Archaeology Odyssey 2:1, Winter 1999

Floating in the Desert

A pleasure palace in Jordan

By Ehud Netzer

For more than a century after the death of Alexander the Great in 323 B.C.E., his heirs, the Seleucids in Syria and Mesopotamia and the Ptolemies in Egypt, fought for control of the portion of southern Israel known as Judea.

Early in the second century B.C.E., a Jew named Joseph stepped into the fray. The son of a Jerusalem nobleman named Tobias, Joseph traveled to Alexandria—the capital of Ptolemaic Egypt and the home of a sizable Jewish community—where he met with the king and was given the right to collect taxes in Syria and Phoenicia. Later, the youngest of Joseph’s eight sons, Hyrcanus, also visited Alexandria and was befriended by Ptolemy V (204–180 B.C.E.). While returning to Jerusalem, Hyrcanus was ambushed by his brothers, who, resenting the young Hyrcanus’s rapid rise to prominence, had apparently cast their lot with the rival Seleucids. In the fight, Hyrcanus killed two of his brothers, along with many of their followers; he then fled across the Jordan River.1 Hyrcanus lived for some time in Transjordan, all the while hoping to return to Jerusalem. But things changed when his father died. The first-century C.E. Jewish historian Josephus, who lived in Rome and wrote in Greek, recounts what happened next:

The elder brothers made war on Hyrcanus, who was the youngest of Joseph’s children, and the population was divided into two camps. And the majority fought on the side of the elder brothers … Hyrcanus, therefore, gave up his intention of returning to Jerusalem, and settled in the country across the Jordan, where he continually warred on the Arabs until he killed many of them and took many captive. And he built a strong fortress, which he constructed entirely of white marblea up to the very roof, and had beasts of gigantic size carved on it, and he enclosed it with a wide and deep moat. He also cut through the projecting rock opposite the mountain, and made caves many stades in length; then he made chambers in it, some for banqueting and others for sleeping and living, and he let into it an abundance of running water, which was both a delight and an ornament to his country-estate. The entrance of the caves, however, he made narrower, so that only one person and no more could enter at one time; and this arrangement he made deliberately for the sake of safety, in order to avoid the danger of being besieged and taken by his brothers. In addition he also built enclosures remarkable for their size, and adorned them with vast parks. And when he completed the place in this manner, he named it Tyre. This place is between Arabia and Judea, across the Jordan, not far from Essebonitis [Tell H\esbon].2

The story ends unhappily. Hyrcanus ruled in Transjordan, Josephus tells us, until the powerful Antiochus IV (175–164 B.C.E.) took over the Seleucid Empire. Fearing that he might be punished for his association with the Ptolemies and for killing Arabs, Hyrcanus “ended his life by his own hand. And all his property was seized by Antiochus.”3

But Hyrcanus’s splendid white “fortress” can still be seen and visited today.

Ancient Tyre (not to be confused with the ancient Phoenician city on the Mediterranean coast of present-day Lebanon) was identified in 1817 as the site of Iraq al Emir, about 10 miles southwest of modern Amman at Wadi Sir, by two officers in the British Royal Navy (Charles Irby and James Mangles). Other scholars later visited Iraq al Emir,4 including the American Howard Crosby Butler, who in 1904 provided a detailed description of the site. Its most striking remains are a series of caves and a monumental building called Qasr el-Abd (Arabic for “Fortress of the Servant”)—both of which are described by Josephus. A large man-made pool—the “wide and deep moat” in Josephus—once surrounded Qasr el-Abd. So the identification of Iraq al Emir as the Tyre referred to by Josephus is almost certain. What remains puzzling, however, is the purpose of the pool. Was it a military installation or “moat” (as indicated in Josephus), a water reservoir or some sort of lake?

An answer to this question came after I began studying the site in 1996. But to understand what function the pool served, we must first understand the building known as Qasr el-Abd.

