Ostrasz and Kehrberg-Ostrasz (2020:402) produced a stratigraphic chart
All these phases in the history of the building were witnessed by the stratigraphical composition of the fill over, inside and outside/along the architectural remains of the monument. In no place inside and along the building were found more than four superimposed distinct layers of fill. Everywhere the upper one was the sedimentary layer composed of greyish dirt, usually a score of centimetres thick. This layer is labelled Stm.0.
Underneath there was the layer of the tumbled masonry. Depending on the place, and on the extent of the stone robbing activity, this layer was from 1m to 4.5m thick. It was composed mainly of the fallen dressed stones of the superstructure of the cavea but often also of a proportion of the dress stones of the outer and transverse walls, and in every case of boulders and stone chips which the builders of the hippodrome used for the construction of the walls (infra:...). All the stones were found immersed in red clayish earth which the builders used as a kind of `mortar' of the masonry (loc.cit). This layer - almost everywhere the main one in bulk - is labelled Stm.1.
In some chambers of the cavea (and in all the stalls of the cavea) the layer labelled Stm.1 lay directly on the `floor' of the chambers (stalls). However, in most chambers there was an intervening layer between the bottom of Stm.1 and the `floor'. In some chambers, or in some places of one chamber, this layer was composed either of greyish soil or of this kind of soil mixed with red earth or the red earth only. This layer of the fill was always associated with intrusive structures built in the chambers or with traces of intrusive activity. This layer is labelled Stm.2.
The lowest layer is the bulk of the red clayish earth of which the builders of the hippodrome formed the platform of the arena and the walking surface around the building and with which they filled in the space within the foundation walls of the chambers. The `floor' of the chambers was just the top of this red earth fill [see n.9]. This lowest layer is labelled Stm.3. In no chamber was there found evidence for any kind of true flooring ascribable to the primary structure of the hippodrome. In chambers E41-E53 the `floor' is the unlevelled surface of rock [see n.8, I.K.].
The hippodrome was already quarried for stone by the end of the 4th C. A number of its seat stones was used for rebuilding (repairing) a stretch of the city wall, which according to an inscription mentioning the event and its date took place in 390 (ZAYADINE 1981a, p. 346).Ostrasz and Kehrberg-Ostrasz (2020:315) report evidence that potters and other craftsmen took over the structure starting at the end of the 3rd century CE. Ostrasz and Kehrberg-Ostrasz (2020:142) suggested the possibility that an earthquake had damaged the structure to such an extent that it could no longer be used for racing.
It is clear that the SW part of the cavea had collapsed at a certain date and that once this happened no races could be held. This occurrence would best explain the reoccupation of and quarrying for stone in the hippodrome. There is no direct evidence for dating the collapse of that part of the cavea but it is tempting to associate it with the earthquake of 363 which affected many sites in Palestine and NW Arabia (RUSSELL 1985, p. 39, 42). This earthquake has not been attested at Jerash so far but the study of the earthquakes which affected Gerasa is only in its infancy.The suggestion of seismic damage stemmed from earlier publications which was later revised by Ostrasz and Kehrberg-Ostrasz (2020:150) where they state that
the building ceased to serve the primary purpose  because of the disintegration of a large part of its masonry and of the arenawhere
the disintegration was caused by the extremely poor foundation of the structure.Foundation problems, including estimates of foundation pressures, are discussed in detail in Ostrasz and Kehrberg-Ostrasz (2020:157). An E-W cross section of a part of the Hippodrome illustrates potential foundation problems
Ostrasz and Kehrberg-Ostrasz (2020) discuss evidence of an "earlier" earthquake to the mid 8th century earthquake; the latter of which produced a significant amount of clear archaeoseismic evidence in the eastern half of the carceres. They indicate that damage observed could have been due to an "earlier" earthquake or stone dismantling (human agency). Ostrasz and Kehrberg-Ostrasz (2020:4) report the following:
The final destruction of the building was caused by earthquakes. The masonry of most of the building collapsed during the earthquake of 659/60; only the carceres and the south-east part of the cavea survived that disaster.Ostrasz and Kehrberg-Ostrasz (2020:36) discussed this possible archaeoseismic evidence further
