The excavation began with the removal of a concrete floor in a modern building used by the Aderet Eliyahu Yeshiva. It revealed two built vaults (L255, L258) containing fill separated by an alley paved with two floors. The lower (L262) and upper (L252) floors were installed in the Byzantine and Crusader periods, respectively. These three spaces were separated by walls oriented along an east–west axis, the tops of which were destroyed when the modern building was constructed. The architectural plan indicates they were constructed contemporaneously, and opened onto a main road that ran to their west, along a north–south axis, the continuation of which extended under today’s Shuq Ha-Zorfim; the alley evidently extended from the main road eastward.
 
The Byzantine Period. Two large flagstones with a gap between them, where a third stone was probably extracted in antiquity, remain from a paved floor (L262; Figs. 2–4) found in the alley between the vaults. It seems that the pavement continued southward, up to the wall that delineated the alley on the south (W28). A light-colored substrate (L263) was exposed beneath the eastern flagstone. It contained two coins: a nummus minted in Carthage in the second half of the sixth century CE (Table 1:17) and a coin that could not be positively identified, but most likely dates to the Byzantine period (Table 1:20). Sherds from the sixth century CE were found in both the light-colored substrate beneath the pavement and the substrate that was exposed where other flagstones had been removed (L260). These ceramic finds include two types of basins characteristic of the period (Fig. 5:1, 2), a Fine Byzantine Ware bowl (Fig. 5:3), Gaza jars (Fig. 5:4, 5), jars (Fig. 5:6, 7) and a cosmetic juglet from the Early Roman period (Fig. 5:8). The size of the flagstones, their thickness and wear pattern, without the typical cracking, are unlike those of the pavement found in the southern part of the Byzantine Cardo (Gutfeld 2012:98). Indeed, the flagstones might have originally been installed in a Roman-period street, dismantled and reused to pave the alley during the Byzantine period. In any event, the pottery and coins found beneath the pavement clearly demonstrate that the alley was built in the mid-sixth century CE at the earliest.
During the Byzantine period, the northern vault (L255) functioned as an elongated underground cavity, as evidenced by the difference in elevations between its floor and the pavement in the alley. Its eastern wall was discernible after removing the fill. However, due to safety problems, it was impossible to expose the vault’s western wall or even examine the wall’s western face. At some point during its use, the vault was apparently bisected by a broad wall (W27; 0.7 × 2.6 m); its eastern part (L253) was plastered and converted into a cistern, whose opening was situated in the vault’s apex, beneath the modern concrete floor. The western part of the vault was excavated to its full depth (c. 3 m). A layer of rubble, probably a floor, was found near the vault’s northern wall. A probe conducted beneath the floor revealed a soil fill containing pottery from the Byzantine period. This fill covered a stone-built water conduit (L261), oriented along a north–south axis, which predates the construction of the vault.
 
The Middle Ages. Remains of a light-colored plaster floor (L252; Figs. 6, 7) abutting a wall oriented along an east–west axis (W25; width 0.4 m, preserved height 0.7 m) were found c. 0.35 m above the pavement of the alley dating to the Byzantine period. Wall 25 was built on top of the flagstones from the Byzantine period (L262); it therefore clearly postdates the pavement. Four coins were discovered while dismantling the floor: two from the Byzantine period (fourth–fifth centuries CE [Table 1:3]; sixth century CE [Table 1:1]); an especially thick coin that probably dates to the Umayyad period (Table 1:4); and a coin from the twelfth century CE, minted in Lucca, Italy (Table 1:2). The latest pottery found below the floor, belonging, among others, to a coarse handmade basin (Fig. 8:2), a fragment of a glazed chamber pot (Fig. 8:6) and a bowl (Fig. 8:7), all date to the Crusader period. A basin fragment adorned with a wavy decoration (Fig. 8:1) and a Fine Byzantine Ware jug (Fig. 8:5), both from the Byzantine period, as well as a jar (Fig. 8:4) dating to the Early Roman period, were likewise recovered from beneath the floor. Also found were fragments of a jug (Fig. 8:3) and a weight (? Fig. 8:8).
The elevation of the plaster floor (753.32 m asl) roughly corresponds with the maximum height of the vault located to its south (L258). Although the southern vault was filled practically up to its ceiling with soil and modern construction debris, the similar levels indicate that the vault and floor might be contemporaneous, and that this vault, like the northern vault, was below ground when the floor was in use during the Crusader period. The soil fill in the vault north of the alley floor (L255; Figs. 3, 9) contained sherds, mainly from the Mamluk period (Fig. 10), a Late Roman coin (Table 1:14) and two Ayyubid coins (Table 1:11, 12). These finds confirm that by the Mamluk period, the northern vault was no longer used and had been turned into a refuse pit.
Three other vaults were documented c. 5 m north of the excavation area: a northern vault (L266; width c. 1.65 m; Figs. 11, 12); a center vault (L265; width c. 1.75 m); and a southern vault (L264; width c. 2.5 m). These vaults can be reached only from the south, through two vaults damaged by modern construction. The documented vaults were found filled almost to their ceiling with soil, such that their width and height could not be measured accurately. While documenting the vaults sixteen coins were discovered with a metal detector, six of which were identified as dating to the thirteenth–fourteenth centuries CE (Table 1:18, 19, 22–25). These coins probably indicate the final phase of use of the vaults. The three vaults, like the vaults that were destroyed and the two vaults excavated to their south, extended along the northern part of the eastern street in the Crusader marketplace, the continuation of what is today Shuq Ha-Zorfim and currently beneath the yeshiva building, forming a row of spaces east of the street which were used as shops.
The level of Plaster Floor 252 (753.32 m asl) in the alley separating the two excavated vaults seems to approximate that of the main Crusader street (what is today Shuq Ha-Zorfim), along which the vaults extended. If this is indeed the case, the presumed height of the documented vaults was c. 3.1 m. Based on the elevation difference between the floor of the three documented vaults and the Crusader-period alley (Plaster Floor 252; Fig. 12), we can conclude that the shops were built in two stories, one above the other: the excavated vaults (L255, L258) belong to the lower story and the documented vaults belong to the upper story. By plotting out a section running along an east–west axis through the three marketplaces and Vault 255 (see Figs. 11, 12), we see that the floor levels of the three streets are similar. The current absolute elevation of Shuq Ha-Qazzavim is c. 755 m asl. In a previous excavation (Margalit 1994), ancient flagstones were found at a depth of c. 1.5 m below the current street, that is to say, at an elevation of 753.5 m asl. This elevation is comparable to the level of the plaster floor dating to the Crusader period. If so, we can conclude that it was the pavement of the Crusader street that was exposed in the excavations on Ha-Qazzavim Street and not that of the Byzantine Cardo.
 
