^{page 298 note 1} This is probably a personal name; the element RA occurs in the personal names on the stele of Manishtusu published by Father Scheil. It is unlikely that the reading is šarra, though a similar form occurs in the Cappadocian texts, the language of which is closely connected with the languge of the Agade period, e.g. in CCT., iv, No. 113383, 11. 19–20, ana kumra… dini; Lewy (ZA., N.F., iv, 244) considers this form a parallel to the status determinatus of Aramaic, but there is a good case for considering it an indeterminatus of the kind found elsewhere in Akkadian, the existence of which is denied by some. Mr. Gadd has suggested to me that ana . . . RA may be due to a mixture of Akkadian and Sumerian idiom by the scribe. But the interpretation really depends on the sense of the letter, and though 11. 4–21, might be addressed to a king they do not necessitate the assumption; LUGAL.RA may have been the king's principal officer in the south. The mention of Sharkalisharri in 28 precludes the addressee being the king if uma be taken as present, the natural view, in 29; if uma is preterite, then one of the alternative explanations given above may be correct. See note 17.

^{page 298 note 2} eruš 2nd s. imp., aruš 3rd s. pret., artš const, nomen agentis; we are justified in assuming an infinitive arašu for this period. The verb is of the u-conjugation, whereas in the First Dynasty of Babylon period it belongs to the i-conjugation. The forms are used by a man whose name points to the Middle Euphrates area, judging from the element Dagan, and may be a local dialect, aruš seems to be a medial form in sequence yahruš, aruš, eruš, eriš (vowel assimilation). The imperative eruš, on the other hand, must derive from huruš, and shows a vowel modification dependent on the softened guttural. Note the form of the sign e, and its clear distinction on this tablet. There can be no question of a scribal error, the assumption to which recourse is so regrettably frequent.

^{page 298 note 3} The generalizing ma, giving the singular a collective significance. The double m, which also occurs in 19, is contrary to the general avoidance of double consonants, and may be due to the special character of ma, see note 8.

^{page 298 note 4} The differentiated use of the negatives in this letter is interesting, a as in Gadd, and Legrain, , Ur Excavations. Texts, i, No. 276Google Scholar (where I should have transliterated and translated a limtud “may he not allot”, from madadu) with the preterite is the negative of the imperative; ula, with present, is the negative of the jussive, ula is also the negative of plain statement with either present or preterite. In the Louvre letter la is used with the subjunctive and with the jussive, and is not to be distinguished from ula by more than the emphasis. The difference in the stress is shown in later texts, e.g. Chiera, , Texts of Varied Contents, No. 55, 1. 21Google Scholar, šumma bitati šašu rabu la inaklcis u šumma ṣihru ula (not ṣihru(u) la) uratta” If those properties are larger (than the given dimensions) he shall not deduct therefrom, and if they are smaller he shall also not add thereto”. The Assyrians used ul and la alternatively with the indicative, and this is true of all later periods; la alone is used with the subjunctive, the weaker form being preferred as in French. But thoughout the history of the Akkadian language ul(a) and la may be parallel and the presence of la in parallelism or apposition is by no means a proof of a subjunctive, as some suppose.

^{page 298 note 5} ana. I take this to be for anna, demonstrative and indeclinable.

^{page 298 note 6} maṣ. I understand this to stand for the later meṣ, which appears in the Cappadocian texts as a declinable noun, but is here indeclinable. Though there is a slight break of the surface I consider the reading certain. The force of the question is: “How am I to do it?”

