835, but its main source is evidently the same. [Footnote: Andrae,
MDOG. 21, 20 ff. 39 ff.; Delitzsch, _ibid_. 52; KTA. 30; Langdon,
_Expository Times_, XXIII, 69; Rogers, 298f; 529.]

But the strongest proof of the existence of this edition is to be
found in the two fragments of clay tablets which are not, like all the
preceding, epigraphical copies of the originals, but form part of the
original itself. [Footnote: Boissier, RT. XXV. 82 ff.] These two bits
are written in the cursive style, and, though their discoverer
believed them to belong to separate documents, the fact that one so
closely supplements the other, and that they have the same common
relation to the other editions, justifies us in assuming that they
really do belong together. At first sight, it might be argued that
they are to be restored from the text of the Obelisk, with which they
often agree verbally. Closer inspection shows, however, that they
contain matter which is not found in that monument, and that therefore
they belong to an earlier and fuller edition, yet the resemblance to
the Obelisk is so close that they cannot be much earlier. On the other
hand, the Bulls inscription can be compared for the events of 854-852
and this has all that our tablets have, plus a good bit more. They
therefore belong between these two editions, and the only time we can
place them is 837. Since the clay tablets so fully abstract the Bulls
inscription wherever the latter is available for comparison, we may
assume that in 857-855 they give the minimum of that inscription. Thus
we have the editions of 845, of 837, and of 829, in a common line of
descent. Although for 857-856, there are numerous verbal coincidences
with the Balawat excerpts, it must be noted that not all the plus of
our tablets appears in that document, and we can only assume a common
source, a conclusion which well agrees with our characterization of
the Balawat inscription as a series of mere extracts. That this common