Horbat Dafit is located in the southern Arabah ~1.5 km. from the Ein Avrona spring. Dolinka (2006) examined the material evidence from excavations performed in the 1980's which were only lightly published. Dolinka (2006) reports that a caravanserai was constructed [at Horbat Dafit] in the 1st century AD and was continuously occupied until the 3rd century CE.

• from Rudolf Cohen in Stern et. al. (1993 v.3)

Situated in the southern Arabah, on a low hill about 1.5 km (1 mi.) west of 'En Dafit (map reference 1500.9001), Horvat Dafit was first surveyed in 1965 by B. Rothenberg (site no. 47 in the Eloth Survey. He found remains of a structure (18 by 18m) from the Byzantine period. In 1983-1984, it was excavated under Cohen's direction, on behalf of the Israel Department of Antiquities and Museums. Three building phases were identified (strata 3- 1). The earliest (stratum 3) was a building (23.7 by 18.2 m) consisting of thirteen rooms built around a central courtyard (10.5 by 8 m). The main gate was in the middle of the eastern wall. The entrance was approximately 2.30m wide; the gate itself was a room (5 by 5.3 m) with three pillars on either side. The pillars supported arches that, in turn, supported a ceiling of stone slabs. On each side of the gate (north and south) was a long room ( 5 by 1. 7 m); two similar rooms were found on the west. The other nine rooms were the same size(c. 5 by 3.5 m). The outer walls of the building were some0.9m thick and the inner walls about 0.5 m thick. Pottery found on the floors of stratum 3 included painted N abate an bowls, oil lamps, cooking pots, and jars from the first century, as well as a few coins of Aretas IV. The rooms of the building continued in use in the second phase (stratum 2). The upper part of the walls was built of bricks. The finds include pottery typical of the second and third centuries and coins from the time of Trajan and Hadrian (117-138).

In the last phase (stratum 1) the walls of the room at the southeastern corner of the building were thickened and it was apparently became a tower ( 5 by 6 m) with an entrance (width, c. 1.2 m) in the east. The rest of the building was turned into a large courtyard. Camel bones were found in the northwestern corner of the courtyard near three cooking installations. The pottery was typical of the third and fourth centuries.

Horvat Dafit was probably a road station on the Nabatean road from the Arabah to Elath, from I:Ia?eva by way ofMo'a, Be'er Menul).a, and Yotvata. It continued in use after the Romans annexed the Nabatean kingdom to the Provincia Arabia in 106. At the end of the Roman period, the site declined in importance and was occupied by a tower and a camping ground.

• Revised chronology from Dolinka (2006)
• Chronology is based on ceramics and/or numismatics
• Dating Matrix of the structure at
  Fig. 4.38
  
  Matrix of the structure at Horvat Dafit, detailing the three loci (i.e. phases) for each room.
  
  Dolinka (2006)
  Horvat Dafit from Dolinka (2006)

Phase	Date	Comments
3	late 2nd- early 3rd centuries CE	major renovation Unfortunately, there were no coins found in either stratified Phase 3 loci or on the surface, so the presence of the `football' flask, the `Amr Type 9 lamp and the Johnson Form XII ungientarium represent the only datable evidence from Phase 3. (Dolinka, 2006:155) Of particular importance for the dating of Phase 3 at Horvat Dafit, was a nearly complete flask (Fig. 4.35), found in Locus 01. Theses vessels are referred to as `football flasks' because of their unique shape, and have been found exclusively in early-3rd century AD contexts at Moa (Erickson-Gini 2005: 51-52, fig. 3.20.12), the pantry at Oboda (Erickson-Gini 2004: 288, fig. 2.39), Shiqmona (Elgavish 1977: pl. II:13) and Mesad Neyarot (T. Erickson-Gini: pers. comm.). (Dolinka, 2006:153) Another significant find from Phase 3 that confirms its date in the late-2nd to early-3rd century is represented by a Nabataean rounded lamp (Fig. 4.36) ... The form corresponds to `Amr Type 9 (`Amr 1987: Pls. 16-17 nos. PL34-35) (Dolinka, 2006:153-154) A final well-dated ceramic specimen found at Horvat Dafit, in the Phase 3 courtyard (Locus 13, Basket 83) is a Nabataean piriform unguentarium (Fig. 4.37) that conforms to Johnson's Form XII (Johnson 1987: 66-67; 1990: 238-239, Fig. 4). These vessels were produced in great quantities during the Severan period, and are attested in the uppermost strata from Room 6 at Oboda, dated to the late-2nd/early-3rd century AD (Erickson-Gini 2004: 286, fig. 2.38). (Dolinka, 2006:154)
2	2nd-3rd centuries CE	rebuilding dating based on numismatics and ceramics (Dolinka, 2006:152)
1	1st century CE	initial construction ended with a presumed early 2nd century CE earthquake dating based only on numismatics (Dolinka, 2006:150)

