For the hamiltonian of a particle of unit mass in a kepler potential:

$$H = \frac{1}{2}\mathbf{p} \cdot \mathbf{p} - \frac{\mu}{r}$$

The angular momentum vector is given by: $\mathbf{L} = \mathbf{r} \times \mathbf{p}$

I know and can show that the poisson brackets of $\mathbf{r} \cdot \mathbf{r}$, $\mathbf{r} \cdot \mathbf{p}$ and $\mathbf{p} \cdot \mathbf{p}$ with any component of the angular momentum vector vanish algebraically, but what is the geometric reasoning behind this? I am trying to develop a better intuition about this. Could someone explain? Thanks!

  • 5
    $\begingroup$ Much like the commutators in quantum mechanics, the Poisson bracket of the angular momentum [or its z-component] with something, XY, tells you what XY gradually transforms into under rotations [under the z-axis, or another axis, depending on the component]. The things you wrote down are scalars, so they don't transform under rotations which is why the Poisson bracket is zero. $\endgroup$ Oct 28, 2014 at 8:54

2 Answers 2


The classical poisson bracket with the generator of any symmetry gives the infinitesimal evolution with respect to that symmetry. The most familiar statement of this is that the time-evolution of any observable $f$ on the phase space is given by

$$ \partial_t f = \{H,f\}$$

Similarily, for a rotation around the $i$-th axis with angle parameter $\phi$, the behaviour under this rotation is governed by

$$ \partial_\phi f = \{L_i,f\}$$

Therefore, if the Poisson bracket $\{L_i ,f\}$ vanishes for all $i$, $f$ is invariant under rotation, i.e. a scalar.

  • $\begingroup$ Note: This is essentially Lubos Motl's comment from above, but this question should not remain unanswered. $\endgroup$
    – ACuriousMind
    Oct 28, 2014 at 11:25

Let me give a further comment (not exactly an answer)

The quantity $\mathbf{r} \cdot \mathbf{r}$ represents the magnitude of the radius as such it does not change under rotation (poisson commutator with ang. momenutm $L$)

The quantity $\mathbf{p} \cdot \mathbf{p}$ represents the magnitude of (linear) momentum, as such it also does not change under rotation

The quantity $\mathbf{r} \cdot \mathbf{p}$ represents the magnitude of action, as such it also does not change under rotation


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.