Popular Science Monthly/Volume 70/May 1907/The Jamaica Earthquake




MAY, 1907




WITHIN nine months three regions in the western hemisphere, geologically closely akin but geographically distant one from the other, have been visited by earthquakes, causing an appalling loss of life and property. In all cases the disasters have been preceded by minor earth-shakings for years, and the areas were known to be in zones of earth-unrest. No warning, however, unless the tremors that occur at irregular intervals every month or two could be counted as such, has characterized these last disturbances. But these tremors must be regarded as the climax of a long-continued yielding to strain which has resulted in a series of minor breakings. This faulting culminated in a great fracturing of the earth's crust and a consequent destructive earth-shaking. The kindred conditions of these different areas appear to be, first, a considerable amount of differential relief only obtained where mountains are associated with marine depths; and, in the second place, where newer and less compacted sediments occur upon these slopes.

For several months previous to the afternoon of January 14, 1907, there had been no noticeable increase in the number or intensity of the customary slight shocks that occur in the Island of Jamaica every month or two. In Weather Report IV. of Jamaica, Mr. Maxwell Hall

PSM V70 D390 Map of the jamaica earthquake.png
PSM V70 D391 Commodore bartlett relief map of the caribbean.png
Fig. 2. Photograph of Comd'r Bartlett's Relief Map of the Caribbean Sea.

has noted some twenty-six minor shocks that occurred from 1880 to 1886, and this number might be regarded as typical of the seismic phenomena in that region. A slight shock was noticed by many in November last, but the memories of the destruction of Port Royal by the historic earthquake of 1692 had been dulled by the interval of two centuries, and the Jamaicans had begun to think themselves in a region of comparative safety. Slight tremors and shocks caused but scant attention or notice on the part of a few of the people. Consequently, when the real cry of 'wolf' came, for the first second or so but few realized the danger. The slight tremor, however, instantly increased to a terrible vibration of the earth that threw down great walls and buildings and inside of a minute transformed the city of Kingston from a prosperous metropolis to a place of destruction and mourning.

In order to appreciate their relative importance and possible influence upon seismic activity, let us notice the topographic, geologic and bathographic conditions that exist at Jamaica.

The etymology of the word Jamaica, originating in two descriptive Indian words meaning 'well wooded and watered' and modified by the Spaniards to 'Xaymaca,'[2] is interesting, taken in connection with the historic topographic description of the island given by Columbus to Queen Isabella on his return from the West Indies—'a crumpled handkerchief picked up by the middle.'

The aptness of the simile can not be questioned when one sees the many steep knife-edged divides (typical 'bad-land' topography) rising abruptly in fifteen miles 7,400 feet to the misty Blue Mountain peaks that tower above the small inland valleys or the narrow plains that
PSM V70 D392 Parish church tower shattered and inclined.png

Fig. 3. Tower of Parish Church shattered and inclined to the East. Stopping of clock by shock.

fringe the seashore. These plains constitute the very small percentage of the island that is fairly level, and it is upon these plains that the larger towns and the larger plantations of bananas and sugar-cane are found. These level areas are made up of alluvial deposits, fans or sheet-wash brought from the adjacent ragged slopes by the rivers in flood time. Upon the rather bare slopes, occasional rectangular patches of light green show the location of small banana farms or 'pens.' But the more abundant and typical tropical verdure is found lower down on the fringing plains. The island has long been known for the dance and variety of its tropical and subtropical products, due to the fertility of the limestone soil and the abundance of the rainfall, which varies largely, however, in the amount, from 10 inches at Port Royal to 126 inches some years in the higher regions.

Geologically, Jamaica is of comparatively recent age,[3] for its basal Blue Mountain series of sediments and intrusives is of late Cretaceous and Eocene times. This series makes up the mountainous backbone of the island, while the later Oligocene limestone overlaps the former series in a thick piedmontal formation covering two thirds of the island. The more recent alluvial and littoral formations were deposited during the period of uniform elevation following, and constitute the fringing plains of the island.

In the structural geology of Jamaica, the earliest axis of folding now evident is the northwest-southeast line of the Blue Mountains, with later east-west foldings along the more ancient line of orogenic movement which outlined the Greater Antilles in early Mesozoic times.[4] The writer has observed transverse faults in the Blue Mountain region, which undoubtedly indicate lines along which fracture may occur.

