There's no end to the ingenuity of these astronomers.

We've now spotted some 300 extra-solar planets, with rate of discovery increasing at an extraordinary rate. Astronomers have only seen one of these planets directly; the rest have all been inferred because of the effect that they have on their parent stars: changing their brightness or making them wobble. Of course, you have to be able to see the stars to do this kind of work, so astronomers can only see extra-solar planets in our local region of the Milky Way.

Until now. Gabriele Ingrosso at the National Institute of Nuclear Physics, in Italy, and pals say that there is a way to spot planets in other galaxies. The trick is to exploit a phenomenon called microlensing in which the gravity of one star focuses the light from a more distant one toward Earth.

The advantage of microlensing is that it works best for more distant objects, so it's ideal for planet hunting in other galaxies. In theory, it should be possible to see Earth-size objects in this way. The disadvantage is that microlensing is a relatively rapid, one-off event that lasts a few days at most. That makes observations difficult to verify.

It's hard to see individual stars like this, let alone planets. Astronomers have so far spotted only about a dozen stars in Andromeda in this way, and plans are afoot to search for lots more.

But get this: the light from one of these Andromedan stars showed a distinct variability that the discoverers attribute to an orbiting companion.

And today, a new analysis from Ingrosso and co shows that this companion has a mass about six times that of Jupiter. That's heading into brown-dwarf territory, but it's also well within planetary territory too.

Which means that we may well have seen our first extra-galactic planet.

Ref: arxiv.org/abs/0906.1050: Pixel-lensing as a way to detect extrasolar planets in M31