calculus independently of the atlas. While here we can take the graph of $|x|,x\in[0,1]$ which will be a smooth $1$-manifold. – Ilya Aug 4 '11 at 14:35
I think the question now is about the notion of submanifold. – Pierre-Yves Gaillard Aug 4 '11 at 15:01
up vote 3 down vote accepted

This is the difference between intrinsic and extrinsic geometries: given a continuous embedding of a topological manifold $X$ into another topological manifold $Y$, it is possible that $X$ can be endowed with a smooth structure, and $Y$ endowed with a smooth structure, but that the embedding of $X$ into $Y$ is non-smooth.

For more details, please see my answer to Is $M=\{(x,|x|): x \in (-1, 1)\}$ not a differentiable manifold? .

share|cite|improve this answer

Here's my interpretation of your question: Suppose $X$ is a topological space homeomorphic to an open subset of $\mathbb R^n$. Can $X$ be given a smooth structure? Phrased this way, the answer is clearly yes because you can just take the open subset as the single chart in the atlas, as you say.

Here's a different question. Suppose $X$ is a smooth manifold which is homeomorphic to an open subset of $\mathbb R^n$. Is $X$ diffeomorphic to an open subset of $\mathbb R^n$? This is false in dimension $4$! There are things called large exotic $\mathbb R^4$'s which are homeomorphic to $\mathbb R^4$ but not diffeomorphic to any open subset of $\mathbb R^4$.

share|cite|improve this answer
Thank you for the explanation. Unfortunately, that's not exactly what I asked about, I've made the question clear now. – Ilya Aug 4 '11 at 14:46

Just to point out that the statement of whether a space M 'is a manifold' can be ambiguous, in that a manifold involves a choice of charts and chart maps, and the choice of these that you are making/assuming may not always be clear. So , I am assuming you're asking about the "induced charts".

If by induced charts you refer to the subspace charts, then the answer is no, meaning that, with the induced (subspace) topology, f(x)=|x| is not a submanifold of $\mathbb R^2$; the definition of submanifold I know of needs for the manifold to look locally like the inclusion of $\mathbb R^n$ into $\mathbb R^m$ ; in your case , you want a neighborhood U of every point (let's choose the problem point 0, and an open set $U_0$), and a chart map $\phi$ with $\phi(U_0)$=$(x_k,0)$

BTW: if your tangent space existed at (0,0) , then, by dimension reasons, it would have to be 1-dimensional, i.e., a copy of $\mathbb R$. Is that what you were asking?

A point that may be worth making is that f(x)=|x| can be made into a manifold, using this fact: If two topological spaces X,Y are homeomorphic, and one of them, say X, is a manifold, then the other one is (can be made into) a manifold too: just compose each chart map by the homeomorphism, to "pullback" the charts. In your case, you can homeomorph f(x)=|x|, so that its corner disappears, and you get a manifold.

share|cite|improve this answer

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

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