PETRA (i) _____, in ecclesiastical writers also al _____), the capital city of the NABATAEANS (q.v.), and the great centre of their caravan trade, is described by Strabo (xvi. p. 779) as lying in a level place, well supplied with water for horticulture and other uses, but encircled by a girdle of rocks, abrupt towards the outer side. The surrounding country was barren, especially towards Judaea ; the distance from Jericho was three to four days' journey, and from Phcenicum cn the Red Sea coast five (see plate VI., vol. vii.). Accord-ing to Pliny (217. H., vi. 144) che little valley of Petra is not quite 2 miles across, and lies at the junction of two roads, from Palmyra and Gaza respectively, 600 miles from the latter. These and other ancient notices leave no doubt as to the identity of the site with the modern Wady Musa in the mountains which form the eastern wall of the great valley between the Dead Sea and the Gulf of Akaba. Wady Musa lies just north of the watershed between the two seas, in 30° 19' N. lat. and 35° 31' E. long.1 Travellers coming up the Arabah usually approach the ruins of Petra from the south-west by a rough path, partly of artificial construc-tion 2; but the natural entrance is from the east down a narrow defile more than a mile long, called the Sik ("shaft"). The Sik is a contraction in the valley of a stream which comes down from the east, ris^pg in a spring now known as the Fountain of Moses (Am Musa),3 and passing between the villages of Elji and Aireh (Palmer). Both these places are ancient; the latter is the fortress Wo'aira of Yakut,4 while Elji, mentioned by Edrisi, is the " Gaia urbs juxta civitatem Petram " of the Onomasticon.b Below these and above the ravine the characteristic rock-cut tombs and dwellings of the Nabatseans begin to appear. But to reach the city proper from these upper settlements one must traverse the whole length of the defile, which is simply a narrow waterway, in some places not more than 10 or 12 feet broad, and walled in by rich brown or red precipices rising from 60 to 120 feet (De Luynes ; Stanley doubles this height) above the stream. In ancient times there was a paved path beside the channel, and remains of an arch spanning it are seen high in the air near the en-trance. Towards the lower end of the gorge, a turn in the dark path and the descent of a side valley admit a sudden flood of light, and here stands the most famous ruin of Petra, the so-called Khazna, or "treasury of Pharaoh," with a rich facade of late Roman style, not built but hewn out of the rose-coloured limestone. The next turn gives room for a rock-cut theatre, and from this point the gorge begins to open out into the little plain described by Strabo, and gives perhaps the most striking view of the multi-plicity of grottoes with elaborate classical facades which line the enclosing mountain-wall. The plain itself is strewn with ruins of temples and other buildings, and stairs once led up the rocky walls to higher structures, of which the most notable is now called the " convent" (Al-Deir). The grottoes are inhabited in cold weather by the Liya-thina Fellahln, who also hold the upper part of the valley, and are so troublesome and extortionate that no thorough exploration of the district has yet been carried out. It is not even known where the torrent-bed leads on leaving the plain . of Petra. De Luynes describes the water as wholly absorbed by the sands near the theatre, but there is an unexplored gorge to the south-west which is the con-tinuation of the valley.
The Nabataeans, as we see from Diodorus, used Petra as a place of refuge and a safe storehouse for their treasures of frankincense, myrrh, and silver before they gave up their nomadic habits. But Petra was not only safe and well watered, it lay close to the most important lines of trade. The modern pilgrim-road from Damascus to Mecca, which has taken the place of the old incense-route, passes indeed a little to the east by Ma'an. But to touch Petra involves no great detour even on this line, and in ancient times, when Gaza was the great terminus of the Arabian trade, Petra was the place where the Gaza road branched off from that to Bostra, Palmyra, and north Syria. The route from Egypt to Damascus is also commanded by Petra, and from it too there went a great route direct through the desert to the head of the Persian Gulf. Thus Petra became a centre for all the main lines of overland trade between the East and the West, and it was not till the fall of the Nabataean kingdom that PALMYRA (q.v.) superseded it as the chief emporium of north Arabia. Many Roman and other foreign merchants were settled here even in the time of Strabo, and he describes the caravans which passed between it and Leuce Come on the Red Sea coast as comparable to armies.
