Proof Positive: How We Used Math to Find Herod’s Palace at Banias

052053

Nestled on a wooded hillside above the cool waters of the Hermon Stream in the Golan Heights lie the ruins of Banias (ancient Caesarea Philippi). They include the remains of a rectangular building, partially hewn out of bedrock. The building is supported by a series of concrete walls decorated with opus reticulatum, a design with diamond or net-like patterns that was common in Roman architecture of the first centuries B.C.E. and C.E. This distinctive decoration caught the attention of Ehud Netzer, the late archaeologist and Herodian architectural expert, who excavated at Banias for two short seasons in 054 1977 and 1978 to determine the structure’s function and builder.^a

Only two other sites from Roman Palestine have substantial remains of opus reticulatum walls: Herod’s Third Palace at Jericho and the so-called Herod’s Monument (or mausoleum) in Jerusalem, just north of the Old City’s Damascus Gate.^b Both were built during the time of King Herod the Great, probably by visiting construction crews from Rome. Netzer logically concluded that the building on the hillside in Banias also should be associated with King Herod. But do these assumptions hold up to the available historical and archaeological evidence? And if the building was the work of King Herod, what do we know about its use?

Interestingly, the site’s Arabic name, Banias, derives from the Greek Paneias (or Paneion)—a shrine dedicated to Pan, the god of nature and woodlands, hunters and shepherds. And, indeed, a large grotto for Pan cuts into the steep mountain at Banias, at the foot of Mt. Hermon. Beside the grotto of Pan are niches hewn in the rock in honor of the gods Pan, Echo, and Hermes. On the lower terrace are the remains of several temples and open-air cult platforms. From a spring below the grotto gushes the Hermon Stream, one of the three sources of the Jordan River, which lingers below in a series of pools.

The Roman historian Josephus records that in 20 B.C.E. Caesar Augustus granted Herod authority over this region, and Herod erected at Banias a temple in honor of the Roman emperor:

Is the hillside building with opus reticulatum walls—located just 100 yards from Pan’s 055 grotto—the temple described by Josephus,^c or is it something else entirely?¹

Herod frequently chose dramatic locations and reshaped the natural environment for his building projects. The architectural remains explored by Netzer at Banias show precisely those signature features. The limited excavations on the elongated terrace about 50 feet above the Hermon Stream revealed parts of a long rectangular hall (measuring about 90 by 40 ft), entered from the south. The north wall of the opus reticulatum building was cut into the natural mountain slope. In the northeast corner of the hall, the wall reaches 16 feet above the bedrock floor, an indication of just how much rock was removed to carve the hall into the cliff face. Both faces of the building’s east wall were finely carved, suggesting there was perhaps another room to the east. More walls to the west and northwest of the hall suggest even more rooms in that direction.

The Banias hall, carved deep into the cliff face and supported with retaining walls, is reminiscent of Herod’s Northern Palace at Masada, where the two lower terraces were hewn directly from the natural rock and buttressed with additional walls. This is another architectural clue that links the Banias edifice to King Herod.

Small and long forgotten, there is another piece of archaeological evidence from Banias that might link the site with Herod. Scattered about the excavations were approximately 170 small geometrically cut tiles and their fragments that once created an elegantly patterned floor known as opus sectile. Meaning “cut work” in Latin, the floor was patterned with triangular, square, and octagonal shapes. Some of the pieces were cut from black bitumen quarried near the Dead Sea, while others were cut from a wide array of imported multicolored marble, alabaster, and granite.

Although Netzer recognized these as parts of a floor, there were no clues as to their original 056 arrangement. Therefore, they were put aside in a box, and it was 35 years before archaeologists took a second look at them. Astonishingly, this “box of rocks” holds vital clues to the identification of the enigmatic structure and its builder.

