by John McPhee
The iron did not hit the water. Charles Ellis, the chief designer, had likened his developing bridge to a hammock strung between redwoods. Addressing the National Academy of Sciences in 1929, he said, “If I knew that there was to be an earthquake in San Francisco and … this bridge were built at that time, I would hie me to the center of it, and while watching the sun sink into China across the Pacific I would feel content with the thought that in case of an earthquake I had chosen the safest spot in which to be.”
Moores and I recross the bridge. South of the south pier, we go down a steep trail to the Pacific beach that comes up from Point Lobos. There are serpentine outcrops above the beach—large blocky hunks in the scaly Franciscan matrix. The closer we come to the pier of the bridge, the more serpentine we see. “It could be part of the basement of the Marin Headlands Terrane,” Moores says. “The seamount, with this serpentine as a basement, is a really fartravelled piece. It’s old and equatorial. It began life way out in the central ocean. What a pile of hash-mash to put a bridge on!” The serpentine is massive, soft and soapy, threaded with asbestos, and below the great bridge it stands up in cliffs. In high positions are concrete gun emplacements built for five-inch pillar-mounted rifles, for six-inch disappearing rifles. There are nudes on the beach. Other men on the beach are sitting upright in little pillboxes that shelter them from the wind. The traffic on the bridge hammers the expansion joints and sounds like the firing of distant guns.
In section more than a mile wide, the serpentine traverses San Francisco from the Golden Gate Bridge to the old naval shipyard on the bay. It underlies all or parts of the Presidio, the University of San Francisco, the Civic Center, Mission Dolores, Haight-Ashbury, Pacific Heights, Hayes Valley, Bayshore. Potrero Hill is serpentine. On Candlestick Hill, near the naval shipyard, we find pillow lavas and red chert. Climbing Billy Goat Hill, above Thirtieth and Castro, we see more pillows under more red chert. Behind a grammar school in Visitacion Valley, at the south end of the city, we scramble up a dark hill of gabbro. Around a pond in McLaren Park, nearby, are diabase boulders.
Chert, pillow lavas, diabase, gabbro, serpentine—item for item, the city seems to be in part composed of the ophiolitic sequence. But here in the Franciscan it is not a sequence. “These rocks are not necessarily from one ophiolite,” Moores comments. “They are probably from all over the globe—bits of ocean crust, of varying age and provenance, collected in the melange.”
It was Andrew Lawson who, in 1895, named the rocks Franciscan. He assumed that they were a conventional formation with traceable stratigraphy—with an eroded structure that could nonetheless be deciphered and spatially reconstructed. One might as well empty a cement mixer and try to number the pebbles in the order in which they entered the machine. On the other hand, Lawson came uncannily close to describing the ophiolitic sequence and to seeing it as ocean crust, and in this respect he was more than half a century ahead of his time. Of the San Francisco sea cliffs that are now described as pillow basalts, Lawson wrote that they resemble “an irregular pile of filled sacks, each having its rotundity deformed by contact with its neighbor.” In 1914, enumerating the igneous rocks of the Franciscan, he came closest to describing the ocean-crustal sequence: “The igneous rocks are genetically allied peridotites, pyroxenites, and gabbros, the first named preponderating and being generally very thoroughly serpentinized; and the second spheroidal and variolitic basalts and diabases.” He mentioned sandstone that had been deposited “upon the sinking bottom of a transgressing sea,” and added that “cherts followed.” Lawson’s sinking bottom was the cooling ocean slab, on its way to the trench.
There are hillsides in downtown San Francisco that are too steep for cars, cable cars, or human locomotion short of rope-andpiton climbing. If you look at a street map, you will see a hiatus in Green Street near the Embarcadero. A few feet west of Sansome, Green ends at the edge of a vacant lot. No one can build there, because the vacant lot is vertical, a cliff of solid rock. “Grungylooking sandstone,” Moores remarks, glancing upward. “Can’t see any bedding. It’s marine sandstone that went into the trench.” Houses line the top of the cliff. An apartment building with cantilevered balconies seems to hang over the edge. Somewhere up there, Green Street continues west.
