search   index   by subject   by year   biographies   books  SF Activities  shop museum   contact

Observations of the San Francisco Earthquake
by Joseph H. Harper

Delivered before the Montana Society of Engineers.
January 11, 1908


Some six weeks after the San Francisco earthquake of April 18, 1906, I prepared an article, describing the impressions produced upon my own mind while the movements were in progress, together with a few interesting exaggerations as related by others who had a somewhat similar experience. The recounting and study of these personal impressions has always had a peculiar fascination for me, and reference to the article referred to will show how I was able to account for my own deception in a manner satisfactory to myself, but I often find it quite impossible to understand how the actual movements., as we now understand them, could have produced the fantastic impressions that have been left upon the minds of many.

A very large majority of those present have no clear conception with regard to the movements. and recall the incident as a confused succession of vibrations, jars and jolts that endured for a considerable period, while of those who did receive and can recall their impressions with regard to the amplitude or interval of the larger movements, I have yet to find two individuals, who were apparently similarly situated, who will tell the same story. Not only the physical environment, but the mental attitude of the individual, with the dim light or the absolute darkness in which many were enveloped, have doubtless all combined to deceive the senses and confuse the mind with regard to what was actually taking place. Every recital of these personal impressions but demonstrates anew the utter untrustworthiness of the unaided senses in estimating the amplitude of the movements involved.

One very naturally assumes that a study of the movements of inanimate objects would throw some light upon the amplitude of the shocks delivered, and one with a larger experience might be able to obtain something of value on this point from the record of cracks and displacements that exist on every hand, but I must confess that my own efforts along this line have been somewhat disappointing.

I have spent many days in studying the effects of the earthquake and the fire upon the City of San Francisco, and in studying the effects of the earthquake at San Jose, Palo Alto and other towns about the Bay, and in the cemeteries and country adjacent, and have made repeated visits to each of the places named, without being able to reach any definite conclusion regarding the amplitude of the larger movement, that which is described by many people as a rocking motion, and which is that portion of the quake that inflicted by far the greater part of the damage sustained.

I am forced to conclude that when the earth shocks become so severe as to cause vibrations far beyond the range of our ordinary experience, our senses fail in comprehension, and I feel that no accurate knowledge of the actual maximum can be obtained except from the record of instruments constructed for the purpose, and further that a seismographic record will not be a reliable index of the actual movements except for a very restricted area about the locality in which it is placed... .

Before going further I will relate more carefully than I have heretofore give my experience during the quake and give the results of my personal observations during that time, eliminating therefrom, as far as possible, those features in which I can now recognize that my senses were much confused, and that through them I was greatly deceived.

Within an hour or two after the occurrence I was asked for an estimate of its duration. I was about to reply one minute, when that time seemed rather long and my answer was 50 seconds. At the time many were estimating it at 5 and some even as much as 10 minutes. I now think that I missed something, in the earlier stages, and that from the first to the tenth second was really a longer than I then thought it. There is a chance for some discrepancy in determining the end of the movement, as the vibrations died out very gradually and may have continued longer than a person whose attention was seriously engaged would notice them. There were a number of lighter shocks during the day, but after the one of the early morning they attracted but little attention, and those continued, separated by irregular and increasing intervals that extended from minutes to hours and days, for many weeks before they finally ceased.

Assuming, then, 50 seconds to be a correct measure of its duration from the first warning note to the last perceptible vibration, and using this as a basis, I will subdivided the period, designating the interval by the numbers of the seconds elapsing as counted from the first warning note.

The prelude, or opening, was a very low rumbling noise, like distant thunder, or an ore car far away in a mining shaft, which began at the first second and gradually increased and became more distinct until about the fifth, the sound gradually growing in volume much as it does when a train approaches one through a tunnel. During this five-second interval there was no movement that attracted my attention, and I attributed the sound to some machine in the basement running at an unusual speed.

At the fifth second a vicious intermittent jolting movement began, and in my estimation continued without cessation at any time to the fiftieth second or the end of the quake. This motion appeared to be mainly in a vertical direction, with an amplitude ranging from 1/3 in. to 3/8 in., and moved with a frequency of about 240 per minute, or about 4 cycles per second. These vibrations seemed to rise in pulsations and die down, to be renewed with greater violence at intervals ranging from 2 to 4 seconds each, this change apparently being accomplished by a variation in their amplitude rather than by any change in their frequency. These movements were forceful enough to make everything loose dance and chatter continuously, they were accompanied by a considerable volume of sound, though they did not rock the building very severely, and finally died away in falling pulsations by a reversal of the manner in which they had arisen. I believe this movement to have been continuous in some degree for the entire 45 seconds for the reason that, between the intervals of the more violent oscillating movements which I will soon describe, the rattling of the chandelier fixtures, the windows and the chattering of some loose objects and the marbletop dresser due to this movement could be distinctly heard, just as both before and after the heavier shocks.

