Johns Hopkins University Applied Physics Laboratory / Carnegie Institution / NASA

In the 19th century, geologists hypothesized that some of Earth’s most dramatic landforms emerged because the planet had shriveled over time, like a juicy grape becoming a raisin. As the Earth contracted, they thought, the ground buckled in on itself, producing towering mountains, steep cliffs, and deep ocean basins.

The theory worked only if they assumed that Earth was covered in a homogenous shell of rocky crust, which, it turned out some years later, wasn’t the case. In the mid-20th century, as theories about colliding plate tectonics took hold, the raisin hypothesis was abandoned.

But it was still a good theory, and it happens to apply to another planet: Mercury.

Mercury, tiny but mighty, the smallest planet in the solar system, is circumscribed to a scorched existence close to the sun. Where Earth has tectonic plates, Mercury has a thin, single crust enveloping its molten-iron core. Mercury is also way more metal than rock; its liquid core comprises more than 60 percent of its volume, compared with our planet’s 15 percent.

“It’s basically a big ball of metal surrounded by a little bit of rock,” says Nancy Chabot, a planetary scientist who studies Mercury at Johns Hopkins University Applied Physics Laboratory.

Humanity got its first good look at Mercury in the 1970s, when Mariner 10, a NASA spacecraft, embarked on a tour of the solar system’s innermost planets. The mission revealed that Mercury produced a magnetic field, an invisible bubble that protects the surface from the worst of the sun’s radiation. Magnetic fields are generated from within, by a planet’s liquid-metal core, still hot from its creation several billion years ago.

Mariner also photographed chains of cliffs and ridges, some as tall as Washington State’s Mount St. Helens, stretching for hundreds of miles. The landforms, known as scarps, fascinated scientists. They set about explaining the scarps’ origins, and quickly ruled out several explanations. Mercury bears the marks of volcanism—long channels carved into the rock by flowing lava—but activity ceased a few billion years ago. It also exhibits signs of an ancient bombardment by objects of various sizes, but such onslaughts produce craters, not mountains.

Scientists examined what they knew about planet formation. The iron cores of all the rocky planets in our solar system, including Earth, are still cooling down from their fiery births 4.5 billion years ago. As liquid iron turns solid, it contracts, creating space for the crust above to contract, too. The ground cracks and thrusts upward, redrawing the landscape. If the planet was covered in a uniform crust, like a lid, the effects of this contraction should remain visible. The scarps, scientists predicted, must be a result of this phenomenon.

With this hypothesis in hand, scientists used Mariner images to calculate the length and height of the scarps. They used the measurements to estimate how much Mercury had shrunk in its lifetime.

But they couldn’t be entirely sure of even their best estimates. Mariner had approached Mercury only three times before zooming back out into space, and the encounters had allowed the spacecraft to photograph just 45 percent of the planet’s surface. To test their predictions, scientists needed a spacecraft to visit the planet and stay for a while.

That spacecraft, named MESSENGER, arrived in 2011. The mission provided global views of Mercury for the first time. “I was like a kid on Christmas morning every single day for three years because there’s always something new to see and always something surprising,” recalls Chabot, who worked on the mission’s imaging system.

The photographs showed that Mercury had undoubtedly shrunk. Thousands of scarps snaked along its surface. Scientists studied about 6,000 of the structures and calculated that the planet’s radius of 2,440 kilometers (1,516 miles) has shrunk between 14 kilometers and 20 kilometers (8.6 miles and 12.4 miles) since its creation.

Scientists have detected such scarps on other worlds in the solar system, like the moon and Mars. But the signs of shrinking are more prominent on Mercury, says Paul Byrne, a planetary geologist at North Carolina State University who led the MESSENGER study.

“I don’t know if Mercury has contracted more than the other worlds,” Byrne says. “I think it’s safe to say that the effects of global contraction have been most pronounced on Mercury.”

MESSENGER eventually went the way of several other spacecraft in NASA history: It was sent crashing into Mercury’s surface. But a replacement is on its way.

Last month, the European Space Agency and JAXA, Japan’s space agency, launched the BepiColombo mission. BepiColombo is a two-for-one deal; when the mission’s spacecraft arrive in seven years, they will split into two orbiters. Like Mariner and MESSENGER before it, BepiColombo should provide even more views of the Mercurian surface.

Byrne says that he doesn’t expect his estimates of the rate of shrinking to change drastically. The contraction of Mercury occurs on tremendous timescales, so it’s unlikely that much would change in a matter of years.

Still, some questions remain about the incredible shrinking planet. Specifically: Is it still shrinking? And if it is, when will it stop?

Solving this mystery would require yet another mission to Mercury, but this time to land on the surface, where temperatures can reach a toasty 430 degrees Celsius (800 degrees Fahrenheit). The spacecraft would need to come equipped with seismometers to listen for Mercury-quakes, and scientists would need to wait a very, very long time for the payoff.

“The only way that we would know for sure that it’s not [contracting] is if we had seismic instruments on the surface that were able to listen for thousands of years and not detect earthquakes that we could attribute to the planet creaking and cooling,” Byrne says. “I don’t think there’s any measure we could make that would reliably tell us for sure the process has ended.”

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.