The presence of the stones belonging to the upper parts of the building used in the passageway of the gate in the period of the intrusive occupancy (supra: THE MAIN GATE) and the presence of the architrave pieces in chamber E2 used there in the same period concurs to strengthen the possibility that before an earthquake finally destroyed the north part of the building there might have occurred an earlier earthquake which partly destroyed the masonry at its upper level. Still, the human factor (dismantling) cannot be ruled out.Ostrasz and Kehrberg-Ostrasz (2020:60) discussed possible archaeoseismic evidence from an "earlier" earthquake again reporting
that before an earthquake ultimately destroyed the gate, the upper parts of the hippodrome were either dismantled or partly destroyed by an earlier earthquake.The assigned date of 659/660 appears to based on earthquake catalog matching. Since Ostrasz and Kehrberg-Ostrasz (2020:4) assign the latest date for activity that preceded the "earlier" earthquake to the 6th century and Ostrasz and Kehrberg-Ostrasz (2020:33) provided a terminus post quem for the following event as the first half of the 8th century, it would seem that archaeological evidence constrains the date of the "earlier" earthquake to the 6th to 7th centuries CE. note.
That the structure was destroyed by an earthquake is evident from the position of the fallen stones in the lowest layer of the tumble; nothing but an earthquake could make the masonry fall so. The amount of the fallen stones in the whole tumble shows that most of the masonry of the structure fell northward, onto the arena. Moreover, there is also evidence for the process itself of the fall. In this respect it has to be noted first that the standing remains of the carceres, that is to say the piers between the stalls, all stand at least two, but none more than three masonry courses high (originally the masonry of the stalls consisted of thirteen courses). Some stones in the standing masonry are slightly shifted from their original position but none was noticed to have lost its verticality. In all, the lowest parts of the masonry of the piers were little affected by the earthquake.Ostrasz and Kehrberg-Ostrasz (2020:29-30) discussed the layer below the earthquake tumble.
The case of the upper parts (originally seven masonry courses high, the course of the imposts of the archivolts included is different. Only one pier (3E/4E) of the east stalls provides full evidence for how its masonry collapsed but it can be maintained (infra) that its example is representative of the situation which, during the earthquake, was found also in the case of the others. All the stones but one of the four upper masonry courses of the north face of the pier (stones 73-82) were found in the tumble. The stones of courses 4-5 (lower) fall closest, immediately against the face of the pier, the stone of course 6 (higher) slightly further from it, and the two stones of course 7 (uppermost) yet further from the pier. The pattern of the falling of the stones of this particular pier is clear. The higher the position of the stones in the masonry the further from the pier they fell. A similar pattern is noticeable in the position in the tumble of the three stones identified of pier 4E/5E (stones 84 - course 3, and 90-91 - course 7) and there is an identical pattern in the tumble of stones of the north face of pier 4W/5W (stones W113, W132, W133-135, W137, courses 4-7). This pattern indicates that the earthquake disturbed fatally not only the static balance of the structure but that it also created the force which projected the masonry (particularly its whole northern vertical layer) forward that is to say northward.
This projecting force is best evidenced by the tumble of the masonry which made up the upper part of the north façade of stalls 1E-4E (courses 8-13, from the level of the spring stones of the archivolts to the level of the crowning cornice). While in place, this part of the façade was about 23m long and 3.3m high, and its surface was about 75m2. After the fall, it covered an area of almost the same length, width (former height) and surface. In the process of falling, it described in the air a curve very close to a quarter of a circle of which the radii of the particular masonry courses were approximately concentric and of which the centre was approximately at the level and face of the top of course 3 of the piers. While the masonry of the north façade stood intact, the top of the comice course was 5.4m, the apex of the archivolts 3.6m and the spring stones of the archivolts were 2m above that level. After the fall, these elements lay at a distance of 5.5 - 6.5m, 4 - 4.4m and 2 - 2.5m, respectively, from the façade. Figuratively speaking, the whole vertical layer of the masonry making up the north façade fell from the vertical to the horizontal position just as a solid platform of a drawbridge would fall, its hinges being at the level of about 2m above ground.