M. de Ballore[5] coincides with Mr. Hill's ideas regarding an eastwest folding for the Antilles in postulating his theory of an anticlinal axis that marks the line of the Greater Antilles and a parallel synclinal belt immediately to the north of Jamaica, which coincides with the Bartlett Deep. In the photograph of a relief map (Fig. 2), the east-west elevation and depression are brought out strongly.

The bathographic relations of Jamaica are significant. We see that Jamaica and the other Antillean islands are but the higher peaks of a lofty and precipitous, but submerged, mountain chain. The tremendous differential relief of over 38,000 feet that exists in places in the Caribbean region apparently coincides with a zone of seismic and volcanic frequency. We know that the crust of the earth is always in a state of tension. This stress may come from the shrinkage of the earth, from the loading or unloading of the earth's surface through erosion or deposition, or from other sources. The resistance is lessened on a relatively steep slope (Fig. 1, b) where the points of application of this lateral pressure at the ends, not falling in the same plane, tend to produce a fracture. When a sudden slip in the adjustment occurs, the resulting jar is transmitted through the earth as earthquake waves.

Port Royal is at the western tip of a narrow seven-mile sand-spit that makes a natural breakwater to one of the finest harbors in the

PSM V70 D394 Northward view of the street in the business section.png

Fig. 4. Street in Business Section looking North. Western covered with débris. Crack in fireproof building, showing angle of emergence waves.

West Indies. When the town was for the most part submerged by the earthquake of 1692, this favorite site was abandoned for the Liguanea plain just across the harbor, and Kingston was founded on the largest of the fringing plains of loosely compacted sands and gravels. And here in this closely built city of 60,000 (and at Buff Bay opposite on the north shore) the destruction by the last earthquake was felt most keenly. Eighty-five per cent, of the buildings were injured or destroyed. Then came Kingston's old enemy, fire, and swept over ten or fifteen blocks of the business and warehouse section. (Figs. 4 and 5.)

The earthquake shock that brought disaster to the island of Jamaica began, according to the regulator of Mr. J. A. Soulette, at 3:33 p. m. Others record its arrival two or three minutes earlier. In various

PSM V70 D395 The narrow harbour street of the burned district.png
Fig. 5. In Burned District; the Narrow Harbour Street, looking east.
places on the island, as reported by local times, its occurrence varied from 3:20 to 3 45 p. m. In the investigation it was found impossible to plot any coseismal lines, for the reason that no accurate coordinated time exists in the island. Since the shock, however, there has been a movement on foot in Kingston to establish a system of accurate timekeeping throughout Jamaica. The shock lasted about thirty-five seconds, varying in length with the location and geological position of the observer. At the east end of the island some noted a duration of sixty seconds; on the north shore a length of ninety seconds, while at other points near by the duration reported was anywhere from five to forty seconds. The slight preliminary tremors were felt immediately before the main shock, and the noise and roar was heard slightly before the coming of the major vibrations. One man, used to earthquake
PSM V70 D396 View along the general penitentiary brick wall.png
Fig. 6. Looking North in General Penitentiary Brick-yard. Upper third of brick chimney, upper part of smokestack (to guys), and upper ring of lime-kiln; together with large part of the brick wall thrown to the east. Fissures made in ground in lower part of view.

countries, hearing the sound from the preliminary tremors, rushed out-of-doors into the street only to be thrown down toward the west by the violent shaking. He dragged an injured companion a hundred feet or so during the slight lessening of the violent shock, and then felt the second climax of a slow undulating character pass underneath. This experience is like the phenomena of double earthquake shocks which have come to Jamaica in past years, and also has characterized many of the sequent shocks. Another man repeated his actions and found that he could jump through the fallen wall of the house and then over

PSM V70 D396 View of the shattered landing place of the general penitentiary.png
Fig. 7. Looking North to Landing Place at General Penitentiary. Massive walls cracked;;a section of the eastern wall was demolished. Upper part of chimney overthrown to east.

