Popular Science Monthly/Volume 69/August 1906/The Investigation of the San Francisco Earthquake




AUGUST, 1906




IT is the natural and legitimate ambition of a properly constituted geologist to see a glacier, witness an eruption and feel an earthquake. The glacier is always ready, awaiting his visit; the eruption has a course to run, and alacrity only is needed to catch its more important phases; but the earthquake, unheralded and brief, may elude him through his entire lifetime. It had been my fortune to experience only a single weak tremor, and I had, moreover, been tantalized by narrowly missing the great Inyo earthquake of 1872 and the Alaska earthquake of 1899. When, therefore, I was awakened in Berkeley on the eighteenth of April last by a tumult of motions and noises, it was with unalloyed pleasure that I became aware that a vigorous earthquake was in progress. The creaking of the building, which has a heavy frame of redwood, and the rattling of various articles of furniture so occupied my attention that I did not fully differentiate the noises peculiar to the earthquake itself. The motions I was able to analyze more successfully, perceiving that, while they had many directions, the dominant factor was a swaying in the north-south direction, which caused me to roll slightly as I lay with my head toward the east. Afterward I found a suspended electric lamp swinging in the north-south direction, and observed that water had been splashed southward from a pitcher. These notes of direction were of little value, however, except as showing control by the structure of the building, for in another part of the same building the east-west motion was dominant.

Fig. 1. Map showing the Position of the Fault which caused the San Francisco Earthquake.

In my immediate vicinity the destructive effects were trivial, and I did not learn until two hours later that a great disaster had been wrought on the opposite side of the bay and that San Francisco was in flames. This information at once incited a tour of observation, and thus began, so far as I was personally concerned, the investigation of the earthquake. A similar beginning was doubtless made by every other geologist in the state, and the initial work of observation and record was individual and without concert. But organization soon followed, and by the end of the second day it is probable that twenty men were working in cooperation under the leadership of Professor J. C. Branner, of Stanford University, and Professor A. C. Lawson, of the State University at Berkeley. At that time and for several succeeding days the ordinary means of communication were so paralyzed or overburdened that no messages passed between these two centers of organization; but as the needs of the hour were patent to all, the work was not prejudiced by the lack of intercommunication.

On the third day after the shock Governor Pardee appointed a State Earthquake Investigation Commission, naming as its chairman the head of the geological department of the State University, Professor Lawson, and including in its membership Professor Branner, of the Stanford University, Professors Davidson and Leuschner, of the State University, Professor Campbell, of the Lick Observatory, Mr. Burckhalter, of the Chabot Observatory, Professor Reid, of Johns Hopkins University and Mr. Gilbert, of the United States Geological Survey. The commission held its first meeting three days later, when the scope of its work was considered and defined, provision was made for circulars soliciting information, an announcement was prepared for the purpose of relieving certain groundless fears entertained by a portion of the community, and two committees were appointed for the general work of observation. To the first committee, with Professor Lawson as chairman, was assigned the determination and study of surface changes associated with the earthquake and the collection of data as to intensity, so that isoseismals, or curves of equal intensity, might be drawn upon the map. To the second committee, with Professor Leuschner as chairman, was assigned the collection of data for the drawing of coseismals, or lines connecting points on the earth's surface reached by the shock at the same instant. Some weeks afterward, when the main features of the earthquake had become known, a third committee was appointed, with Professor Reid as chairman, to consider the relations of the earthquake phenomena to certain problems in geophysics, or the science of the inner earth.

The work of these three committees is still in progress, and will not be completed for several months. The actual drawing of isoseismals and coseismals can not be performed until a large body of observations have been compiled and studied, and the geophysical problems are as yet only imperfectly formulated; but of the physiographic phenomena, or the disturbances of the earth's surface, so much is known that it has been thought advisable to prepare a preliminary report. This was submitted to the governor on the third of June, and has been issued as a pamphlet of twenty pages. The expenses of the commission are being met by the Carnegie Institution.