Filbert Street, a couple of blocks away, is similarly interrupted. The escarpment is a little less sheer. A roadway is out of the question, but the street turns into a staircase. Houses are on both sides of the steps, some of them dating from the mid-nineteenth century. A few are hutlike, with stovepipe chimneys and sagging windows. It is reported that in April, 1906, when the temblor destroyed the municipal water system and the helpless city was being razed by fire, these houses were saved by sheets, blankets, tablecloths, and bedspreads soaking wet with wine. The pale painters and coughing poets who once lived here are gone now, replaced by readers of Barron’s Weekly. Their cliff dwellings are charming past the threshold of envy, and they look ten miles over water.
We climb three hundred and eighty-seven risers, not to mention intercalated terraces and ramps. We are climbing, actually, Telegraph Hill. Moores refers to it as “a single large thick turbidite bed.” In the rock behind the stairway, marine sandstones are interbedded with shales. The sands and muds probably derived from North America but slid out far enough to get into the trench.
At the summit of the hill is Coit Tower, and we go on up that as well—for a dollar, in an elevator. Coit Tower is two hundred and ten feet tall, and its observation terrace is five hundred feet above the sea. The view incorporates the city, the bays, the shoreline suburbs. We look down upon absurdly straight thoroughfares rollercoasting the precipitous hills. Moores says, “The hills have risen rapidly and have therefore eroded steeply. They’re still rising rapidly. San Francisco streets were drawn on paper, without regard to geology or topography. There is one reaction. You laugh.”
Skyscrapers ascend the apron of Nob Hill. South, from the tower, the view passes over them as if they were stalagmites. The eye is stopped by the San Bruno Mountain ridgeline, altitude thirteen hundred feet, fencing off the peninsular city. San Bruno Mountain, like Telegraph and Nob and Russian hills, is a large loose piece of marine sandstone within the Franciscan melange. In the opposite direction, the view crosses the Golden Gate, passes over the Marin Headlands, and is again stopped, this time by Mt. Tamalpais, another block of float sandstone, twice as high as the San Bruno ridge. Francis Drake, the English pirate, two years shy of being knighted by the queen, spent the winter of 1579 camped beside a Pacific beach close to the base of Mt. Tamalpais. Neither he nor anyone in his crew went up the mountain to look around, to discover the three hundred and fifty square miles of protected water close to their encampment. The probable explanation is fog: the cold and almost quotidian sea fog that will overlap the coastal land when the air of California is otherwise cloudless; the fog that fosters the growth and survival of redwoods; the fog that conceals the Golden Gate Bridge and brings out the sounds of tubas.
In the summer and fall of 1769, sixty-four Spanish soldiers walked north four hundred miles from San Diego in search of Monterey Bay. Having no more to go on than a navigator’s description a hundred and sixty-six years old, they failed to recognize Monterey Bay; and they kept on walking, another hundred miles, until they came up against a large coastal mountain. On a clear day, they climbed it. They were fourteen miles from the Golden Gate, but they could not see water there. All they could see was the ocean to their left and, ahead of them, an endless reach of mountains. Forty miles up the coast they could see Point Reyes. Several soldiers were sent ahead to blaze a trail to Point Reyes and prove that it was not Monterey. Their exact route is not clear. It is more than probable that they went far enough to be stopped by the deathly currents of a narrow strait set against unfriendly cliffs, and that—from one of the numerous hills—they became the first Europeans to look upon San Francisco Bay. Soldiers back at the main encampment, nearly starving, climbed San Bruno Mountain hunting deer, and saw the bay.
We, on the tower, have the city of San Francisco spread around us, whereas the soldiers, from similar positions, looked out upon a confused topography of deep swales, creased gullies, and high dunce-cap hills. The hills were all but treeless. They were matted with bush monkeyflowers, woolly painted cups, coffeeberries, Christmasberries, bush lupine, and poison oak. In the swales and gullies were wax myrtle, arroyo willows, coast live oaks, creek dogwood. Large parts of the future city were covered with marching dunes, restrained by dune tansy, coyote shrubs, and sand grass. The scouts probably shrugged. In any case, their mission had failed.
Seven years after the discovery of the bay, thirty-three families with the intent of settling beside it trekked eight hundred miles north from the part of Mexico that now is Arizona. Their leader, Juan Bautista de Anza, went ahead of them and examined the terrain. On top of the serpentine cliffs quite close to what is now the southern approach of the Golden Gate Bridge, he erected a cross. This was to be the site of the citadel (presidio). He erected another cross a few miles away, beside a small lake. This was to be the site of the mission. When the settlers arrived, they pitched their tents by the lake. Their seventh day of residence was the Fourth of July, 1776.
