The Fortresses at En Haseva
Rudolph Cohen

The Biblical Archaeologist, Vol. 57, No. 4 (Dec., 1994), pp. 203-214

(The Fortresses at En Haseva Rudolph Cohen, The Biblical Archaeologist, Vol. 57, No. 4, p 203-214, 1994)

See also: (The Iron Age Fortresses at En Haseva, Rudolph Cohen, 1995 AD)

Large scale excavations at 'En tlaseva have provided one of the most important recent discover-ies in the Arabah. Excavators have ex-posed a succession of fortresses—from the Byzantine and early Islamic period to the Iron Age—occupying this crucial crossroads of ancient commerce. In both Roman and Iron Age periods, the tlaseva fortress was among the most immense in the region. Excavation is finally darifying the true nature of the site, which has been known and visited for over a century.

A History of Early

Research at the Site

As early as the nineteenth century, re-searchers touring the area of `En

Haseva, one of the most abundant springs of the Arabah, noted remains beside the spring (Tin Husuv; map ref. 1734 0242). A. Musil visited the area in 1902 and prepared a sketch of the square fortress, which measured 120 x 120 feet, and had projecting corner tow-ers (1907207-208, figs. 144-145). He saw an additional multi-roomed struc-ture adjacent to the fortress in the south, as well as the remains of a bathhouse in the east. Musil identified the fortress with the caravanserai (an inn serving caravans) which Hasta mentioned in the Notitia Dignitatum (Seeck 1876:73). In 1930, the fort—and its original ground plan—were damaged. F. Frank (1934:254) visited the site in 1932. A. Alt (1935:6) identified the large structure at

`En kla,seva as the Roman fortress Eise-ba on account of the similarity between the Arabic and Greek names. The name Eiseba is mentioned only in the Beer-sheva Edict (Alt 1935:31). N. Glueck (1934-1935:17-20,115) concluded that the ruin was a Nabataean caravanserai

Aerial view of klaieva. The southern wall of the Roman fortress runs across the upper left portion of this north-oriented view. The two eastern towers of the latest Iron Age fort stand at the right. Part of the wall of the earlier Iron Age fortress is visible at the photo's right-hand margin. The photograph encompasses the bottom half of the top plan presented on page 204. (All photographs courtesy of the author and the Israel Antiquities Authority.)

Biblical Archaeologist 57:4 (1994) 203

The southern wall of the Roman fortress, Stratum 2, stretches ca. 46m towards one of the fort's four projecting towers. Against its inside face, builders constructed an ashlar wall, probably to support a set of stairs.

also used later by the Romans. He, like Alt, identified the site with Eiseba, in-cluded in a list of Negev towns and the yearly taxes levied on each by the Byzantine authorities. In 1950, during a study trip, B. Mazar found a small number of Iron Age sherds as well as decorated Nabataean and Roman-Byzantine sherds (Aharoni 1963:31). Relying on these finds, Y. Aharoni pro-posed identifying cEn Haseva with both biblical Tamar and Roman Tamara (Aharoni 1963; contra Alt 1935; Gichon 1976:80-81 and esp. fn. 4). B. Rothen-berg (1967:123-125,162-165) found only Roman-period remains during his survey in 1960.

The Excavations

In 1972, on behalf of the Department of Antiquities and Museums, salvage ex-cavations at cEn Haseva concentrated on the southwestern corner of the Ro-man fortress with square towers. Subse-quently, excavators turned their atten-tion to the southwestern square tower and a dump area containing numerous Nabataean sherds (Cohen 1972). These excavations were renewed in 1987 and continued intermittently unti11991 (Cohen 1988/89b; 1991) under the aus-pices of the Israel Antiquities Authority (IAA). Beginning in 1993, large scale

204 Biblical Archaeologist 57:4 (1994)

excavations at the site have been direct-ed by the author and Y. Yisrael, on be-half of the IAA, with funding provided through the Negev Tourism Develop-ment Administration. The excavations employ 50-60 workers from Yeruham. The excavators have uncovered two underlying fortresses dating to the Ju-

The Roman Fortress (Strata 2a-2b) The most obvious and substantial re-mains at cEn Haseva are those of the Roman fortress. The building of the Roman fortress at Haseva occupied two clearly distinct stages. Stratum 2a of-fered a typical square fortress (castellum ca. 46 x 46 m) with four projecting tow-

known in the Arabah. The Emperor Trajan probably founded it after the annexation of the Nabataean Kingdom to the Roman Empire (106 cE), thus establishing Provincia Arabia.

