be impassable from the rise of the water in the marsh, we should conclude it would be amply sufficient. As there is no other part of the gulf in which it could, at any time, be absolutely necessary to pass over the sand-hills at all, we are at a loss to imagine why the army of the Bey, and that of his Roman predecessor, should have given themselves so much trouble in crossing them. Immediately after the marsh commences pasture land, and after five hours’ journey from Sachreen, we arrived at a place called Gartubbah, where we found some Arab tents, and established ourselves for the night. The next morning we proceeded on to Braiga, where we were led to expect, from the report of our Arab guides, that we should find a harbour full as good as that of Tripoly. Braiga has been a strongly-fortified post, as appears from the remains of several well-constructed and spacious castles which have been erected there. On the western point of the bay which constitutes the mersa (or harbour) is some tolerably high land, on which one of the forts has formerly stood; but which is now so much destroyed and encumbered with rubbish, as to offer little interest on examination. Along the same range of hills are other remains of building, originally connected with this fort, part of which we were induced to excavate, but found the chamber which we cleared to have been merely a storehouse for grain, or a reservoir for preserving water. It had been excavated in the rock, on the top of the range, and may be considered as offering an excellent example of the durable quality of the cement employed by the Romans in its formation[21]: for the stone in which it had been excavated had crumbled away, and left the cement with which the interior had been coated standing upright in its original position, in defiance of the storms of wind and rain which must have frequently assailed it from the sea. We found some Greek and Roman characters traced in the interior, and the representation of a ship and a palm-tree, of which copies will be found annexed, together with plans of the forts and of the chamber excavated. The surface of the cement on which these objects had been sketched was as smooth and as perfect as it could have been at any time, and we were in hopes, when we first saw the drawings, that others would be found on further excavation, and probably some inscription in Greek or Latin, by which we might have dated these productions. No other drawings or letters however were found, and we were obliged to content ourselves with taking copies of those described, and in making the plan of the chamber.