But the actual color filling the windows and skylight of their rover was much more reddish than that, and even on clear mornings after heavy storms, Sax had never seen it look anywhere near as blue as a Terran sky. He thought about it more. Another effect of Mars’s light gravity was that the air column lofted taller than Earth’s. It was possible that the smallest fines were effectively in suspension, and had been blown above the altitude of most clouds, where they escaped being scrubbed out by storms. He recalled that haze layers had been photographed that were as much as fifty kilometers high, well above the clouds. Another factor might be the composition of the atmosphere; carbon dioxide molecules were more efficient light scatterers than oxygen and nitrogen, and Mars, despite Sax’s best efforts, still had much more CO2 in its atmosphere than Earth did. The effects of that difference would be calculable. He typed up the equation for Rayleigh’s law of scattering, which states that the light energy scattered per unit volume of air is inversely proportional to the fourth power of the wavelength of the illuminating radiation. Then he scribbled away on his lectern screen, altering the variables, checking handbooks, or filling in quantities by memory, or guesswork.

He concluded that if the atmosphere was thickened to one bar, then the sky would probably turn milk white. He also confirmed that in theory the present-day Martian sky ought to be a lot bluer than it was, with its scattered blue light about sixteen times the intensity of the red. This suggested that fines very high in the atmosphere were probably reddening the sky. If that was the correct explanation, one could infer that the color and opacity of the Martian sky would for many years be subject to very wide variation, depending on weather and other influences on the cleanness of the air. . . .

And so he worked on, trying to incorporate into the calculation skylight radiance intensities, Chandrasekhar’s radiative transfer equation, chromaticity scales, aerosol chemical compositions, Legendre polynomials to evaluate the angular scattering intensities, Riccati-Bessel functions to evaluate the scattering cross sections, and so on— occupying the better part of the drive to Arena Glacier, concentrating hard and steadfastly ignoring the world around him and the situation in which he now found himself.

• • •

Early that afternoon they came to a small town called Bradbury, which under its Nicosia-class tent looked like something out of Illinois: treelined blacktop streets, screened-in porches fronting two-story brick houses with shingle roofs, a main street with shops and parking meters, a central park with a white gazebo under giant maples. . . .