a low fence and get into the street in about forty seconds. The increase and decrease of the tremors are so gradual that it is very difficult for an observer to tell just when the shock comes and when it ends. From the majority of the testimony it is evident that in this disaster the movement quickly reached the major climax in about ten seconds, then lessened in intensity for about ten more, then gently swelled to a second and minor climax and disappeared in a total of about thirty-five seconds. While there were apparently. no preliminary shocks at Jamaica, there have been many sequent vibrations of the earth, more or less severe. The press has chronicled one on February 23, which was the strongest since the earthquake, and another one also was noted on March 22. Mr. Maxwell Hall[6] has noticed some eighty shocks after the main shock on January 14 to February 5, several of them shaking the whole island, while others were of local extent. On the early morning of January 28 one small shock awakened me instantly by a slight shaking of my cot in the tent in which we were sheltered. The continuance of the motion gave one a sense of insecurity and unsteadiness, and brought with it a slight tinge of dread and nausea. My first impression upon waking was of a rushing, whistling sound from the southwest; it increased and passed overhead, rapidly lessening and disappearing. It was very similar in sound to the approach and passing of a large flock of ducks flying low. Then from the race-course, only a quarter mile distant and only a short time quieted, came the cries of the frightened negroes and the howls of the numerous dogs with which Kingston is cursed, and the crowing of the many roosters in the trees—as they did about every hour during the night. The shock felt on board the moving Port Antonio train produced a feeling as if the coaches were running upon the sleepers and at the same time swaying so much that it seemed as if they would topple over to the southeast. No damage, however, was done to any of the rolling stock or to the roadbed. In none of the many tunnels was any displacement observed. A man driving on the road suddenly felt his vehicle thrown in an angling position across the road and it seemed difficult for the horse to keep its footing. It was observed, however, that motion sometimes counteracted the vibration of the ground and made the latter imperceptible.

The sketch map (Fig. 1) shows by the isoseismal lines the relative intensity of the shock at Kingston as compared with other places on the island. It has seemed rather strange that the most intense destruction should happen to occur just where a large number of buildings are found. But in the case of Kingston, the gravelly foundation in proximity to the epicenter readily accounts for the destruction.

PSM V70 D398 Gables destroyed by the shock.png
Fig. 8, a. East and West Gable Ends destroyed by Shock.
From the data available, the dependence of earthquakes in intensity upon topography is well emphasized. Loosely compacted fringing and alluvial plains extended the intensity farther than the more compact and elastic mountain regions. Not only do these less elastic plains give a greater amplitude to the waves and cause greater destruction, but apparently the earth-waves are affected by plains indented in hills as sea-waves change their direction in entering the arm of a bay. The arrows (Fig. 1, a) indicate generally the direction of the wave motion. In the middle of the Hope River Valley at Mona Plantation an observer noticed the motion pass him and then saw the landslide occur at the mouth of the river to the southward. As the wave passed over the cane-fields, a motion was observed similar to that produced in a field of grain by the wind. The direction here was at right angles to the path of the wave-motion only five miles away at Kingston, situated on the western slopes of Long Mountain. The motion approached the island from the southwest, changing on the land its direction and intensity with the change in the nature of the material through which it passed. In the lower part of the city of Kingston the path of the movement was well marked by the overthrowing of walls, piers, statues, monuments, large chimneys and a similar movement toward the east of even large marble slabs covering graves (Figs. 3-9). Northward from the city the motion appeared to come more from the south, and the northern walls showed the greatest damage; and westward, the path of motion appeared to swing so that it came from Kingston. The absence of any large buildings, away from the villages and cities, made the plotting of directions rather difficult, for the lightly-built mud-wattled huts were not affected by the shock and tests by hearing are very unreliable. But there was a general diminution in intensity away from Kingston;
PSM V70 D399 Crushed east and west walls.png
Fig. 8, b. East and West Walls crushed.

decreasing rapidly eastward, less rapidly westward and still less so to the north. Haiti did not feel the shock, neither was it felt at Colon or at Grand Cayman, 175 miles west, but Santiago, 120 miles north, experienced a slight shock.