At about the tenth second there came a crash as of something shattered, and the falling of the broken parts.

I shall probably never know what caused the crash; my first thought was that the elevator had been run into the sheaves, but a moment later, realizing that an earthquake was upon us, I raised myself to a sitting posture, and was certainly wide-awake from this time on, though I made no further move till all had passed and everything was again quiet.

I have some misgivings with regard to the preceding time intervals as I have given them, but from this moment on I can speak with more assurance.

The vertical vibratory movements I have just described were at this time in full swing and increasing in amplitude and violence at almost every rising pulsation. These shocks, though more violent, did not differ greatly from some I had before experienced, and had this been all might have endured for a full minute and passed without doing any great amount of damage. But this was not all.

From the twentieth to the twenty-sixth second, for an interval of about 6 seconds, there appeared a violent reciprocating movement of tremendous energy, much stronger in its horizontal than in its vertical component, having an interval of about 90 per minute or 1.5 cycles per second, which was superimposed upon the lighter vibrations and while in progress wholly dominated all other movements. An exhibit that aptly illustrates my idea of the manner in which these different vibrations are imposed upon one another may be seen by observing the ripples that always play over the surface of the larger waves when the water is rough.

These heavier shocks were not of constant energy but seemed to progress in three distinct pulsations, each separated by an interval of comparative quiet, the change, as was noted with regard to the lighter vibrations, being effected by a variation in amplitude rather than in the frequency of the movements. For the first three or four shocks the motion appeared to be in a general easterly and westerly direction; then, after a pause of a moment or two, and another change in direction, some three or four additional shocks, the first of which seemed more violent than any of the preceding, were delivered in an easterly and westerly direction, when these in turn died away and disappeared among the lighter movements that I think had been in progress all the time, and that continued to the end of the disturbance.

Words fail utterly when I attempt to convey an adequate idea of the dynamic energy manifested in each of these larger movements, as they appear vicious in the extreme, and forceful to a degree far beyond anything I had ever experienced or imagined as possible. The blows were delivered as intense, instantaneous and resistless shocks, of startling severity and limitless power, crushing everything that offered sufficient resistance, and rending and tearing everything that did not bend to their requirements. The building rocked and swayed in an alarming way, while at every lurch there came the sound of groaning lumber, the creaking of nails and spikes being drawn, the snapping of painted woodwork, all of which, when joined by the rattle and jar of everything loose, filled the room with a confused din that was disconcerting in the highest degree.

For reasons heretofore given, I cannot estimate the amplitude of these larger movements with any degree of assurance, but I cannot see how the effects produced by them in my immediate neighborhood could be accomplished without an actual movement of at least 2 inches, and indeed for some of the exhibits it would seem that considerably more than this would be required.

You will understand that what I have written has special reference to what occurred in our own apartments, that is, on the fifth floor of a very well-constructed frame building standing on a good brick foundation at 808 Bush Street, but I imagine it will fairly describe the experience of a large majority of those who were similarly situated about the city of San Francisco.

In estimating the magnitude and intensity of the movements in any locality, one must depend largely upon the character of the effects they produced, and these when found are generally more of less illusive, and very unfortunately are often quite contradictory.

The lighter shocks, those whose main component was vertical, I imagine were substantially the same in all parts of the city and were but slightly modified at elevations above the ground, while with the larger of horizontal movement a marked difference seems apparent between different points in the same locality, and I feel quite certain they were often greatly magnified, or much modified, dependent upon the character of the structure in which the observer was situated and his elevation above the ground. Generally they appeared to be amplified in some degree on the upper floors of most buildings, but it does not follow that the higher the floor the larger they became, for if the building were tall enough, one or two nodes would usually appear in the vibrations, which must greatly modify the character of the movements.

