these things, and had made such previous preparations as would enable me, in a great measure, to correct any mistakes I might commit in making an entry. I had drawn out, as I have already stated, sketches of the two most important maps of this province, and had marked particularly where they differed from each other, or from any other map I had seen. These I carried constantly with me, and referred to them every few minutes. I could thus see at a glance where they were defective or inaccurate, and on every such occasion I inserted the necessary correction in the sketches themselves as well as in my note-book. That my own map is free from errors I by no means affirm; but such future travellers as may traverse this interesting country will, I am confident, at once admit that it is far more full and accurate than any yet published.

We left 'Ary at 11:30, and, having learned that our servants had gone on in advance, we set out at a rapid pace towards Mujeimir, which now appeared before us on the eastern declivity of a conical tell in the plain. The valley of 'Ary extends up on the left to the base of the mountains, about 3 miles distant, and on its southern side is a low ridge of rocky mounds, a spur from the main chain, extending a considerable distance into the great plain, and shutting in the view in front. We reached Mujeimir in twenty minutes; the houses are all of stone and of considerable antiquity, like all the others in this region. We felt disappointed in not finding our servants here, as we had been told they were awaiting us; but we supposed they had followed the main road to Busrah, which runs on the western side of the tell. We

rode on without dismounting, and, after passing the houses, commenced to ascend the easy slope of the ridge above alluded to. At 12.30 we reached the small village of Wetr, situated on its summit. Some distance to the south-east we observed the much more extensive ruins of Ghussân on a tell. Abulfeda, in his Historia Anteislamica," mentions a tribe or family of Arabs which came from Yeman, in southern Arabia, to this country, several centuries before Mohammed, and whose chiefs were called Melûk Ghassan, Kings of Ghassân." They probably took their name from this place. It was one of these princes who was above referred to as having erected the castle of Suweideh.

The vast plain southward now opened up before us, dotted thickly with deserted cities and villages. That broad black belt in front, with the massive towers and battlements rising up in the midst of it intermixed with tapering column and minaret, is Busrah. Jemurrîn,

Keires, Burd, Ghûsam, and a host of others, are seen on each side; while on the summit of yon graceful hill on the eastern horizon is the ancient castle of Sulkhad, and there, in the intervening valley, are the wide-spread remains of Kureiyeh.

Here we found our servants, and sent them forward along the straight road to Busrah; while we, turning a little to the right, galloped across some rich fields to a large and massive building called Deir Zubeir. We reached it in ten minutes. It is a square structure, with thick stone walls, and has probably been latterly used, as

• Ed. Fleischer, p. 128.

Around it

its name would seem to imply, as a convent. are clustered a few houses, with stone doors and roofs, but the whole is now deserted. From hence we rode towards Jemurrîn, lying between us and Busrah. Soon after leaving the Deir we struck the Roman road which has already been referred to as running in nearly a straight line from Suweidah to Busrah. The pavement is in some places quite perfect, and the line of the road, extending across the fine plain as straight as an arrow, is clearly marked. This, I believe, is still the road laid down in the Pentinger Tables. The next station after Chanata is Rhose, and the distance between the two places is twenty miles; now the distance of Busrah from Kunawât is exactly twenty Roman miles. It is easy to understand how the carelessness of a transcriber or the ignorance of a copyist might make Rhose out of Bostra.

Following the line of the Roman road for about fifteen. minutes we reached the brow of the Wady Zêdy, a deep, narrow, and rugged ravine, extending across the plain like a huge fissure. In the bottom is a small stream flowing lazily over its rocky bed. A fine Roman bridge of three arches here spans it, crossing which we rode up to the village. Jemurrîn stands on a gentle eminence on the south bank of the Wady. It is of considerable extent, and contains the ruins of some large and handsome buildings. A lofty square tower beside the bridge was the first that attracted our notice from its resemblance to the tombs we saw in Kunawât. In front is a tablet for an inscription, but we did not stop to examine it. Burck

Reland, Palæs. p. 421.

hardt states that it contains a long Arabic inscription, with the Greek name pnλixos over it.2

Leaving Jemurrîn we followed the road which leads. thence to the eastern extremity of Busrah; and a few minutes' gallop brought us up to our servants. At 1.40 we stood beside the ruins. Our servants, who had proceeded at their usual pace, were thus an hour and a quarter in going from Wetr to the city. The whole distance from 'Ary is a little over six miles. Our first object was to procure a house, or at least some apartment we could call our own, for the approaching Sunday. Mahmûd soon found a small room in the sheikh's house, where our baggage was speedily stowed away and arrangements made for ourselves. The people we had already met with here, though not actually uncivil, showed us no attention or kindness; and some of them even regarded us with scowling looks. We resolved to be on our guard, but to take no notice of incivility; and we at once engaged the brother of the sheikh to guide us over the ruins. This we deemed necessary to save us from insolence and abuse.

That the reader may be able to follow me in my attempts to describe the various ruins of this great city, I have constructed a plan, partly from general measurements, but chiefly from rough sketches made from several positions; it will serve to show the relative positions of the more important ruins and the form of the city.

In form the walled city was almost rectangular; as nearly as I could estimate, a mile and a quarter in length,

2 Trav. in Syr. p. 106.