The opus sectile tiles found in the opus reticulatum building at Banias are consistent in size, shape, color, material, and craftsmanship with those found in Herodian structures. The black bitumen was quarried near Nebi Musa, just northwest of the Dead Sea. The imported stones were the same ones popular in opus sectile floors in villas and palaces in ancient Rome and Pompeii—pink, gray, red, purple, and greenish marbles and striped alabasters from Greece and Asia Minor; pink granite from Egypt; and bright yellow stones from Tunisia. The bedrock within the Banias hall was leveled to create the opus sectile floor, but archaeologists found no tiles still in place. Only a few impressions of large square tiles remain in the ancient gray plaster along the north wall, but they do not match the squares, triangles, and octagons excavated at the site.

So where do we go from here? We turn to mathematics!

When I (Frankie Snyder) was invited in 2013 by Rachel Bar-Nathan, a member of the original excavation team, to help solve the puzzle of the Banias floor, I noticed that the square tile impressions along the north wall at Banias are the same size as those around the outer edges of the triclinium (dining room) in Herod’s Third Palace at Jericho. Could there be more similarities between the two floors that may help us reconstruct the vanished floor at Banias?

057

A close examination of the opus sectile floors in several Herodian palaces revealed five characteristics, or “rules of construction,” that all these floors have in common:

1. Measurements are based on the Roman foot (11.65 inches).

2. Tile shapes include squares, triangles, hexagons, octagons, rhombi, and rectangles.

3. Each side of a tile borders a tile of a contrasting color.

4. The patterns are popular Roman patterns or close variations.

5. The pattern can be easily repeated over a large area.

When I applied these observations to the tiles and fragments from Banias, I was able to create a working pattern (see sidebar below for the step-by-step process). Although it is not a reconstruction of any actual segment of the original floor, this pattern works and shows how the original floor probably looked.

This proposed opus sectile is remarkably similar to the central pattern of the triclinium in Herod’s Third Palace at Jericho. While the Banias pattern alternates black triangles with multicolored squares, triangles, and octagons, the Jericho example alternates black triangles with multicolored squares, triangles, and rhombi. Both patterns also include groupings of tiles that form squares and octagons. In the Banias pattern, the squares are made up of smaller tiles, and the octagons are left 058 whole. In Jericho, the octagons are divided into smaller tiles, and the squares are left whole. Finally, the Banias pattern is created by repeating and overlapping eight-pointed stars, while the Jericho pattern has repeating and overlapping four-pointed stars, creating an amazing optical illusion.

These two unique floors are very similar in overall design, pattern generation, and materials. They may have been conceived and created by the same artist and produced by the same masonry team. And since the Jericho palace demonstrably belonged to King Herod the Great, the floor at Banias seems to indicate that Herod was responsible for that construction as well.

Ehud Netzer wondered whether the opus reticulatum building at Banias was Herod’s temple to Augustus mentioned by Josephus. He reconstructed the building as a rectangular temple with a small portico opening onto a colonnade. It was approached from the south by an arched stairway leading up the hill from the level area of the Hermon Stream below. Based on its opus reticulatum walls, Netzer dated the Banias building to the time of King Herod the Great, after 15 B.C.E. But because the Banias opus sectile floor shows so many similarities to the triclinium floor in Herod’s Third Palace at Jericho, we suggest a different interpretation: Rather than a temple dedicated to the emperor, the Banias hall was likely a triclinium or reception hall.

There are several other similarities between Banias and Jericho, besides their opus reticulatum walls, opus sectile floors, and arched stairways. The Banias building is about the same size as the Jericho triclinium, though somewhat narrower. The colonnade in front of the Banias hall overlooks the beautiful Hermon Stream, while the colonnade in front of the Jericho triclinium overlooks Wadi Qelt. Finally, the layout of the Banias hall, with its many rock-hewn walls and vast, leveled terrace, suggests that the building was part of a much larger complex.

It seems likely then that the opus reticulatum building at Banias was not a temple but rather part of a Herodian palace or royal residence. As such, we would argue that the temple described by Josephus is still awaiting discovery.

Archaeology and mathematics may have identified a forgotten palace of Herod the Great at Banias (ancient Caesarea Philippi). The marvelous floor design that once decorated an early Roman monumental structure at the site has parallels in other Herodian palaces. Explore the similarities and learn about the ingenious step-by-step reconstruction that made the identification possible.