Few people look near the sun in the sky; I do. My reward is that sometimes I spot a halo around the sun. At sunset, I sometimes see two bright patches of color on either side of the sun, called sun dogs or parhelia. These phenomena are caused by sunlight shining through water ice crystals falling through Earth’s atmosphere, and they occur dozens of times every year. I love to spot these halos around the moon as well. Their color and shape always brighten my spirits, although I know that seeing these halos often means rain is in the forecast. When the Viking 1 and Viking 2 spacecrafts landed on Mars in 1976 and took images with their primitive digital cameras, I wondered if these images would show halos around the sun. Then I realized that the snowflakes (ice crystals) on Mars might not be the same shape or composition as snowflakes on Earth. So, being a scientist, I decided to investigate further. I went to the library to find out the shape of Martian snowflakes. I searched for days with no luck. I suddenly realized that I had made a great discovery—no one knew the shape of Martian snowflakes! Here was my chance: I could grow Martian snowflakes in my laboratory and become the first person to know the shape of a snowflake on Mars. I enlisted the help of a physics resource engineer, Clarence Bennett, and together we set off to grow and photograph Martian snowflakes.

Snowflakes on Earth are made of water ice crystals and they all have six sides. (Scientists say that snowflakes have hexagonal symmetry.) Over a century ago, Wilson Bentley collected and photographed snowflakes by adapting a microscope to a bellows camera. He published his images in a book, Snow Crystals, and showed many different shapes: thin flat hexagonal plates, hexagonal prisms shaped like an unsharpened pencil, and the usual six-armed snowflakes that I try to catch on my tongue. Although all of these shapes have the same hexagonal symmetry, they have different crystal habits. Scientists discovered that the shape of a snowflake depended on the temperature and humidity of the air from which it formed. Warmer air with high humidity led to delicate six-armed snowflakes, while cold, dry air at higher elevations produced flat-sided hexagonal plates or prisms.

Martian snowflakes, I realized, would probably be made of carbon-dioxide (CO₂) ice, since 95 percent of the atmosphere of Mars is made of CO₂, and the temperature sometimes sinks below CO₂’s freezing point, an extremely cold 148 kelvin (-193°F/-125°C). When Clarence and I began our experiment, scientists already knew that carbon-dioxide ice crystallized with cubic symmetry. However, they didn’t know its crystal habit when it formed under Martian temperatures and pressures.

Ice crystals on Earth (under a microscope)

In order to grow Martian snowflakes, Clarence and I built a "Mars chamber." It had a vacuum pump that reduced the pressure inside to simulate the pressure on Mars, which is less than 1 percent of the pressure on Earth. (The average surface pressure on Mars is 6 millibars; the atmospheric pressure on Earth is 1 bar, or 160 times more than that on Mars.) Our Mars chamber also had a liquid-nitrogen plumbing system to cool the inside to under 148 kelvin and to allow carbon dioxide to crystallize. The chamber was made of strong steel, with a thick Lexan (a very strong plastic) window that could withstand the tons of force exerted on it by the Earth’s atmosphere (that is, the air outside the chamber) when the chamber was evacuated. It even had a cool circular handle to "dog" (crank open) the access hatch. When I spun the handle to seal the door, I felt like the submariners I had seen in movies.

Growing a snowflake turned out to be harder than we expected, but after months of growing frost, we eventually produced a snowstorm in the chamber and were able to collect and photograph the carbon-dioxide snow crystals we had grown under Martian conditions. The photographs showed beautiful cubes of ice with every corner neatly clipped off to form a triangle. This shape is called a cubeoctahedron—a cross between an octahedron and a cube. A cubeoctahedron has cubic symmetry, so it fit the known symmetry of carbon-dioxide ice.

For a while, Clarence and I smiled and reveled in the knowledge that we were the only ones who knew the shape of snowflakes on Mars. But, being scientists, we quickly returned to work to discover if we could be wrong. I decided to calculate the position of ice crystal halos around the sun when viewed from Mars, all the while hoping that the Viking landers would have photographed these halos. In order to calculate the halo’s position when seen from Mars, I needed to know the shape of Martian snowflakes, which Clarence and I had discovered, and the index of refraction of Martian carbon-dioxide ice, information I found from other researchers’ work.

When I see a halo around the moon, I reach out my arm and "touch" the moon with the tip of my thumb. Then, stretching my hand as wide as I can, I "touch" the inside edge of the halo with the tip of my little finger. The common halo is one handspan from the moon, about 22.5 degrees in angle.

Since I already knew the position of halos made of hexagonal ice crystals, I decided to try and figure out what the position would be of halos made of the cubeoctahedronal ice crystals that might exist on Mars. Using the researchers’ data and my crystal shapes, I calculated that one of the possible ice crystal halos on Mars should have been about 26 degrees from the sun; a second one would have been about 38 degrees from the sun. The smaller halo on Mars would be slightly over one handspan from the sun. The larger halo would be slightly less than two handspans, just a bit smaller than the angular size of a rainbow, which is 42 degrees. (As with all measurements, my measurement had a possible range of error.)

I called up the Viking project at NASA’s Jet Propulsion Laboratory and asked Steve Wall, an engineer on the Viking imaging project, if he had seen any ice crystal halos in their photographs. He said no, and then asked an amazing question: "Where should we look?" I told him, which resulted in a scientific search through all of the Viking photos. Alas, they spotted no ice crystal halos. However, in 1996, the Mars Pathfinder lander also took pictures of the sky. These photos showed high, thin, ice crystal clouds, but still no halos.

Now, in 2005, two more rovers covered with cameras are on Mars. Perhaps this time the cameras will photograph a halo around the sun and bring back to Earth proof of the shape of Martian snowflakes. The rovers also have a camera with a powerful magnifier on it, equivalent to a geologist’s hand lens. This is the perfect tool for examining Martian snowflakes. If the rovers are caught in a snowstorm, they may be able to photograph Martian snowflakes, and I’ll be able to compare their images to mine. If this happens, I’ll be in heaven. Or at least on Mars.