Architects and engineers were not less prompt and energetic. To

Fig. 2. Fault Topography between Tomales and Bolinas Bays; looking northwest. The general slope toward the left has been interrupted by a slight uplift of the part at the left. The pond occupies a hollow thus produced.

the men who plan and direct construction in the earthquake district of California it was important to know what materials and what structural forms best withstood the shock, and they immediately began the study of earthquake injuries and of instances of immunity from earthquake effects. In that part of San Francisco where the earthquake injury was most serious the shock was quickly followed by fire, which destroyed much of the evidence, but many important observations were made in the brief interval. The study of structural questions, like the study of natural phenomena, was at first individual only, but afterward was aided by organization. Committees were appointed by various professional societies, national and local, and were charged with the investigation of specific structural questions, and the results of their labors will find place not only in the transactions of the societies, but in revised building regulations and in important modifications of municipal plants for lighting and water supply. Various bureaus of the national government have also taken part in the structural studies, sending experts to San Francisco and other localities of exceptional earthquake violence.

The Japanese government promptly sent to California a committee of investigation headed by Dr. Omori, professor of seismology in the University of Tokyo, and composed otherwise of architects and engineers. The first conference of these visitors with the state commission warranted the suggestion that we may find it as profitable to follow Japanese initiative in the matter of earthquake-resisting construction as in that of army hygiene.

The following sketch of the physical features of the earthquake is based chiefly on the body of data gathered by the State Commission:

An earthquake is a jar occasioned by some violent rupture. Sometimes the rupture results from an explosion, but more commonly from the sudden breaking of rock under strain. The strain may be caused by the rising of lava in a volcano or by the forces that make mountain ranges and continents. The San Francisco earthquake of April 18 had its origin in a rupture associated with mountain-making forces. A rupture of this sort may be a mere pulling apart of the rocks so as to make a crack, but examples of that simple type are comparatively rare. The great majority of instances include not only the making of a crack but the relative movement or sliding of the rock masses on the two sides of the crack; that is to say, instead of a mere fracture there is a geologic fault. After a fault has been made its walls slowly become cemented or welded together, but for a long time it remains a plane of weakness, so that subsequent strains are apt to be relieved by renewed slipping on the same plane of rupture, and hundreds of earthquakes may thus originate in the same place. From the point of view of the geologist the displacements of rock masses are the primary and important phenomena: the faults are incidental phenomena, of great value as indices of the displacements; and the earthquakes are of the nature of symptoms, serving to direct attention to the fact that the great earth forces have not ceased to act.

A faulting may occur far beneath the surface and be known only by the resulting earthquake; but some of the quake-causing ruptures extend to the surface and thus become visible. The New Madrid and Charleston earthquakes are examples of those having deep-seated origins the Inyo and San Francisco of those whose causative faults reached the surface of the ground.

The general character of California earthquakes was so well known that when the dispatches told of a severe shock at San Francisco no American geologist had a moment's doubt that it was caused by a fault movement, and among those specially conversant with the structure of the affected district attention was immediately directed to several fault lines, with the expectation that one or more of them would show the marks of fresh dislocation. Mr. Ransome prepared a prophetic article

Fig. 3. This Fence, Previously Continuous and Straight, was broken and parted by the earthquake fault, the offset being 812 feet. The line of fault, concealed by the grass, crosses the ground from left to right, touching both the dissevered ends of the fence.

in which he indicated the lines most likely to be concerned.Nat. Geog. Mag., Vol. 17. 1906, pp. 280-296. Professor Branner stated in an interview that he had immediately made a forecast of the locality of the origin and that it had proved to be correct, and Mr. Fairbanks went at once to a zone of 'earthquake topography' with which he was already acquainted, and found a fresh rupture in the expected place.