Cracks in buildings, which at Kingston dip some 50 degrees east, are always perpendicular to the path of the emergence of earthquake waves. Hitherto, the intensity area and epicenter have been regarded as synonymous. But the dip of the angling cracks at Kingston points to a locus of disturbance much to the west of that city, while the lines of isoseismals indicate the intensity area in the eastern half of Kingston. It may readily be imagined, then, that the area of greatest destruction may not be directly above the focus. Suppose a highly elastic rock is there situated, and some distance away is found a plain of loosely-formed material. The destruction in the latter area will far exceed that in the former in spite of its favorable location. Until we register the actual amplitude, wave-length and period and, with the elasticity of the rock underneath, calculate from the more readily-discerned data on adjacent but less elastic media the changes that have occurred in the wave-motion, it will be difficult to determine with accuracy in a region of rocks of widely varying elasticity the location of epicenters. For outliers of rock in plains must deflect, refract and reflect wave-motion and even shadow areas in these plains. The only conclusion then is that the eastern end of the Liguanea plain was the nearest area to the real epicenter that by nature of material would give the greatest amplitude to the destructive epifocal waves. Further, the angle of emergence at Kingston coordinated with the proximity of a probable epicenter, together with the limited area of disturbance, indicates a shallow origin of about three miles.

PSM V70 D400 Crushed walls of the jamaica club.png
Fig 9. Showing the East Walls of the Jamaica Club. Crushing of the walls by the roof and of the first story by ceilings and partitions. The majority of the first stories remained intact.

PSM V70 D400 Damaged statue of queen victoria.pngFig. 10. Statue of Queen Victoria, twisted an Eighth Turn Counter-clockwise from the South.

The line of intensity of the earthquake destruction apparently extended to a greater distance northward than to the east or west. For at Buff and Annotta Bays on the north shore, the destruction was but little less than at Kingston. Furthermore, the shock was felt at Santiago to the north and not at Haiti to the east or on land to the west of Jamaica. The inference is that the locus of the disturbance originated in a line of north-south faulting rather than in an area of less linear extent. The north-south fault-lines extending throughout the island, as noted before, and some probable fault lines extending in a similar direction through Cuba (marked by sharp valleys) may indicate in a general way the direction of possible faulting at the present time. It might be noted that this line of faulting lies at a considerable angle with the general trend of the Antillean folding. The beautiful mountain road from Kingston to Newcastle was in the line of greatest intensity. But though spurs showed considerable destruction and in places the road slipped off the face of the steep slopes (Fig. 11), or portions of the hills slipped down on the road carrying it away or obliterating it by landslides in many places, yet the destruction was caused more by the unstable position of the road, or of these masses of earth, rather than by the intensity of the shock. At Newcastle, moreover, the buildings for the most part were not damaged to any great extent, except as their location on a terraced slope or on the crest of a short divide would place them in a position of unstable equilibrium. Similar destruction might be caused by a severe rainstorm,

PSM V70 D401 Destroyed carriage road to newcastle in the blue mountains.png
Fig. 11. Destruction of the Beautiful Carriage Road to Newcastle, built on the steep slopes of the Blue Mountains.

or, in the northern countries, by frost action as well as by earthquake waves.

From the investigation of the many cracked walls at Kingston, the amplitude of the wave motion (as one might expect on alluvial foundations) was considerable. Spaces from half an inch to two inches were left in massive walls. Floors and ceilings were pulled from the shallow supports in many cases and caused destruction in more instances than would have been necessary had there been greater foresight used in the manner of building. From an open circular well of masonry some twenty feet in diameter water was thrown up some eight feet and over the northeastern lip of this well. A brick pier in a fence was thrown to the eastward beyond its arc, some two thirds the length of its radius. At the same place large slabs of marble were moved along
PSM V70 D402 Fissure belt at the base of the palisadoes.png
Fig. 12, a Looking East along Belt of Fissuring at Base of the Palisadoes, showing one of several parallel fault planes in the sand, with craterlets of mud.
PSM V70 D402 Closeup of the faulted belt.png
Fig. 12, b. Looking Southeast across Faulted Belt.

on their cement base to the eastward some three inches or more in spite of the attendant friction. The amplitude was probably less than an inch at Kingston.