Two factors contributed additionally to this pattern of collapse for which the earthquake was, of course, instrumental. One was the tectonics of the piers and especially of the upper parts of the carceres. As all other parts of the hippodrome, they were built of dressed stones on the outside while the inside was filled with boulders and stone chips set on earth. In consequence, the masonry was not cohesive in its entirety; a slightest disturbance of the static stability of the structure could (and did) immediately detach the dressed stone facing from the inner `core' of boulders, stone chips and earth. The other factor was the physical condition of most stones in the lowest courses of masonry of the piers. As in the case of the lowest courses of masonry in most parts of the hippodrome, these stones deteriorated in a much greater degree than the stones of the upper courses (for the reasons cf. infra:...). They lost most of their resistance to pressure of the masonry above; any movement of the structure combined with the pressure of that masonry could not fail to make them disintegrate instantly.
All the above considered, the process of collapse can be reliably reconstructed. The earthquake caused the structure momentarily to lean forward (northward). In that instance and in that position two things occurred simultaneously: the force of gravity made the masonry of the north façade detach itself from the inner core and the deteriorated stones making up the lower courses of the face of the piers gave way, as the support for the upper parts of the façade. In this situation the masonry could not fail to collapse. However, the gravity force alone could have made the stones of the masonry fall roughly vertically and in a rather haphazard order. They did not fall so. Instead, they described in the air a part of a circle and fell `orderly' and far from their vertical position. This shows that apart from the force of gravity there was another force, the force which catapulted the stones first horizontally before the force of gravity `pulled' them down onto the ground. This ejecting force must have been created in the moment of leaning of the whole structure forward and this shows in turn the leaning occurred instantaneously and violently.
Considering the fact that the structure fell northward it must be assumed that during the earthquake the ground under the structure moved upward at its south side and/or downward at its north side in a split second and with a great force (speed). That movement made the structure lean violently which created the force catapulting the stones forward. This force naturally increased in direct proportion to the height of the structure as is clearly witnessed by the position on the ground of the fallen masonry of the upper parts of the north façade of the carceres. To make it all happen as it happened, the earthquake must have been extremely strong.
The fallen stones show the direction of fall of the carceres. It has been observed that `During an earthquake the columns, pilasters, and walls of structures have a tendency to collapse in the opposite direction of the quake's epicenter or hypocenter.' (Russel 1985: 51-52) Accordingly, the directional pattern of collapse of the carceres indicates that the epicentre or hypocentre of the earthquake which destroyed the structure was to the south of Gerasa. The reconstruction of the process of the collapse points to a forceful earthquake. The recent studies of the earthquakes in the region of Palestine and northern Arabia from the 2nd throughout the 16th century elucidate the stronger and weaker earthquakes known in that period and region. Accordingly, both phenomena - the directional pattern of collapse and the strength of this earthquake - are, then, additional evidence (beside the deposit sealed by the tumble) for dating the occurrence (infra).
The stone tumble contained no ceramic or coin deposits. It was only the excavation of the top layer of the ground underneath the tumble that yielded the ceramic and coin material (Compendium B: Kehrberg 1989, 2004 and 2016a). The surface of the ground sealed by the tumble in front of the stalls was about 140m2 (about 7m by 20m). This surface was not level, that is to say it was not the original top surface of the arena.Ostrasz and Kehrberg-Ostrasz (2020:31-32 also discussed earthquake collapse in the western half of the carceres (stalls 1W-5W) where, for a variety of reasons, archaeoseismic evidence was not as rich in details but where most of the collapse, as with the eastern stalls, fell northward. Ostrasz and Kehrberg-Ostrasz (2020:33) provided a terminus post quem of the 1st half of the 8th century CE for the archaeoseismic destruction and suggested that one of the mid 8th century earthquakes was responsible.
