This small fragment in the Yale Babylonian Collection preserves the lower right corner of a tablet with entries of the first tablet of Diri. Eckart Frahm suggested that this fragment can eventually be considered a commentary on this lexical series. Like some Aa commentaries, the layout of the text is tabular with four subcolumns. On CCP 6.2.7.B, for instance, which probably contained a commentary on Diri VII, explanations are appended as a separate column. This appears not to be the case here, since the right edge is partially preserved. Noteworthy, however, is the constricted presentation of Akkadian equivalents, which are given as running text and separated by cola. The cola are also used to demarcate overflowing text of the subcolumns (r 6). A similar feature, although not as extensively, can be found in two further manuscripts of Diri I:
- BM 38592+ BM 64190 (154 × 135 × 30 mm) = ms. A is a large multi-column tablet that contained the whole text of Diri I. The layout follows that of the present tablet fragment with four subcolumns per entry. Occasionally Akkadian equivalents are combined in one line and separated by cola.
- VAT 17218 = ms. J. is a fragment from Babylon, which preserves subcolumn 4 and also combines Akkadian equivalents in one line separated by triple cola.
Whether commentary, excerpt or not, the present fragment is particularly important because the section on the logogram group ud.du is much better preserved than on ms. A. Civil offers a transliteration of the traces in subcolumn 4 of ms. A, which have been collated in April 2018. The logogram group ud.du is also treated in Aa III/3, 145–191. There, ud.du is, for instance, equated with ullulu ša išāti in line 167 followed by padû. Our text, on the other hand, has petû ša bābi, for which see also Antagal D, 62: è = min(petû) šá ká, and Astrolabe B (KAV 218) A ii, 26–27 and 35: ká / zu.ab ta-è = ba-⸢ab⸣ ⸢ap⸣-si-i ip-pat-te.
It is unclear how much of Diri I was originally preserved on our fragment. Even with several Akkadian entries combined per line it seems unlikely that the tablet originally contained the whole of tablet I (i.e., 360 entries).