The speed of the various waves in this earthquake can only be approximated. During a slight shock that occurred afterwards, of about one third the intensity, from an interrupted telephone conversation from Kingston to Port Antonio, it was estimated that the wave traveled about two thousand feet per second. As yet no data have been available concerning the breaking of the cables, and as to the exact time or speed as marked by such fractures. The Panama cable was broken in two places, one four miles and the other some twenty miles offshore from Bull Bay, but so covered was it with débris that a couple of miles
PSM V70 D403 Closer view of the submergence at port royal.png

Fig. 15. Nearer View of Submergence at Port Royal, looking south. Most of the area now covered by water in the photograph was formerly land.

or so of the cable had to be abandoned. The preliminary tremors were heard before being felt and probably were slower than sound-waves. With the increase of speed that comes with the augmentation of intensity of earthquakes, it is probable that the rate of the major vibrations was about ten thousand feet per second.

As has been previously stated the shock was a double one; the first climax apparently came from the west, while the second one, less disputive and more undulating in its character, apparently came more from the southward of Kingston. These two directions of vibration resulted in an almost universal gyratory movement of columns, statues, piers, sections of brick chimneys, and even of buildings, in a counterclockwise (Fig. 10) fashion.

Geologically, earthquakes often are not very important. In the case of the earthquake at Jamaica, however, there apparently was a

PSM V70 D403 Twisted rails and tilted buildings at the victoria battery in port royal.png
Fig. 10. Twisting of Rails and Tilting of Buildings in Victoria Battery, Port Royal, by Subsidence.

zone of Assuring and subsidence from a hundred yards to three hundred yards in width (Fig. 1, a, AA). It started at the western part

PSM V70 D404 Submerged palms and battery casemates of old port royal.png
Fig. 14. Looking East toward Port Royal Showing Submerged Palms and Battery Casemates (in line with tower). Old Port Royal formerly extended northward from the point at the left.

of the city of Kingston, ran along the water front encircling the harbor, and continued along the line of the Palisadoes, reaching its greatest destructive effect at Port Royal. One arm of this Assuring followed up the River Cobre to the carriage road. From soundings taken by the kindness of Mr. Charlton Thompson, harbor master, it was ascertained that in several places along the edge of the harbor, the bottom had sunk from old soundings of a fathom and a half to over six fathoms, and that on the harbor side of the base of the Palisadoes a series of step-faults reached a maximum depression at the shore to the north of four fathoms (Figs. 12 and 13). This zone of disturbance continued, as far as could be traced, in an interrupted line along the Palisadoes, and caused a maximum depression at the western tip of Port Royal, where the buildings were tilted by the sinking and a hundred yards or more of land were submerged to a depth of from eight to twenty-five feet (Figs. 14-16). This Assuring of the earth was caused by the repeated tearing apart and closing of the earth's crust, accompanied generally by the ejection of water, sand and mud, sometimes to the height of three or four feet, but the subsidence prevented the forming of any cones about these craterlets. The sands first thrown up were afterwards covered by a layer of mud.

To account for the unique line of Assuring and subsidence is difficult. It was noted that considerable disturbance took place at the shore line where the earth vibrations were refracted in changing from the medium of one elasticity to a medium of a different elasticity. But the middle portions of the harbor were stable and the channel was unchanged, though a beacon light near Fort Augusta was broken off. In this limestone country, solution by underground waters might be sufficient to account for the sinking of a small area like the harbor ac Kingston. But the harbor did not sink—only a small encircling zone, and that located either on the shore or slightly offshore. The continuous tearing apart and closing of these fissures, covering a few hours' time as it did in some instances, might account for the hydraulicking of the loosely compacted sands and gravels in the zone of Assuring, and allow subsidence. Again, ground-waters may have caused considerable solution of the limey constituents where the waters entered the harbor. No theory as yet satisfactorily accounts for this peculiar subsidence. At the eastern end of the harbor at Rock Fort a considerable change in underground drainage was observed, where a small spring was increased to a stream eight feet wide and six inches deep.

