Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

I don't understand why the following is true, explanation would be greatly appreciated! Suppose $$E(x,y)=\gamma x^{2n}+{y^2\over a}$$ where $\gamma >0, a>0,n\in \mathbb Z^+$.

And we define $$\alpha = {1\over 2\pi}\oint y\,\,\,dx$$ where the path is where $E, \gamma$ are constants.

Why then does it follow that $$a^{n\over n+1}E=\gamma^{1\over n+1}\left({n\pi\alpha\over f}\right)^{2n\over n+1}$$ where $f=\int_0^1(1-u)^{1\over 2}u^{1-2n\over 2n}\,\,du$?

I first tried making $y$ the subject then substitute its expression in terms of $E$ and $x$ into the integral of $\alpha$. Here I can take $E, \gamma$ to be constants, since they are so on the path. This gives $$\alpha = {\sqrt a\over 2\pi}\oint (E-\gamma x^{2n})^{1\over 2}dx$$. But what next? Or perhaps there is a different approach to begin with?

share|cite|improve this question
Welcome to Math.SE! This is a strange calculations indeed; it makes one wonder in what context you encountered it. Can you provide a reference? – user53153 Jan 4 '13 at 3:14
@PavelM : Thank you for the welcome! It is from applied mathematics where $\alpha$ is usually called the action variable and $E$ is the Energy/Hamiltonian. $y$ is the conjugate momentum of $x$. – user55116 Jan 4 '13 at 3:18
up vote 0 down vote accepted

One possible approach is to use Green's formula, but I think you are already on the right track. In the integral $\oint (E-\gamma x^{2n})^{1\over 2}dx$ the square root should be treated with care, since it's negative half of the way. (The path is oval-shaped, symmetric with respect to both axes). Let's focus on the upper half of the path: the lower half will contribute exactly the same amount, because both $y$ and $dx$ will have opposite sign there. Thus, the upper half of integral is $\pi \alpha/\sqrt{a}$ in your notation

The variable $x$ ranges between $\pm (E/\gamma)^{1/(2n)}$, but by symmetry we can consider $$2\int_0^{(E/\gamma)^{1/(2n)}} (E-\gamma x^{2n})^{1/2}\,dx$$ instead. Now introduce the substitution $u=(\gamma/E)x^{2n}$. Since $x=(E/\gamma)^{1/(2n)}u^{1/(2n)}$, we have $dx=(E/\gamma)^{1/(2n)}\frac{1}{2n}u^{\frac{1-2n}{2n}}du$. Thus, the integral becomes $$2\int_0^1 E^{1/2}(1-u)^{1/2}(E/\gamma)^{1/(2n)}\frac{1}{2n}u^{\frac{1-2n}{2n}}\,du = \frac{E^{\frac{n+1}{2n}}}{n\,\gamma^{\frac{1}{2n}}}f $$ We have arrived at $$\frac{\pi\alpha}{\sqrt{a}} = \frac{E^{\frac{n+1}{2n}}}{n\,\gamma^{\frac{1}{2n}}}f$$ which, if there is any justice in the world, should be equivalent to your equation.

share|cite|improve this answer
Indeed! Thank you so much, Pavel! – user55116 Jan 4 '13 at 4:06

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.