I won't waste your time by first writing a lot of words. Let's get right to the way cool picture:
That image is of the galaxy cluster CL0024+1652 (go look at the higher resolution version -- it's very pretty!), a galactic city located a whopping 5 billion light years away! That means the light we see from this cluster left it five billion years ago, so we're seeing this structure as it was when the Universe was just 2/3 its present age. Almost every small object in that image is a galaxy, and all of them are held sway by the cluster's gravity, orbiting the center like bees flying around a beehive.
It has long been thought that every large object in the Universe is surrounded by a halo of dark matter -- unseen, mysterious, yet profoundly influential in the life of normal matter. Dark matter (or just DM for short) gives off no light, and does not interact with normal matter directly-- a cloud of it could pass right through you and you'd never know. But, like regular old matter, it has gravity, and that can betray its presence.
I've gone over this before -- Einstein postulated that gravity from matter bends space, like a bowling ball on a bed bends the mattress. Light will follow that bend in space the same way a marble rolled across the bed will curve from the bowling ball's dip. If there is some massive object out there in space, and some galaxy beyond it, the light from the more distant galaxy will bend as it passes by the intervening material. We see that as a distortion in the shape of the galaxy. This is called gravitational lensing, and can be used to map out the location of dark matter. So even though we cannot see DM directly, we can see its effects.
Clusters are a rich hunting ground for DM: it's thought that DM helped normal matter form into large structures when the Universe was young. So clusters should still have lots of DM around them. It was also thought that in general, the DM halo around clusters would be roughly spherical, or maybe slightly elongated like a football. More exotic shapes weren't really expected.
Until now, that is. CL0024_1652 is surrounded by a ring of dark matter, as shown in that Hubble image! (Update: this is also available at Hubblesite.org). Astronomers looked at the cluster, and very carefully mapped the distortions of background galaxies by the gravity of the cluster. What they found, after applying some fiendish math and physics to the observations, was that ghost ring of dark matter. Mind you, the image is not directly of the DM itself, but is a map of its location using those gravitational distortions. Here's what the cluster looks like in just visible light:
If you look carefully at it (or grab a higher-res version) you can see blue arcs of light all around it. Those are images of more distant galaxies distorted by the cluster's gravity. It's distortions similar (but weaker) than those that astronomers use to map to the DM.
This ring of dark matter is totally cool. When I first heard about it a few days ago, my first thought was, "Hmm, must have been some sort of collision between clusters to do that." Since DM doesn't interact with normal matter except through gravity, I knew that there is no way for matter to sculpt the DM through the usual methods (pressure, winds, things like that which give gas clouds such gorgeous shapes). But sometimes, when a small galaxy collides with a big one, the gravitational interaction can totally whip the big galaxy into a frenzy, leaving it with a ring shape. My favorite example of this is Hoag's object:
That's a single galaxy, though (if it were at the cluster distance, it would be just one of the dots in the cluster image). The DM ring around the cluster is far, far larger-- it's something like 2.6 million light years across! For comparison, our own Milky Way Galaxy is about 100,000 light years across. The nearest big spiral galaxy to us is over 2 million light years away, so that ring is BIG. Sometime in the past, maybe one or two billion years ago, the cluster suffered a mighty collision with another cluster, and their mutual gravitational dance puffed out the DM halo into that ring shape (you can see a diagram of this on the European Hubble website).
Interestingly, the astronomers who discovered the ring did some research and discovered a paper claiming that the cluster did in fact have a collision with another cluster: it appears that there are two separate groupings of galaxies in the cluster, which implies a collision. Evidently, we are seeing this event along our line of sight; it's as if the colliding cluster was aimed right at us (or, I suppose, directly away from us). From the side, the ring would look more rectangular, probably, like a barrel or a life preserver seen from the side. It just so happens that our location gives this event its dramatic appearance.
This ring will prove important to astronomers for many reasons. For one, it gives us insight on how dark matter can be shaped by normal matter. We don't understand the nature of DM very well at all, so anything like this can only be helpful in honing the theories. For another, this is a bright, dense, well-observed cluster, so we can learn quite a bit about it. The more we understand the cluster, the less we have to guess about its DM halo. For a third, this is the first time a halo of DM has been seen to be so differently shaped from the gas and other mass in the cluster. It can be studied separately from the normal matter, making that task in some ways easier.
And the best thing about it? It's unexpected! Surprise! It's always nice in science when things go your way, and predictions pan out. But it's even better when you process the data, and a big fat bogie like this is sitting in the middle of it. That means there's more to learn, more to know, more to understand. And that's the very essence of science!