^136. The Akkadian version of the incipit is written on the upper edge of the obverse, ahead of its usual place following the Sumerian version. There are several other examples of this practice in first millennium bilingual Sumerian-Akkadian tablets (see Gabbay Reference Gabbay2014: 235, with n. 54). As has been suggested by Cavigneaux and Ismail (Reference Cavigneaux and Ismail1998: 6), this practice is likely to have functioned as a means of quick reference within tablet collections or “libraries”.

^137. Lines 137–38 are good examples of highly divergent understandings between the Old Babylonian and late bilingual versions. In the Old Babylonian version the speech of Ninurta's divine weapon, the Šarur, clearly directs Ninurta not to go to battle; line 137 reads: ešemen ^dinana-ke₄ á-zu ba-ra-ni-zi “To the play of Inana, do not raise your arm!”. The late bilingual version of these lines directs Ninurta to do the opposite, i.e. to go to battle. Note Seminara's (Reference Seminara2001: 444) understanding of these lines as an “ideological” translation, reflecting a greater emphasis on Ninurta's martial aspect in the first millennium.

^138. I interpret the Sumerian postposition -e as a vocative. Cf. the opening line of Tablet I in the late version: Lugal-e u₄ me-lám-bi nir-ĝál “O king, a storm the glory of which is noble”. On the use of vocative -e see Woods (Reference Woods2000: 322–23), with earlier literature. Although it appears to be attested infrequently in early Sumerian literature, vocative -e may be considered characteristic of late Sumerian; it is, for example, a feature of the post-Old Babylonian (Emesal) Sumerian Eršahuĝa prayers (Maul Reference Maul1988: 6). The Sumerian of this line shows substantial differences with manuscript i₂ (BM 38838), such as the introduction of the -/ene/ plural suffix. Use of this suffix is irregular here, according to Old Babylonian conventions, as it is normally attached exclusively to personal nouns. In addition, the Akkadian version of this line does not reflect an understanding of “battle” as a plural. The -/ene/ plural suffix is used with non-personal nouns in the bilingual Šamaš-šuma-ukīn inscription VR 62, 2 (Jacobsen Reference Jacobsen, Cogan and Eph‘al1991: 284).

^145. There are clear traces of NIM for ^d a-nim in our tablet. The traces of this sign in manuscript t (4R² 23, 2) fit NIM better than van Dijk's proposed reading of -t]i ^?.

^146. The traces before ur₄-ur₄ could potentially be SA, known from the Old Babylonian version; note the proposed reading pes₁₀-sa in Geller (Reference Geller1985: 217). However, the traces fit NI better than SA. In addition, a conjugation prefix is expected at this point before the reduplicated verbal base. The sign UR₄ is very close to RI in Late Babylonian script; however, in this context, and with the parallel Late Babylonian manuscript u (BM 47892) the reading of these signs is clear. The usual form of UR₄ features a descending diagonal which ends the sign in a triangular point. The scribe of this tablet has a tendency to draw the diagonals very close or touching the horizontals, making the distinction between RI and UR₄ less distinct. See for example, the form of the similar sign TUK in bu-tuq-ti, in the Akkadian of this line.

^147. For the image of mountains rushing (ḫiāšu) towards Ninurta, see the bilingual version of line 125 in Angim (Cooper Reference Cooper1978: 76–77). By the introduction of negation, the meaning of this line is transformed. In the Old Babylonian version Ninurta's fearsomeness is the reason why he rushes to the mountain. In the first millennium version, Ninurta's fearsomeness is the reason why the mountains do not rush towards him.

^175. The newly attested bilingual equation between the ṣarṣaru snake and its Sumerian equivalent conforms to the lexical evidence (CAD Ṣ: 115).

^176. The clear ŠU sign in im-šu-ur confirms as correct the emendation in manuscript u: im-šu!(text: KU)-ur proposed in Geller (Reference Geller1985: 217–18). In manuscript i₂ I see ub-˹bil˺ (collated), confirming our tablet.

^177. The first GIG sign may possibly be read as sim_x. Cavigneaux (Reference Cavigneaux1987: 45) discusses this reading, the primary evidence for which is the Old Babylonian writing sim_x-ma. However, I know of no first millennium examples of such a writing. The Akkadian of this line probably omitted the sign -mi in sim- < mi > . The parallel Late Babylonian duplicate manuscript u is written syllabically (sim-ma). But the possibility of a construct state (sim marṣi) cannot be excluded. The use of the N stem of letû “to split”, which differs from the G stem of the only other manuscript (i₂) for this line, clarifies that the body of the earth became split as a direct result of being dried out, not as a separate action of the Asakku.

^179. In manuscript i₂ my collation confirms our tablet. In manuscript i₂ I see: ˹šà-ge-túm˺ ì-˹bal-bal˺ ka[lam-ma ság ba-ab-du₁₁] // ˹qer˺-[b]é-e-tum uš-b[al-kit-ma māta šuāta ispun]. The equation between qerbētu “environs, district” and šà-túm is well attested. The transformation from the Old Babylonian to late Sumerian, appears to be based on homophony, resulting in the shift ti > túm. This technique is well attested elsewhere in Lugal-e (Seminara Reference Seminara2001: 421–29).

^180. This line is central to the argument of K. P. Foster (Reference Foster, Milano, de Martino, Fales and Lanfranchi1999), who understands the Asakku as a personified volcano. For the various interpretations of Asakku, see Foster (Reference Foster, Milano, de Martino, Fales and Lanfranchi1999: 28), with earlier literature.

