We should feel obliged to insist upon this conclusion, even in the absence of any remains of a harbour in the situation indicated. Our task, however, is not so arduous; for manifest traces of such a harbour have been identified. In the first place, traces of a harbour in the district above mentioned came to view in 1819, and were then officially noted by so competent an authority as the Patriarch Constantius.[1174] In that year a great fire burned down a large part of the Turkish quarter near Yeni Kapou—Tulbenkdji Djamissi—and brought to light a portion of an ancient circular enclosure around that quarter. The discovery excited considerable attention, and the patriarch was specially instructed by the Turkish Government of the day to examine the wall and report the result of his investigations. Accompanied by two distinguished members of the Greek community, the prelate proceeded to the scene of the conflagration, and found a wall built of huge blocks of stone, about seven feet long, four and a half feet wide, and over a foot thick. The stones were carefully hewn and placed in three tiers; the blocks in the two lower tiers being the ordinary limestone found on the banks of the Bosporus, while the blocks in the highest row were of marble from the Island of Marmora. The territory enclosed by the wall presented the appearance of a great hollow which had been filled in, since the Turkish Conquest, and raised to afford ground for building. All that the patriarch saw convinced him that he stood upon the site of one of the ancient harbours of the city. The wall has disappeared, as the excellent building material it provided rendered natural. But other remains of a harbour at this point, the complement of those discovered by the patriarch, have been recognized, and can, to some extent, be still distinguished. Off the shore in front of the territory enclosed by the wall described above is a mole formed with boulders (marked “Molotrümmer” on Stolpe’s map of the city), similar to the mole before the old harbour at Koum Kapoussi. At a point about half-way between Koum Kapoussi and Yeni Kapou, there is a wide gap in this mole, dividing it in two unequal parts, and forming a passage through it. The shore[1175] opposite the gap was, until the construction of a quay in 1870 for the Roumelian railroad, a sandy beach extending back to the foot of the city walls. The portion of the walls at the rear of the beach was, however, not Byzantine; but a piece of Turkish work[1176] inserted between the Byzantine walls on either hand to close an opening which gave admittance to the area occupied by the quarter of Tulbenkdji Djamissi.