sides of the Jezzîn valley are in contrast, for on the west the hills are thickly wooded and less steep, while the eastern hills, though terraced and planted on their lower slopes, culminate in headlands of solid rock which round out from the cliffs like huge towers.
One of these projections is more rounded than the rest and is not unlike the first rude fashioning of the head of some gigantic statue. It is called the Castle of Fukhredeen Ma'an, from the Druze chief who held it against the Turks. Notices of this castle have appeared in Churchill, Thomson, &c., but I have not seen a full description. The approach is somewhat difficult but not at all dangerous, for a rough path steeply climbs up a break in the cliff where enough soil clings to support brambles and wild fig. We could easily understand, however, how different would have been the attempt to get up to the castle had it been held by an enemy. A natural ledge of rock runs around the centre of the headland, and this was probably what suggested the use of the place as a stronghold. The approach to the ledge from the continuation of the cliffs to the north was probably guarded by a drawbridge, for, on reaching the level of the ledge, we found a deep cut in it, about 15 feet across. The entrance was also once guarded by a tower of masonry, of which we saw the foundations. The ledge begins as a narrow groove in the face of the cliff, but further on has a width of 12 feet. Towards its southern end the rock of the cliff above projects over the ledge, beyond the precipice below, making a roof which slants upwards and outwards. The length of the ledge is about 500 feet; at first, after leaving the "drawbridge," it slants upwards but later runs quite horizontally. Where the ledge slants the edge is stepped down, as if to secure stones which may have formed an ancient parapet. In the cliff back of the ledge, opening from it, are two chambers hewn in the solid rock. In the floor of the ledge, as well as in the wall of the cliff, there are lines of small square holes, as if for beam-ends, tier upon tier, as I found at Ma'lula. This suggests that a second storey of wood was built against the cliff' above the ledge and explains several steps hewn in the cliff high above the ledge floor leading to a door hewn in the face of the cliff. These steps probably connected this chamber with the second storey. We need not suppose that such a second storey above the ledge was very substantial; it may have been constructed like the watch towers erected on stout poles in the Lebanon vineyards to-day, and floored with boards and brush. We counted five openings to chambers in this upper level, one of which was walled up with nicely-cut white stone. As these openings were in the face of the cliff some 15 feet or 20 feet above where we stood, we could not get up to examine the masonry or see whether the chambers were connected by galleries.
In the rock floor of the ledge two pear-shaped excavations had been made, large at the bottom and narrowing to a small mouth, 16 feet deep. These were evidently used for storing corn, as they are of the same shape as the Bedawin corn-pits. The one we examined was nicely plastered. Another square excavation served for a pool. Water was brought