It was here at the Rock Fort penitentiary quarry that a guard gave me the only reliable account of a sea wave. After a few moments had elapsed and the convicts had run from the landslides on the face of the quarry and gathered around him for protection, the sea retreated for a hundred feet and then advanced inward upon the shore about sixty feet in a low wave a couple of feet high. Ocho Rios, near
PSM V70 D406 Width of sunken belt between palisadoes and kingston.png

Fig. 13. Looking towards Kingston, across Harbor from Base of Palisadoes, showing width of sunken belt. Soundings of four fathoms were taken where the tree-tips emerge from the water, formerly near the old shore-line.

St. Anne's Bay, on the north shore, also had its harbor emptied for about seventy-five yards, after which a small incoming wave was followed by gradually lessening oscillations. A careful search ten days later along the other places of the harbor and coast line, however, revealed no trace of any sea wave, even of slight degree.

Thanks to the energy of the department in charge of the waterworks and to the good fortune that caused no important breaks in the system, Kingston was shut off from its water supply for only two hours. Some of its cement reservoirs situated near a large wrecked school building showed no damage. The pipe that carries the city's sewage eastward to the sea at the base of the Palisadoes, however, was broken at several places along the zone of Assuring, and its linear extent, like that of the water pipe along the Palisadoes, was marked by rifting in the earth. A prompt repairing of the breaks in these two systems undoubtedly saved the city from an outbreak of destructive pestilence.

Arches in buildings apparently withstood the shock to a notable degree, whether transverse or parallel to the line of the earthquake motion. Generally when built in houses they preserved the parts around them. The Institute, a building in which some two hundred delegates had assembled in the first session of the West Indian Agricultural Conference, is built on two lines of arches at right angles to each other. The Institute was damaged, but withstood the shock. The great destruction of brick buildings in Kingston was doubtless due to the fact that poor mortar and dry bricks were used in the construction. The mortar generally appeared to be rather porous and usually the cracks in the wall followed the mortar, though at Up Park Camp, where the bricks were laid in cement mortar, the cracks passed through the bricks.

The streets were narrow (Figs. 4 and 5), so that the falling wall of even a two-story building would block the street, and many persons escaped from falling buildings only to be crushed in the choked narrow streets. A cement floor may help preserve a building from destruction. In many cases it could be seen that if the floors had been well tied to the walls and the walls themselves held at the corners, a great lessening of the destruction would have resulted. On account of the white ants foreign woods are, unless creosoted, difficult to use, but some frame houses showed but the slightest effect of the earthquake shock. The 'barrack' or 'noggin' structure, much used in earthquake countries, apparently suffered nearly as much as other brick walls.

Jamaica lies in a region of great differential relief and consequent stress. The earthquake was confined in its area of greatest destruction to small limits upon alluvial detrital material, where the amplitude was increased to bring about this effect, varying with the heterogeneity of material. The origin of the shock was comparatively shallow and the earthquake was local in character. While there was a general distinct rotary motion induced by two components of the vibrations, the major component came from a westerly direction. There were few evidences of sea waves, but there was a unique zone of Assuring and subsidence about the harbor of Kingston. Finally, the disasters at San Francisco, Valparaiso and Kingston should teach the lesson that in the case of cities located in a danger zone (where there are many recurring shocks of slight degree), there is always a possibility of the coming of a disastrous shock; that certain types of buildings should be built and streets laid out with that possibility in mind; that water, sewage and lighting systems should be planned in sections, and that as far as possible a city should not be located nor large edifices erected upon uncompacted rocks and soils.

  1. The writer desires to acknowledge his indebtedness to Dr. Charles D. Walcott, formerly director of the U. S. Geological Survey, and to J. D'Aeth, assistant director of Public Works; Mr. Maxwell Hall, resident magistrate; Mr. Charlton Thompson, harbor master, and to many other official and private citizens of Jamaica for their cordial cooperation and aid in the prosecution of the investigation.
  2. 'Handbook of Jamaica,' 1906, p. 23.
  3. 'The Geology and Physical Geography of Jamaica: a study of a type of Antillean development,' Robert T. Hill, Bull. Mus. Comp. Zool., Vol. XXXIV., Geol. Series, Vol. IV., September, 1899, p. 421.
  4. Ibid., p. 164.
  5. 'Tremblements de Terre,' F. de Montessus de Ballore, 1906, Fig. 63.
  6. Personal communication to the writer.