The San Francisco earthquake was caused by a new slipping on the plane of an old fault which had been recognized for a long distance in California, and in one place had been named the San Andreas fault. Associated with this fault is a belt of peculiar topography, differing from the ordinary topographic expression of the country in that many of its features are directly due to dislocation, instead of being the product of erosion by rains and streams. One of its characteristics is the frequent occurrence of long lines of very straight cliffs. Another is the frequent occurrence of ponds or lakes in straight rows. The tendency of erosion is to break up such cliffs into series of spurs and valleys and to obliterate the lakes by cutting down their outlets or filling their basins with sediment. Fig. 2 shows one of the fault-made ponds. This line and zone have been recognized by California geologists through a distance of several hundred miles. It was to this line Fig. 4. Diagrams Illustrating the Dislocation Causing the San Francisco Earthquake. The upper represents an earth block 100 feet square and 25 feet thick, with indication of the position of the fracture. The lower shows the relation of its two parts after faulting. that attention and expectation were especially directed, and it was on this line that the surface evidence of new faulting was actually found. The new movement was not coextensive with the line as previously traced, but affected only the northwestern portion; and, on the other hand, it extended farther to the northwest and north than the old line had previously been recognized. The accompanying map represents only the line along which the recent change occurred. From a point a few miles southwest of Hollister it runs northwestward in a series of valleys between low mountain ridges to the Mussel Rock, ten miles south of the Golden Gate. Thence northwestward and northward it follows the general coast line, alternately traversing land and water. The farthest point as yet definitely located is at Point Delgada, but the intensity of the shock at the towns of Petrolia and Ferndale probably indicates the close proximity of the fault and warrants the statement that its full length is not less than three hundred miles. South of Point Arena its course is direct, with only gentle flexure, but the data farther north seem to imply either branching or strong inflexion. Opposite San Francisco its position is several miles west of the coast line, and it nowhere touches a large town.

That which occurred along this line was a differential movement and permanent displacement of the rock and earth on the two sides of a vertical crack. The principal displacement was not vertical, but horizontal. If one thinks of the land to the east of the crack as stationary, then the change may be described as a northward movement of the land west of the crack. If the land to the west be thought of as stationary then the land to the eastward moved toward the south. It is probable that both tracts shared in the movement, the eastern shifting toward the south and the western toward the north. Perhaps the nature of the change can be more readily understood by reference to Fig. 4, which represents an ideal block of the earth's crust, 100 feet square on the surface and 25 feet deep, before and after its division and dislocation by the earthquake-causing fault.

Wherever a fence, road, row of trees, or other artificial feature following a straight line was intersected by the fault its separated parts were offset, and an opportunity thus afforded for measuring the amount of change. The measurements range in the main from 6 to 15

Fig. 5. A Faulted Road near the Head of Tomales Bay. The nearer and more distant parts of the road were originally in one line—a continuous, straight road. The present offset is twenty feet.
feet and have an average of about 10 feet. At one place (Fig. 5) a road was offset 20 feet, but in this case the underlying ground was wet alluvium and part of its movement may have been due to a flowing of Fig. 6. Ordinary Appearance of the Earthquake Rift where it traverses Firm Turf. the soft material. There was also some vertical change, but this was not everywhere in the same direction and its amount was comparatively small. At many points the land west of the fault appears to have risen one or two feet as compared with the land at the east.

The surface manifestation is not usually a simple crack, but a disturbed zone a few feet broad, the earth within the zone being split into blocks which show more or less twisting or rotation. In some places the zone is slightly depressed below the adjoining surfaces, and elsewhere slightly elevated. Other disturbances of the surface were associated with the earthquake, but the track of the central fault has a character of its own, a character with which the field workers soon became familiar, so that it could be clearly identified. It came to be distinguished in their conversation and note-books as 'the rift.' For considerable distances the rift is single, but elsewhere it is more or less divided, the parts lying within a few rods of one another and being approximately parallel. There are also branches parting from the main rift at various angles and gradually dying out in the adjacent country, and in some of these the belt of disturbance is broad and complicated (Fig. 7). There are also outlying cracks occurring within a mile or two of the central rift and having irregular courses, and these may probably be referred to the same general system of rock strains.

Other cracks are distinctly secondary in character; that is to say. they are not due directly to the stresses and strains by which the fault was made, but are results of the earthquake itself. The jar constituting the earthquake, or in technical language the earthquake wave, as it travels through rock and earth produces temporary compressions and other strains, and these often occasion cracks at the surface. Where the material is elastic such secondary cracks merely open and close, leaving the ground with its original form; but where it is inelastic and incoherent, as in the case of young alluvial formations and artificial fillings, some of the cracks opened by passing waves do not close again, but remain as permanent vestiges of the shock. Closely associated with these secondary cracks in soft ground are permanent changes in surface form. At the head of Tomales Bay, for example, a broad tract of soft ground between high and low tide was thrown into low ridges, with cracks along their crests, and these remained until destroyed by wind waves. In San Francisco considerable tracts of 'filled' land were shaken together and thus made to settle a few feet, and were at the same time slidden several feet toward the bay (Fig. 9).