San Francisco is on the North American side of the San Andreas Fault, barely. The fault comes in from the ocean at Mussel Rock and goes straight down the San Francisco Peninsula, southeast. It runs close beside Skyline Boulevard above South San Francisco, San Bruno, Millbrae, Burlingame. The “skyline” is one of several Coast Range ridges that are separated from others by the depression of the bay. Five miles south of the city proper, Moores and I left Skyline Boulevard one December day and climbed through a subdivision to an elevation that was close to a thousand feet. We walked a hundred yards or so to a lookoff above a finger lake. It was three miles long and a tenth as wide, very straight, trending north forty degrees west. It lay in what resembled a small rift valley. Called San Andreas Lake, it lay in the trace of the fault. Here, actually, was where the fault was given its name—by Andrew Lawson, in 1895, who thought he was describing a local feature when in fact it extended more than seven hundred miles. Rock of the fault zone, frequently mashed, erodes easily, and the erosion leaves a groove in the terrain. The groove ran on as far as we could see, and included a second and much longer lake, called Crystal Springs Reservoir. Both lakes were man-made—a term, in this milieu, that women might be pleased to accept. In California, the San Andreas Fault is used as a place to store drinking water. In ditches and pipes the water travels a hundred and fifty miles from Hetch Hetchy Reservoir, in the Sierra Nevada, which lies in a valley adjacent to and equal to Yosemite. In 1913, environmentalists led by John Muir lost America’s first great conservation battle when the valley of the Hetch Hetchy was dammed. Above San Andreas Lake, we could see over a flanking ridge and down through a deep pool of sky to San Francisco International Airport, at the edge of the bay. One after another—up through the pool slowly—747s were rising.
The San Andreas and Crystal Springs reservoirs were built before the turn of the century. On April 18, 1906, when the fault-zone surface in northern California was ruptured for nearly three hundred miles, the parallel sides of the reservoirs slid in different directions. The motion here was eight feet. The Pacific side moved north. The rupture went straight up both lakes, but the dams did not break. Nor have they since. They held in March, 1957, when a 5.5 temblor epicentered near Mussel Rock tore open a smaller reservoir. And they held in the Loma Prieta earthquake, of October, 1989.
“There’s a seismic gap in here that did not get filled by that last event,” Moores remarked. He meant that in 1989 in this stretch of the San Andreas Fault—well north of the epicenter—the two sides did not move.
The idea of the seismic gap first occurred to the seismologist Akitsune Imamura, in Tokyo, more or less at the time of the great San Francisco earthquake of April, 1906. As he studied Japanese earthquake records, which went back hundreds of years, Imamura arranged them graphically in zones of time and place. Where he found quiescent stretches—unfilled areas of his charts—he could see that they had been temporary, as pressure built to fill them. He could see that Tokyo—for what was then the time being—was in a large quiescent zone. In 1912, he began warning the public that the Tokyo gap was soon to be filled. He said that its size suggested to him a severe shock. Essentially, no one was interested. Imamura repeated his warnings for eleven years. The response remained as empty as the gap. In 1923, a hundred and forty thousand people died as Imamura’s gap, in a couple of minutes, closed.
In 1906, when fewer than a million people lived near San Francisco Bay, an estimated three thousand died as a result of the earthquake. The population now exceeds six million, and the much publicized fact that the region is traceried with active faults—that the San Andreas system is not just one trace but a whole family of faults in a stepped and splintering band a great many miles wide—has done nothing to discourage the expanding populace from creating new urban shorelines and new urban skylines and so crowding the faults themselves that the faults’ characteristic landforms are obscured beneath tens of thousands of buildings and homes. In addition to troughs and sag ponds, the motion of transform faults produces scarps, scarplets, saddles, notches, kerncols, kernbuts, and squeeze-up blocks. Streambeds are offset. Alluvium is ponded. Undrained depressions form, and parallel ridges and shutter ridges. Springs appear, and oases. Scarce has the geology made these features afresh when earthmovers move in to move them out, preparing the ground for structures even less permanent. One has to travel to remoter parts of the San Andreas Fault to see its full range of geomorphic features. In San Mateo County, the first county south of the city, nearly all such features have been obscured or destroyed by housing since 1945. Greater San Francisco, the most beautiful urban landscape in the United States, just will not be inconvenienced by a system of sibling faults. Less than a year after the major earthquake of 1989, a modest (fourteen-hundred-square-foot) twobedroom house in the Marina, the most devastated residential district in San Francisco, could be had for five hundred and sixteen thousand dollars, a fall of barely ten per cent from pre-earthquake prices.
