pottery, bronze implements, and other remains almost identical with the most characteristic relics found in the terremare of the Po Valley. The objects disinterred from the third, or lowest bed, consisted of the refuse of flint and obsidian workings, hut-foundations, stone hearths, potsherds, etc. all common relics of the Neolithic Age. In the course of systematic excavations important structural features piles, platforms, huts, moat, rampart, etc., came to light, showing that the middle bed had been the site of a true terramara settlement. One of the huts was rectangular, 15.50 metres long by 5 broad, and had a kind of apse at the west end, and a portico at the other, with a kitchen opening on the central room. The walls were of wickerwork plastered over with clay, the foundations of which still remained, as well as portions of the clay castings of the wickerwork. In the kitchen were found hearthstones of earthenware and remains of a stove or cooking apparatus (Fig. 156, No. 6). The stove was covered with a perforated top, and had a square opening near its base. Remains of other two similar stoves were observed. Beneath the floor of this chamber the platform of an older palafitte was detected, thus proving that successive occupations, on the principle of raising the habitable area so as to get clear of the accumulated debris, were practised by the terramaricoli of Taranto, precisely the same as by those of the Po Valley.
According to Quagliati's description of his researches the station on the Scoglio del Tonno had all the structural features of the terremare north of the Apennines, viz., a moat, rampart, contrafforte (partly of wood and partly of stone), a palafitte, and finally a street 2 metres wide, running from east to west, supposed to be the Decumanus. These features were particularly well marked on the west side, which had a north and south direction. The moat, measuring about 5 metres wide and 3 deep, was partly excavated in the soft rock, and had a portion of its outside lined by a roughly built limestone wall.
From the structural analogy of this terramara to those of Emilia, Pigorini argues (B.P., xxvi., p. 12) that it supports his theory, viz., that whatever the nature of the site may be, whether on a hill or on the plain, there are certain characteristic features that are always adhered to.