Certain changes, very conspicuous to the observer who drove about the country, are closely associated with roads. A side-hill road is

Fig. 7. A Zone of Earthquake Fracture where it crosses a Road near Bolinas.

usually constructed by excavating a notch in the natural slope and piling the excavated material in an embankment at the outer edge of the notch. In course of time, and especially during rainy seasons, the embankment at the outer edge of such a road settles and has to be built up as a matter of repair. Portions of the bluff on the up-hill side of the notch are also apt to fall away, taking the form of small landslides, which have to be removed from the road as a rule after every rainy season. The earthquake precipitated many changes of this sort. Along all side-hill roads in the immediate vicinity of the rift a crack was developed between the embankment and the original soil against which it rested, and this crack often assumed formidable dimensions (Fig. 10); in fact its magnitude was found to be a convenient index of the local violence of the earthquake in regions where buildings are rare. Landslips from the bluffs margining the roads (Fig. 11) were also very numerous, in many instances stopping traffic until repairs could be made. And there were many landslides on a larger scale, the earthquake initiating movements which might otherwise have been delayed for years or even centuries. Some of these landslides fell into streams, dammed their waters and created temporary lakes.

Fig. 8. Cracks caused by the Shaking of Marshy Ground. The comparatively firm road embankment preserved the cracks better than the bog.

Other disturbances of water supply were more directly connected with the earthquake. At several points large volumes of water were squeezed from the ground during the agitation, causing brief but violent torrents, and one of these brought with it so much sand as to constitute a sort of sand eruption. There are reports also that certain springs have received a permanent increase in volume, a result which would naturally follow from the modification of underground circulation by the cracking of rock and earth.

Wherever the shock was specially strong there was considerable injury to trees; some were overturned, others broken near the ground, and yet others broken near their tops. A number of large redwood trees standing on the line of the rift were split from the ground upward, the basal portions being faulted along with the ground they stood on.

In the systematic survey of the earthquake area the relative intensity is being estimated by means of the records of various physical

Fig. 9. Street Scene in San Francisco, showing effect of the earthquake on filled ground. The distant part of the street probably retains its original level and position. Nearer by the ground has settled several feet and has also moved to the left.

effects. In the immediate vicinity of the fault road-cracks and cracks in alluvium are large and numerous; many trees were broken or overturned; there were many landslides; half of the wooden buildings of any village or hamlet were shifted horizontally, often with serious injury; buildings and chimneys of brick or stone were thrown down; during the shock men. cows, and horses found it impossible to stand, and fell to the ground; and some persons were even thrown from their beds. In a general way all these evidences of violence diminish gradually with distance from the fault on either side. The rate of diminution, with exceptions to be mentioned presently, may be expressed by saying that at live miles from the fault only a few men and animals were shaken from their feet, only a few wooden houses were moved from their foundations, about half the brick chimneys remained sound and in condition for use, sound trees were not broken, and no cracks were opened which did not immediately close. At a distance of twenty miles only an occasional chimney was overturned, the walls of some brick buildings were cracked, and wooden buildings escaped without injury; the ground was not cracked, landslides were rare, and not all sleepers were wakened. At seventy-five miles the shock was observed by nearly all persons awake at the time, but there were no destructive effects; and at two hundred miles it was perceived by only a few persons.