^{page 298 note 7} If the use of the signs da and ta in this letter be compared it may be thought that there is a method discernible, da is used (1) where it would appear in later script, e.g. danata/dannata, anadakum/anaddakkum, anadanukum/anaddanukum, danadanu/tanaddanu, tida/tida or tide; (2) it stands as the second syllable where (is doubled, e.g. adda/atta;) (3) it stands for the ta of later orthography in the initial position, danadanu/tanaddanu, danazar/tanassar. On the other hand, ta is used between vowels where t should not be duplicated, danata/dannata. In the Louvre letter the only instances of da fall under head (3). This suggests that the differentiation is between aspirate and unaspirated consonants, not that between fortes and lenes. But zudarib/suterib is difficult to avoid, and I can find nothing in the use of the other consonantal signs to suggest any system. I mention the matter in order to avoid any hasty generalizations, and to show that there is good ground for keeping a transliteration which differentiates different signs.

^{page 299 note 8} The rendering of this line is doubtful. I first thought that maggati might be pl. ace. of a noun connected with maqatu, and that the phrase might be compared to the later ṣuiṣṣbu niduta; it might then be rendered “cause the fallen habitations to be inhabited”. But the reduplicated consonant is unusual, and another view has suggested itself, danata, the permansive, requires some link with the imperative; that points to the use of ma as a disjunctive copula found in later texts, e.g. the Cappadocian. Thus in Lewy, , Die Kultepetexte der Sammlung Hahn, No. 6, 1. 12Google Scholar, where the sense of the letter, and particularly 1. 25, require the reading šuma inumi NN. ilikani ma garum ana išrišu iduar “If, when NN arrives, then the merchants' court makes a claim for his tithe”, ma begins a line and is appositional in force, aggati should then be a country or district, and is parallel to aggide KI in the Sumerian column of Poebel, Historical Texts, No. 34, and to the later orthography Akkad, so far as the doubled consonant is concerned. The use of ti for the later di is usual. The historical inscriptions, of course, all have A-ga-de, but a phonetic spelling in a letter is intelligible. The absence of the determinative is disturbing but not conclusive, and the sense is good. For the use of disjunctive ma see further Gadd and Legrain, op. cit., No. 275, 1. 29.

^{page 299 note 9} The use of zu where later orthography has šu points clearly to a pronunciation of the Shaph'el forms with sin for shin. The sibilants of the Agade period have been discussed by Thureau-Dangin, , RA., xxiii, 28Google Scholar. The whole question of the sibilants in cuneiform needs reconsideration, see JEA., xi, 238–9. I doubt whether the phenomena can be reduced to order in a land of mixed population where the people mostly wrote as they spoke. The same phenomena occur to-day.

^{page 299 note 10} The plural of KI.KAL does not require the repetition of KI. The meaning of the word is by no means clear. Both KI.KAL and KISLAH (KI. UD) are rendered nidutu or teriqtu; the latter is presumably from rēqu, and means land that has been cleared of buildings or cultivation. In many cases nidutu seems to mean waste land, but a discussion of KISLAH by Götze in Kleīnasiatische Forschung, i, 194, Anm. 1, proves that in the Hittite texts the word is used of a place where grain was stored. In the present text KI. KAL. KAL and agammu are obviously to serve some useful purpose, and since the cattle would naturally water at the agammu, it is natural to suppose that they will find fodder at the KI.KAL.KAL; that the word refers, in fact, to the išpiki kenutim of 1. 38. Now in the descriptions of the seven devils it is said that they jump about ina nigiṣi irṣiti and lie down ina niduti irsiti, and the parallelism suggests that as the former means “in a cleft in the ground” the latter means something like “in a hole in the ground”. Similarly these devils ina niduti irṣiti ittenenbu “rise out of a hole (?) in the ground”; tebu could hardly be used of a flat surface. Grain, of course, is not stacked in Iraq, for obvious reasons, but stored in jars or simply in holes below ground, and I suppose nidutu is ground lying fallow and used for storage in silos in this way.

