Use Morera's Theorem to prove the following: Let $\Omega\subset_{\text{open}} \mathbb{C}$,$z_{0}\in\Omega$. Suppose $f$ is continuous in $\Omega$ and analytic in $\Omega \backslash \{z_{0}\}$. Then $f$ is analytic in $\Omega$

I am not really sure how to proceed. Morera's theorem states:

Theorem: Morera If $g$ is a continuous function in the open disc $D$ such that for any triangle $T$ contained in $D$: $$ \oint_{T} g(z) \, dz = 0 $$ Then $g$ is holomorphic in $D$.

  • $\Omega$ is an open subset of $\mathbb{C}$. So at $z_{0}$ there is some open disk of radius $r$ contained in $\Omega$, $D(z_{0},r)\subset \Omega$. If I can show that Morera's thoerem holds on this disk, then I am done, since holomorphic functions are always analytic - so is $f$ is holomorphic on $D(z_{0},r) \supset \{z_{0}\}$ then it is analytic at $z_{0}$.

  • I know need only concern myself with triangles $T$ such that one of the vertices is $z_{0}$, since all other cases follow automatically from analyticity (holomorphicity) of $f$ and Cauchy's/Goursat's theorem(s).

  • However, how can I evaluate these integrals??

  • $\begingroup$ Also note $(z-z_0) g(z)$ is holomorphic (if $g$ is only bounded near $z_0$ look at $(z-z_0)^2 g(z)$) thus it is analytic thus $g(z)$ is meromorphic with a pole at $z_0$ and its boundedness implies the pole is of order $0$ ie. $g(z)$ is analytic. For Morera it is the same idea except you'll look at $G(z)=\int_a^z g(s)ds$ which is holomorphic thus analytic and so is $G'(z)$. $\endgroup$ – reuns Jun 18 at 21:48

It's not true that you need only consider triangles with one vertex $z_0$. Indeed those are not a problem: the integral over a triangle with one vertex $z_0$ is the limit of integrals over triangles that $z_0$ is outside of, which are $0$ by Cauchy.

But you can't use Cauchy/Goursat when $z_0$ is inside your triangle. To handle those, break up such a triangle into three triangles, each of which has $z_0$ at one vertex. The integral over your triangle is the sum of the integrals over the three sub-triangles (as the integrals over the interior edges cancel). And by the previous paragraph, the integrals over those triangles are $0$.

enter image description here

  • $\begingroup$ What is the formal way in which I would define that limit? - I can see intuitively that it should follow imeadiately from the continuity of $f$ there shouldn't be a discontinuity on the boundary, but how would I show this rigrously? $\endgroup$ – thesundayscientist Jun 18 at 21:43
  • $\begingroup$ If $\Gamma_p$ is a positively-oriented triangle with corners $a, b, p$, and $f$ is continuous on a region containing $\Gamma_p$ for $p$ in a neighbourhood of $z_0$, then $\lim_{p \to z_0} \oint_{\Gamma_p} f(z)\; dz = \oint_{\Gamma_{z_0}} f(z)\; dz$. This follows from uniform continuity of $f$ on a compact set. $\endgroup$ – Robert Israel Jun 19 at 1:52
  • $\begingroup$ Really sorry, but from where did we get the idea that $f$ is uniformly continuous on compacts? $\endgroup$ – thesundayscientist Jun 19 at 9:45
  • 1
    $\begingroup$ Heine-Cantor theorem: a continuous function from a compact metric space to a metric space is uniformly continuous. $\endgroup$ – Robert Israel Jun 19 at 12:09
  • 1
    $\begingroup$ He was talking about the triangle. $\endgroup$ – cmk Jun 20 at 11:40

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.