Congress could not have stated its July 15, 1870, law more clearly: The superintendent of the Naval Observatory was “to contract for the construction of a refracting telescope of the largest size, of American manufacture, at a cost not exceeding fifty thousand dollars.”

It is rather extraordinary to think that the result of this stipulation should have anything to do with Bob Harrington’s livelihood as he sits in his office on a busy April morning speaking into his telephone: “—No—not even the fifteen-minute ones—I could barely even see there was an image there.” For the past half-dozen years Harrington has headed the Equatorial Division of the Astrometry Department of the United States Naval Observatory near Washington’s Rock Creek Park. Astrometry is the measurement of the positions of celestial bodies, and Harrington’s chief tool is that telescope, a long-focus refractor with a 26-inch lens.

The instrument won its congressional approval largely through the zeal of Simon Newcomb, one of the greatest astronomers of his day. Newcomb was also pushing for a particular builder: Alvan Clark, who, with his two sons, had already completed an 18.5-inch refractor that in its time was the world’s largest. The Clarks were not eager to try to grind a lens much larger than that, but in 1868 an English firm turned out a 25-inch refractor, and that was a challenge that could not be ignored. The Navy’s would have to be 26 inches.

This actually meant the fabrication of two 26-inch disks, one of flint glass and one of crown glass. A single lens will focus each wavelength of light at a different distance—as in the familiar prism—and the two different types of glass partially cancel out this effect in each other. Almost nobody on earth knew how to make such a large piece of flawlessly clear flint glass, and in the end Alvan Clark went to Chance Brothers of Birmingham, England. It took several tries, but at last, in 1871, two satisfactory blanks arrived at Clark’s shop in Cambridge, Massachusetts. There followed the most painstaking work, a year and a half spent first grinding the lenses into shape on steam-driven turntables, then polishing them to impeccable clarity.

The telescope took up its post in November 1873 and four years later made history when Asaph Hall, Newcomb’s successor, used it to discover the moons of Mars. In 1893 the instrument was moved from its original site—on swampy lowlands, where Potomac fogs often clouded the view—to its present spot in northwest Washington.

“Since 1961 the telescope has really had only one job,” says Harrington, leading us out into the narcotic splendor of Washington springtime and uphill toward a domed building. “It was built to measure the motion and orbits of planets. But now it’s all binary stars.” Double stars—pairs held together by gravity that revolve around a common center—tell us everything we can know about stellar masses. The telescope reveals their changing orientation toward each other; it was through binaries that astronomers established the relationship between the mass of a star and its luminosity.

Inside Asaph Hall, appropriately (and conveniently) named for the discoverer of the Martian moons, we peer through a pair of glass doors at a round room forty-five feet in diameter. Harrington points to what appears to be the ceiling. “That’s really the observatory floor, and normally we’d ride up on it—it’s the largest licensed elevator in the city of Washington.” Right now, though, it’s not operating because the dome is under repair. This allows me to look at where a massive concrete pier mates untidily with the floor, the seam apparently caulked with sand. Harrington explains: This is the telescope’s mounting, and it is entirely disconnected from the building. It reaches down twenty-five feet to rest on bedrock, and the city’s constant vibrations lap about it like a brook around a boulder.

We climb a cast-iron spiral staircase and step out onto the floor beneath the steel dome; in the center of the room, the telescope points its 32.5-foot barrel skyward. It’s a modern-looking form, but something about its riveted solidity puts it in a different age from the electronics around it. Harrington nods to an instrument covering the eyepiece. “Up until last November we still took visual sightings, because there was no other way to measure the smallest close binaries, even with the help of photography. But they’ve just developed this. It’s a speckle interferometer—lets you use a computer for the whole process.”

So now the only step that requires the help of the human eye is the initial training of the telescope through a small finder scope bolted to its side. But all the marvelous new ancillary instrumentation is there merely to take down what it is told by a machine that was built’ at the time of the first transcontinental railroad.