^{page 300 note 11} This accusative to denote the intention of a verbal action is not easily paralleled in Akkadian. It may be classed as an accusative of extension in origin, delimiting as it were the action of the verb. For the accusative of extension see e.g. Driver, , Letters, No. 64, 18–21Google Scholar, ina ali (Kl)-ki 1 šiqlam kaspam riški ukal lu še-am ša šiqlim riški ukal” I will support you in your city to the extent of one silver shekel, or a shekel's worth of barley”. Accusatives in Akkadian have much broader uses than is generally recognized. The double accusative, for instance, after verbs of making, the existence of which has been denied (JRAS., 1926, p. 288), is to be found in one of the best literary texts extant, Sargon's Eighth Campaign, 1. 401, where there is mention of various statues ša (m) I štarduri mar Išpuen eru ZUN bit (ilu)Haldia ana eqi utirruma ištapuk ṣirššun “into which Sarduris son of Ispuinis had turned the copper in Haldis' temple by means of moulds by pouring it into them”. In fact, a special study of the accusative in Akkadian is badly needed.

^{page 300 note 12} The sign šik is quite clearly written; there is no need to assume an error of the scribe or the copyist. It is possible to read sik here and derive the verb from nasaku “to choose”, but the value ši occurs in Hammurabi's Code, 1. 12, and is preferable; the verb is then še'u “to search ”.

^{page 300 note 13} This word is unknown to me. Is it a preposition governing the noun? And is it in that case comparable to išbi, of which Ungnad has collected instances in MVAG., 1915, Heft 2? The sign for bi in that word is different. Or is it by any chance for šuba”i?

^{page 300 note 14} The ordinary usage does not require the preposition, see Thureau-Dangin, , RA., xxiii, 26Google Scholar, but the insertion is intelligible. Note that by the time of Sharkalisharri the form is already eli, not al, as previously.

^{page 300 note 15} For the subjunctive directly dependent on umawithout conjunction see Thureau-Dangin, loc. cit., 27.

^{page 300 note 16} I understand alakam to be 3rd s. preterite energicus with dative suffix of the 1st person pronoun, and compare aruš Others may have other explanations.

^{page 300 note 17} The interpretation of this u depends upon the tense of uma; if, as I translate, the verb is present, then u is the copula, and introduces a new address to the recipient. If uma is preterite the sequence is not clear.

^{page 300 note 18} For išpiku “grain-dump”, see Gadd, Tablets from Kirkuk, No. 43, Chiera, Texts of Varied Contents, No. 51.

^{page 300 note 19} I know no other instances of a verb mat(t, d)ak(q, g)u with which this word can be connected. If the word madaktu, not yet satisfactorily explained—the latest attempt may be found in AJSL., xli, 274—belongs to a root madaku “to encamp”, then muduk may be 2nd s. imper. ii, 1, and mean “make to camp”, with double accusative.

^{page 301 note 1} PSBA., 1901, May, and Old Testament in the Light of Historical Records, 477.

^{page 301 note 2} Sumerian and Babylonian Psalms, No. xxv.

^{page 301 note 3} Cambridge Ancient History, i, 424. Hommel, , Ethnologie und Geographie, 421Google Scholar, thinks the reference may be to the Kassites.

^{page 302 note 1} MAŠ.KI is not otherwise known in Southern Babylonia and I suggest with due reserve that MAR.KImay be meant.

^{page 302 note 2} Amurath to Amurath, pp. 134–5. The spelling there given, Shetateh, is an error which is perpetuated on the survey maps, though implicitly corrected by Reuther, Ukhaidhir.

^{page 304 note 1} Journal of the Manchester Egyptian and Oriental Society, xvi, pp. 31 ff., and further references there.

^{page 308 note 1.} Jensen reads the name of this deity Baba, and is supported by Thureau-Dangin, , Homophones sumiriéns, p. 40Google Scholar.

^{page 308 note 7.} Presumably a female proper name.

^{page 308 note 8.} This appears to describe the lady's relation to the governor for whose life the object is dedicated, but I do not understand the phrase.

^{page 308 note 9.} dumu seems to stand for dumu.sal as in Gadd, and Legrain, , Ur Excavations. Texts, iGoogle Scholar, No. 12, or No. 17. al alone is a peculiar personal name; one expects al-la, but the scribe is given to contracted forms.