At the end of the third or the begin-ning of the fourth centuries CE, during the reign of Diocletian, the fort flour-ished once again. The projecting towers were added at that time (Stratum 2a). This Late Roman fortress was part of the Diocletian frontier system (Berchem 1952; Bowersock 1971; 1983:138-147; Gichon 1967; 1980; Graf 1987; Avi-Yonah 1966:118-121; Roll 1989:252-260) that depended principally on towers, forts, and fortresses erected along strategic roads. The inhabitants of the southeast-ern frontier of the Roman Empire, be-tween the Euphrates and the Red Sea, were Arab tribes called Saracens (Graf 1978; Parker 1985; Gichon 1986; Mayer-son 1986) who conducted razzia, with the basic aim of taking booty from the merchants and travelers who crossed the desert areas and whose destination was important commercial and tradecities, including Jerusalem and Gaza.

The track south from Jerusalem would have first reached Elusa, continued on to Tamara (Haseva), positioned south of the Dead Sea (Lacus Asphal-tides), and led hum there to Transjordan, Rababatora, and Philadelphia.

Archaeological investigation indi-cates that 'En tla5eva stood at an impor-tant junction with roads leading west, northeast, and south. The western route ascended via Macaleh `Aqrabim road, past the forts and towers of Rogem Safir, Horvat Safir and Mesad Safir (Cohen 1983b), until it reached Mamshit.

The road to the northeast stretched toward the area of the Dead Sea and Mesad Boqeq, where M. Gihon (1971) uncovered a smaller (ca. 17 x17 m) but similar fortress. On the modern high-way between the Dead Sea and Dimona, Gichon also excavated a fortress (ca. 38 x 38 m) at Qasr el Juheiniye and, following Alt's proposal (Alt 1935:34), identified it as ancient Tamar (Gichon 1976). From here, it seems there was a Roman road to Zoar, south of the Dead Sea.

Traveling south from cEn tlawva the road followed the Arabah Valley to Yot-vata where Z. Meshel excavated a for-tress similar in plan and size (ca. 40 x 40 m). The fort at Yotvata contained a very important discovery among others: a Latin Imperial inscription. Dating to the time of Diocletian, the inscription was found outside and opposite the eastern facade of the Roman fort, in front of its gate (Roll 1989).

cEn 1:laeva seems to have been, therefore, an important military and administrative center in the Roman period significant enough to have left an impression in ancient sources. But do we know its ancient name? As op-posed to Gichon's identification of the fortress at Qasr el Juheiniye with Tamara, the author believes that the Tamara described in the ancient sources is in fact cEn Haseva, as Aharoni proposed long ago (Aharoni 1963). cEn Haseva does find mention in Eusebius (late third, early fourth century cE.) Eusebius describes Tamara an Haseva) as a day's march from Mamshit and states that "today it is a military guard post" (Onomasticon led. Klostermanni 8:8). The fortress at 'En Haseva-Tamara, was mentioned in the Tabula Peutingeriana (Miller 1916:773; Aharoni 1963:33-37) as a stronghold along the road leading south from Jerusalem. Tainarar En Haseva is mentioned in several other ancient sources: Ptolemaios, Geograpiae V 15, the Notitia Dignitatum (Seeck 1876: 74), and in The Madaba Mosaic Map (Avi-Yonah 1954:42-43, Pl. 4).

The Roman fort fell into disuse during the second half of the fourth century CE,perhaps as the result of the earthquake of 363 CE which destroyed Petra and several other sites (Russell 1980; Ham-mond 1980). Over its remains Stratum 1 offered scanty and unidentifiable rem-nants, including pottery of the sixth-seventh centuries cE.

The Nabataean

Caravanserai (Stratum 3)

The foundations of the Stratum 2b Ro-man fortress were probably built over the remains of a Nabataean caravan-serai (Stratum 3), possibly similar in its

square ground plan and size to those uncovered by the author at Mo'a and Sha'ar Ramon, sites along the Petra-Gaza Road (Cohen 1982; 1987). Decorat-ed pottery and coins of Nabataean kings were collected on its floors. It is possible then to surmise that during both the Nabataean and Roman periods there existed a route connecting Haseva (Tam-ara) with Moca and continuing from there southward along the Arabah to Mesad Beer Menuha (Cohen 1983c), Yotvata, H. Dafit (Cohen 1984), and, finally, to Aila (Elath).

Roman period elongated jar and jug.