The point where the reversal took place was made manifest by the shattering of a story or two, while at other points the walls would escape comparative injury. There were so many buildings to be seen about the city with one story, usually between the fourth and eighth, badly damaged, while those above and below were apparently unharmed, that this feature became quite noticeable. It is generally believed that the shocks were much more severe on what is known as made ground, or land that has been reclaimed by filling in along the water front, than they were in sections where a rock foundation was obtainable. It is true many buildings on an elastic foundation suffered severely, but it is also quite true that many frail structures so situated were marvelously preserved. My impression is that the most intense shocks were delivered from the firmest foundations been greater where a more elastic foundation or method of construction prevailed. I think the shocks lacked something of the dynamic energy that is evinced where the foundations were more solid or the structure less elastic.

Volumes might be written descriptive of the havoc wrought by this terrible visitation which in a few moments wrecked the city and started the fires which eventually destroyed it, shattered the nerves of a large majority of its population and profoundly impressed all by a momentary exhibition of limitless power, but I have time only to note a few cases which occur to me as most interesting.

One is early impressed by the abundant evidence of the most erratic damages, and still more unaccountable immunities, that are exhibited side by side on every hand. In our own apartments (on the fifth and upper floor) the motion was certainly severe, as a steam radiator that naturally stood 2.5 inches from the wall was thrown over until it inclined at an angle of 25 degrees from a perpendicular, though the iron fittings by which it was connected were so rigid that I found it impossible to replace it unassisted, and yet in these rooms nothing was broken and no other furnishing was seriously disturbed, while in the rooms below nearly everything movable was overturned, many objects were broken and much damage done. Mrs. Harper explained this by saying that in our room almost everything stood out from the side walls while in the rooms below many of the furnishings were set close against them, and in this remark I have found an explanation for much apparently fitful and capricious destruction noticed in all parts of the city. In the simple conditions here presented, with the results observed, will be found a key that will explain the collapse of the City Hall, the destruction of much property in San Francisco and the fall of many buildings at Palo Alto, Agnew, San Jose and other points about the bay.

The lesson to be drawn, and to which I shall again refer, when briefly stated, is that objects must be separated by an interval so great that they will not collide, or they must be bound together so they will move as one.

Marble-topped furniture tables in particular, seemed to be marked for destruction. and they were, literally by the hundred, fired from the walls across the apartment and were very frequently disfigured or broken. The fire mantels and marble wainscoting were interior decoration for the destruction of which the shocks appeared to have a peculiar penchant, and when we recall the manner in which those fittings are usually fastened, it is easy to understand how the first movement might separate them from the wall and the second send them shattered and broken far into the room. In like manner a plastered surface, if at all loose, would be easily broken and thrown into the room, but I think firm walls of good mortar usually came through without serious damage. Though I have noticed many marvelous escapes, bric-a-brac was usually scattered in wild confusion, small flat objects were often completely overturned, pictures were thrown down, while the spectacle of one, though two or more feet across, with its face turned to the wall, was not an unusual occurrence.

I have found the cemeteries a fruitful field for the study of eccentric movements. In Laurel Hill [cemetery in San Francisco] I noted two similar granite shafts, separated by least than 200 feet, one of which had been turned to the right upon its base some 15 to 20 degrees, while the other, upon the same style of a foundation, had been turned as far to the left.

Among many of like character I noted a monument, with pedestal, shaft and capital, in which the shaft had turned in one direction upon the pedestal, while the capital had turned in the opposite direction upon the shaft on which it rested to a much greater extent, and both movements were attended with an adverse lateral motion of the members.

There is no apparent regularity in the direction in which the upper stones have moved upon the lower. A northwesterly and southeasterly direction may be most frequent, but I have noted these movements toward every point of the compass. There was nothing in the movement that gave me an impression that it was rotating, but a number of witnesses have stated that among the other movements they detected a circular or twisting tendency. I think the tendency to rotate is the resultant of two or more components from the other movements, and the manner in which they were combined in any particular locality probably determined the direction in which the object affected would turn. There remains, however, a difficulty in understanding how a capital can turn in one direction while the shaft upon which it rests is turning in the opposite direction upon its foundation.

The most perfect illustration of this tendency to rotate which I have anywhere noted was seen in a brick chimney about 2 feet in height, and the only one that I could see standing in that locality, which made a quarter turn to the right by the movement of each individual brick upon the one on which it rested, with their final arrangement almost as perfect as could be secured had a mechanic been employed to place them there.

Though I believe a measurement of the horizontal component of the earth’s movement would greatly exceed any instrumental record that I have ever heard mentioned, there are many tall, slender and apparently unstable objects about the city that still stand erect and uninjured.

