The spring of ‘En Nun emerges south of the Ginosar Valley, on the northwestern shore of the Kinneret. Nahal Nun flows from it south for about 200 m, and then turns eastward and flows to the lake. The Ginosar Valley is situated in the delta of Naḥal Arbel, Nahal Zalmon and Nahal ‘Amud, and it was cultivated in the Roman period and thereafter. Josephus Flavius describes it thus: “its soil is so fruitful that all sorts of trees can grow upon it… it supplies men … with grapes and figs continually, during ten months of the year, and the rest of the fruits as they become ripe together through the whole year” (Wars III:10.8). The valley was also an optimal place to grow sugar cane. It is flat, with high summer temperatures and no frost in winter. The springs at the foot of the mountains, including ‘En Nun, were water sources used to irrigate the fields and for the sugar processing installation discovered at Ḥorbat Minnim (Stern 1999:49, 84).
This was the first excavation at the site, and two excavation squares were opened (A and B), c. 80 m apart. In Area A was a round pool built over the spring in the Roman period, from which water was drawn to irrigate fields to the east. At some point, not before the Byzantine period, the pool was reinforced with an embankment. In the Ottoman period, the pool was cleaned and reused. In Area B, a small part of a dam built in Nahal Nun was uncovered. The dam was apparently built after the pool went out of use, perhaps in the Early Islamic period.
 
Area A. The outline of a round pool (diam. c. 35 m; Figs. 2, 3) built above the spring of ‘En Nun could be seen on the surface prior to the excavation. The excavation took place within the pool, immediately to its north and south. The perimeter wall of the pool (W101; height 4 m) was built of two rows of basalt stones coated on the side facing the pool with a thick layer of hydraulic plaster; only 1 m of the plaster survived on the uppermost part of the wall. On the northern side of the pool was an embankment (L106, L112; Fig. 4) abutting the outer face of W101, built to prevent the great mass of the water from breaching the pool’s wall. The embankment was built of a series of square cells that were filled with basalt rocks mixed with mortar consisting of plaster, charcoal chips and sherds. The bonding material had dripped into the bottom of the embankment, filling the spaces between the stones and solidifying into a conglomerate too hard to excavate manually (Fig. 5). Adjacent to the embankment cells on the north, another reinforcement was added made of a cluster of basalt stones (L102) bonded with the same mixture. This conglomerated cluster, which had cracked over the years, seems to have been the last reinforcement added to the embankment. The few sherds found among its stones date from the Byzantine period (fourth–seventh centuries CE), setting a terminus post quem for the embankment’s construction.
A floor (L109, L110) abutting W101 revealed two different construction methods: along the wall of the pool and up to 1.5 m from it, the floor was made of a layer of compact travertine with potsherds, covered by a layer of flat basalt stones, pebbles and sherds (L109; Figs. 6, 7). The sherds, all very worn due to their waterlogged environment, were dated to the Roman period—the time of pool’s construction. The central part of the floor comprised large basalt pebbles (L110) set on a thick, dark, impermeable layer of clay, devoid of sherds (Fig. 8).
A small drainage channel (L113; Figs. 9–12) was built into the embankment south of the pool (L115) and cut through the southern wall of the pool. A fragment of an Ottoman period smoking pipe was recovered between the stones of the channel wall. This smoking pipe, dated to the eighteenth century CE, seems to set a terminus post quem for the construction of the channel. It appears that in order to build the drainage channel the outer face stones of W101 and part of the embankment 115 were dismantled, to create the opening and drainage channel. The channel’s opening was not found due to a later addition of concrete at its upper end, but it is clear from the height differences along the channel that it drained water from the pool southward, rather than feeding it; thus, it was probably installed to empty the pool for maintenance and cleaning of the colluvium that accumulated in it.
 
Area B yielded a segment of an elongated, east–west dam (L200; length c. 12 m, width 1.5 m; Figs. 12, 14) traversing the streambed of Nahal Nun; only a small part of this dam was preserved, resting against the eastern bank of the streambed. An opening in the dam (L203) set for water passage could be closed by a shutter, which was set into hewn grooves in the basalt rocks on the northern side of the opening (Fig. 15). The southern face of the dam was coated with hydraulic plaster only from the water passage to the eastern bank of the stream (Fig. 16). Thus, it seems that after the water flowed through the passage from north to south, it did not continue farther downstream to the south, but rather was diverted and channeled eastward along the southern face of the dam to fields east of the stream. The western part of the southern face of the dam, west of the water passage, was not plastered (Fig. 17); this part of the dam was built at an angle that indicates that it rested on a foundation, but it did not survive. In a probe (L202) excavated south of the western part of the dam were layers of compact travertine devoid of sherds, perhaps the dam’s foundation. The water passage was blocked by accumulations of soil and colluvium containing worn sherds from the Early Islamic period. No finds were discovered near the dam that could date the time of its construction.
 
The pool was built during the Roman period directly over the spring. The foundations of the wall surrounding the pool and the floor abutting the wall also dated from the Roman period. Presumably, the water pressure on the pool’s wall eventually breached it, and the embankment was subsequently built to support the wall. Based on the few sherds discovered among the stones of the embankment, it seems to have been constructed no earlier than the Byzantine period. Although we know that sugar production was well developed in the Ginosar Valley during the Crusader, Ayyubid and Mamluk periods, no ceramic finds from these periods that would have indicated the continued use of the pool were found. Thus, the pool may have gone out of use prior to these periods, perhaps during the Early Islamic period.
The pottery discovered in the opening of the dam is worn and meager, and as it originated in colluvium, it can date neither the dam’s construction nor its use. Nevertheless, it seems to have been built after the pool went out of use, because both served the same purpose—channeling water to fields to the east. If the pool indeed went out of use in the Early Islamic period, the dam may have been built to replace it and channel at least some water to fields east of the streambed before it flowed into Lake Kinneret and was lost.
It seems that during the Ottoman period, in the eighteenth century CE at the earliest, a drainage channel was built in the pool to clean and reuse it (Fig. 18).
The Ginosar Valley, extending north of the excavation area, was cultivated throughout all historical periods during which the area was inhabited. The proximity to Lake Kinneret enabled the distribution by boat of agricultural products to all the settlements around the lake shore. Flour mills operating in the estuary of Nahal Zalmon, about 1 km north of the excavation, served also as sugar mills, in which sugar cane was crushed, and the sweet liquid was pressed from it (Stepansky and Abu-Raya 1994). Thus, there is no doubt that the ancient inhabitants of the valley tried to harness the water from ‘En Nun to irrigate the surrounding fields.