Biblical Archaeologist 57:4 (1994) 207

Viewed from above its western half, the gate to the Iron II fortress shows off its first entry way and chamber. Each of the gate's total of four chambers measure 2.5 x 3.3 m. Between the gate's two halves runs a 4m wide passageway.

The Iron Age Fortresses (Strata 4,5)

Stratum 4

Only the eastern section (ca. 36 m long) and two projecting towers (ca. 14 m apart) have thus far been cleared in the Stratum 4 fortress. The square south-eastern tower, completely uncovered, measures 11 x 11 m. Its outer walls reach approximately 1.5 m in width.

Pottery retrieved from the floor be-longs to the seven-sixth centuries BCE, which suggests that this fortress was built during the reign of Josiah and de-stroyed at about the same time as the First Temple in Jerusalem, in 586 BCE.

Stratum 5

The Stratum 5 fortress is square, 100 x 100 m, covering approximately 1 hec-tare (2.5 acres). While some sections of the walls were destroyed completely, others still stand 2 m high. The fortress's offset-inset wall was 3 m wide at each offset and 2.5 m wide at each inset, with three projecting corner towers. The in-sets and offsets were spaced 8-10 m apart. An outer rampart and moat are currently being exposed. Several case-mate-rooms have been cleared along the northern and eastern sides of the parapet; their inner walls are 2 m wide. The southeastern and northwestern towers, which protrude about 3 m from the wall's line, have also been uncov-ered. It is not clear, however, if there is a fourth tower at the northeastern corner.

Corner of the tower of the late Iron Age fortress, Stratum 4. With exterior walls over 1 m in width, the square, 11 x 11 m tower occupies one of the corners of the last Iron Age fortress at Haseva.

208 Biblical Archaeologist 57:4 (1994)

The fortress-gate stood near its north-eastern corner. The gate complex mea-sures ca. 15 x 12.8 m. The western section has not yet been completely exposed. The walls, ashlar-built, are preserved to a height of ca. 3 m and are especially im-pressive in the quality of their construc-tion and state of preservation. Between the gate's four piers (ca. 2.5 m in width), the gate passageway narrows from 4.8 m on the outside to 4 m on the inside. Two identical chambers, 2.5 x 3.3 m, stand on each side of the passageway, thus demonstrating that this is a four-chambered gate, common in fortifica-tions in Israel and Judah in the ninth—eighth centuries BCE (Stern 1990). This gate resembles in plan the fot ti ess-gate at Tell el-Kheleifeh, differing only in orientation: the 'En Haseva fortress-gate faces the road approaching from the north, while the Tell el-Kheleifeh gate sits on the southern side, facing the sea.

View of the end of the unexcavated part of the western gate, in section, show-ing its well-preserved pier.

(Above.) The four-chambered gate to the fortress, Stratum 5, viewed from the north through the gate into the fort. The gate complex covered an impressive ca. 15 x 13 m, a size commensurate with the over-all dimensions of the fort. Built according to a plan common throughout Israel and Judah in the Iron Age, Haseva's gate most nearly resembles the fortress gate at Tell el-Kheleifeh.

210 Biblical Archaeologist 57:4 (1994)

Technically speaking, the architecture at `En Haseva consists of two principal parts. With its gate area, the northeastern section is similar in plan and is alone as big as the Tell el-Kheleifeh foi Less. Re- mains of an outer-gate are also being uncovered at En Haseva.

The plan of the fortress in this stratum is complex and unique, reflecting archi-tectonic elements from two Iron Age fortress types: the square fortress with a solid offset-inset wall, like the fortresses at Tel 'Arad (Strata VII-IX) (Aharoni 1981:6-7), Tell el-Kheleifeh (Strata II-IV; Glueck 1939; Pratico 1985), and Horvat Toy (Cohen 1985; 1988/89a); and the fortress with projecting towers, like the middle and upper fortresses at Tel Kadesh-Barnea (Cohen 1981; 1983a) and the fot tiess at Horvat Uzza (Beit-Arieh

1986), which is rectangular. It should be noted that the 'En Haseva fortress in this stratum was surrounded by both a solid outer wall and a row of casemate rooms.

This immense fortress is one of the largest known in the Negev and adja-cent regions. It covers four times the area taken up by such considerable fortresses as that at Tel 'Arad, Horvat Toy, Horvat Uzza, and Tell el-Kheleifeh, coming close in size to fortified cities of this period-Beersheva, for example, which also ex-tends over an area of about 1 hectare (2.5 acres; Aharoni 1973:75, 80).