Ceramic deposit. (see Compendium B: Kehrberg 1989-2006, fc 2018)
Stm.2, Stm.3, and possibly Stm.1 - 1600 potsherds, 2 intact lamps and 62 lamp fragments. Most pieces are fragmentary and worn, especially the lamp fragments. A very small proportion of the material (%)20 dates from the lst throughout the 3rd century, the bulk (%) dates from the 4th throughout the 6th century, and the remainder (%) dates to the 7th and 8th centuries. In the first group, the proportion of the sherds and lamp fragments dating to the 3rd century is the least. In the second group, the proportion of the material dating to the 4th, 5th and 6th centuries was found to be roughly equal, respectively, and so was the material in the third group dating to the 7th and 8th centuries.
Finally, the excavation yielded evidence for dating the collapse of the carceres. The latest potsherds and lamps found in the area sealed by the tumble are of the Umayyad period. The latest coin underneath the tumble is datable to the first half of the 8th century. The sealed deposit contained no artefacts of a later date. Of all the material, the coin provides the relatively strictest terminus post quem for the destruction of the carceres - the first half of the 8th century. The terminus is based on the evidence ex silentio of the material of a date later than of the first half of the 8th century, but this evidence can securely be accepted as reliable considering other parts of the monument (supra....).
Ostrasz (1989) found archeoseismic evidence at various parts of the hippodrome which they attributed to a mid 8th century CE earthquake.
The archaeological context of the excavated sections of the cavea was found to be the same almost everywhere. On the outside of the remains of the outer and podium walls, and contiguous to them, was the stone tumble of the upper parts of the walls. The inside of the chambers was filled mainly with the tumble of the stonework of the cavea proper (seat stones and voussoirs of the stepped arches which supported the seating tiers) and with a number of stones of the outer wall. In many chambers the position of the stones displayed clearly that the stonework collapsed during an earthquake. The tumble was subsequently quarried for stone. The quarrying was very extensive; only a small proportion of the stones which made up the particular parts of the masonry was left in the tumble. The parts of the masonry which survived the disaster were also robbed of stones.Ostrasz (1989:137-138) discussed the chronology of destruction.
The stratigraphy of the fill in the chambers was very simple. In most chambers there was only one stratum (from 2 to 4 m thick) over the `floor' level: masonry tumble composed of dressed stones, boulders and rubble, all immersed in earth. 7 The tumble lay directly on the `floor' which in chambers E40-E55 is the unlevelled surface of rock and in all others the top of the fill within the foundation walls of the chambers. The fill itself is another, the lowest stratum. Is is composed of thick layers of earth and thinner and irregular layers of stone chips. In some chambers there was an intervening thin layer of earth and rubble between the top and bottom of the two strata mentioned above. The tumble outside the outer wall lay on top of a residual layer from 0.3 m to 0.8 m thick. Underneath, there is the same kind of earth with which the space within the foundation walls of the chambers (and the arena) is filled. The masonry tumble outside the podium wall lay directly on the surface of the arena. 8
The archaeological context of the carceres was very similar to that of the cavea. On both sides of the remains in situ and contiguous to them, as well as inside the staffs, there was the tumble of the upper parts of the masonry destroyed by an earthquake (fig. 4). Most of the masonry collapsed northwards, on to the arena. The bulk of the tumble was not disturbed by quarrying for stone and every stone retained its tumbled position. The tumble lay on the surface of the arena.Fig. 4
Lowest stratum of tumble
The excavated sections of the hippodrome displayed clearly that the building was finally destroyed by an earthquake. The best attested examples were found in the carceres, in chambers E40-E43 and E25-E28 (currently under excavation), and in the neighbouring church of Bishop Marianos. The coins and the ceramic material from the deposits sealed by the tumble provided evidence for dating the occurrence. No material dating beyond the Umayyad period was found in any of the deposits. The latest coin from the deposit under the tumble of the carceres is datable to the first half of the eighth century and the latest ceramic material found in it dates to the eighth century (Kehrberg 1989: 88). The latest coins recovered from under the tumble in chambers E40, E41, E42 and E43 were minted in 383-395, 498-518, 575/6 and between 527 and 602, respectively. The latest pottery, lamps and lamp fragments from the same deposits date to the seventh century. The only coin found under the tumble of the church of Bishop Marianos was minted in the first half of the eighth century and the objects are dated to the same period (Gawlikowski/Musa 1986: 149-153).