Petra is a Greek name which cannot have been that used by the Semitic inhabitants, and from Josephus (Ant., iv. 7, 1 ; 4, 7) and the Onomastica (ed. Lag., p. 286 sq.) it may be concluded that the natives called the place Rekem (Dpi), a designation probably derived from the variegated colours of the rocks about Wady Musa, to which all travellers refer with admiration.' But Petra had yet another ancient name familiar from the Bible. The Biblical Sela (generally with the article 5?l>Dn), a city of Edom (2 Kings xiv. 7 ; Isa. xvi. 1 ; also Judges i. 36, where E. V. has " the rock " ; perhaps also Isa. xlii. 11), appears to be identified with Petra by the LXX., and certainly is so by the Onomastica. Petra, in fact, or the "rock," seems to be simply a translation of Sela, but a somewhat loose one,for the Hebrew name, corresponding to the Arabic Sal', is properly a hollow between rocks, just such a place as Petra is. The fortress of Edom, according to Obadiah 3, lay "in the clefts of the Sela," and seemed impregnable. And that the name of Sela survived the Nabataean occupation is known from Yakut, who places a fortress Sal' in Wady Musa (comp. Noldeke in Z.D.M.G., xxv. 259). Petra, therefore, was a city before the Nabataeans, and, occupying one of the few cultivable spots in the dis-trict, probably never wholly ceased to be inhabited. This identification disposes of another which was accepted alike by the Jewish and Christian Aramaic versions of the Old Testament, and, passing from the Aramaeans to the Arabs, has given rise to the modern names Fountain and Wady of Moses (comp. Yakut, iv. 879). According to these versions Rkem, Rkam, or more precisely Rkem of Gaia (that is, Elji), is Kadesh Barnea, where flowed the waters of Strife or " well of judgment" (Gen. xiv. 7 ; Num. xx. 1 sq., xxvii. 14), where Moses struck the rock. This view is ably supported by Greene (The Hebrew Migration from Egypt); others identify Kadesh with Ain Kadis (Kudais) on the south border of Judaea.
Petra survived the fall of the Nabataean kingdom, and indeed most of the buildings may be dated from the 2d and 3d centuries. It appears from coins that Hadrian took it into favour and gave it his name. But Palmyra absorbed its trade with the Persian Gulf, and long before Islam the great incense-route was deserted and left Petra, like the more southern Nabataean city of Egra (Hijr), to fall into ruin. The ruins were an object of curiosity in the Middle Ages, and were visited by Sultan Bibars (Quatremere, I.e.). The first European to describe them was Burckhardt, and since his time they have often been visited. See the descriptions, plans, and views of Laborde and Linant, Arable Petree (Paris, 1830-34) ; the Due de Luynes, Voyage à exploration à la mer morte, &c., Paris, s.a. ; Palmer, Desert of the Exodus, vol. ii., 1871 ; Stanley, Sinai and Palestine ; Guerin, Terre Sainte, 1883. (W. R. S.)
705-1 The latitude and longitude are taken from De Luynes's map. Ptolemy, who, according to Olympiodorus, spent some time in Petra, and doubtless owes to this fact his excellent information about the caravan-routes in Arabia, gives the latitude, with surprising accuracy, as 30° 20'.
2 Compare Diod., xix. 97, who describes the Nabatsean fortress it was not a town at the time in question (312 B.C.), for the Nabatseans were still nomads when they were attacked by Antigonusas ascended to by a single artificial path.
3 This seems to be the fountain mentioned by Nowairi, ap. Quatre-mere's Melanges, p. 84, which flowed with blood and was changed to water by Moses. The name Od-dema, which gave rise to this legend, may possibly be a relic of the old name of Edom.
4 Perhaps also the Irani, DTJ?, of Gen. xxxvi. 43.
5 See Tuoh's Genesis, 2d ed., p. 271, note.
Arabia Petraa is not properly Stony Arabia, but the Arabia of which Petra is the centreri Kara Hirpav 'Apa/3ict of Agathemerus.
The rock-hewn city of Rakim (Istakhri, 64 ; Geogr. d'Abulf., Fr. tr., ii. 2, 5), which Schultens (Ind. Geog. in Vit. Sal.) proposes to identify with Petra, is a different place, close to 'Amman (Mokaddasi, p. 175).
The above article was written by: Prof. W. R. Smith.