The two most important archaeological excavations of Iraq al Emir were an American expedition, directed by Paul Lapp in 1962 and 1963, and a French expedition, from 1979 to 1985. The French team, directed by Ernest Will and the architect François Larché in collaboration with the Jordanian Archaeological Service (represented by Fawzi Zayadine, who still serves in the department), concentrated mainly on the building, documenting its architectural remains and producing a reliable series of reconstruction drawings; they also partially restored this magnificent building.

According to the French team, Qasr el-Abd was a two-story, rectangular building about 125 feet long, 62 feet wide and 40 feet high. On all four sides, it was decorated with reliefs of lions, panthers and eagles—Josephus’s “beasts of gigantic size.” Except for two panthers that served as fountains on the ground level, these reliefs were carved only on the building’s upper-story walls. The building was entered through a portico on the north; from there, a visitor could either continue into the ground floor or ascend to the second floor by means of an elaborate staircase. A second portico on the south did not serve as an entrance; it was probably designed to give architectural balance to the building. This south portico was also flanked by a staircase leading to the upper floor.

The partly preserved ground floor consisted mainly of four central rooms, each about 23 by 12 feet, surrounded by corridors enclosing the building. The 12-foot-wide corridors were illuminated by 17 large windows (7 on the east and west walls, and 3 on the south side), which also provided the inner rooms with their only source of light.

The French team did not leave us a plan for the upper story. But they did restore the building’s outer walls, which contained an abundance of windows—one after another with only small pillars separating them. The French team’s final report tentatively suggests that the second floor contained a few dwelling rooms as well as halls for entertaining guests.

Qasr el-Abd was built on top of a large rectangular platform, 215 feet long and 150 feet wide. The platform is higher than the surrounding ground, which was leveled and filled with water in antiquity. This leveled area covers almost 15 acres; it was bounded by the slopes of hills on the west, on the north, and partially on the east. The south side contained a 500-foot-long dam.

We estimated that over 250,000 cubic yards of earth were used to build this ancient dam. Creating the pool that surrounded Qasr el-Abd, therefore, was no simple task—prompting us to ask why it had been done in the first place.

One possibility is that the original builders wanted to make use of the reflective property of water—as is done in reflecting pools around the world. To test this theory, we built two models, one on a 1:100 scale and a later one on a 1:500 scale, in which mirrors replaced the water. The results of these tests were astonishing: The Qasr el-Abd water palace came to life, with reflections of its animal reliefs and porticoed facades shimmering in the mirror.

The architects of Qasr el-Abd must have planned this monumental building so that its lovely reliefs and limestone walls would be reflected in the pool.

We have no doubt that the main—perhaps even the sole—function of Qasr el-Abd was to provide lavish entertainment. It was not built as a water reservoir or as a military refuge; it was a pleasure palace. To a certain extent, it resembles an installation built by Herod the Great about a century and a half later. At the foot of Herodium, an elaborate palace complex built into a man-made mountain about 10 miles southeast of Jerusalem, Herod built a small round pavilion at the center of a large artificial pool. Guests would have been shuttled to the pavilion by boat, for a pleasant afternoon diversion. As at Lower Herodium, visitors to Qasr el-Abd could only reach the water palace by boat.

Our interpretation of Qasr el-Abd as a pleasure palace, rather than a military installation, inspired us to look into the mystery surrounding the building’s second floor. According to the archaeological evidence, this upper story might have contained only one division wall, toward the north end of the building. Drawing on this evidence, and on the French team’s plan of the ground floor, we can now suggest a reconstruction of the upper floor. At either end of the building were balconies. From the north end, one passed through the balcony into a broad entrance hall (12 feet deep and 55 feet wide); the entrance hall led into a large central dining hall (67 feet by 17 feet), or triclinium, probably constructed in a basilical plan, with a central aisle enclosed by parallel colonnades.

Open and spacious, well lighted by the numerous windows built into Qasr el-Abd’s outer walls, the triclinium would have held many dozens of guests—it was undoubtedly the centerpiece of this exotic building.b This large dining hall, we believe, was covered by a flat roof, possibly with towers at the corners. The roof probably served as an observation deck, allowing guests to enjoy spectacular views of the northern hills, palace gardens and large pool.

