From May 2005 to November 2006, five seasons of excavation were conducted at Horbat Zerifin (Sarafand al-‘Amar; Permit Nos. A-4490, A-4684; map ref. NIG 1860/6517; OIG 1360/1517), prior to the construction of a wholesale marketplace. The excavation, on behalf of the Antiquities Authority and financed by the Tenuva Company, was directed by A. Kohn-Tavor, with the assistance of A. Oren, A. Nesher and Z. Bar-Or (area supervision), E. Bachar and S. Ya‘aqov-Jam (administration), A. Hajian and T. Kornfeld (surveying), T. Sagiv (photography), M. Avissar (ceramic reading) and N. Katsnelson (glass).
The site, around which the current Zerifin compound has developed, is located on the main Jaffa–Ramla–Jerusalem road, c. 3 km northwest of Ramla. The site is situated along the fringes of a hamra hill that protects it from the south, on the southern border of the fertile Lod Valley. Eshtori Ha-Parhi identified the village of Sarafand al-‘Amar with Zerifin of the Mishnah and this identification is accepted by most scholars, although no remains in the recent excavations that can be attributed to a settlement from this period were identified. Therefore, I suggest that Sarafand al-‘Amar is not the Mishnaic Zerifin and the origin of the Arabic name, which means ‘Sarafand the Living’ or ‘the Built’ derives from the Byzantine name of the place.
The site was severely damaged in the 1950s when Ha-Argaz Factory was built. All the archaeological strata in the western part of the site were removed and in the eastern part the remains reach a depth of 2.5–3.0 m where it seems all of the strata have survived, apart from the upper layer. The southern part of the site was not damaged and the buildings remained as they were, surrounded by remains of later period activities.
A limited trial excavation in the southern part of the site was conducted in 2004 (Permit No. A-4112), revealing an industrial installation and buildings from either the Ottoman period or the British Mandate. Backhoe probes that were cut throughout the entire area during the current development work, aided in estimating the size of the site in its different periods. Eighty squares were opened in ten excavation areas (Fig. 1); Areas A–F in the damaged section and Areas G, H and J in the southern part of the site.
The destruction that transpired at the site precludes defining its original area. However, judging by aerial photographs from the First World War and according to the excavations and trial probe remains, the site’s expanse seems to have been 15–20 dunams. The results of the excavation point to a medium-sized village whose habitation sequence spanned 1,500 years. Eleven strata, dating from the Late Byzantine period until the modern times, were exposed. The dating of the strata is based on a preliminary identification of the potsherds and changes are likely to occur during the processing of finds.
Stratum XI, the Late Byzantine Period (sixth century [?]–middle of seventh century CE).
The beginning of settlement at the site was exposed in limited sections, primarily due to their great depth (up to 3 m below surface). The limited number of buildings from this period is probably due to their continued use in the next stratum. Part of a building that was constructed on virgin soil was exposed in Area F. The date ascribed to the building is likely to represent the time it went out of use, possibly in the middle of the seventh century CE. Sections of a building, apparently the southern wing which was dismantled and its stones robbed in the Abbasid period (below), were excavated in Area J. Since its northern part was removed in the 1950s, it is difficult to gain a coherent plan. The excavated part indicates that the building was oriented east–west and had at least three lengthwise spaces that were divided into rooms. A foundation in the western part was surmounted with a row of pillars, standing a single course high. A marble column found nearby probably stood on one of the pillars. The thick, gray-plastered floors of the rooms were adapted to the natural topography, hence their level was not uniform; the western floors were c. 1 m higher than the eastern ones. A few remains, dating to the Byzantine or Umayyad periods, which probably belonged to one of the early phases of the building, were found below the floors. The building was covered with a thick layer of fill that contained a large quantity of roof tiles and pieces of marble pavement in a variety of shades, including a chancel screen fragment bearing a carved cross (Fig. 2) and a gilded, glass tessera. It seems the public building was connected to a church, which was probably built in the Byzantine period and used until the Abbasid period, perhaps in a different capacity, thereafter dismantled and covered over. Throughout the site, numerous architectural elements, mainly of marble and probably associated with the church, were found in secondary use in later strata or fills. Noteworthy among them were marble chancel posts, columns and capitals, as well as a fragment of a Late Roman Ware, Type 3 vessel and the base of a bowl adorned with a stamped cross.