 
The Coins
Robert Kool
 
Forty coins were found in the excavation. Twenty-six were identified, although the identification of three of these (Nos. 4, 20, 21) is uncertain (Table 1).
 
Table 1. The Identified Coins
No.
IAA No.
Locus/
Basket
Ruler, Mint
Dating (CE)
Comments
1
139453
252/2503
Justin I, Thessalonica
518–522
 
2
139454
252/2515
Lucca, Italy
1099–1200
 
3
139455
252/2516
Byzantine
4th–5th centuries
 
4
 
252/2517
Umayyad, Egypt (?)
697–750 (?)
Thick coin
5
139456
253/2524
Mamluk
14th–15th centuries
Half coin
6
139457
253/2525
sir Nāsir al-Dīn Hasan, Damascus
1347–1351
 
7
139458
253/2527
Al-Kāmil, Muhammad I, Egypt
1218–1237
 
8
139459
254/2522
Theodosius I
383–395
 
9
139460
254/2523
Al-Kāmil, Muhammad I, Egypt
1218–1237
 
10
139461
255/2520
Byzantine
383–395
 
11
139462
255/2535
Saladin, as vassal of Nūr a-Dīn, Damascus
1171–1173
 
12
139463
255/2536
Al-Mansūr Muhammad I, Hamāh
1191–1220
 
13
139464
256/2519
sir Nāsir al-Dīn Muhammad 
1310–1341
 
14
139465
259/2538
Byzantine
364–367
 
15
139466
260/2548
Justin II
572/573
 
16
139467
260/2554
Constans II or Arab-Byzantine
641–670
 
17
139469
263/2560
Byzantine, nummus
500–550
 
18
139470
264/2563
sir Nāsir al-Dīn Hasan, Damascus
1348/1349
 
19
139471
265/2567
Mamluk (?)
14th–15th centuries
 
20
 
265/2568
Byzantine minima (?)
4th–5th centuries (?)
 
21
 
265/2569
Mamluk (?)
14th–15th centuries (?)
 
22
139472
265/2571
Al-Ashraf Nāsir al-Dīn Sha‘bān II, Damascus
1363–1377
 
23
139473
265/2573
James II, king of Cyprus, Cyprus
1460–1473
Half coin
24
139474
265/2576
Al-Zāhir Sayf al-Dīn Barqūq, Alexandria
1390–1399
 
25
139475
265/2577
sir Nāsir al-Dīn Muhammad, Cairo
1320/1321
 
26
139476
(W32) 2539
Umayyad
697–750
 
 
The limited excavation offers a glimpse at two periods when extensive construction work took place near the Cardo — the main road that crossed Jerusalem from north to south. The architectural, ceramic and numismatic finds discovered in the row of excavated and documented vaults confirm the commonly held assumption that the three Crusader markets were built along the route of the ancient Byzantine Cardo. Although only the row of shops on its eastern side was exposed and not the street itself, it is clear that the row of vaults, as well as the alley that separated the two excavated vaults, opened onto it. By integrating the excavation finds on the map of its surroundings and by checking the elevation of today’s three marketplaces, we can reconstruct the eastern street of the Crusader market, Shuq Ha-Zorfim, which originally continued north, as do the two adjacent market streets to its west, and was not shorter, as it is today (Zelinger and Erlich 2013).