For which real $\alpha,\beta>0$ is there a single valued branch $f$ of the analytic function $f(z)=z^{\alpha}(1-z)^{\beta}$ such that $f$ is defined on $\mathbb{C}\backslash[0,1]?$

I know that the argument of $f$ must not have any ambiguity when circumventing the points 0 and 1. Expressing $f$ in exponential form I get $$f(z)=e^{\alpha(ln|z|+iarg(z))+\beta(ln|1-z|+iarg(1-z))}$$.

I know $e^z$ is $2\pi i$ periodic, does this imply $2\pi\alpha+2\pi\beta=2\pi n$, where $n\in\mathbb{Z}?$


Yes, it means the condition is $\alpha + \beta \in \mathbb{Z}$.

One way to see it is to follow the change of the argument along small circles around the branch points, where we have a change of $2\pi \beta$ at $1$, and a change of $2\pi \alpha$ at $0$, and these must add to an integer multiple of $2\pi$ in order for the image of e.g. a dogbone contour to be a closed path.

Another way to see it is to look at the potential third branch point, $\infty$. Consider the function

$$g(z) = f\left(\frac{1}{z}\right) = z^{-\alpha} \left(1-z^{-1}\right)^\beta = z^{-(\alpha+\beta)}(z-1)^\beta$$

defined on $\mathbb{C}\setminus [1,+\infty]$. There is a holomorphic branch of $f$ on $\mathbb{C}\setminus [0,1]$ if and only if there is a meromorphic branch of $g$ on $\mathbb{C}\setminus [1,+\infty]$. The condition for the latter is easily seen to be $\alpha+\beta\in\mathbb{Z}$, so that $g$ has at worst a pole in $0$, and not a branch point.


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.