^181. The equation between matīma “ever, always” and me-da (or me-dam in manuscript i₂ for this line) is well known. I know no parallel for the variant me-dè in our tablet. sa₅ = ṣarāpu is not attested lexically, but the use of sa ₅ as a logogram for the verb ṣarāpu “to dye red” is attested (Farber Reference Farber1977: 91).

^182. My understanding of the Sumerian of this line partly follows Gragg (Reference Gragg1973: 25). Seminara (Reference Seminara2001: 272) interprets the Akkadian verb here, I think correctly, as ra’ābum “to shake, tremble”. The Akkadian translation understands íb as equivalent to agāgum “to be(come) furious”, contrary to its original function as a vocalic prefix (/i/) plus pronominal element (/b/) before the verbal base (íb-dúb). dúb = ra’ābum “to shake, tremble” is not attested, but note the well attested equation between dúb and râbu “to tremble, quake” in bilinguals (CAD R: 55–57). A confusion between the two verbs may explain the equation in our passage. The form ir-mu-um-ma could potentially be a pret. of ramāmum “to roar” plus -ma. The spelling differs in the parallel mss. x (BM 38433) and n₂ (BM 46971 + 43974), where it appears as ir-’u-um-ma. ra’ābum (CAD R: 2–3) is used together with agāgum in other contexts, and it is also used in passages where gods are the subject. Thus, it fits the context of our passage better than ramāmum. The shift from /m/ > ’ is known from certain contexts in late periods (see GAG³ §31d; Mayer Reference Mayer1992: 45–51). Evidence of the reverse shift ’ > /m/ is admittedly rare, but attestations in Late Babylonian literary texts have been compiled in Jiménez Reference Jiménez.2017: 279.

Colophon. The formula ina me-˹reš˺-tú la ú-šá-ka-áš “he must not knowingly/deliberately allow it (the tablet) to go missing” occurs, in variant forms, in Late Babylonian colophons from Uruk, Borsippa and Babylon. The particular spelling of the formula in our tablet is known from tablets which most probably come from Achaemenid Borsippa (see for this, and for discussion of the formula in general, Jiménez Reference Jiménez., Kogan, Koslova, Loesov and Tishchenko2016). The formula occurs together with the direction for the tablet to be returned to its master within a specific time scale, in Late Babylonian colophons from Uruk (Hunger Reference Hunger1968: nos. 91, 96, 97; CAD M/I: 140). I am not aware of any parallel for such a direction allowing only half a day; other examples in Late Babylonian colophons stipulate longer periods for the return of tablets, such as two days, the “same day” (ina ūmīšu) or “the same evening”(?) (ina kakkabī(TE)-šú; see Farber Reference Farber1987: 31, n.22). The phrase ina mišil ūmīšu could alternatively be interpreted as “midday” (cf. CAD M/2: 129). It is unclear which of these alternatives is intended in this context. The reading of SAR-tum as ḫubtu “robbery” (Borger Reference Borger1970: 166) cannot be excluded, but I prefer the reading sartu “falsehood, dishonesty” based on the context of our tablet. The phonetic complement to ĝeš in inaššī(ĝeš-ši)-šú supports the proposed reading of ĝeš as našû “to lift, carry”, doubted in Hunger (Reference Hunger1968: 13). See the discussion of this reading in Jiménez Reference Jiménez., Kogan, Koslova, Loesov and Tishchenko2016: 230, n.6.

¹ I would like to thank M. J. Geller, E. Jiménez, S. M. Maul and an anonymous reviewer, for their helpful comments on an earlier draft of this paper. I would also like to thank E. Jiménez and G. Nicolet for sharing their unpublished work with me. Of course, I remain responsible for all errors. BM 48053 is published here by permission of the trustees of the British Museum.

² See the introduction by J. E. Reade in Leichty (Reference Leichty and Reade)1986: xxxii).

³ The possibility cannot be ruled out that some of these unplaced small fragments, which are all of similar colour and script, may have been mistakenly attributed to BM 48053. For example, the larger fragment on the right side (see below) reads ˹mu˺-šal-˹lim˺, followed by mu-šal-lim(-)˹x˺, on the next line. A succession of lines involving such repetition does not fit with what is currently known of Lugal-e Tablet IV.

⁴ This explanation was proposed to me by I. L. Finkel. The tablet was baked on 12 May, 1971. In its current state the tablet measures 9.7 cm × 7.5 cm. See the entry for this tablet in the British Museum online catalogue at http://www.britishmuseum.org/research/collection_online/.

⁵ Other new sources include five Old Babylonian Nippur mss. in Peterson (Reference Peterson2011: 64–68). An Old Babylonian ms. from Mari (TH00-T062a + 00-T2081 + 00T211a = lines 118–124, 285–296) is unpublished, but a photograph and description appears in Nicolet (Reference Nicolet2015: II 224). Collations of Old Babylonian Ur mss. appear in Ludwig (Reference Ludwig2009: 17–21). New copies of Middle Assyrian mss. h (VAT 9306), d₁ (VAT 9710) and n₁ (BM 122625) can be found in Wagensonner (Reference Wagensonner, Selz and Wagensonner2011; Reference Wagensonner and Marti2014). Note also the post-Old Babylonian extract fragment from Nippur (lines 683’–87’; Peterson Reference Peterson2013).