Bay 1 (Figs. 5: Section 1–1; 6: Section 3–3).The bottom part of a rectangular-shaped cistern (L100; length 5.4 m, width 2.25 m, preserved height 3.45 m) was discovered. The sides and floor of the cistern were coated with gray plaster (thickness 0.15 m), applied to a core (thickness 0.3 m) that contained soil and potsherds. The sides were built of two rows of medium stones bonded with gray plaster and small stones.
The southern side of the cistern was higher than the other sides and supported a vault. The opening of a conduit that was dug in Bay 2 (below) and blocked in a later phase was identified in the middle of this side.
Fragments of pottery vessels from the Ottoman period were found in the eastern side of the cistern (L101), including potsherds of black clay characteristic of Gaza Ware and other pottery vessels that dated to the seventeenth–nineteenth centuries CE.
Bay 2 (Figs. 6: Section 3–3, 7). Four strata (I–IV) were identified and are described below from the earliest to latest.
Stratum IV. A cross-vault (L306; length 4.15 m, width 2.2 m) is ascribed to this layer. The excavation of the vault was suspended before its complete exposure. The vault was built of medium-sized fieldstones and dark gray plaster. A lighter color plaster was noted in several places within the vault, indicative of localized repairs that were done to the vault, as well as its long period of use. The fill excavated inside the vault (thickness c. 1.7 m) contained pottery from the Ottoman period, including fragments of tobacco pipes and a potsherd bearing a Persian inscription (see Rubanovich below). The date of the vault’s construction is unknown.
Stratum III. A channel (W3; min. length 2.6 m, width 0.6 m, depth 0.8 m) is ascribed to this layer. Its sides were built of meticulously hewn medium and large ashlars that were not plastered. The channel was covered with large stone slabs and extended to the east and west beneath later walls. The bottom level of Channel 3 was higher than the floor of Vault 306 of Stratum IV (Fig. 7: Section 1–1). An opening (width 2.2 m, height 0.5 m) in the middle of the channel’s northern wall conveyed water to the cistern in Bay 1; the opening was blocked in a later phase.
Stratum II. A wall (W5) attached to the northern part of Vault 306 eastern wall belongs to this stratum. Wall 5 adjoined the vault’s wall at an oblique angle. The wall, preserved six courses high, was built of medium-sized ashlars with a layer of gray plaster between them (thickness c. 1 cm).
Stratum I. Two walls (W7, W8) that formed the northwestern corner of a space, whose nature h was unclear, belong to this stratum. The two walls were built of medium-sized fieldstones and gray plaster mixed with pieces of animal bones and eggshells. Each course of W8 comprised two rows of medium stones with a thick core of earth between them (width 0.9 m). Wall 7 (length 3.1 m, width 0.9–1.0 m, preserved height 1.3 m), survived by six courses, was attached in a later phase to the upper part of the western wall of Vault 306 (W6).
Bay 3 (Fig. 6: Section 2–2). Below the present eastern wall of the building, fill was excavated and three courses of an ancient wall were noted (W2; min. length 1.5 m, height 1.2 m) on which the present wall was constructed. Wall 2 protruded west from the western side of the present wall. The wall was built of large meticulously dressed ashlars with a layer of gray plaster between them. Potsherds from the Ottoman period were found in the fill excavated near W2; however, they are insufficient for dating the wall.
Bay 4 (Fig. 5: Sections 1-1 and 2-2). Dark wet soil fill (thickness 1. 2 m) was excavated in the northwestern bay. Below the fill, medium and large fieldstones were discovered. Between the stones was soil and traces of gray plaster and ash (L404), probably the upper part of a vault. The vault continued to the south and was related to a channel (L411; min. length 2.1 m, width 0.75 m, depth 0.8 m) built of carefully dressed ashlars. The channel, which was not plastered, continued west, beyond the limits of the IDF House.