In spite of their amplitude and complexity it is still possible that a person standing upon one foot in the middle of a room would have felt the movements less than one who had a larger surface in contact with the floor, as in the end, without question, the sum of them all practically balanced one another, and the ultimate motion, if there were any, was probably small and the new position gradually attained.

I note three familiar objects that illustrate this feature. one being the Dewey Monument in Union Square, the second, the Native Sons’ Monument at the entrance of Mason and Turk into Market Street, and the third, two columns which many will remember at the southwest corner of the City Hall, at the intersection of Grove and Larkin streets, which stood for many months capped across by a span of the architrave which originally rested upon them. The two objects last named both illustrate the universal character of the movement, as the projecting cove at the base of the Native Sons’ shaft had been spalled and broken off quite impartially in all directions, while the ornamental castings at the base of the City Hall columns had been broken on all sides.

Further curious and interesting effects of the fire and earthquake might be described, as they are in evidence at every hand, but features of more practical value to the engineer and structural artist relate to the character of material and methods of construction that can best withstand a constant war with the elements and an occasional assault of earthquake, fire or flood.

It is now apparent that, from the moment of the shock, which occurred at 5:15 a.m., the destruction by fire of a large portion of the business section of the city was foreordained, for before leaving my room at about six o’clock I had noted four separate and distinct fires, and there were really six, I have since been told, all of which looked threatening in the extreme and were apparently at that time beyond control.

I am not equipped to discuss in a technical manner the stresses and strains involved, or the size and proportion of members required to secure stability in large municipal structures of recent years, as my practice has usually lain along other lines of work, but my experience in San Francisco, and my observations since, have developed a few persistent notions of what it is desirable to attain, and some regarding the conditions it is necessary to avoid in buildings designed to stand the shocks of an earthquake and endure the fires that may follow.

There are two properties in available building material, rigidity and elasticity, that we may invoke, and these may be employed to produce two very dissimilar structures, either of which will sustain shocks without injury, but the ideal building will result from a judicious combination of the two rather than from the exclusive use of either. A granite block and a rubber globe will both endure the roughest usage without injury, but the first accomplishes the result almost wholly by transmission. while the latter largely absorbs the force of the blow.

To build strong enough is a solution of our difficulties in but one direction, for not only should the structure stand, but it should possess a property that will absorb the energy, modify the movements and thus in some measure protect the occupant and the interior furnishing. In small buildings elasticity is of comparatively slight importance, and in all timber work is naturally attained without special effort, but in larger structures, and particularly in all tall buildings where this property seems most desirable, the best results cannot be secured unless careful provision therefor be made in the design. A large part of the immunity from damage observed in wooden buildings is doubtless due to their inherent elasticity, and to the fact that many of them were placed upon a frame underpinning from 2 to 4 ft. in height, usually lightly enclosed and often left entirely open. This style of foundation is much in vogue in California, and in San Jose, where the quake was probably stronger than in San Francisco, and where I saw many such buildings, I cannot recall one that had sound walls and was well built that sustained any serious damage unless the foundation had failed, and in that case they were without exception very badly wrecked.

For tall buildings the modern steel frame is the only style of structure to be considered, and in the San Francisco test they gave a very excellent account of themselves.

As usually designed, the attachment between the posts and girders appears weak, and the bracing between all vertical and horizontal members entirely insufficient to safely withstand the burden of a shock, but I have a suspicion that the large riveted corner plates and the rigid diagonal bracing now being used on some of the buildings under construction is a departure on the opposite extreme.

Much is claimed for "reinforced concrete," and if concrete be properly reinforced it is without question a most excellent material, combining in a marked degree the desirable properties of strength and elasticity, but I cannot avoid an impression that this material is just now being badly overworked. Concrete in dams, bridges, retaining walls, foundations and all sub-work fills an important field, and in recent years its intelligent reinforcement has enabled us to economize greatly in quantity and to use it for many purposes for which it is not otherwise suitable, but I do not feel that any method of reinforcement that I have yet seen will render it a desirable material for the construction of supporting members in buildings that are more than four, certainly not more than six, stories in height.

Concrete construction is usually monolithic in character, and in this respect is admirably adapted to withstand earth shocks, but one naturally speculates on what might occur if an important vertical, say one of the lower columns, should cleave on a plane approaching a diagonal through the member. With a rupture of this character we would hardly expect the reinforcement to carry the entire shear strain, and a very serious situation is presented.

