An artist’s rendering of a comet (which is very similar to an asteroid) getting torn apart near a white dwarf. Many white dwarfs have rings of dust around them as well, more evidence of a planetary system orbiting them. Credit: NASA/JPL-Caltech More info i
An artist’s rendering of a comet (which is very similar to an asteroid) getting torn apart near a white dwarf. Many white dwarfs have rings of dust around them as well, more evidence of a planetary system orbiting them. Credit: NASA/JPL-Caltech

The first evidence of planets around another star was found in… 1917?

Contributed by

When was the first evidence for an exoplanet — a planet orbiting another star — truly found?

If you ask some people, they'll say 1995, when 51 Pegasi b was found. But that's a planet orbiting a Sun-like star. Up until recently I would've replied 1992, when the planet PSR B1257+12 b was discovered (two more, given the designations c and d, were found two years later). These planets orbit a pulsar, the super-dense neutron star remnant of a star that exploded. That's weird, but still counts.

You can certainly even make the case that it was in 1988, when a planet was found orbiting the star Gamma Cephei A (one of the two stars in a nearby binary system). That discovery was retracted, but then later confirmed, a story so peculiar that I think I'll write more about it in a later post.

But even those dates are off. Way off. It turns out that the first material evidence of a planetary body orbiting another star was seen in 1917! The fun part is that it wasn't recognized at the time. It couldn't be. And it sat, unnoticed, in a vault for nearly a century.

Oh, this is a good story.

In that year of 1917, the Dutch-American astronomer Adrian van Maanen published a short paper announcing the discovery of two different stars with "high proper motion"; that is, stars that appear to be moving across the sky with faster than normal speed. He had actually taken a glass photographic plate of a different star for a different reason, but when he compared it to an older plate taken of the same part of the sky he noticed that a very faint star in the field had noticeably moved.

All stars move relative to one another as they orbit the galaxy, but the motion we see is usually very small. They're moving rapidly, yes, but the apparent motion is diminished by distance (just like when you look out the window of a moving vehicle, nearby trees whiz by but distant buildings appear to move more slowly; it's perspective). So high proper motion stars are of interest because it usually means they're relatively close to the Sun.

One of these two stars, now called van Maanen 2 (or vM 2), was found to be about 14 light-years away; indeed, pretty close. But the thing is, it's faint. Moreover, a spectrum taken of vM 2 for van Maanen by astronomer Walter Adams indicated it looked like an F star, a star hotter and more massive than the Sun … and far more luminous. At that distance, it should be one of the brightest stars in the sky! But in fact it's incredibly faint, a hundred times fainter than the faintest star you can see with your naked eye.

The original glass plate with the spectrum (shown in negative) of vM 2 on it (right), with the inset showing the lines associated with elements in the star's surface (left). The thick dark lines above and below are a reference spectrum of a known object, for comparison to the star's spectrum. Credit: Carnegie Institution for Science

This was a paradox, unexplained until the mid-1920s. It turns out the star is a white dwarf, the core of a star like the Sun after it dies. When a star runs out of fuel in its core, a complicated series of events occur that blow off the outer layers of the star, revealing the very hot, very dense core. This material might have half or more the mass of the Sun, but squeezed into a ball the size of the Earth. That's why the star is so faint! It's hot, and it glows, but it's so small that even at 14 light-years away you need a decent telescope to spot it at all.

So in 1917 the spectrum of the star confused van Maanen … but it turns out there was a huge secret encoded in that data. A spectrum is the light of the star broken up into individual colors, like a rainbow but with hundreds or thousands of colors instead of seven (my episode of Crash Course Astronomy: Light goes into more detail on how this works). This can reveal a lot about the object, including its temperature, velocity, spin, and (most importantly in this case) its composition. At the time, it wasn't really possible to interpret the spectrum correctly.