Stratum X, the Umayyad Period (mid seventh–mid eighth centuries CE).
The public building in Area J continued to be used during the Umayyad period. Sections of buildings were found in Areas A, A1, E and F and debris from pottery (mostly jars) and glass workshops, as well as layers of fill that dated to the Umayyad period were exposed in Areas B, C and D. An installation, which contained a minimum of three pools placed in a row (max. preserved height 1.7 m) and whose function is unclear, was found in Area E. The walls of the installation, built of poured gray bonding material, were plastered. A pipe of ceramic sections was incorporated in one of the walls. Part of the installation was cast underground and part, like other installations at the site, was higher than the ancient surface. More foundations of buildings and installations were found in this area; some had probably an extended usage during the Abbasid–Fatimid periods. A building that was partly preserved a single course high and contained a square pool was found in Area A1. Most of the masonry stones were robbed for secondary use and some of the walls were found beneath collapse, which was abutted by occupation levels from Stratum IX, the Abbasid period (eighth century CE). Other buildings dating to the Umayyad period were probably also erected in the Byzantine period. A building, preserved 1 m high, had a vault in one of its spaces and was probably a water cistern, was found in Area F. Part of it continued in use during the Abbasid period and another part, which was destroyed at the end of the Umayyad period, was filled with crushed pottery vessels and sealed with stone collapse (Fig. 3). A small depression that was indented when the vessel was leather hard occurred on the base of some of the vessels; in some cases, the hole was fully perforated while in others it was smoothed over with clay. The destruction of the buildings was probably local but it may be connected to the wars fought between the Abbasid and Umayyad dynasties. Another possibility is that the destruction was caused by the earthquake of the year 749 CE, although it is thought that this earthquake caused no damage in the region of Ramla. The destruction layer attributed to this earthquake was recently discovered in the nearby Mishmar David excavations (E. Yannai, pers. comm.).
Stratum IX, the Abbasid–Fatimid Periods (mid eighth–eleventh centuries CE).
This stratum was exposed in all the excavation areas and probe trenches; hence it is estimated that the ancient settlement reached its zenith during these periods. The areas along the fringes of the site (G, H, J) revealed brown layers of fill that yielded potsherds from this period and installations (see below). Parts of buildings were exposed in Areas A, B, C, E and F. A large part of a building, constructed in the second half of the eighth century CE and went out of use in the eleventh century CE, was discovered in Area B. Five rooms in a row that were probably the southern wing of a courtyard building were exposed. One side of the rooms’ walls was built of ashlar stones, whereas the other side consisted of fieldstones, some of which were coated with a thin layer of light plaster. The floors were of tamped earth and many of the building’s components were made of poured gray bonding material and small fieldstones, similar to the bonding material of the walls. An installation for storing liquid was in one of the rooms; nearby was a channel and below it a cesspit. Two corner pilasters that apparently supported an arch were found in an adjacent room. On the floor of this room were c. 10 jar lids of a type that is unknown from other sites. On the floor of the public building in Area J was a similar lid, incised with an Arabic inscription that is not yet deciphered (Fig. 4). Below the floor of this room were three jars that functioned as underground storage installations; a similar phenomenon was found in Areas A and F. One of the building’s rooms was divided in a later phase into two separate rooms. The building is a typical example of the construction in this and the preceding period. Buildings that seem to have been built in the Umayyad period and continued to be used, often with some modifications, during the Abbasid and Fatimid periods, were discovered in Areas F and J. Part of a building in Area F was canceled in a proper manner, whereas in another part the floors were raised and changes were made to its interior. The building from Stratum XI exposed in Area J was modified in the Abbasid period. The area between the pilasters was closed by a thin wall, which may point to a change in the function of the building. It seems that the floors of the building were replaced in this period; below the Stratum XI floors were potsherds that dated to the Abbasid period and these were also found in the thick fill that covered the building. The orderly plundering of the walls’ stones and the architectural elements related to the church, which were also found in the large fill that covered the building bear witness to the intentional dismantling of the public structure and perhaps reflects the transition from Christianity to Islam.