Fragments of pottery vessels from the Ottoman period and a fragment of a roof tile were found in the fill inside the channel. Based on the thickness of the roof-tile fragment, it probably belonged to the Roman period; although it might be from the Byzantine period. The time when the channel was constructed is not known; however, it is clear it continued to be used at the time when Vault 404 was used.
Bays 5 and 6 (Figs. 5: Sections 2–2 and 3–3, 8). A water reservoir that had a cross-vault (L405; length 5.7 m, width 1.35 m, depth 4.6 m) was excavated. The sides and floor of the reservoir were coated with white-pale yellow plaster (thickness 7 cm). A depression (diam. 0.7 m, depth 0.15 m) was discerned at the bottom of the cistern, near the northwestern corner. The cistern’s ceiling was built of small and medium fieldstones with gray plaster and soil in-between. The original opening of the cistern was discovered close to the northwestern corner of Bay 6. The square opening (0.35×0.35 m; Fig. 9) was built of three courses of carefully dressed medium-sizes ashlars with a thin layer of gray plaster between them. The cistern’s northern side (W1; Fig. 10) was built of large stones, one of which had drafted margins, and a door socket that was originally utilized as a threshold stone and was incorporated in the wall in secondary use.
Wall 1 was abutted from the south by a white plaster floor (L407), of which only a small section was preserved (c. 0.5×1.2 m). Fragments of pottery vessels from the Early Islamic period were found below the floor (L408), indicating that the cistern and the floor that abutted W1 did not predate this period. Because of the limited scope of the excavation, the possibility should not be negated that the cistern’s opening, the floor and Wall 1 are later than the Early Islamic period.
A wall (W110; Fig. 5: Section 3–3) separated Bay 6 in the west from Bay 2 in the east. Fill containing fragments of pottery vessels dating to the seventeenth–eighteenth centuries was excavated next to the western side of W110; it is possible that this fill was brought after the building was constructed. Wall 110 was abutted by a white plaster level, perhaps a floor (L203; elevation 740.44 m; Fig. 5: Section 3–3). Dark brown fill was discovered below the floor and an ancient wall (W10) was identified below the fill. The wall was built of large ashlars and was integrated in W110. The potsherds from the layers of fill above and below Floor 203 are for the most part from the seventeenth and eighteenth centuries CE and therefore its exact date is not clear.
The Coins
Donald Tzvi Ariel and Ariel Berman
Sixteen coins were found, of which ten could be identified in general; however, none of the coins can be precisely identified.
Description (dates CE)
Weight, Dimensions
Æ, fourth century?
0.89 gr., 15×17 mm
Æ, 450–550
0.43 gr., 10 mm
Æ, Follis, Byzanto-Arab, 645–670
0.43 gr., 15×20 mm
1*: Type E
, Fragments of a dirham, Mamluk, fourteenth century
0.53 gr., 10 mm
Æ, fals, Mamluk, fourteenth century
1.77 gr., 17 mm
Double strike
Æ, fals, Mamluk, fourteenth–fifteenth centuries
0.62 gr., 15×17 mm
Æ, fals, Mamluk, fifteenth century
3.60 gr., 20 mm
Æ, fals, Mamluk, fifteenth century
3.68 gr., 20 mm
Æ, fals, Mamluk, fifteenth century
1.39 gr., 11×15 mm
Coin-like jewelry, Ottoman
0.08 gr., 13 mm
*SICA 1: S. Album and T. Goodwin. 2002. Sylloge of Islamic Coins in the Ashmolean 1: The Pre-Reform Coinage of the Early Islamic Period. Oxford.
Based on the finds discovered in the fill it appears that the building was constructed in the early nineteenth century CE. It was built on ancient remains, such as Floor 407 that abutted Wall 1 and Channel 411, which are ascribed to the Late Byzantine–Early Islamic periods. Presumably, Channel 3 also belonged to an earlier period, but no finds that could date it were discovered.
A Persian Inscription on a Pottery Vessel
Julia Rubanovich