The finds prove that the south-east part of the cavea stood high in the seventh century and the carceres and the church still stood high in the first half of the eighth century. The lack of material dating after the middle of the eighth century shows that this part of the building was either abandoned or destroyed at, and never occupied after, this date. The archaeological context of the finds in the church clinches the matter. It shows that...the church remained in use to its end.(Gawlikowski/Musa 1986: 141), that is until the earthquake which must then have occurred about the middle of the eighth century.
Only one earthquake is securely attested in the region of ancient Palestine in the eighth century and this is the earthquake of 748 (747) (Russell 1985: 39, 47-49). It is also well attested at Jerash (Bitti 1986: 191-192; Crowfoot 1929: 19, 25; id., in Kraeling 1938: 221, 242, 244; Parapetti 1989a: passim; Parapetti 1989b: passim; Rasson/Seigne 1989: 125, 151; Seigne 1986: 247; Seigne 1989: passim). The hippodrome of Gerasa is yet another well attested example of that disaster.
Arch damage at the Hippodrome is evident from various photos taken during excavations
Seismic Effects include
It is clear that the SW part of the cavea had collapsed at a certain date and that once this happened no races could be held.
Possible seismic Effects include
The masonry of most of the building collapsed
there might have occurred an earlier earthquake which partly destroyed the masonry at its upper level. Still, the human factor (dismantling) cannot be ruled out.
the upper parts of the hippodrome were either dismantled or partly destroyed by an earlier earthquake.
On the outside of the remains of the outer and podium walls, and contiguous to them, was the stone tumble of the upper parts of the walls.
The inside of the chambers was filled mainly with the tumble of the stonework of the cavea proper (seat stones and voussoirs of the stepped arches which supported the seating tiers) and with a number of stones of the outer wall.
masonry tumble composed of dressed stones, boulders and rubble, all immersed in earth
tumble of the upper parts of the masonry destroyed by an earthquake
Most of the masonry collapsed northwards, on to the arena
The amount of the fallen stones in the whole tumble shows that most of the masonry of the structure fell northward, onto the arena.
In all, the lowest parts of the masonry of the piers [of the carceres] were little affected by the earthquake.
Figuratively speaking, the whole vertical layer of the masonry making up the north façade fell from the vertical to the horizontal position just as a solid platform of a drawbridge would fall, its hinges being at the level of about 2m above ground.
apart from the force of gravity there was another force, the force which catapulted the stones first horizontally before the force of gravity `pulled' them down onto the ground. This ejecting force must have been created in the moment of leaning of the whole structure forward and this shows in turn the leaning occurred instantaneously and violently.
|Collapsed Walls||VIII +|
|Collapsed Walls||VIII +|
|Collapsed Walls||VIII +|
|Collapsed Arches||VI +|
Ostrasz and Kehrberg-Ostrasz (2020:146-147) reporoduced an earlier article by Antoni Ostrasz in 1991 which reports on the discovery of skeletons beneath collapsed masonry which they tentatively attributed to an earthquake in Late Abbasid/Early Mamluk time. This was corrected in the 2020 report - see the final bracketed paragraph below.