But things have changed in the past century. When we look at the spectrum with modern eyes (and modern equipment), we see a very big surprise.

white dwarf spectrum

The spectrum of van Maanen 2 measured using modern instruments. The brightness (vertical axis) versus wavelength/color (horizontal, stretching from roughly blue to green) for a white dwarf is usually devoid of features, but the sharp dips indicate the presence of heavy elements like iron, calcium, and magnesium: evidence for a planetary system. Credit: Napiwotzki et al. via Zuckerman

Normally, the surface of a white dwarf is almost entirely hydrogen or helium. That's in part because the gravity at the surface is fierce, hundreds of thousands of times stronger than the Earth's! The upper layers of a white dwarf are a weird kind of fluid, allowing heavier elements to sink below the surface rapidly, leaving only lighter elements behind.

And that's the surprise: The spectrum of vM 2 shows the presence of heavy elements like iron, calcium, and magnesium. These should sink rapidly, in 100,000 years or less. But the star is billions of years old! Those elements should be long gone from the surface … unless something is resupplying them.

Over the past 30 years or so many other white dwarfs have been found with heavy elements in their spectrum. What could they all have in common? Astronomers have had a thought about this, but only one seems to fit all the data: asteroids.

An artist’s rendering of a comet (which is very similar to an asteroid) getting torn apart near a white dwarf. Many white dwarfs have rings of dust around them as well, more evidence of a planetary system orbiting them. Credit: NASA/JPL-Caltech

An artist’s rendering of a comet (which is very similar to an asteroid) getting torn apart near a white dwarf. Many white dwarfs have rings of dust around them as well, more evidence of a planetary system orbiting them. Credit: NASA/JPL-Caltech

We now know that a large fraction of stars in the sky have planets orbiting them, and that multiple planet systems exist as well. White dwarfs form from stars very much like our own Sun, so it's natural to assume they'd have planets as well.

So why asteroids? The amount of material seen in the white dwarfs' spectra is consistent with that amount of mass. And you have to remember, a white dwarf is what happens to a star after it turns into a red giant and ejects its outer layers. This supremely messes up the planetary system! It changes the gravity of the star around which they orbit. That can throw the planetary orbits into chaos and cause them to collide. This creates a huge amount of rubble, and sometimes the gravity of a planet can toss those asteroids toward the white dwarf. The fierce gravity then rips the asteroid into dust that settles onto the surface. In at least one case (the white dwarf GD362), the chemical composition of the material seen in the white dwarf spectrum matches that of asteroids in our own solar system.

All of these ideas are rather new and strewn around in several different journals. When a field reaches a point where it helps to have all this information collated, an astronomer usually does a "review" of it, an overview that has the links to all the different papers. Astronomer Jay Farihi was doing just that for all this evidence of white dwarf contamination when he saw van Maanen's paper. Suspicious, he contacted the Carnegie Observatory so he could look at the original spectrum, the glass plate with the data taken by Adams.

When Farihi saw it, he knew what he had. The spectrum, sitting in that file drawer for almost 100 years, showed evidence that the star van Maanen 2 had eaten its planets. It was the very first observation we know of showing evidence that planets had once existed (and may yet still exist) around another star.

And as if that’s not enough, remember that van Maanen's original 1917 discovery of the star vM 2 was by accident! He wasn't looking for this star, but basically just happened to notice it. It caught his attention and he decided to follow up. The rest is, literally, history.

Imagine! In 1903, light left this star on its way to Earth. 14 years later it left a faint dot on a glass plate coated with emulsion and exposed to the sky. An astronomer saw it by accident, and then had another astronomer take a spectrum of it, painstakingly recording a streak of light fluctuating in brightness across another emulsion-covered glass plate. A notice was written and published, while the glass plate sat in a vault. It remained there for nearly a century while technology and physics passed it by, while humans retained their curiosity but applied more advanced tools to satisfy it.

And then, one day in a subsequent century, another human's curiosity motivated him to uncover it. And what he found in that smear of light was photographic evidence of an ancient planetary system, likely long ago destroyed as its sun itself died, yet stirring even in death.

Stories like this give me chills. We have been observing the heavens for millennia, but only recording that evidence photographically for the past century or so. What else lies within these archives of human exploration, momentous discoveries just waiting to be uncovered?