L. 83 f.; G. Smith, _Disc_., 76; Budge-King, 155 ff., Le Gac,
172; the very fragmentary Obelisk, Le Gac, 207 ff.; KTA. 25; MDOG. 20,
21 ff.; 21, 15 ff. King, _Supplement_, no. 192, 470,
1805. Hommel. _Zwei Jagdinschriften_, 1879, with photographs;
Andrae, _Tempel_, 86 ff.]

Much the same situation as regards the sources is found in the reign
of his son Shalmaneser III (860-825). Aside from a few minor
inscriptions, our main source is again the official account which has
come down to us in several recensions of different date. The process
by which these recensions were made is always the same. The next
earlier edition was taken as a basis, and from this were extracted,
generally in the exact words of the original, such facts as seemed of
value to the compiler. When the end of this original was reached, and
it was necessary for the editor to construct his own narrative, the
recital becomes fuller, and, needless to say, becomes also a better
source. If, then, we have the original from which the earliest portion
of a certain document was copied or abstracted, we must entirely cast
aside the copy in favor of the contemporary writing. This would appear
self evident, but failure to observe this distinction has led to more
than one error in the history of the reign. [Footnote: The majority of
the inscriptions for the reign were first given in Layard,
_Inscriptions_, and in the Rawlinson publication, cf. for first
working over, Rawlinson, JRAS. OS. XII. 431 ff. The edition of
Amiaud-Scheil, _Les inscriptions de Salmanasar_ II, 1890, though
without cuneiform text, is still valuable on account of its
arrangement by years, as well as of its full notes, cf. also
Winckler-Peiser, KB. I. 128 ff. The one edition which is up to date is
N. Rasmussen, _Salmanasser den II's Indschriften_, 1907, though
the same may be said of the selections in Rogers, 293 ff.]