The building
The building is located along David Street, the main market street and the spine of the city, at the very heart of the Old City, where the four quarters meet (Fig. 1). On the north, it borders with the Saint John Muristan Garden of Peace and the courtyard of the Martin Luther School, and on the west—Muristan Street (Crown Prince Frederick Street). The current entrance to the building is from the south, through one of the arched openings of the bazaar along David Street. The building served as a fruit and vegetable market until 2000, but nowadays it is underutilized, as only part of it is used as a storeroom for the adjacent shop.
The building (650 sq m; Fig. 2, 3) comprises a large, one-storied hall, with a small, second-story unit on its northwestern side that belongs to the Order of St. John and was used to house an ophthalmic hospital between 1949 and 1960 (Figs. 2:A; 4); the floor separating the two stories has was removed recently. The hall consists offour longitudinal bays divided into 11 units by 17 massive, square and quadrangular piers (I–III, V–VII, IX–XI, XIII–XVIII, XX, XXI) and three or four elongated rectangular piers along the south perimeter of the hall (VIII, XII, XIX and possibly IV), which support the roof. The southwestern unit of the building is a closed room with the tomb of a wely, Sheikh el-‘Abayen (الشيخغباين; Fig. 2:B).
The hall is roofed with eleven groin vaults with pointed arches, more or less uniform in height (4.5 m from the floor to the crown of the vaults). The vaults are constructed with small and irregular stones bonded with whitish mortar; the majority of the vaults were coated with plaster when first built. They are organized in three rows and are separated from one another by shallow, transverse arches. The transverse arches between Piers VI and VII, and VI and X protrude downward about 0.5 m beyond the surface of the vault, possibly indicating repairs following the collapse of the original module (Fig. 5). Three round and square openings at the haunch of the vaults (oculi) in the eastern part of the building allowed air and light to enter the hall (Fig. 6).
The piers and the transverse arches bear features typical of Crusader masonry: large ashlars with high quality surface-finish, some still bearing the diagonal chisel marks and mason’s marks despite the many layers of plaster and paint that covered them over the centuries. The three eastern piers (I–III), placed about 5 m apart, are capped with a quirked cavetto impost molding, i.e., a concave molding with a curve from which the vaulting springs (Fig. 7). Piers X and XIV were renovated—parts were dismantled and then rebuilt as door jambs—perhaps in a later period (Fig. 8). The lower part of several of the piers exhibit small drilled tethering rings.
The hall is paved with rectangular dressed limestone flagstones of various sizes set in rows (Fig. 3). The enclosing walls of the hall (see Fig. 2), spanning between piers and rising up to the top of the arches, are very likely to be a later construction, as they are poorly built of various stones with earth fill, and abut the piers; these walls contain windows and openings (e.g., Fig. 4).
The Excavation
A narrow strip (2× 8 m; Fig. 9) was opened along the east face of Pier VI, in the eastern part of the building, where the tile floor (L1) was undisturbed and sealed the lower archeological layers. The excavation reached a depth of over 3 m beneath Floor 1 without reaching bedrock.
Stratum III. Five courses of Pier VI were revealed below Floor 1 (depth 3.5 m; Fig. 10). The pier courses are built of large, well-dressed stones (0.5 × 0.5 × 1.0 m), implying that these were not foundation courses, but rather part of the pier itself and visible above the original floor level. Two partially exposed walls abut Pier VI: W12 on the east and W13 on the north. Both walls are built of medium-sized field stones interlaced with small stones and bound together with earth; the southern face of W12 was coated with a thin layer of gray plaster. Wall 12 (length 2.5 m, width 1.3 m, preserved height 2.2 m), running east–west, continues east towards Pier II, and probably abuts it (Fig. 11). Wall 13 (length 0.9 m, height 2 m; Figs. 9: Section 2–2; 11; 12) runs north–south in the western edge of the excavation; its western face was not exposed, and therefore its width is unknown. As it continues northward, it likely abuts Pier VII. These walls are thus either contemporaneous with the piers or of a later date.
A water channel (L18), only a small part of which was exposed (0.9 × 0.9 m), abuts the northern face of W12 (Fig. 12). Built of small field stones bonded with grayish mortar, it was inserted into a fill (L17), which contained pottery shards, among them two carinated bowls (Fig. 13:1, 2) and jars (Fig. 13:8) from the twelfth and thirteenth centuries CE, and an Ayyubid coin (1171–1250 CE; IAA 145782). The fill above the water channel consisted of brown earth with small stones (L9) and contained a large amount of metal slag, animal bones and pottery shards. These included a carinated bowl (Fig. 13:3) and a cooking pot (Fig. 13:7) from the mid-twelfth century CE, as well as carinated bowls (Fig. 14:1, 2), a flask (Fig. 14:6) and a tobacco pipe (Fig. 14:8) from the Ottoman period (sixteenth–nineteenth centuries CE).
