call I. To be sure, the Arabian expedition already occurs in B, but I
has also sections which appear only in A, and which therefore probably
date later. The one indication that points to its being later than A
is the fact that, while A ascribes these actions to his generals, our
document speaks of them in the first person. [Footnote: K. 2802;
G. Smith, 290 ff.] Still later are the Beltis [Footnote: II R. 66;
G. Smith 303 ff.; S. A. Smith, II. 10 ff.; cf. I. 112; Jensen,
KB. II. 264 ff.; Menant, 291 ff.] and Nabu inscriptions, [Footnote:
S. A. Smith, I. 112 ff.; III. 128 ff.; Strong, RA. II. 20 ff.] though
as these are merely display inscriptions, the date matters
little. Here too belongs J in spite of its references to the
accession. [Footnote: K. 2867; S. A. Smith, II. 1 ff.; cf. Olmstead,
_Bull. Amer. Geog. Soc._, XLIV. 434.--The various British Museum
fragments, cited in King, _Supplement_, seem to be of no special
importance for this study as they are duplicates with few variants.]
And to this very late period, when the empire was falling to pieces,
is to be placed the hymn to Marduk which speaks of Tugdami the
Cilician. [Footnote: S. A. Strong, JA. 1893, 1. 368 ff.]

We have already crossed the boundary which divides the really
historical narratives from those which are merely sources. Among the
latter, and of the more value as they open to us the sculptures, are
the frequent notes inscribed over them, [Footnote: Scattered through
the work of G. Smith, cf. also Menant, 287 ff.] while a number of
tablets give much new historical information from the similar notes
which the scribe was to thus incise. [Footnote: K. 2674; III R. 37;
G. Smith, 140 ff.; S. A. Smith, III. 1 ff. K. 4457; G. Smith, 191
ff. K. 3096; G. Smith, 295 ff.] The Ishtar prayer is a historic
document of the first class, the more so as its author never dreamed
that some day it might be used to prove that the king was not