I am reading Theorem 9.20 (Flowout Thoerem) from Lee's Introduction to Smooth Manifolds, Second edition.

A part of the theorem states the following:

Let $M$ be a smooth manifold and $S$ be a $k$-dimensional embedded submanifold in $M$. Let $V$ be a smooth vector field on $M$ which is nowhere tangent to $S$. Let $\theta:\mathcal D\to M$ be the maximal flow associated with $V$, and let $\mathcal O=(\mathbf R\times S)\cap D$. Let $\Phi=\theta|_{\mathcal D}$.

Then $\partial/\partial t\in \mathfrak X(\mathcal O)$ is $\Phi$-related to $V$.

First Question: What is $\partial/\partial t$? Is it defined as $$ d(id\times \varphi)_p\left(\frac{\partial}{\partial t}\Big|_p\right)=(1; 0,\ldots, 0)\in \mathbf R\times \mathbf R^k $$ where $id\times \varphi$ is a smooth chart on $\mathcal O$? If yes then we should also check that this is well-defined.

Second Question: What is the theorem trying to say geometrically? I can verify the details given in the proof but am having no insight as to what is really being said.



The vector field $\partial/\partial t$ is just the standard coordinate vector field on $\mathbb R$. By means of the canonical identification of each tangent space to a product manifold with the direct sum of tangent spaces of the factors, we can view it as a vector field on $\mathbb R\times S$. It's canonically defined, so there's no need to check that it's well defined.

As for what the theorem means geometrically, I tried to suggest the geometric meaning in the illustration that comes with it (see below). The idea is that you start with a $k$-dimensional submanifold $S$, and then flowing out along integral curves of $V$ gives you a certain $(k+1)$-dimensional submanifold containing $S$, to which $V$ is tangent.

(BTW, there's a typo in your transcription of the theorem statement -- the definition of $\mathcal O$ should be $(\mathbb R\times S)\cap \mathcal D$, not $(\mathbb R\times S)\times \mathcal D$.)

enter image description here

  • $\begingroup$ You made a typo in the same too! It is $(\mathbf R\times S)\cap \mathcal D$, not $(\mathbf R\times S)\cup \mathcal D$. $\endgroup$ – caffeinemachine Jul 12 '15 at 22:14
  • $\begingroup$ I see. This has a nice geometric meaning. My main hindrance was a lack of understanding of what $\partial/\partial t$ means. This clarifies it. $\endgroup$ – caffeinemachine Jul 12 '15 at 22:18
  • $\begingroup$ Good grief. Fixed now. Thanks. $\endgroup$ – Jack Lee Jul 12 '15 at 22:37

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.