Plans

Discussion

Dolinka (2006:130) reports that Phase 1 ended with the earthquake of the early-2nd century AD, and several of the rooms within the structure exhibited collapse of the architectural elements as well as ashy layers associated with that event. Dolinka (2006:135-136) further reports that the earliest levels of [phase 2] are characterized by mudbrick collapse and/or building debris (e.g. Locus 46), cleaning of the interior of some of the rooms (e.g. Loci 23. 26 and 50), repair to damaged walls (e.g. Locus 45), and reconstruction of the main entrance and gate (Locus 27). The site appears to be well dated except for in some loci. Dolinka (2006:155) noted a difficulty in discerning the division between Phases 1 and 2 in some loci which, it was suggested, was due to activities associated with the reconstruction of the gate after the earthquake, which likely disturbed the floor levels from those loci.

Effect	Location	Image(s)	Description
collapse of architectural elements	several of the rooms Fig. 4.4 Plan of the Phase I structure at Horvat Dafit (courtesy IAA Archives, Rockefeller Museum). Dolinka (2006)		several of the rooms within the structure exhibited collapse of the architectural elements as well as ashy layers associated with that event. - Dolinka (2006:130)
mudbrick collapse building debris Cleaning of the interior repairs and reconstruction	Loci 46, 23, 26, 27, 45, and 50 Fig. 4.4 Plan of the Phase I structure at Horvat Dafit (courtesy IAA Archives, Rockefeller Museum). Dolinka (2006)		the earliest levels of [phase 2] are characterized by mudbrick collapse and/or building debris (e.g. Locus 46), cleaning of the interior of some of the rooms (e.g. Loci 23. 26 and 50), repair to damaged walls (e.g. Locus 45), and reconstruction of the main entrance and gate (Locus 27). - Dolinka (2006:135-136)

• Earthquake Archeological Effects chart
  Earthquake Archeological Effects (EAE)
  
  Rodríguez-Pascua et al (2013: 221-224)
  of Rodríguez-Pascua et al (2013: 221-224)

Effect	Location	Image(s)	Description	Intensity
collapse of architectural elements - collapsed walls	several of the rooms Fig. 4.4 Plan of the Phase I structure at Horvat Dafit (courtesy IAA Archives, Rockefeller Museum). Dolinka (2006)		several of the rooms within the structure exhibited collapse of the architectural elements as well as ashy layers associated with that event. - Dolinka (2006:130)	VIII+
mudbrick collapse building debris Cleaning of the interior repairs and reconstruction	Loci 46, 23, 26, 27, 45, and 50 Fig. 4.4 Plan of the Phase I structure at Horvat Dafit (courtesy IAA Archives, Rockefeller Museum). Dolinka (2006)		the earliest levels of [phase 2] are characterized by mudbrick collapse and/or building debris (e.g. Locus 46), cleaning of the interior of some of the rooms (e.g. Loci 23. 26 and 50), repair to damaged walls (e.g. Locus 45), and reconstruction of the main entrance and gate (Locus 27). - Dolinka (2006:135-136)	VIII+

The archeoseismic evidence requires a minimum Intensity of VIII (8) when using the Earthquake Archeological Effects chart of Rodríguez-Pascua et al (2013: 221-224).