Here I present two cases.

1) I first consider the keyhole contour. For instance, if I want to find $\int_{0}^{\infty} \frac{{\text{d}}x}{(x+a)^2 \sqrt{x}}$, when I consider the piece of the contour going from $\infty$ to 0, I replace $x$ an $xe^{2i\pi}$ in the square root, and the justification is that the argument of $z$ along this contour is zero but in the fourth quadrant, so we $2\pi$.

2) I next consider the dog-bone contour to find $\int_{-1}^{1} \frac{\text{d} x}{\sqrt{1-x^2}}$. I easily find the branch points of $\frac{1}{\sqrt{z^2-1}}$, and knowing that the circular bits go to zero, the contour integral is $\int_{1}^{-1} \frac{\text{d}x}{\sqrt{e^{i\pi}(1-x^2)}}+ \int_{-1}^{1} \frac{\text{d}x}{\sqrt{[e^{-i\pi}(1-x^2)}]}$.

I am unable to understand where these phases come from. I get that in the second case, the phase factors introduce a minus sign in the integral to get a form usable for the contour integral. However, why did we choose $\pi$ and $-\pi$? Firstly, these are multiplied to the entire terms in the square roots, rather than to the $x$’s alone. Additionally, the range of values used seem to be from $-\pi$ to $\pi$ instead of 0 to $2\pi$, and if we use the quadrant argument, each time it crosses the imaginary axis and reaches a different quadrant, the phase should be different! Could someone please resolve my conceptual misunderstanding? Thanks!


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.