An unexpected, and to say the least, dramatic discovery was made in the course of excavation in chamber W2. The upper part of the chamber was (and its lower part still is) filled with tumbled stones of the cavea (mainly the seat stones and voussoirs of the stepped arches). Human skeletal remains were found under the removed upper part of the tumble and within the tumble. This is not the case of a burial. In the north-east corner of the chamber, in an area 1.5m by 1m large and at approximately the same level, were found five skulls, all cracked, with parts missing. Directly over the skulls there were hand and arm-bons, even rib-bones and at the level of the skulls lay some vertebrae. In this area and at this level no pelvis or leg-bons were found. In the middle of the chamber there are remains (left in place) of another skeleton. In the extreme opposite part of the chamber, close to the podium wall, there were recovered from under and from within the tumble the pelvis, leg, arm and rib-bones (all at approximately the same level) of at least two individuals. No skulls were found above or beside these remains. There are, then, the skeletal remains of at least eight individuals discovered so far in the chamber. The lower part of the tumble was left in place to be excavated in the spring of 1991.
There seems to be only one plausible explanation [but see comment below, I.K-O] for the condition in which the skeletal remains were found: the individuals were killed by a sudden collapse of the cavea and such a collapse could be caused by nothing else but an earthquake. The five individuals in the north-east corner and the one in the middle of the chamber were obviously caught by the disaster inside the chamber. However, the two individuals whose remains were found in the opposite part of the chamber seem to have been surprised by the earthquake while being in the cavea and seem to have caved in the chamber together with the tumble; their skulls may be found in the lower layer of the tumble.
So far, there is no evidence for dating the occurrence. It is expected to be found when the occupation level of the chamber is reached. [see below, I.K-O] However, some tentative suggestions may be advanced already at this stage.
The earthquake occurred in the period of reoccupation of the hippodrome. This is evidenced by a well preserved intrusive doorway built within the original doorway of the chamber - a feature found in most excavated chambers of the building (Ostrasz 1989a: 55 and Fig. 2). The terminus post quem for the reoccupation is a date in the first quarter of the fourth century or, possibly, even slightly earlier (supra) and this is the terminus post quem for the disaster. However, a much later date should be considered. In 748(647) AD ab earthquake destroyed the south-east part of the hippodrome (Ostrasz 1989a: 75) but considering the situation found in chamber W2 it seems rather dubious that this earthquake was responsible for the collapse of the masonry of the chamber. The fact that the bodies of the people killed in this disaster were not recovered from the rubble for burial bespeaks a period of a great decline of the Gerasene community in every respect. What is presently known of the history of Gerasa in the last decades of the Umayyad period is not compatible with such a degree of decline.The recent students of the history of Gerasa tend to view Gerasa of the Umayyad period as an important urban centre. A tendency of overstressing the importance of Gerasa in that period is detectable but there can be no doubt that Gerasa of the Umayyad times was still a centre of some substance. For an early view on the subject cf. Kraeling 1938: 68-69. Of recent studies cf. in the first place Gawlikowski (in press and 1986: 120-121). Also: Bitti (1986: 191-192), Schaefer (1986: 411-450); Zayadine (1986: 18-20; Naghawi (1989: 219-222).43The date of this earthquake may, therefore, be as late as a date in the Late Abbassid or even the Early Mamluk periods.A sedentary community at the site of ancient Gerasa is attested to have occupied, perhaps intermittently, the North Theatre in the Late Abbassid and Mamluk periods. Cf. Bowsher, Clark in F. Zayadine (ed.), Jerash Archaeological Project 1981-1983, I. Amman: 237, 240-241, 243, 247, 315. The situation found in chamber W2 fits a picture of such an occupation rather than that in the earlier periods. [ see above comment, I.K-0]44.
[We completed excavation of W2 and W3 in 1993 retrieving conclusive evidence correcting the preliminary interpretation for the cause of death posited in this article; see Ostrasz 1994, and Compendium B: Kehrberg and Ostrasz 1997; 2016b, for the dating and identification of the event: the mass burial of about 200 mid-seventh century plague victims. The tumble relates indeed to the 748 earthquake, I.K.]