A noteworthy phenomenon that occurred at Zerifin in general, and in the Abbasid–Fatimid periods in particular, is the multitude of plastered installations. These, which occurred in a variety of sizes and plans, were mostly built of poured gray bonding material and coated with gray hydraulic plaster that contained potsherds. An installation in Area A1 that comprised two adjacent pools linked by a pipe is noteworthy. Its southern floor (1 × 2 m, depth 0.8 m) was 0.5 m higher than the northern floor (2.0 × 5.3 m, depth 1.5 m). A plastered bench adjoined the western and northern walls of the northern pool and in the southwestern corner was a narrow staircase that facilitated the descent into the pool. The installation was partly built above surface and partly dug into the ground and lined with a cast wall. Elsewhere in the site four other circular pools were discovered; two were partly excavated (diam. 4–5 m, min. depth 2 m). Their walls were coated with gray hydraulic plaster that contained potsherds and in the center of the floor was a depression for collecting liquids. Although refuse from the Ottoman period was found in the pools, they were ascribed to the Abbasid-Fatimid period, based on their construction. The purpose of the installations is yet unknown; however, it appears as though they were used as processing and industrial installations that provided some product to nearby Ramla. Built in a similar manner to the plastered installations were eight vaulted cesspits, the likes of which are known from Ramla. One of the pits was dug in the fill of Area B and four were discovered in Area H; some of them were unstable and built of fieldstones. They were an anomaly in being located beyond the limits of the settlement and not close to a building. A collapse dating to the end of the Fatimid period, which was found in some of the rooms in Areas B and F, was probably connected to the earthquakes that struck Ramla in the eleventh century CE.
Stratum VIII, the Crusader Period (twelfth–thirteenth centuries CE).
A few remains were discovered. A habitation level that sealed the Abbasid–Fatimid layer was found in Area F. The northern pool (above) in Area A1 contained a hearth and a section of an ashlar-built wall that had a sealed opening beneath the Ottoman layers. The building to which the wall belonged was destroyed in a mighty conflagration that left in its wake a layer of ash on the wall and the floor, which contained numerous iron nails. Other remains of collapse were found in a small section of Area J. In Area D was part of a building whose walls were built of two rows of ashlar stones with a core of dry-built fieldstones. This new method of construction at the site continued in later periods. The fill color in this layer and afterward was gray, in contrast with the brown color that typified the earlier fill layers. The Crusader structure in Area D was destroyed in a heavy collapse that covered the floors. The destruction can either be attributed to Salah ed-Din conquest in the year 1187 or to the Mamluk conquest in 1263. This destruction layer and the refuse pits from the Ottoman period contained numerous luxurious Crusader vessels that attest to the economic prosperity of the settlement at this time.
Stratum VII, the Mamluk Period (thirteenth–sixteenth centuries CE).
Buildings from this period were exposed in Areas B, D and E. A wall in Area D was founded on the collapse of the destroyed Crusader walls. A cellar in Area B, which penetrated the Abbasid-Fatimid stratum, was lined with neatly arranged ashlar stones and paved with pebbles that were overlaid with a collapse layer replete with numerous pottery vessels, some of which were intact. Mold-made bowls decorated with script (IAA Reports 26, Type I.1.7.1) were found, as well as a complete mold-blown bowl of purple glass that is decorated with white marvered trails (Fig. 5). Several sections of buildings and floors that were characterized by poor construction and the secondary use of architectural elements were found in Area E.
