Statement of the problem:

For the conformal mapping $f:\mathbb{D}\to\mathbb{D}$ , we suppose that the domain $f(\mathbb{D})$ is convex. Prove that for $\mathbb{D}_r=\{z\in \mathbb{C}:|z|<r\}$ the domain $f(\mathbb{D}_r)$ is convex

My approach: I considered the function $$g(z)=f^{-1}(tf(z)+(1-t)f(0)), t\in[0,1]$$ This function is holomorphic , maps the unit disc to itself, and $g(0)=0$. So we can apply Schwarz's lemma. After using it, we have that $$|f^{-1}(tf(z)+(1-t)f(0))|\leq|z|$$ which means that $$g(\mathbb{D}_r)\subset \mathbb{D}_r,\Longrightarrow f^{-1}(tf(z)+(1-t)f(0))\in \mathbb{D}_r \Longrightarrow tf(z)+(1-t)f(0) \in f(\mathbb{D}_r)$$ so all of these line segments belong to our domain, but in order to show that the domain is convex I need to prove that this happens for every $f(z),f(w)$ and not only for $f(z),f(0).$ At this stage I considered the function $$\phi^{-1}\circ g\circ \phi:\mathbb{D}\to \mathbb{D}, \phi(z)=\frac{z-w}{1-\bar{w}z},z,w \in \mathbb{D}$$ and i applied Schwarz's lemma again and I tried to prove it for all $z,w \in \mathbb{D}$ but got stuck.

I would be grateful if you give me just the smallest possible hint and not a whole solution.

Thanks in advance


This is Study's theorem. You can find a proof, e.g., in Duren's book Univalent functions. If we may assume that $f$ is defined and $C^1$ on the closed disc $\bar D$ the proof is quite easy: Look at the images $\gamma_r$ of concentric circles $\partial D_r$ $\,(0<r\leq1)$, given by $$\gamma_r:\quad t\mapsto w(t):=f\bigl(r e^{it}\bigr)\qquad(0\leq t\leq2\pi)\ .\tag{1}$$ Such a curve $\gamma_r$ is convex iff its tangent argument $$t\mapsto\theta(t):={\rm arg}\bigl(w'(t)\bigr)={\rm Im}\bigl(\log w'(t)\bigr)$$ is monotonically increasing. Therefore the convexity condition amounts to $$\theta'(t)={\rm Im}\left({w''(t)\over w'(t)}\right)\geq0\ .\tag{2}$$ If $f(D)$, hence $\gamma_1$, is convex then $(2)$ translates via $(1)$ into a condition involving $f'$, $f''$ on $\partial D$. The maximum principle then guarantees that this condition holds througout $D$, and this in turn implies that all $\gamma_r$ are convex.

  • $\begingroup$ I searched and found it was Study's theorem. I fully understand your answer ! Thanks for that. I found a later paper though were it states that Radό proved the theorem using Schwarz's lemma. Do you have any idea on this ? $\endgroup$ – Beslikas Thanos Nov 12 '18 at 10:58
  • $\begingroup$ Found it. I was close , yet so far... Thanks for your insight $\endgroup$ – Beslikas Thanos Nov 12 '18 at 11:45

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.