One of the most striking features of our solar system is how flat it is. All the planets (now that Pluto is no longer a member) orbit the Sun in the same plane; if you could see the solar system "from the side" all the planets would look like they fall in a line.
What could cause this? Astronomers centuries ago came up with some theories (for example, a passing star ripped matter away from the Sun which formed planets in a plane-- but why then does the Sun rotate in that same plane?) but they didn't work out. Finally, it was theorized that the Sun and the planets all formed from a flat disk, collecting junk and growing.
This disk theory has become a virtual fact in astronomy over the years, with all sorts of indirect evidence to support it. But the direct evidence was lacking. You'd need to see a planet that is sitting in the disk of material around it, and that's hard to find. We see lots of disks surrounding stars, but the stars are too young to have formed planets. We see gaps in some disks, evidence of planetary disturbances, but we can't see the actual planets so we can't be sure they orbit in the same plane as the disk. And where we see planets well enough to determine their orbital tilt, the stars have long since blown away their disks.
That's all changed now. Astronomers using the Hubble Fine Guidance Sensors -- extremely high-resolution instruments usually used to lock Hubble onto target, but can also be used to observe bright stars -- have watched the star Epsilon Eridani for several years. They have seen the star wobble over that time, indicating a planet is there.
The planet has been known for some time. But the extremely nice data generated by the FGSs allows a lot of information to be gathered about the planet. For example, they have found that the period of the planet is 6.85 Â± 0.03 years, the semi-major axis (the orbit is elliptical, so this is the long radius of the ellipse) is 1.88 Â± 0.20 milli-arcseconds on the sky (this is an incredibly small number, roughly equal to seeing a car on the Moon or a U.S. quarter 2500 kilometers away), and the tilt of the orbit is 30.1 Â± 3.8 degrees.
But wait! Eps Eri has a disk (it's the picture above, from data announced in 2002)... and it turns out the tilt of the planet's orbit is consistent with that of the disk.
Pretty cool. It's not proof that the theory is correct, but it's suggestive. Had the planet's orbital tilt not been consistent with the disk tilt, the theory would have been in trouble. As it is, it strongly supports the theory.
I'll note that many young earth creationists like to say that astronomers don't know how solar systems form, and we've never seen planets forming. That's baloney. I've seen that claim on many creationist sites, even long after we've seen disks around other stars, and all the other supporting evidence. Think they'll take that statement down from their sites now? Yeah, me neither.
A few more neat things about this. Eps Eri is pretty close, just 10.5 light years away (a bit more than twice as far as the nearest known star to us). It's a bit lighter weight than the Sun, with about 0.83 times the mass, and is slightly cooler and redder. The planet is about 1.5 times the mass of Jupiter, and its orbit takes it as close to Eps Eri as the Earth is to the Sun, and out as far as Jupiter (incidentally, the FGS data indicate the possible presence of another planet with a 50 - 100 year period that remains otherwise undetected as yet). When the planet is closer in to the star it gets brighter, and easier to see. Astronomers are hoping to be able to actually get an image of the planet when it next is at periapsis (closest approach to the star), which is in early 2007. This is a very tough observation, since the planet will be so close to the star that the star's light will drown it out. The separation is only going to be .3 arcseconds, which strains even Hubble's ability to split them.
Years ago, when I worked on Hubble, we wondered if we could spot Jupiter if it were orbiting Alpha Centauri. I worked really hard on a simulation to see how our camera could observe that, and found in the end we just barely might be able to see it. But I didn't think we could positively identify it. We never tried; it would take too much time and effort for too low a chance of payoff. We still don't know if there's a planet around the nearest known star, which is kinda funny: we know a lot about a planet more than twice as far away.
I thank Dr. Fritz Benedict, who led the team who did this work, for the use of the diagram above.