Fig. 10. Road Crack caused by the Earthquake.
Fig. 11. Small Landslide on the Uphill Side of a Side-hill Road.
A number of exceptions to this gradation of intensity are connected with tracts of deep alluvial soil, especially if saturated with water, and with tracts of 'made ground.' The great destruction in the low lying part of San Francisco, eight miles from the fault, is directly connected with the fact that much of the ground there is artificial, the area having been reclaimed from the bay by filling in with sand and other materials. The severity of the disaster at San Jose, twelve miles from the fault, has been ascribed to the deep alluvial soil on which the town stands, and many other local peculiarities seem to admit of the same explanation. It is necessary also to distinguish carefully between earthquake intensity and destructive effect, because injury to property was conditioned by mode and material of construction no less than by intensity of vibration. But after making due allowance for differences in natural foundation and for differences in the resisting power of buildings, there remain various anomalies for which satisfactory explanation has not as yet been found. The natural foundation of Oakland is similar to that of San Jose, and its distance from the earthquake
Fig. 12. Water Tank thrown from its Pedestal by the Earthquake.
Fig. 13. School-house at Point Reyes Station, near Inverness; shifted horizontally two and one-half feet by the earthquake. The corner here shown was slipped from the foundation and rests directly on the ground.

origin is about the same, but the injury to its buildings was decidedly less; and Santa Rosa, standing on ground apparently firmer than that at Oakland or San José and having a somewhat greater distance from the fault, was nevertheless shaken with extreme violence.

It is too early to discuss these anomalies. With the data now in hand it seems to be true that there are outlying tracts of high intensity surrounded by areas of relatively low intensity; and these features, if they shall be fully established, will doubtless affect in some important way the general theory of the earthquake.

One of the chief uses of time observations in connection with most earthquakes has been to determine the position of the origin. As the elastic wave travels outward in all directions from the initial point it reaches successively points on the earth's surface which are more and more remote. Coseismal lines, or lines of simultaneous arrival, are, therefore, closed curves circling about the region of the initial fracture. In the case of the San Francisco earthquake this particular function of the coseismals is not required, because the fracture is visible at the surface; but they are not therefore without value. It is not to be supposed that the yielding of the earth occurred at the same instant throughout the entire extent of the fault plane. We should assume, rather, that the fracture, beginning at some point, was extended thence to the remainder of the tract, a certain amount of time being consumed in its propagation. When the time data have been collected and studied, it may be possible to determine at what point the fracture began and at what rate it was extended. It is hoped also that when the time records and intensity records shall have been systematically discussed there may result some conclusion as to the depth to which the fault extended and also as to its subterranean form.

Mention has already been made of the question whether the permanent dislocation or change of absolute position involved in the faulting was divided between the tracts of land on the two sides or was confined to one or the other of them. At first sight it would appear that the only thing susceptible of actual determination is the relative displacement, and that the absolute displacement, or the real movement with reference to the earth as a whole, must remain a matter of theory only. Nevertheless, it happens that in this particular instance the real changes in geographic position are not only susceptible of determination, but are actually to be investigated. To illustrate the problem, let XY represent, in ground plan, a portion of the fault line, and let ABB'C be the original position of a straight line intersected by the fault. Assuming for the moment that the dislocation was equal on the two sides of the fault, then the line AB was carried to the position DE, and the line B'C to the position FG. We may think of the distances BE and B'F as each equal to five feet. The dislocation of five feet pertains to every point near the fault line, but it is not supposable that the same dislocation affects points at a great distance from the fault. At some remote point, for example Z, in the direction B'C, there was no displacement. If B'C and FG were both produced in that direction they would be found not precisely parallel, but would eventually coalesce. How far the undisturbed region Z may be from the fault line is a matter of pure conjecture, but we may plausibly assume that the transverse dimension of the area affected by the displacement Fig. 14. diagrammatic plan of a Portion of the Earthquake Fault, illustrating changes in geographic position. is of the same order of magnitude as the length of the fault line and is measured by hundreds of miles. If this assumption is correct, then throughout a great region in central and northern California all points have experienced a change in geographic position, the change in the vicinity of the fault being of about five feet and the amount diminishing toward the northeast and south-west. If the only determinations of latitude and longitude within this area were of the ordinary approximate character, it would be impossible to measure the changes in geographic position theoretically accomplished by the fault; but it fortunately happens that the region is traversed by two belts of the triangulation of the United States Coast and Geodetic Survey, one being a system of triangles for the control of the coastal map work, and the other the elaborately measured transcontinental belt. The region thus contains several scores of points whose coordinates have been determined with a high degree of precision, and it is possible by the redetermination of these positions to measure the dislocations which have taken place in connection with the earthquake. As all topographic and hydrographic maps of California are dependent for their latitudes and longitudes upon the positions given by this triangulation, and as there is reason to believe that many of these positions have been disturbed by a measurable amount, the superintendent of the Coast Survey has determined to repeat so much of the work of triangulation as may be necessary in order to redetermine the geographic positions. And it is proposed to carry this work far enough eastward to connect the redetermined points with stations that may safely be regarded as quite beyond the effect of the recent fault. When this has been accomplished much light will be thrown on the nature and distribution of the strains which were relieved by the dislocation along the fault line, and it will be possible to say definitely whether the original displacement involved the territory on both sides of the fault or on one side only.

