* A fragment of an anonymous list, discovered by Delitzsch, puts the expedition against the Tabal in 837 B.C. instead of in 838, and consequently makes the entire series of ensuing expeditions one year later, up to the revolt of Assur-dain- pal.  This is evidently a mistake of the scribe who compiled this edition of the Canon, and the chronology of a contemporary monument, such as the Black Obelisk, ought to obtain until further light can be thrown on the subject.

** For the site of Khashmur or Khashmar, cf. supra, p. 35, note 3.  The other localities cannot as yet be identified with any modern site; we may conjecture that they were scattered about the basin of the upper Diyalah.

*** Parsua, or with the native termination Parsuash, has been identified first with Persia and then with Parthia, and Rost still persists in its identification, if not with the Parthia of classical geographers, at least with the Parthian people.  Schrader has shown that it ought to be sought between Namri on the south and the Mannai on the north; in one of the valleys of the Gordysean mountains, and his demonstration has been accepted with a few modifications of detail by most scholars.  I believe it to be possible to determine its position with still further precision.  Parsua on one side lay on the border of Namri, which comprises the districts to the east of the Diyalah in the direction of Zohab, and was contiguous to the Medes on the other side, and also to the Mannai, who occupied the southern regions of Lake Urumiah; it also lies close to Bit-Khamban, the principal of the Cossaean tribes, as it would appear.  I can find only one position on the map which would answer to all these requirements:  this is in the main the basin of the Gave-rud and its small affluents, the Ardelan and the sources of the Kizil-Uzen, and I shall there place Parsua until further information is forthcoming on the subject.

Skirting Misi, Amadai, Araziash,* and Kharkhar, and most of the districts lying on the middle heights of the table-land of Iran, he at length came up with Ianzu, whom he seized and brought back prisoner to Assyria, together with his family and his idols.

* Amadai is a form of Madai, with a prothetical a, like Agusi or Azala, by the side of Guzi and Zala.  The inscription of Shalmaneser III. thus gives us the first mention of the classical Medes.  Araziash, placed too far to the east in Sagartene by Fr. Lenormant, has been located further westwards by Schrader, near the upper course of the Kerkha; but the documents of all periods show us that on one side it adjoined Kharkhar, that is the basin of the Gamas- ab, on the other side Media, that is the country of Hamadan.  It must, therefore, be placed between the two, in the northern part of the ancient Cambadene in the present Tchamabadan.  Kharkhar in this case would be in the southern part of Cambadene, on the main road which leads from the gates of