The inscription appears on the upper part of the vessel, at the point of connection with the wide neck of the jar (Fig. 11). The inscription is only partially preserved. It is written in classical Persian, in the nasta‘līq handwriting, typical of Persian manuscripts from the second half of the fourteenth century onward (Yūsofī, olām-osayn. 1990. "Calligraphy". Encyclopaedia Iranica IV: 696–97). Below are the remains of the text of the inscription in Persian script and in transliteration:

ی بینی

دستیست که در کردن یار بودست

- ī bīnī

dast-ī-st ki dar kardan-i yār būda-st


The inscription preserves one complete line that constitutes the fourth and final line (mira‘), as well as the vestiges of the third line, of a rubā‘ī (quatrain). Originally, in conformity with the rules of the genre, the poem consisted of a quatrain, whereby the first, second and fourth lines rhyme according to the “a-a-b-a” rhyming pattern and end with the poetic device of radīf, a kind of epiphora, i.e., the repetition of a word, phrase or personal pronoun.


The line originates from a rubā‘ī ascribed to Abū al- af ‘Umar ibn Ibrāhīm (1048–1131/2) known as Khayyām (literally: “tent builder”) or Khayyāmī (literally: “son of tent builder”; for the discussion of his nick-name, see: de Blois, F. 1994. Persian Literature. A Bio-Bibliographical Survey Begun by the Late C.A. Storey. Vol. V, part 2: Poetry ca. A.D. 1100 to 1225. London: 356–357; for the ascription of this particular rubā‘ī to ‘Umar-i Khayyām, see Csillik B. 1934. The Principal Manuscripts of the Rubá‘iyyát of ‘Ummar-i-Khayyám in the Bibliothèque Nationale, Paris. Vol. I: Original Texts of the MSS Suppl. Pers. 1417., Anc. Fonds 349, and Suppl. Pers. 823. London: 3, 83, nos. 11, 142). The transliteration and translation of the entire poem is as follows:


īn kūza chu man ‘āshiq-i zār-ī būda-st

v-andar alab-i rū-yi nigār-ī būda-st

v-īn dasta ki dar gardan-i ū mībīnī

dastī-st ki dar gardan-i yār-ī būda-st


This jug was once a plaintive lover, like myself,
And was in pursuit of the face of a beauty;
This handle that thou seest upon its neck
Was an arm that [lay] around the neck of a beloved.


In some manuscripts the second line appears in a variant form (see: Hidāyat 1342/1963. Tarāna-hā-yi Khayyām. Tehran: 91;  E‘tessam-zadeh A.-G. 1931. Les Rubaiyat d'Omar Khayyam. Tehran: 33; Csillik 1934: 121, no. 108; Rosen F. [tr.]. 1930. The Quatrains of ‘Omar Khayyām. London: 29, no. 20), as follows:


dar band-i sar-i zulf-i nigār-ī būda-st

He was captive in the loop of a beauty's curl


‘Umar-i Khayyām, born in the city of Nishapur in the eastern Iran, was a polymath—an astronomer, mathematician and prominent philosopher. Even though there is no mention in contemporary sources of ‘Umar-i Khayyām having composed poetry in the Persian language, by the mid-thirteenth century a considerable number of quatrains - mostly  hedonistic, fatalistic and even anti-Islamic in nature - had been ascribed to him. Modern scholarship is rather unanimous in rejecting the ascription of the rubā‘īs to ‘Umar-i Khayyām, customarily treating the corpus attributed to him as a collective work by numerous authors of different periods (de Blois 1994: 363–366). At the same time, it should be noted that the rubā‘ī discussed here, appears in the earliest manuscripts (e.g., Bibliothèque Nationale, Paris, Suppl. Pers. 1417, dated to 1475) and unlike other quatrains, it was never attributed to some other poet (Csillik 1934: xxviii–xxix).


The inscription on the original vessel presumably extended around the entire upper part of the jar and contained four complete lines, whereby each line was separated from the other by means of a simple geometric design, a kind of twisted X. At the same time, given the high degree of textual fluctuation and variability, characteristic of the rubā‘ī genre in general, the possibility that the original text of the jar differed from the one presented above cannot be ruled out.


In terms of handwriting, I shall mention the absence of the second line above the letter “g” (ک instead of گ), which is typical of Persian manuscripts from about the tenth till the fifteenth centuries. However, since this is a single occurrence in the inscription, it remains uncertain whether we are dealing here with a period-specific orthography which could aid in dating the production of the vessel, or with an idiosyncrasy of a particular artist. Moreover, one should keep in mind the discrepancy which must have existed in modes of spelling in manuscript versus pottery production.


As for the dependability of the inscription upon its source, the meter of the only complete line is flawed (yār instead of yār- ī). It seems that the error stems from the lack of poetic expertise on the part of the artist who was not receptive enough to the required rhythmic pattern; alternatively, it can be explained by the insufficient physical space available to the artist, which forced him to omit the indefinite article. Assuming that the text on the jar was identical to the variant presented above, the selection of this particular rubā‘ī to adorn the jar is certainly remarkable for its direct connection between the subject of the poem and the jar as a physical object. In the rubā‘ī, the jar, its handle and the neck of the jar metaphorically signify a lover. The location of the inscription at the point where the body of the jar meets the neck, is not fortuitous, allowing the wordplay with the word “neck” (gardan) that appears twice in the quatrain; hence “the neck of the jar”= ”the neck of the beloved”. This rubā‘ī might  have been incorporated into the “inventory” of poetic verses used by potters to decorate their vessels. Thus, in at least one manuscript (Bibliothèque Nationale, Paris, Suppl. Pers. 823; the part including the rubā‘īs of ‘Umar-i Khayyām is dated to 1528), the poems are grouped according to a thematic principal, whereby the theme of the “jar” is represented by a corpus of twelve poems (see Csillik 1934:118–120, nos. 99–110; and the discussion therein: xlix–xl).