This article is a sample from DISCOVER's special Extreme Universe issue, available only on newsstands through March 22.

Also see the video from DISCOVER's Mysteries of the Cosmos event, in which Perlmutter was a panelist.

Few scientists can say their work forever changed how we see the universe. Saul Perlmutter is one of them, for his central role in the 1998 discovery of dark energy. That invisible energy, which accounts for a whopping 73 percent of everything in the cosmos, is stretching the fabric of space and could cause a runaway expansion of the universe. Through his groundbreaking research, the then 38-year-old physicist at Lawrence Berkeley National Laboratory in California basically turned our model of the universe on its head.

Scientists had long assumed that atoms—the constituent parts of stars, planets, and people—dominated the universe. Now it is generally accepted that matter makes up merely 5 percent, its share dwarfed by the mysterious antigravity energy that is propelling space apart. (The remaining 22 percent of the cosmos is so-called dark matter, unrelated to dark energy except in its ability to defy all current methods of detection.) Scientists had also long assumed that the universe would either slow infinitely or eventually stop expanding and collapse in on itself. Perlmutter’s findings have forced them to consider that it might instead expand away into nothingness or, worse, end in a “big rip” as the ingredients of stars and galaxies are literally pulled apart.

Since 1998 Perlmutter has worked to refine his measurements of the accelerating universe and the dark energy causing it. Theories abound about the nature of this elusive energy, and Perlmutter is hotly pursuing observational evidence to help find the answer. He spoke to DISCOVER about his strange discovery, the latest ideas about dark energy, and the projects that have the best shot at making sense of this cosmic mystery.

What was the underlying motivation behind the research that led to your discovering dark energy in 1998?
In the 1920s, Edwin Hubble showed that the universe is expanding. But the very next thing out of people’s mouths was more questions: Will it keep expanding? Could it stop expanding? Maybe it could turn around and collapse. How do we know the universe will last forever? These are the obvious things you want to know when you say we live in a changing, expanding universe. And the way you answer questions about the future is by looking at the past.

How do you approach such a complex problem as the history of the universe’s expansion?
The basic idea is that when you look at farther and farther distances, you’re looking further and further back in time. There were some very early papers in the 1930s that proposed using supernovas—really, really bright exploding stars—to measure the universe’s expansion because it appeared there was consistency in how bright they got. If every supernova had almost exactly the same brightness, then you could use how bright it appeared from Earth to measure its distance. But it turned out that the more you looked at supernovas, the wider the variety you saw, and that consistency disappeared. It wasn’t until the 1980s that scientists realized there are subgroups of supernovas, and that one of them, called Type Ia, is very consistent in its brightness. Fortunately, it’s also the brightest of the group, so it’s the one you could follow farthest away.

How did those supernovas reveal the way the universe is expanding?
We used these Type Ia supernovas as our distance indicators. Then you want to know how much the universe has expanded since each explosion occurred. There’s a really convenient way of getting that. The supernova sends out almost all its light in a specific wavelength of blue. But as that blue light travels, it gets stretched exactly as the universe stretches, so it looks red [with a longer wavelength] by the time it reaches us. How red the light looks tells you exactly how much the universe has expanded since the explosion of that supernova. Looking at different supernovas, you should be able to figure out how much the universe has expanded since, for example, 5 billion, 3 billion, then 1 billion years ago, and you would see how that expansion has changed over time. The expectation was that over time the universe’s expansion would be slowing down due to the gravitational attraction of all the mass of all the stuff in the universe. As it turned out, we found that the universe’s expansion was actually speeding up.

Why is it so significant that the universe is expanding faster and faster?
It suggests that the universe is not just a single-parameter story. It can’t just be mass that is causing the change in expansion; the only thing mass can do is slow everything down. So we immediately knew that there was something else in the story. It turns out that most of the stuff in the universe is in the form of some energy in the vacuum that has an odd repulsive property. It makes space reproduce faster, accelerating the expansion of the universe. We don’t know what it actually is, but for now people use the term dark energy as a placeholder to describe the attributes of this mystery.

Was there a moment when the enormous implications of your research really hit you?
Well, it’s funny. This had to be the slowest aha in history, an aha spread out over several months. And the reason is that these are really complex data analysis jobs and there are many steps you have to calibrate and get all straightened out before you get those nice, final data points. On the other hand, there was the very first time I went out to give a talk and present the data. After the talk a famous cosmologist, Joel Primack, stood up and said he just wanted to point out to the physicists in the audience that this is an amazing, absolutely flabbergasting result. I think at that moment I felt the extra sense of ah, that’s right, this is really shocking.