Here is Prob. 19, Chap. 6, in the book Principles of Mathematical Analysis by Walter Rudin, 3rd edition:

Let $\gamma_1$ be a curve in $\mathbb{R}^k$, defined on $[a, b]$; let $\phi$ be a continuous 1-1 mapping of $[c, d]$ onto $[a, b]$, such that $\phi(c) = a$; and define $\gamma_2(s) = \gamma_1(\phi(s))$. Prove that $\gamma_2$ is an arc, a closed curve, or a rectifiable curve if and only if the same is true of $\gamma_1$. Prove that $\gamma_2$ and $\gamma_1$ have the same length.

Here is the link to a Math SE post of mine where I've included all the relevant definitions.

Prob. 18, Chap. 6, in Baby Rudin: Analysis of Some Curves in the Plane

My Attempt:

As $\phi$ is a continuous, bijective mapping of the compact set $[c,d]$ onto $[a, b]$, so $\phi$ has an inverse $\phi^{-1}$, which is a bijective, continuous mapping of $[a, b]$ onto $[c, d]$, by Theorem 4.17 in Rudin.

And, as $\gamma_2 = \gamma_1 \circ \phi$, so we have $\gamma_1 = \gamma_2 \circ \phi^{-1}$.

Now suppose that $\gamma_1$ is an arc; this means that $\gamma_1$ is 1-1, and since $\phi$ too is 1-1, so is the composite $\gamma_1 \circ \phi$; that is, $\gamma_2$ is an arc.

Conversely, suppose that $\gamma_2$ is an arc; this means that $\gamma_2$ is 1-1, and as $\phi^{-1}$ too is 1-1, so is the composite $\gamma_2 \circ \phi^{-1}$; that is, $\gamma_1$ is an arc.

Thus we have shown that $\gamma_1$ is an arc if and only if $\gamma_2$ is an arc.

Am I right?

Now suppose that $\gamma_1$ is a closed curve. Then $\gamma_1(a) = \gamma_1(b)$. And as $\gamma_2(c) = \gamma_1(\phi(c)) = \gamma_1(a) = \gamma_1(b)$.

What next?

Conversely, suppose that $\gamma_2$ is a closed curve. Then $\gamma_2(c) = \gamma_2(d)$; that is, $\gamma_1(\phi(c)) = \gamma_1(\phi(d))$.

What next?

How to show that $\gamma_1$ is rectifiable if and only if $\gamma_2$ is rectifiable?

And, how to show that $\gamma_1$ and $\gamma_2$ have the same length?

  • $\begingroup$ You have to prove that since $\phi$ is a continuous bijection, then you have that $\phi(d)=b$. That would finish your proof on the case the curves are closed. $\endgroup$ – Sergio Enrique Yarza Acuña Aug 10 '17 at 13:49
  • $\begingroup$ @SergioEnriqueYarzaAcuña but how to show this? I've thought about it but have not been able to. $\endgroup$ – Saaqib Mahmood Aug 10 '17 at 13:53
  • $\begingroup$ @SaaqibMahmuud: the map $\phi$ is better than a continuous bijection, it's a homeomorphism, which together with the assumption that $\phi(c)=a$ gives you $\phi(d)=b$. $\endgroup$ – Jack Aug 10 '17 at 14:20
  • 1
    $\begingroup$ I appreciate your effort for lots of your lengthy proof-verification type questions. However, there are quite a few of them answered but with no feedback at all: no comments, no votes. For instance, this, this, and this. With all due respect, do you not care about your old questions any more? $\endgroup$ – Jack Aug 10 '17 at 14:44
  • 1
    $\begingroup$ @SaaqibMahmuud Suppose that $\phi(d)\neq b$. Then, $a<\phi(d)<b$. But since $\phi$ is a bijection, $\exists x\in(c,d)$ such that $\phi(x)=b$. However, since $\psi$ is continuous, by the intermediate value problem, there exists $y\in(a,x)$ such that $\phi(y)=\phi(d)$. This contradicts that $\phi$ is a bijection. $\endgroup$ – Sergio Enrique Yarza Acuña Aug 10 '17 at 15:44

The argument for the one-to-one (arc) part could be written in a shorter way: the composition of one-to-one functions is one-to-one. Note that in this part, the continuity of $\phi^{-1}$ is not needed (but yes, it would be needed later).

For the "closed curve" part, you have already observed that $\phi$ is a homeomorphism (continuous bijection with continuous inverse); note that by assumption $\phi(c)=a$ one must have $\phi(d)=b$. (This can be done using properties of homeomorphisms.) Thus we see that $\gamma_1(a)=\gamma_1(b)$ if and only if $\gamma_2(c)=\gamma_2(d)$.

Finally for the "same length part", observe that $\phi$ and its inverse $\phi^{-1}$ gives a one-to-one correspondence between partitions $\{s_i\}$ of $[a,b]$ and $\{t_i\}$ such that $$ \sum|\gamma_1(s_i)-\gamma_1(s_{i-1})|=\sum|\gamma_2(t_i)-\gamma_2(t_{i-1})|. $$ Hence the two curves must have the same length by definition.

  • $\begingroup$ can you please rigorously show how $\phi(d) = b$? $\endgroup$ – Saaqib Mahmood Aug 10 '17 at 15:02
  • $\begingroup$ That deserves to be posted as another question if you are looking for a fully detailed proof. It would be a good question to ask! To be short: a homeomorphism maps boundary to boundary. $\endgroup$ – Jack Aug 10 '17 at 15:19

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.