Visiting California after the 1906 earthquake, Harry Fielding Reid, of Johns Hopkins University, conceived the theory of elastic rebound, which is also known as the Reid mechanism. It describes the mechanics of fault motion. It preceded by sixty years the larger story of where such motions can lead. In a couple of hundred miles of the San Andreas fault trace, nature’s hints to Harry Reid were not faint. He saw offset crop rows, tree lines, and fences. He found tunnels, highways, and bridges misaligned. Reid decided that elastic strain must have accumulated for years in the rock below until a moment came when the strain surpassed the strength of the rock, causing an abrupt slip, which released the stored energy.
Because the slip followed the direction—the strike—of the rupture, the San Andreas was a strike-slip fault. It was also known as a wrench fault. Not until the discovery of plate tectonics would it also be called a transform fault. The magnitude of the slip—the jump—diminishes with distance from the epicenter. In Marin County in 1906, a dirt road was severed where it crossed the Olema Valley at the head of Tomales Bay. It sprang apart twenty feet. That, or something near it, was the earthquake’s maximum jump. The epicenter was underwater, not far away. The shaking lasted a full minute in 1906—four times as long as the shaking in the event of October, 1989, which released about one thirty-fifth as much energy.
Tomales Bay, long and narrow, resembles San Andreas Lake, and is also in a trough directly on the fault. Standing on its shore, you are impressed not only by its fjordlike dimensions but even more by the complete dissimilarity of the two sides. A tan cotton sock on one foot and a green wool sock on the other could not represent a greater mismatch. On the east side of Tomales Bay are bald unpopulated hills, straw brown in most seasons, and a scatter of lone oaks. Over the west shore of the bay—above the small riparian towns—is a dark-green vegeta
ted ridge, a comparative jungle, which expresses the geology beneath it. The rock of the east shore is Franciscan melange, and presents at its surface a typical Coast Range demeanor. The west side is, for the most part, granite. In age, the two formations are millions of years apart, but, more to the point, they are different in provenance as well. The granite on the west side of Tomales Bay, like the granite under the sea off Mussel Rock, broke away from the southern Sierra Nevada and has travelled north along the fault at least three hundred miles, an earthquake at a time.
Similar offsets line the great fault. A Cretaceous quartz monzonite on the east side of the fault in San Bernardino County mirrors a Cretaceous quartz monzonite on the west side of the fault near San Luis Obispo. An Eocene sandstone on the east side of the fault near San Luis Obispo appears to be the same as an Eocene sandstone on the west side of the fault in the Santa Cruz Mountains. The volcanics of Pinnacles National Monument, on the west side of the fault at the latitude of Monterey, evidently broke away from a chemically identical formation that remains on the North American side of the fault some two hundred and fifty miles south. (Wedged into a tight canyon at Pinnacles with his entire family one day, Moores found himself intoning, “Fault, don’t move now.”) A Cretaceous gabbro on the east side of the fault in Santa Barbara County closely matches a Cretaceous gabbro on the west side of the fault in Mendocino County, three hundred and sixty miles away. Included in the gabbros in both places are bits of rare purple amygdaloidal andesite. On the peninsula south of San Francisco is a piece of a structural basin that seems to have broken away from the San Joaquin Basin, two hundred miles down the fault. In the nineteen-sixties, when all this motion was beginning to be understood as the steady movement of the Pacific and North American plates sliding past each other, a foundation hole was dug for a nuclear power plant exactly in the trough of the San Andreas Fault at Bodega Head, fifty miles up the trace from San Francisco. Half the hot fuel rods would have ended up in the Tropics and the other half in Alaska, but environmentalists halted the project. The Salinian Terrane, sliding past North America with San Diego aboard—and Big Sur and Salinas and Santa Cruz—will, in time, carry Los Angeles to San Francisco. Meanwhile, a part of northern Salinia is the Point Reyes Peninsula—the granite west of Tomales Bay. Looked at from the air, the Point Reyes Peninsula seems about as disjunct from the rest of California as Saudi Arabia is from Africa, and for the same reason: a boundary of lithospheric plates.