A Newly Identified Roman Military Bath at ‘Ayn Gharandal (Jordan)

Project Title: 2009 ‘Ayn Gharandal Survey and Preservation Project
Directors: Robert Darby (University of Missouri-Columbia)
Erin Darby (Duke University)
Surveyor: Jim Bucko (Independent Researcher)
Ceramicist: Andi Shelton (North Carolina State University)
CAD and 3-D Modeling: Peter Carrasquillo (North Carolina State University)
Affiliation: American Center of Oriental Research (ACOR), Amman


Location and History

The site of ‘Ayn Gharandal is located c.100 km north of the gulf of Aqaba and some 40 km south-west of Petra on the eastern edge of the Wadi Arabah. The ruins rest alongside a modern paved road running east from the nearby Dead Sea highway. The presence of an artesian spring in the mouth of the nearby Wadi Gharandal, which still flows abundantly today, presumably served as the reason for human occupation at the site.


‘Ayn Gharandal and its surroundings were visited by many of the early twentieth century European explorers to the region (1). Alois Musil was the first to record the ruins of a Roman castellum at Gharandal in 1902 (2). Musil also noted in his visit the presence of two additional buildings near the fort, a water reservoir to the east and an unidentified structure lying southeast of the fort. A little over a decade later T.E. Lawrence passed through ‘Ayn Gharandal as part of the British Palestine (Wilderness of Zin) Survey where he noted the presence of two structures, presumably the fort and water reservoir previously recorded by Musil.

Recent investigations of the site by the author and others revealed that two rooms of a bath complex associated with nearby the remains of a Late Roman fort or castellum had fallen victim to illegal digging. As a result, a portion of the building’s architectural remains were unearthed, including several well-preserved and heavily plastered walls, exposing them to the extreme environmental conditions of the Wadi Arabah. The exposed structure was at risk to not only the elements, but potential stone robbing of its well-cut ashlar blocks.

Preliminary Findings Concerning the Bathhouse and its Water Supply

I. Bathhouse

In the two looted rooms of the bathhouse, the interior faces of several walls were visible. This includes some portion of the N, S, E, and W walls from the Northern Room and the N and W walls of the Southern Room. Several of the walls still exhibited remains of concrete for the vaulting of the structure. Although much of the structure remains buried in the surrounding dunes, it was possible to discern a plan of the two exposed rooms.

The Southern Room of the structure was badly disturbed, but the internal face of its northern wall was still covered in thick plaster extending over three courses of roughly cut stones. The Southern Room’s western wall also maintains its plaster veneer, but it is poorly preserved and still largely buried. Regardless, we identified at least one voussoir in this wall, justifying the conclusion that the Southern Room was barrel vaulted.


In the Northern Room, the walls were considerably better preserved. The northern and southern walls of the room are constructed of undressed stones and chinkstones. The southern wall, the central cross-wall between the Northern Room and Southern Room, contains a doorway leading to the Southern room. The western wall contained a row of springers, proving that the barrel vault covered both the Northern and Southern Rooms. Finally, the eastern wall was the best preserved, with at least two visible courses of well-cut stones. Its plaster is in fine condition with many layers, including a finished outer veneer. Two voussoirs (on the south end of the wall) exhibit diagonal tooling. On the north end, at the corner with the northern wall, another voussoir is placed vertically (rather than horizontally). In all four corners small amounts of ash were visible. This is possibly due to chimney flues from the heating system of the bath.


Musil’s plan (above), drawn in 1902, includes three structures. The largest structure is correctly identified as a Roman fort, but Musil labeled the structure to the east of the fort as a reservoir. It is now clear from our work that Musil’s “reservoir” is actually the bathhouse. When we compared our plan with Musil’s, the external walls of the Northern and Southern Rooms, as well as the E-W wall, all appear to correlate with Musil’s “reservoir” plan.

2009-Plan-of-‘Ayn-GharandalHowever, Musil’s plan orients the bathhouse due north, while our new plan shows that the bathhouse is oriented, like the fort, to the NE. Further, Musil’s drawing preserves at least three more major walls connected to the baths that are not currently visible. If Musil’s drawing is accurate, the bathhouse is considerably larger than is presently discernible and may be well-preserved in several locations.

The Bath area yielded a large quantity of pipes (both water and flue) and tiles. The largest concentrations of tubuli were found adjacent to the north and east walls of the Northern Room. This suggests the possibility that the pipes were intact in the walls prior to their disturbance by the modern looters. The hypothesis is further supported by the large number of flue fragments recovered with plaster and concrete adhering to the surface.

Finally, the large number of burnt hypocaust tiles suggests that either the Northern Room or both the Northern and Southern Rooms were related to the heating system of the bath, either the caldarium or tepidarium. Future excavation of these rooms will help clarify their function.

Although no exact parallel for the water pipes has been identified thus far, they appear most similar to Fourth century pipes found at Humayma and Petra (Petra Pool Project). The flue pipes are consistent in form, although slightly larger, with those found at Humayma. The pottery from this area dates mostly to the Late Roman/Early Byzantine period, but this could be misleading due to the disturbed nature of the structure.

II. Aqueduct

Only upon arrival at the site for the 2009 season did we notice a line of low earth and stone mounds, running from the north east corner of the fort toward the mouth of Wadi Gharandal. The structural debris and undulating topography suggest the presence of piers for an elevated aqueduct system. The aqueduct line was disturbed in one section, but it is traceable for the majority of the distance between the fort, bath, and spring with mounds visible at regular intervals and heights consistent with a stone built arcade.  The total length of the aqueduct is ca. 190.3 m, and ca. 0.78 m wide. Our preliminary findings suggest that the aqueduct provided both the bath complex and the fortress with a constant source of water from the nearby ‘Ayn or spring.


  1. Musil 1907: 193-97; Wooley and Lawrence 1914: 14-15; Frank 1934: 231-32; Glueck 1935: 39-40.
  2. Musil 1907. See also Kennedy & Riley 1990; Kennedy 2004; which reproduce Musil’s original plan

All images © property of the ‘Ayn Gharandal Project


L'identifiant "Balnéorient" regroupe deux des membres du projet éponyme, Th. Fournet (Ifpo) et B. Redon (HiSoMA), en charge de la mise en ligne des billets. Ces derniers sont cependant souvent le fait d'autres participants au programme. Ils sont dans ce cas signalés comme auteur.

More Posts

1 réponse

  1. 8 février 2010

    […] This post was mentioned on Twitter by researchnewsinla and Hypothèses billets, Peter Nathan. Peter Nathan said: RT @researchnewsinl: A Newly Identified Roman Military Bath at ‘Ayn Gharandal, Jordan (Balnéorient) http://bit.ly/bhmzwL […]

Laisser un commentaire

Votre adresse de messagerie ne sera pas publiée. Les champs obligatoires sont indiqués avec *