Apollonia Arsuf - History of Excavations

Seventeen seasons of excavations have been carried out so far at Apollonia-Arsuf, under the direction of Prof. Israel Roll of the Institute of Archaeology of the Tel-Aviv University.

The first season was carried out, in 1977, in three areas located inside and in the middle of the walled city (B-D). It was a salvage excavation on behalf of the Israel Department of Antiquities, which lasted for almost eight months and focused on Area B, located opposite the city gate.

Consecutive compound structures were found there on both sides of a narrow street oriented north to south which made part of an overall urban scheme carried out during the reign of the Umayyad caliph Abd al-Malik (685-705 C.E.). The entire complex served as a market street and the structures on both sides of the street were used as shops and food-stalls. During the second and third seasons (1980-1981) the north-east corner rooms of the Roman villa were uncovered in the south, outside and below the Medieval city-wall. A trial trench was opened upward, towards the city-wall, which provided a clear sequence of strata, reflecting the main occupation periods, from Roman times until the Mamluk destruction.

In 1982, the status of the excavations at Apollonia-Arsuf changed. From that point on, it became an academic excavation, on behalf of the Institute of Archaeology of Tel Aviv University. It also became a study excavation for the students of classical archaeology from the Department of Classics. As the excavation became an academic enterprise, the areas of excavation were not imposed upon us any more, as was the case in the previous salvage excavations.

We were now able to choose the working areas according to criteria of professional interest. At the same time, we also realized that the time had come to formulate a "strategy of excavation" at Apollonia-Arsuf. Thus far, work had been carried out in the north (Areas A-C, in 1950) and Area F, in the centre and in the east of the walled town (Areas B-D), and in the south (Area E).

To attain an overall archaeological picture on all four cardinal points of the site, an excavation was needed in the west. Therefore, a new area was excavated, inside and close to the western city-wall (Area H), from 1982-1984 (the fourth, fifth and sixth seasons), with the support of the Walworth Barbour American International School in Kefar Shemaryahu.

The early levels of occupation in Area H belong to the Persian period and above them, badly preserved walls of structures from the Hellenistic, Byzantine and Medieval periods were uncovered. When the excavations reached the lower end of the city-wall, we found that in Area H too, as in Area E, the wall was built on a deliberately laid bed of beach sand.

As we noted the same phenomenon in some of the walls of the market street in Area B, it became clear that we were in evidence of a peculiar method of wall-bedding that was used extensively in Early Islamic period Arsuf. This method seems to have the dual purpose of attenuating and absorbing shocks from earthquakes, while also allowing for the efficient drainage of rainwater in order to preserve the walls.


The next four seasons of excavations (the seventh, eighth, ninth, and tenth seasons) were carried out from 1990 through 1993, with the support of the Municipality of Herzliya. During this time, the entire eastern wing of the Roman villa has been uncovered. The fill above it included a substantial number of round Roman lamps of the type current in 2nd and 3rd century C.E., from which the usual pagan and erotic figurative scenes were intentionally broken away. This act, of which we were already aware from previous seasons, seemed to be the work of monotheistic believers, of either Samaritan or Jewish faith.

The eleventh season of excavations took place in the summer of 1996, this time as a contract excavation, initiated by the legal owner of the entire area of Apollonia-Arsuf, the Land Administration of Israel. The request was to undertake a series of trial excavations in the area that extends from the Crusader city-gate (Area J) to the east, with the purpose of determining the maximum extension of the ancient site, and defining its eastern limits.

In Area J, excavations revealed two construction phases of the city-gate, both from the Crusader period. In Area K, the only element of the Byzantine church preserved in its entirety has been uncovered - its underground reservoir. In the rest of the expanse, up to ca. 350 m. to the east of the city-wall, 28 trial trenches were opened, by means of mechanical tools. In one of the trial trenches, four burials were located, of Early Byzantine times. As these burials were necessarily located outside the city, their discovery provides the first tangible proof regarding the limits of the settled area of Apollonia in Late Antiquity. This area can be estimated today at ca. 280 dunams.

From 1998 to 2000, during the twelfth through the fourteenth seasons, the excavations were organized as a joint venture of the Institute of Archaeology of Tel Aviv University and the Universidade Federal do Rio Grande do Sul of Porto Alegre, Brazil. The Municipality of Herzliya provided full support. During the first term of 1998 work was carried out only in Area E, and was focused on the structure of the Roman period.

The second term was invited and funded by the Municipality of Herzliya as a first step towards preparing the site as a public archaeological park. Therefore, the excavations focused initially on fully uncovering the Roman structure in the south (Area E), and then on the large-scale unearthing of the Crusader castle in the north (Area F).

In Area E, the entire Roman structure was fully uncovered. This permitted clear identification of it as a typical Roman villa of the peristyle type. Moreover, as the western facade of the dwelling faces towards the sea, it seems clear that this was a villa maritima. Such a villa could belong to a foreign rich merchant, or to a local magnate who seemed to undergo Romanization. In a later phase the entries to several rooms were intentionally blocked in a rather negligent manner, which seems to indicate that several of the living rooms had been turned into storage rooms for more modest owners. Then, the whole complex suffered a sudden and violent destruction, which was apparently caused by a devastating earthquake, which occurred in 113/114 C.E., or 127/128.

In the Crusader castle (Area F), large scale excavations promoted by Parks and Nature Authority were undertaken during many months of the years 1998, 1999 and 2000 and practically all the components of the Castle's latest phase have been uncovered. These include the following elements described from the inside outward: 1) a central irregular courtyard surrounded by broad halls ; 2) the main pentagonal fortification complex, including a gate protected by two semi-circular towers with an additional semi-circular tower on each side in the east, and a donjon and two square corner towers in the west ; 3) an external fortification with an outline largely corresponding to that of the main fortification but at a lower elevation and with a wider opening ; 4) a broad, deep moat bounded by an outer retaining wall ; 5) a system of vaulted halls and retaining walls along the cliff at the western facade of the fortress, which largely collapsed falling to the base of the cliff ; 6) a port facility, of which the foundations of two side breakwaters (northern and southern) have been preserved.

Early in 2002, the last large scale season of excavations was conducted in the Crusaders fortress (Area F), focusing primarily upon the space east and south of the fortress gate between the main fortifications and their outer walls. During the same year, excavation in the southern part of the site (Area E) extended and deepened the exposure of a row of the Early Islamic rooms abutting the city wall from within. These rooms were restored during the Crusader period and served first as dwellings, then as craft work shops and drainage area. Finally, looking toward the Mamluk siege, the walls of the rooms were leveled and the space between them filled with earth and stones up to the height of the tops of the truncated walls. A thick plaster floor abutting the fortification wall was laid over the entire area. Thus, a firm inner rampart was added to the fortification while at the same time providing direct access to the wall for the first time. This did not prevent the site's conquest and destruction by the Mamluks, signs of which were abundant in the upper layer throughout the area.

In 2003 and 2004, excavation was renewed in the southern part of the city, in a new area extending east of the Roman villa (Area P). At the bottom of the excavation in the north of the area were found 2.70 m wide foundations constructed of field stones held together with mortar, penetrating deep into a Byzantine hamra layer. The robber trench above the foundations is clearly visible, indicating that they supported an ashlar wall that was robbed in its entirety, apparently part of a monumental structure of the Roman period. Inside the fortifications were found a group of thick-walled grenade-shaped pottery vessels. Such vessels served as incendiary devices of the class known as "Greek fire". As they were found near the city wall in a context associated with the end of the Early Islamic period, they would appear to have been utilized by the Muslim defenders of Arsuf against the abortive Crusader siege of 1099.