A further cheek is to be afforded through the observations for astronomic latitude at Ukiah. The observatory at Ukiah is between 25 and 30 miles in a direct line northeast of the fault. In connection with the general dislocation it was presumably moved toward the southeast and its latitude diminished by several hundredths of a second. This is one of an international series of observatories established in approximately the same latitude but in different longitudes, for the purpose of determining variations in the position of the earth's axis of rotation. If the observations at Ukiah were studied, alone it might not be possible to separate the result of a small change in the observatory's position from the effects of the migration of the axis; but by combining the Ukiah data with those furnished by the other observatories of the system, it is probable that the effects of the two causes can be discriminated.

The most important practical results of the various earthquake studies will probably be afforded by the engineers and architects, and will lead to the construction of safer buildings in all parts of the country specially liable to earthquakes; but the geologic studies of the State Commission are not devoid of economic bearings. In the city of San Francisco and adjacent parts of the peninsula on which it stands the underlying formations include several distinct types, and the district is so generally occupied by buildings that the relations of the several formations to earthquake injury can readily be studied. Such a study is being made with care and thoroughness, and one of its results will be a map of the city showing the relation of the isoseismals, or lines marking grades of intensity, to tracts of solid rock, to tracts of dune sand in its natural position, to upland hollows partially filled by grading, and to old swamps, lagoons and tidal marshes that have been converted into dry land by extensive artificial deposits. The information contained in such a map should guide the reconstruction and future expansion of the city, not by determining the avoidance of unfavorable sites, but by showing in what areas exceptional precautions are needed, and what areas demand only ordinary precautions.

Another economic subject to which the commission may be expected to give attention is what might be called the earthquake outlook. Must the citizens of San Francisco and the bay district face the danger of experiencing within a few generations a shock equal to or even greater than the one to which they have just been subjected? Or have they earned by their recent calamity a long immunity from violent disturbance? If these questions could be answered in an authoritative way, or if a forecast could be made with a fair degree of probability, much good might result; and even if nothing more shall be possible than a cautious discussion of the data, I believe such a discussion should be undertaken and published. Of snap judgments there has been no lack, and the California press has catered to a natural desire of the commercial public for an optimistic view; hut no opinion has yet been fortified by an adequate statement of the pertinent facts. Among these facts are the distribution of earthquake shocks as to locality, time and severity in California, and also in the well-studied earthquake district of Japan; the relation of the slipping that has just occurred to the geologic structure of the coast region; the relation of other fault lines to the hay district; and the relation of the recent shock to a destructive shock that occurred in 1868. If a broad and candid review of the facts shall give warrant for a forecast of practical immunity, the deep-rooted anxiety of the community will find therein a measure of relief. If a forecast of immunity shall not he warranted, the public should have the benefit of that information, to the end that it shall fully heed' the counsel of those who maintain that the new city should he earthquake-proof. In any case, timidity will cause some to remove from the shaken district and will deter others who were contemplating immigration; but such considerations have only temporary influence and can not check in an important way the growth of the city. The destiny of San Francisco depends on the capacity and security of its harbor, on the wealth of the country behind it, and on its geographic relation to the commerce of the Pacific. Whatever the earthquake danger may he, it is a thing to be dealt with on the ground by skillful engineering, not avoided by flight: and the proper basis for all protective measures is the fullest possible information as to the extent and character of the danger.

  1. Published by permission of the director of the United States Geological Survey and of the chairman of the California Earthquake Investigation Commission.