Strata VI, V, IVa-b, III, the Ottoman Period
The formidable presence of artifacts from this period is evident in all the excavation areas. It was difficult to date the strata in this period due to the multitude of layers and our limited knowledge regarding the typological development of the Ottoman ceramics. In Areas A and A1 a complete strategraphic sequence from the Ottoman period was found and the layers in the other areas are correlated according to it. Due to the absence of physical continuity between the excavation areas, the distinction between the strata is based at this point only on stratigraphy and construction characteristics. In Stratum VI pebble floors were found over a large area that seals the Abbasid-Fatimid layer. Habitation levels (height c. 0.4 m) accumulated on the floors. In Stratum V there was a building that was only partially exposed. The tops of the walls of this layer were used by the builders of Stratum IV as part of a proper stone foundation on which a tamped earth floor was deposited. Stratum IV was exposed across a rather large area and there were two phases to it: an early phase (IVb; Fig. 6) that is characterized by ashlar stone floors and well-built walls and a later phase (IVa) that reflects a significant decline that is apparent by the closing of spaces and the blocking of streets by means of meager partitions built of ashlar stones. Stratum III survived in very few places; most of it was removed in 1950. Sections of stone foundations survived in which the tops of walls from earlier layers were incorporated. A small hoard of coins that dates to the beginning of the nineteenth century which was concealed inside a cloth pouch beneath the floor (Fig. 7) is ascribed to this stratum. The Late Ottoman village is clearly visible in aerial photographs that were taken in 1917 and its dense plan was documented in 1946 within the framework of the “Village Survey”. The village mosque, which was used until its destruction in 1948, was built in the Ottoman period, and possibly even earlier. An Arabic inscription that was carved on the bottom side of a marble abacus in secondary use was fit in place in the mosque’s outer wall. Based on a preliminary decipherment by M. Sharon, the inscription dates to 962 AH (1460 CE), that is, the end of the Mamluk period. Throughout the entire Ottoman period numerous refuse pits were dug into the earlier strata and installations from earlier periods were also used for this purpose; all of these have yielded a wealth of artifacts. The Ottoman strata are characterized by a large quantity of pale gray sandy fill that was found along the fringes of the village in Areas G and H, which sometimes reach a height of more than 1.5 m. These layers of fill probably represent some sort of unidentified industrial activity or, on the other hand, especially large volumes of refuse.
Stratum II, the Time of the British Mandate
The nucleus of the village from Stratum III continued in use, preserving the traditional construction. A map of the village from 1946 credibly represents the development of the village during the British Mandate, until its abandonment in 1948. Buildings with courtyards were erected around the Ottoman village during this period, some of which were excavated in Area G. The construction combined cement and ashlar stones, which were probably in secondary use. The extensive use of Marseilles roof tiles (the ‘heart’ kind) and painted tile floors is quite evident. The destruction layer of the village reached a thickness of 1 m and contained a wealth of artifacts, including ceramic table ware, glass and metal vessels and agricultural tools. The only survived building was the village saqiye well (diam. 3.5 m, depth 20 m; Fig. 8) that was lined with ashlar stones and had a vaulted superstructure. It included the wheel assembly that turned two bucket belts and a short channel, leading to a pool (6 × 6 m) from which irrigation channels extended. Numerous clay ‘bottles’ found near the well were apparently used to ease the superstructure roof’s weight. Based on the iron pipes and the Marseilles tiles incorporated in the pool, the well was definitely built prior to the British Mandate era.
Part of the village cemetery where c. 75 graves were discerned was exposed in Area H; the graves were not excavated. These were simple cist graves aligned east–west and built of stones in secondary use. Incorporated in some of them were fragments of cement that indicate the graves should be dated to the time of the British Mandate. The densely packed graves, up to eight graves in a 25 sq m area and sometimes placed one atop the other, indicate the prolonged use of the cemetery, which may have had its beginning in the Ottoman period. According to the village map the cemetery extended to the southwest and indeed, several cement-built tombs standing in their place are visible beyond the excavation area.
The excavations at Horbat Zerifin greatly contribute to our knowledge of the settlement history in the region of Ramla. The proximity of Zerifin to the city of Ramla and to the road leading to it was a major factor in the village’s prosperity. Besides being a farming village, Zerifin was also a station on the main road. Especially interesting is the interrelationship between the village and the adjacent city. It seems that the numerous installations, whose nature still needs to be investigated, supplied agricultural or industrial produce to the city. It is clearly apparent that the material culture of the village, albeit certain differences, was very similar to that of the city. In light of the few, in situ Umayyad finds at Ramla, the early strata at Horbat Zerifin provide the first stratigraphic knowledge from the founding phase of Ramla. The excavation at Horbat Zerifin, which is one of the most extensive projects conducted in a village from the Early Islamic period, supplied much information regarding the rural village in its various periods. The intensive stratigraphic sequence and destruction episodes are likely to offer insight into the settlement processes during the last fifteen hundred years, particularly in the region of Ramla and throughout the country, in general.