The pottery found in this tomb offered no very distinctive features, being coarse and unglazed, but the numerous fragments of ostrich egg-shells, coloured and scratched with rough patterns in bands, also pointed to a Phoenician origin, or at least to a race of wide mercantile connection: and in those days the Phoenicians were the only people likely to combine in their commerce ostrich egg-shells and ivory. We also found small shapeless pieces of oxidised metal, brass or copper. There were no human bones in the upper chamber, but those of a large animal, presumably a horse. The chamber immediately beneath was much more carefully constructed; it was exactly the same length, but was higher, being 6 feet 7 inches, and the passage was wider. It was entirely coated with cement of two qualities, the upper coat being the finest, in which all round the walls at intervals of two feet were holes sloping inwards and downward. In similar holes, in one of the other tombs we opened, we found traces of wood, showing that poles on which to hang drapery had been inserted. The ground of this lower chamber was entirely covered with a thin brown earth of a fibrous nature, in appearance somewhat resembling snuff; it was a foot in depth, and evidently the remains of the drapery which had been hung around the walls. Prior to the use of coffins the Phoenicians draped their dead,[6] and amongst this substance we found traces of human bones. Thus we were able to arrive at the system of sepulture employed by this unknown race. Evidently their custom was to place in the upper chamber broken utensils and the body of an animal belonging to the deceased, and to reserve the lower chamber for the corpse enshrouded in drapery. For the use of this upper chamber our parallels are curiously enough all Phoenician. Perrot gives us an example of two-storied tombs in the cemetery of Amrit, in Phoenicia, where also the bodies were embedded in plaster to prevent decay prior to the introduction of the sarcophagus, reminding us of the closely cemented lower chambers in our mounds. A mound containing a tomb with one chamber over the other was in 1888 observed in Sardinia, and is given by Della Marmora as of Phoenician origin. Here, however, the top of the tomb is conical, not flat, as in our mounds, which would point to a later development of the double chamber which eventually blossomed forth into the lofty mausolea of the later Phoenician epoch, and the grandiose tombs of Hellenic structure. Also at Carthage, that very same year that we were in Bahrein, _i.e._ 1889, excavations brought to light certain tombs of the early Phoenician settlers which also have the double chamber. In answer to Perrot's assertion that all early Phoenician tombs were _hypogea_, we may say that as the Bahrein Islands offered no facility for this method of sepulture, the closely-covered-in mound would be the most natural substitute.