During whose reign was the Stratum 5 fortress at 'En Haseva built? Thus far, the scant ceramic material recovered makes it difficult to assign a date. It was probably built in the eighth-seventh centuries BCE, during the reigns of King Uzziah and his successors. An examina-tion of the relations between Judah and Edom as they are described in the Bible reveals several possibilities. Amaziah, the son of Joash, diligently fortified his kingdom both from within and without, and, after he instituted reforms in the army, went to war with Edom. He de-feated the Edomites in the Valley of Salt, in the northern Arabah, and then went on to conquer Sela (2 Kgs 14:7). He re-named it Joqte`el, and settled descen-dants from the Tribe of Judah there. Wasthe large fortress at `En Haseva estab-lished during Amaziah's rule, and was it from here that he set out against the Edomites? Or was the foitiess built during the reign of his son Uzziah (2 Kgs 15:1), the powerful and active king who "built Elot and restored it to Judah" (2 Kgs 14:22; 2 Chr 26:13)? Or perhaps this fot Ness was built during the reign of Jehoshaphat when "there was no king in Edom, a deputy was king" (1 Kgs 22: 48), and when in an un-successful attempt to repeat Solomon's achievements, "Jehoshaphat made ships of Tarshish to go to Ophir for gold, but they did not go, for the ships were wrecked at Ezion-geber" (1 Kgs 22:49; Bartlett 1989:115-116).

Casemate rooms in the wall of the central area of the Iron Age II fortress, Stratum 5. The walls of Iron II Haseva combined solid offset-inset with casemate wall construction. The walls enclosed the largest fort of the Negev, equal in size to the regional adminis-tration center Beersheva.

Pottery vessels from the Iron Age II fortress, Stratum 5, include these carinated bowls and oil lamp. Unfortunately, ceramic finds remain too scant to permit accurate dating of the structure's construction.

Another possibility is that this fortress was established in the course of the Israelite/Judahite retaliatory campaign against Mesha, King of Moab (2 Kgs 3:4-15), whose rebellion against the King of Israel is mentioned in the Stele of Mesha (Bartlett 1989:116-122; Dearman 1989). The large Stratum 5 fortress may have served as the deployment center for this invasion. The ground plan of the Stra-tum 5 fortress at `En Haseva has several features in common with that of the for-tress uncovered at Tel Jezreel (Ussish-kin and Woodhead 1992), an important administrative center in the Israelite Kingdom.

The dimensions of the Stratum 5 for-tress at 'En Haseva and the evidence of the intense building activity there reflect its strategic importance. It sat on the road which followed the Arabah from north to south leading to Elath and the Red Sea, and defended the area opposite the Edomite mountains to the east. The resemblance between the plan of the 'En Haseva fortress and that at Tell el-Kheleifeh (Strata II-HI) is not surprising since it appears that both were built at the same time.


The finds from the Roman and Iron Age fortresses at 'En Haseva support Aharoni's proposal to identify the site, which was a major fortress on the south-eastern frontier of the Judaean Kingdom, both with biblical Tamar (Ezek 47:19; 48:28) and with Tamara mentioned in the Roman and Byzantine sources cited.

1 Kgs 9:17-18 states that "Solomon built Gezer, Beth-boron the Lower, Baalath and Tamar in the wilderness, in the land." Some scholars believe a mistake was made here and that, in fact, the reference is not to Tamar but to Tadmor, referred to in the parallel description in 2 Chr 8:4: "Tadmor in the wilderness" (Gray 1970:248-249). This problem will be solved if remains dating to the time of Solomon are uncovered at 'En Haseva. This is likely, given the parallel between the histories of 'En Haseva-Tamar and Kadesh-Barnea. In Ezekiel, these two sites on the southern border of Israel are mentioned in proximity.

The strategic location of `En Haseva-Tamar is obvious, sitting as it does at the intersection of four major routes — one leading south to Elath, one traveling east to Edom, a third leading north to Jerusalem, and the fourth leading west through Macaleh `Agrabim to the central Negev area. The remains of the im-pressive fortresses testify to the impor-tance of the site during the Iron Age and, later, in the Nabataean and Roman periods. Aharoni believed that the sys-tem of road-fortresses was established during the time of the First Temple peri-od in the Negev and the Arabah. These

1988/ 'En Haseva-1987. Excavations and 89b Surveys in Israel 7-8:52-53.

1991 'En F.laseva— 1988/1989. Excavations

and Surveys in Israel 10:46-47.

Dearman, J.A.

