0
$\begingroup$

To be more precise, as sets, $S\subset M$, and S has structure such that it is a manifold on its own right but not a submanifold of the manifold $M$?

  • 0
    Might be mistaken, but have you heard about [irrational winding of a torus](https://en.wikipedia.org/wiki/Linear_flow_on_the_torus#Irrational_winding_of_a_torus)?2017-02-24
  • 1
    Or take the mapping $(-\pi/2,\pi/4)\to\Bbb R^2$ given by $t\rightsquigarrow \big(\cos(2t)\cos t,\cos(2t)\sin t\big)$, whose image is ... a figure 6. :)2017-02-24

1 Answers 1

2

Take $M = \mathbb R^2$. Take $S$ to be the image of $i: \mathbb R \to M$ given by $i(t) = (t^2, t^3)$. Then $S$ has a cusp at $(0,0)$, so it does not inherit a manifold structure from $M$. However, $S = \mathbb R$ as a set, which has an obvious manifold structure of its own.

  • 0
    Right: In the smooth category. In the topological category it would be a wild sphere in $R^3$.2017-02-24
  • 0
    How is $S=\mathbb{R}$ as a set? should not S be a curve in $R^2$?, its y component equals set R, no doubt.2017-02-25
  • 0
    And S is not a manifold, right?2017-02-25
  • 0
    @jack $S = \mathbb R$ as a set because $S$ is the image of an injective function of $\mathbb R$.2017-02-25