At a depth of 2 m beneath Floor 1 was a stone pavement (L27; Fig. 15) of irregular slabs with grayish mortar that abutted W12 on the south. A small probe under Pavement 27 revealed a layer of gray fill (L28) above a layer of fieldstones (L29), which served as a foundation for W12 and the pavement. A coin-like lead object (Fig. 16; IAA 145774) was found in Fill 28. It bears depictions on both sides: a fisted hand on one side, and a face on the other. Two tiny piercings indicate that this object was likely attached to a garment or some other object; its date is unknown. Fill 28 also contained a fragment of a cooking pot (Fig. 13:6) from the twelfth century CE and a coin from the late Roman period (324–497 CE; IAA 145790). A second floor (L25) abutting W12 was laid above Pavement 27. Floor 25 is a whitish plaster floor (thickness c. 7 cm) dotted with small stones. Both floors were damaged by pits dug in Stratum II (L14, L16; below), thus preventing us from assessing the exact connection between pier VI and Floors 25 and 27.
Stratum II. Floor 25 was covered by fills (L23, L24) containing medium-sized stones and pottery shards, including a jar (Fig. 13:9) from the eleventh–twelfth centuries CE; slip- painted ware, including a bowl (Fig. 13:4); a frying pan (Fig. 13:5); a jar decorated with a geometric design (Fig. 13:10); a vessel commonly dubbed a ‘grenade’ (Fig. 13:11) from the twelfth–thirteenth centuries CE; and a lamp (Fig. 13:12) from the thirteenth–fourteenth centuries CE. The fills also yielded two Ayyubid coins from the reigns of Saladin and Al-‘Adil Saif al-Din Abu Beker (1174–1198 CE; IAA 145786, 145787). Also found in the fills were the bones of cattle (Bos taurus), sheep and goat (Capra/Ovis) and a jaw of a rare specimen—a parrot fish (Scarus persicus) from the Red Sea. One of the cattle bones exhibited a pathology: it was enlarged due to repetitive stress applied to it over time. Such repetitive stress could be caused by hard duty labors, such as plowing.
Two superimposed plaster floors (L20, L22) were found above Fills 23 and 24. Two waste pits (L14, L16) cut through these floors, through Fills 23 and 24 below them and penetrated down to Floors 25 and 27. They contained large amounts of metal slag and an antler of a red deer (Cervus elaphus) bearing cutting marks. Since no datable finds were found, we suggest that the pits were installed during the Ottoman period.
Floors 20 and 22 were sealed by a bright red fill (L11), possibly the traces of a tabun or of a furnace, as suggested by the large amount of metal slag it contained. The fill also yielded two Ayyubid coins from the reigns of Al-Zahir Ghazi and Al-Nasir Saladin Yusuf (1169–1216 CE; IAA 145783, 145785), and one coin from the Mamluk period from the time of Al-Nasir Muhammad (1329–1330 CE; IAA 145784).
Fills containing earth and small stones (L4, L6, L10) were found above Fill 11. These contained pottery, including a basin (Fig. 14:4) and a jug (Fig. 14:5) from the Ottoman period (the sixteenth–twentieth centuries CE), and a Mamluk coin (IAA 145781). A round installation (L5), constructed of small field stones, was built into these fills. A coin dating from the reign of Al-Nasir Muhammad (1293–1341 CE; IAA 145780) was found within the installation. Above Fill 4 was a grayish fill with fieldstones (L3), which covered W12. Fill 3 contained a fragment of a pipe from the Ottoman period (Fig. 14:7) along with three Mamluk-period coins (IAA 145776, 145777, 145779) and a large quantity of horse-like (Equid) and sheep/goat (Capra/Ovis) bones; one of the horse bones exhibited a pathology caused by pressure on the horse’s pelvis.
Stratum I. The pavement still in use today (L1) comprises square stone tiles of various sizes, which covered the entire compound, abutting the piers and sealing the fore-mentioned fills and floors. A fragment of an Ottoman-period bowl (Fig. 14:3), as well as a Mamluk-period coin (IAA 145775) and a British Mandate coin (1927 CE; IAA 822423), were found in the makeup of a fill that served as a bedding for the paving (L2), marking 1927 as a terminus post quem for the pavement; the bedding also contained sheep bones.
The excavation results indicate that the hall was originally built in the Crusader period (Stratum III). Since it is not clear whether Floor 27, the lowest floor uncovered in the excavation, is the original floor of the hall, the height of the vaults was seven meters at the very least, making the hall a truly impressive and vast space. During the medieval period, Walls 12 and 13 and possibly additional walls as well were erected between the piers, at least in the eastern part of the hall. These may have been later partition walls within the huge hall, or part of an installation, such as a pool or a cistern. Conversely, as they abut two piers, these walls may have serves as supporting beams for the piers, so as to prevent the collapse of the building during an earthquake. Indeed, this part of the hall stands to this day fully preserved, while the vaulted area to its north has all but collapsed over the centuries, as nineteenth-century photographs reveal (Pringle 2007:200, Fig. CII).