Other facilities enabled guests to take in Qasr el-Abd itself. Small boats that ferried people to the palace were probably docked at the northern side of the pool; these boats were likely used for pleasure excursions as well. A promenade, about two-thirds of a mile long, ran around the pool. At one point on the promenade, at the northern fringe of the pool, are the remains of a structure that might well have served as an observation deck. A tour of the promenade and a climb to this lookout post would have provided superb views of the “floating” palace.

Qasr el-Abd and its reflecting pool were only the most recent, and lavish, features of a large estate, covering at least 150 acres, belonging to the Tobiads—a Jerusalem clan, to which Hyrcanus belonged, that descended from either Hyrcanus’s grandfather, Tobias, or from an earlier Tobias who lived in the time of Nehemiah (mid-fifth century B.C.E.). The association of the Iraq al Emir estate with the Tobiads is almost certain. The Qasr el-Abd pavilion, as we have seen, provides an exact match with Josephus’s description of Hyrcanus’s building projects. Moreover, many of the pavilion’s details replicate Alexandrian architecture, and we know that members of the Tobiad family, including Hyrcanus, visited Alexandria and gained the favor of the Ptolemies.

There’s another, conclusive piece of evidence. The estate included a number of caves cut into the hill north of Qasr el-Abd. Josephus describes these caves as rock-cut chambers for banqueting and sleeping, with very narrow entrances. The entrances of two caves at Iraq al Emir are inscribed with the name “Tobias” in Aramaic—confirming the identification of the site. These two caves have an upper hall, perhaps used for banqueting, and a lower, cellar-like level. (The entrances, however, are not as narrow as Josephus suggests.) Another cave was probably used as a stable, capable of sheltering about 120 horses. According to the Zeno Papyric and other historical sources, the Tobiads were known for their use of cavalry.

This large estate was entered from the west (the direction of the Jordan Valley) through a monumental gate, situated almost 400 feet from Qasr el-Abd. This 33-foot-high gate, excavated by the French team, was carved with animal reliefs, like the water palace. A second gate may have stood at the estate’s eastern end, on the road from Amman.

Scattered about the estate are numerous leveled areas. These may have been the sites of ancient buildings, long since cannibalized for other purposes. But it is also possible that these leveled grounds once contained gardens—the “enclosures” and “vast parks” referred to by Josephus.

What was, in antiquity, the most prominent building on the estate is now covered by the village of Iraq al Emir. A portion of the building was excavated by Paul Lapp during the early 1960s. Among the site’s various strata, the most striking one (Stratum III) dates to the beginning of the second century B.C.E. This early second-century B.C.E. building appears to have contained courtyards, colonnades and several rooms decorated with frescoes. These remains, along with evidence from earlier surveys, suggest that most of the hill where the village now stands was once covered by a huge building, measuring 330 feet by 245 feet. We believe that this structure was the principal villa or palace of the Tobiads; it may well have served as the estate’s central building prior to the arrival of Hyrcanus.

Another large structure may have been built not far from Qasr el-Abd, near the pool’s northeastern corner. The site contains various architectural remains and a large leveled area, about 295 feet by 230 feet. We believe this site once held a palatial building that was plundered and destroyed in antiquity. If a building was erected on this site, it was probably constructed together with the pool and its large water palace. Our tentative suggestion is that Hyrcanus put up a large building on this site to house guests and that Qasr el-Abd itself was used for special, exotic entertainment. But only future excavations can prove or disprove this hypothesis.

The crowning achievement of the Tobiad estate, Hyrcanus’s opulent pleasure palace, was likely never actually completed—as indicated by unfinished remains at the site. The growth of Seleucid power in the East, however, not only put an end to Hyrcanus’s many building projects; it caused him to take his life.

We would like to thank Michael Sasson for building the models, Gabi Laron for his photographs, Kris Laureys for her drawings, and Judit Gertner for her general assistance in the study. This study could not have been undertaken without the superb work done by the French team, in particular François Larché’s drawings and restorations.