A column of brick similarly situated would possess the virtue of failing slowly; it would disintegrate on a series of vertical and horizontal planes, and will usually carry its burden for a considerable time though badly shattered.

The argument that concrete is cheaper and can be placed in position with ordinary labor has largely disappeared when it is used in forming the important members of a large building, for in addition to generous reinforcement, the best of selected material must be used and a care in every detail of construction realized that cannot be attained without employment of men with large experience and special training in such work.

But poor material cannot be held responsible for anywhere near all of the failures that occurred in and about San Francisco, but poor workmanship was responsible for many, and a radical defect in the design directly responsible for some of the most lamentable. A heavy dome upon a light skeleton of steel, projecting far above the body of a building, unless securely bound thereto by a substantial and elaborate system of bracing, is a form of architecture that simply courts disaster.

The imposing City Hall in San Francisco and the barely completed gymnasium at Stanford were both examples of this character, and both were seemingly wrecked much in the same manner, by battering themselves down through unequal vibrations between the buildings and their respective towers. The assertion is frequently made that the material used in the City Hall and the buildings at Stanford was very poor, a remark that caused me to examine the work with more than usual care. The material is well above the average in grade, most of the mortar used was of good quality, some in fact very excellent; and I found none worth noting at either point that was positively poor.

One would naturally anticipate that in a region where earth-shocks were not infrequent special care would be taken to bond all work securely, but here a different policy seemed to have prevailed, for I have never seen as much poorly bonded work, nor such a quantity of work without even a pretense of bonding, as was exposed in San Jose and Palo Alto, and in general about the city and bay of San Francisco.

A style of construction much in vogue with the architects here, and used extensively upon many large buildings, consisted of a veneer or facing of dressed stone backed by a wall of brick, concrete or rubble work, but instead of being thoroughly bound together, the two courses were often continued for a story or more without the use of a single tie that I was able to discover, while what the mason work is termed a "header" was apparently unknown.

Much damage resulted from overturned chimneys, the falling of cornice and other ornamental members and the throwing of firewalls upon adjoining property, but danger from this source will be very greatly reduced on the buildings now going up, for as reconstruction proceeds one can note many provisions for supporting and anchoring these ornamental projections that were formerly regarded as unnecessary.

After many large fires, engineers and architects have often found it difficult to determine what proportion of the damage done to stonework should be attributed to the fire and how much to the water that was used, but in San Francisco abundant opportunity has been offered to study the effects of fire without water upon structural material, and in my judgment the use of water can add but little to the destructive effects of a great conflagration.

As in Baltimore, Chicago and other severe fire tests, so in San Francisco, the common ordinary brick of good quality, I think, comes through with a far better record than was made by any other material in general use. The opportunities for observing it after a fire test were not as numerous as one would expect, owing to the abnormal proportion of wood that entered into the construction of the city and owing again to the indifferent character of both brick and mortar so frequently used, which failed to support the walls after the floors and the beams had been burned away. There were, however. a few large and well-built brick buildings in the city, and so far as my observation goes, every wall of good common brick well laid in cement mortar stood through both the earthquake and the fire without serious damage... .

To build stronger than your neighbor is a safe proposition unless he is entertaining the same idea, but the spirit is too selfish to be taken seriously. The necessity, however, of paying some attention to what is being done on adjoining property was abundantly manifest at San Jose and other towns as well as in the city, for the crushing of a low or weak front, or the forcing of it into the street, by the pressure of taller or stronger structures on each side, was a manner of failure very frequently observed, while on the other hand, if the lower building was substantial enough to hold its own, the taller structure would be badly damaged in that portion projecting above the general level.

The conditions above named, with the inherent elements of weakness in conspicuous evidence, are so frequently encountered in all business districts that I would like to suggest some easy way of avoiding trouble, but the only efficient method that occurs to me is to leave a small open space between the buildings, and this plan, I presume, will generally be rejected. A small space, an inch or less, would generally be sufficient, and there are many ways in which it could be covered and made to appear respectable.

Much damage was accomplished by the thrust of rafters and truss members that threw the top of side walls out and caused them to fall. It seems hardly necessary to suggest that an effectual remedy for this consists in strengthening the chords and securing them at the foot of the trusses in a manner that they may sustain any thrust that may come upon them, thus making the truss a substantial tie and an element of strength instead of weakness.


Return to the 1906 Exhibit

Return to top of page