1989 Historical Reconstruction and the Mesha Inscription. Pp. 155-210 in Studies in the Mesha Inscription and Moab, edited by A. Dearman. Archae-ology and Biblical Studies, 2. Atlanta: Scholars Press.

Eph'al, I.

1982 The Ancient Arabs: Nomads on the Borders of the Fertile Crescent: 9th-5th Centuries B.C. Leiden: E.J.Brill.


Onomasticon. E. Klostermann ed. Leipzig 1904.

Frank, E

1934 Aus der 'Araba I. Zeitschri ft des deutschen Palfilestina-Vereins 57:191-280.

Gichon, M.

1967 The Origin of the Limes Palaestinae and the Major Phases in its Develop-ment. Pp. 175-193 in Studien zu den Mil itargrenzen Rams. Koln-Graz: Bohlau Berlag.

1971 Das Kasten En Boqeq. Bonner

Jahrhucher 171:386-406.

1976 Excavations at Mezad Tamar — "Tamara" 1973-75. Preliminary Re-port. Saalburg-Jahrbuch XXXIII:80-94.

1980 Research on the Limes Palaestinae: A Stocktaking. Pp. 843-864 in Roman Frontier Studies 1979. Papers Presented to the 12th International Congress. British Archaeological Reports International. Series 71. Edited by Hanson, WS. and Keppie, L.J.F. Oxford: BAR.

1986 Who were the Enemies of Rome on the Limes Palaestinae. Pp. 584-592 in Studien zu den Militargrenzcn Roms III. 13th Internationaler Limeskongress, Aalen 1983. Stuttgart: Kommissions Verlag K. Teiss.

Glueck, N.

1934- Explorations in Eastern Palestine, II. 1935 Annual of the American Schools of

Oriental Research XV. New Haven:


1939 The Second Campaign at Tell el-Kheleifeh (Ezion-Geber: Elath). Bul-letin of the American Schools of Oriental Research 75:8-22.

Graf, D.E.

1978 The Saracens and the Defense of the Arabian Frontier. Bulletin of the Ameri-can Schools of Oriental Research 278:1-26.

1987 The Roman Road System of Arabia Petraea [Abstract, The 88th General Meeting]. American Journal of Archaeol-ogy 91:319.

Gray, J.

1970 1 & II Kings: A Commentary. 2nd ed. Old Testament Library Philadelphia: Westminster Press.

Hammond, P.C.

1980 New Evidence for the Fourth Century CE. Destruction of Petra. Bulletin of the American Schools of Oriental Research 238-65-67.

Mayerson, P.

1986 The Saracens and the Limes. Bulletin of the American Schools of Oriental Re-search 262:35-47.

Meshel, Z.

1989 A Fort at Yotvata from the Time of Diocletian. Israel Exploration Journal 39:228-238.

Miller, K.

1916 Itineraria Romana. Stuttgart: Strecker and Schrader.

Musil, A.

1907 Arabia Petraea II Edotn . Vienna: A. Holder.

Parker, S.T.

1985 Romans and Saracens: A History of the Arabian Frontier. ASOR Dissertation Series 6. Winona Lake, IN: Eisenbrauns.

Pracico, G.D.

1985 Nelson Glueck's 1938-1940 Excava-tions at Tell el-Kheleifeh: A Reap-praisal. Bulletin of the American Schools of Oriental Research 259:1-32.


1932 Geographiae. Edited by J. Fisher Codex Urbinas Graecus 82. Bibliothecae Vaticanae, Lipsia

Roll, I.

1989 A Latin Imperial Inscription from the Time of Diocletian Found at Yotvata. Israel Exploration Journal 39:239-260.

Rothenberg, B.

1967 Negev. Archaeology in the Negev and the Arabah. Givatayim-Ramat Gan. (He-brew).

Russell, K.W.

1980 The Earthquake of May 19,363 CE. Bulletin of the American Schools of Orien-tal Research 238:47-64.

Seeck, 0, ed..

1876 Notitia Dignitatum. Berolini.

Stern, E.

1990 Hazor, Dor and Megiddo in the Time of Ahab and under Assyrian Rule. Israel Exploration Journal 40:12-30.

Ussishkin, D. and Woodhead, J.

1992 Excavations at Tel Jezree11990-1991: Preliminary Report. Tel Aviv 19:3-56.

Yeivin, S.

1960 Did the Kingdoms of Israel have a Maritime Policy? Jewish Quarterly Review 50:193-228.



By Steve Rudd: Contact the author for comments, input or corrections.

Click to View