During the Late medieval and Ottoman periods (Strautum II), soil gradually accumulated within the hall (Fills 23, 24). Floors 20 and 22 were fixed above these accumulations, and a tabun or a furnace may have been installed into it during the Mamluk or the Ottoman period, as evident by the red color of Fill 11 and the abundant metal slag found in it and in Pits 14 and 16. It seems that the tethering rings in the piers were drilled during this phase of use, since a large quantity of animal bones were found in Fill 11, suggesting that the hall was used to house and treat animals. Conversely, the numerous animal bones may be associated with the slaughterhouse and tannery that operated in the nearby ruins of St. Mary Latin (Berkovich and Re’em 2016:213n13).  
Without further investigation, it cannot be determined whether the entire ceiling of the hall is original, although it seems that it underwent various changes and repairs over time. In any case, the presence of oculi indicates that at some point there was a need for such openings that will allow air and light into the hall.
A survey and excavations conducted begining in the 1870s, by Conrad Schick (1902) and other archaeological explorations, within the Muristan compound (Kenyon 1974:227–231; Re’em et al. 2011), indicate that the vaulted hall, where the excavation took place, was originally the southeastern part of a huge Crusaders-period L-shape vaulted enclosure. The enclosure (c. 6000 sq m) consisted of more than 100 piers and extended all the way to the abbey of St. Mary Latin on the north; to today’s Muristan Street (Crown Prince Frederick Street) on the west, and—if we follow the piers marked on Schick’s plan—even farther west, up to the church of St. John the Baptist; to the three markets on the east; and to the façade running along David Street in the south. All that remains today of this vaulted enclosure are several of the southern vaults, partially occupied by the bazaar shops along David Street, and one bay, today’s Muristan Street (Crown Prince Frederick Street)—the main entrance leading from the south into the modern Muristan compound. We assume that many additional vaults still survive, hidden in the modern shops along the street. The rest of the remains are either buried under the modern Muristan compound structures or were destroyed and removed prior and during their construction in the second half of the nineteenth century CE. Several surviving pier stumps can still be seen in the Muristan Garden of Peace and the courtyard of the Martin Luther School. Clearly, these formed with the piers found in the hall one, continuous structure. The farthest free-standing pier in the Garden of peace, directly to the north of the hall, bears cavetto molding as do the easternmost piers (I–III)—further proof that the hall and the piers in the garden and the school courtyard belong to a single architectural unit (cf. Berkovich and Re’em 2016:201–202; for recent arguments to the contrary, see Pringle 2007:204; Heinzelmann et al. 2014:171–174; Boas 2017:21–22, 29).
As the vaulted enclosure provided a very large space that could have easily held many hundreds of people or large quantities of supplies and provisions, it was obviously an important part of the Hospitaller quarter. This assumption is further reinforced by the substantial water facilities, cisterns, conduits, drains and tunnels that were identified in the excavations under the hall (Warren 1884: Pl. L; Warren and Conder 1884:256–260; Pringle 1997). Nevertheless, scholars have assigned the southeastern hall a secondary role, and some even place it outside the Muristan compound altogether. Several of these scholars follow Schick’s suggestion that the hall contained the Knights Hospitallers’ quarters and the stables for their horses (Schick 1902:50; King 1931:66–67; Benvenisti 1970:70; Barber 2001:153). Others, possibly recalling its more recent uses, see it as primarily a commercial space; Patrich (1984:14–15), for example, suggests that its entire area was occupied by shops and by the poultry market mentioned in La Citez. Similarly, Bahat (1990:91) has suggested that this part of the hall was an independent unit, separated from the Hospitaller complex by an alley running along the western façade of St. Mary Latin down to David Street, just like the Muristan Street (Crown Prince Frederick Street) does today.
On the other hand, in his study of Crusader Jerusalem, Boas (2001:88, 143–144, 159–160) has argued that the southeastern hall was an integral part of the Hospitaller quarter. According to his interpretation, most of the vaults running along David Street were occupied by large, two-storied shops, each including a storeroom and living quarters, while an open market occupied the southeastern corner of the vaulted hall, where the current excavation was held. The vaults extending northward, toward the abbey of St. Mary Latin, formed a separate unit, which probably housed the hospital wards for female patients. Murphy-O’Connor, who agrees that this part of the Hospitaller quarter was more significant than previously thought, has offered a somewhat different interpretation: the vaults along David Street were used as warehouses, while the northern extension, together with the second floor, above the warehouses, was used for the residential spaces of the Hospitaller convent (Murphy-O’Connor 2012:280–282).
When we consider the location of the hall and the vast space it occupies, as well as the height of its vaults and the quality of its construction—as indicated by our excavation results—there is no doubt that this was a public building meant to hold large numbers of people or a large amount of supplies. The excavation results along with a reexamination of the medieval historical records and a study of Conard Schick’s original papers, kept at the PEF Archives in London, lead us to the conclusion that the building formed an integral part of the palacium infirmorum of the Hospital of St. John—the first and most famous public hospital of the Knights Hospitaller (Berkovich and Re’em 2016).