SELAH, or rather SELA' (Sbp, 'rock,' with the article in 2 Kings xiv. 7, y9ort the rock ;' Gr. 47 Ilirpa, Petra, which has the same signification as Selah ; sometimes plural, al II&pac), the metro polis of the Edomites in Mount Seir. In the Jewish history it is recorded that Amaziah, king of Judah, slew of Edom in the valley of Salt ten thousand, and took Sclah by war, and called the name of it Joktheel unto this day ' (2 Kings xiv. 7). This- name seems however to have passed away with the Hebrew rule over Edom, for no further trace of it is to be found ; and it is still called Selah by Isaiah (xvi. r). These are all the certain notices of the place in Scripture ; for it may well be doubted whether it is desig,nated in Judg. i. 36 and Is. xlii. r, as some suppose. We next meet with it as the Petra of the Greek writers, which is merely a translation of the native name Selah. The earliest notice of it under that name by them is connected with the fact that Anti gonus, one of Alexander's successors, sent two ex peditions against the Nabathxans in Petra (Diod. Sic. xix. 94-98). For points of history not imme diately connected with the city, see IDUM,EA ; NEBAIOTH. Strabo, writing of the Nabathmans in the time of Augustus, thus describes their capi tal :--` The metropolis of the Nabathmans is Petra, so called ; for it lies in a place in other respects plain and level, but shut in by rocks round about, but within having copious fountains for the supply of water and the irrigation of gardens. Beyond the enclosure the region is mostly a desert, especi ally towards Judxa ' (Geog. xvi. p. 906). At this time the town had become a place of transit for the productions of the east, and was much resorted to by foreigners (Diod. Sic. xix. 95 ; Strabo, /. e.) Pliny more definitely describes Petra as situated in a valley less than two miles (Roman) in ampli tude, surrounded by inaccessible mountains, with a stream flowing through it (Hist. Nat. vi. 28). About the same peiiod it is often named by Jose phus as the capital of Arabia Petrxa, with which kingdom it passed under the immediate sway of the Romans in the tinae of Trajan, whose successor Hadrian seems to have bestowed on it some advan tage, which led the inhabitants to give his name to the city upon coins, several of which are still ex tant (Mionnet, Med. Antiques, v. 587 ; Eckhel, Doctr. Num. ii. 503). In the 4th century Petra. is several times mentioned by Eusebius and Je rome ; and in the Greek ecclesiastical Notitix of the 5th and 6th centuries it appears as the metro politan see of the third Palestine (Reland, Palast.
pp. 215, 217) ; the last-named of the bishops is Theodorus, who was present at the council of Jerusalem in A.D. 536 (Grim' Christ. iii. 725). From that time not the slightest notice of Petra is to be found in any quarter ; and as no trace of it as an inhabited site is to be met with in the Arabian writers, the probability seems to be that it was destroyed in some unrecorded incursion of the desert hordes, and was aftenvards left unpeopled. It is true that Petra occurs in the writers of the era of the Crusades ; but they applied this name to Kerek, and thus introduced a confusion as to the true Petra which is not even now entirely removed. It was not until the reports con cerning the wonderful remains in Wady Musa had been verified by Burckhardt, that the latter traveller first ventured to assume the identity of the site with that of the ancient capital of Arabia Petr-xa. He expresses this opinion in a letter dated at Cairo, Sept. r2th, 18'2, published in '8E9, in the preface to his Travels in Nubia ; but before its appearance the eminent geographer, Carl Ritter, had suggested the same conclusion on the strength of Seetzen's intimations (Erdkunde, 117). Burckhardt's view was more amply de veloped in his Travels in Syria, p. 43r, published in 1822, and received the high sanction of his editor, Col. Leake, who produces in support of it all the arguments which have since been relied upon—namely, the agreement of the ancient de scriptions with this site, and their inapplicability to Kerek ; the coincidence of the ancient specifi cations of tbe distances of Petra from the Elanitic Gulf and from the Dead Sea, which all point to Wady Musa, and not to Kerek ; that Josephus, Eusebius, and Jerome testify that the Mount Hor where Aaron died was in the vicinity of Petra ; and that to this day the mountain which tradition and circumstances point out as the same, still rears its lonely head above the vale of Wady Musa ; while in all the district of Kerek there is not a single mountain which could in itself be regarded as Mount Hor, and even if there were its position would be incompatible with the recorded Journey ings of the Israelites (Leake's Preface to Burek hardt's Travels in Syria, pp. vii.-ix. ; Robinson's Pa/atine, 576-579 ; 653-659).