¹ See Fiey, J. M., Assyrie Chrétienne II, 493–524Google Scholar for the history of Nebi Yūnus, especially its Christian connections; and also briefly in Sarre, F. and Hetzfeld, E., Archäologische Reise im Euphrat-und Tigris-Gebiet II, 206–207Google Scholar. The present mosque is described by ed-Dawachi, Sa'id in Sumer 10 (1954), 250–266Google Scholar (Arabic section), 17 (1961), 100–112 (Arabic section), and 22 (1966), 75–78 (Arabic section).

² This article is based on a thesis done at London University, 1964–7, under the supervision of Professor Seton Lloyd and with further assistance, especially on linguistic matters, from Professor D. J. Wiseman. To both I am most grateful for all their advice and help, and also to the British School of Archaeology in Iraq for a grant which enabled me to complete this paper.

³ For a general account of the history of this tell see Thompson, R. Campbell, Iraq 1 (1934), 95–104CrossRefGoogle Scholar.

⁴ Royal Asiatic Society, Proceedings of the twenty-ninth anniversary meeting of the society (1852), xliiiGoogle Scholar.

⁵ I R 35, No. 4. See Gadd, C. J., The Stones of Assyria, 82Google Scholar.

⁶ Rassam, H., Asshur and the land of Nimrod, 4–7Google Scholar, and Gadd, op. cit., 88–89 and 92.

⁷ Gadd, op. cit., 91–92.

⁸ Royal Asiatic Society, loc. cit., xlii–xliii.

⁹ Sumer 10 (1954). fig. 1 and pp. 110–111Google Scholar.

¹⁰ Sumer 12 (1956), 9–37Google Scholar.

¹¹ For the most part the Late Assyrian palace texts follow a standard format, namely: (i) preamble; (ii) foundations; (iii) component parts of the palace; (iv) rooting; (v) doors, doorways and the decoration thereof; and (vi) mural decoration. Nevertheless, for reasons that will be evident below, those dealing with the Nebi Yūnus arsenal are more varied in their arrangement.

¹² Borger, op. cit., 59 l. 40, and Piepkorn, op. cit., 86 1.64. On this term see Piepkorn, op cit., 8711. 43, ZA 42 (1934), 174 n. 4Google Scholar, and Iraq 21 (1959), 39 n. 1Google Scholar.

¹³ Borger, op. cit. 62.11. 42–43.

¹⁴ Luckenbill, op. cit., 128 l. 39 and 131 l. 55. See also an inscription of Aššur-rēš-iši I which refers to bīt ša-ḫu-ri ša bīt ku-t[al-li] (E. F. Weidner, Die Inschriften Tukulti-Ninurtas I. und seiner Nachfolger (AfO, Beiheft 12), 56 §63 l. 4).

¹⁵ Luckenbill, op. cit., 133 l. 85. In a slightly earlier text he refers to the city gate lying adjacent to Nebi Yūnus as abul ekal māšarti (ibid. 113 l. 2). In the case of Fort Shalmaneser, the arsenal of Nimrud, Shalmaneser III, its founder, simply refers to it as ekallu ‘palace’ (Iraq 21 (1959), 38 l. 1CrossRefGoogle ScholarPubMed, and 25 (1963), 52 l. 1); and it is first called an ekal māšarti by Esarhaddon (Borger, op. cit., 34 l. 42).

¹⁶ Borger, op. cit., 59 ll. 42–46. See also Luckenbill, op. cit., 128 ll. 39–40 and 131 ll. 55–56.

¹⁷ Luckenbill,op. cit., 128 ll. 41–45 and 131 ll. 56–59.

¹⁸ Ibid. 128 ll. 46–47 and 131 l. 60. Although at this period the Tigris may well have followed a course different from that of to-day, it is evident both from this passage and also from the fact that this line of the city walls was pierced by a series of gates (ibid. 113 vii 94 to viii 5) that it did not actually flow along the wall, but that there was an intervening tract of pasture land (usallu, elsewhere described by Sennacherib as qaq-qa-ri ú-sal-li šá ul-tu mal-di nāri, ibid. 129 ll. 49–50). In the various accounts of his South-west Palace on Küyünjik Sennacherib also refers to two other rivers, ^nārḪusur (ibid. 105 l. 2, et passim), the Khosr, which still flows along the southeast side of Küyünjik and thence into the Tigris, and a larger stream which is no longer existent, ^nārTebiltu (ibid. 96 ll. 74–76, et passim). See also on this subject Thompson, R. Campbell and Hutchinson, R. W., A Century of Exploration at Nineveh, 122–124 n. 1Google Scholar.

¹⁹ tamlū ‘building platform’ is first used by Aššur-uballiṭ I (1364–1330 B.C.) (KAH I, 64 l. 11Google Scholar) and then throughout the Middle and Late Assyrian periods, in many cases in a cognate construction with the II¹ or III¹ of malū, e.g. Iraq 14 (1952), 33 l. 25Google Scholar, and Luckenbill, op. cit., 129 l. 50. Compare the Hebrew millō' (Koehler, L. and Baumgartner, W., Lexicon in Veteris Testamenti Libros, 527Google Scholar, and Kenyon, K., Jerusalem: Excavating 3000 Years of History, 50–51Google Scholar).

²⁰ AJSL 27 (1910), 188–189Google Scholar. The variant tikpu is used exclusively in the texts of Ashurnaṣirpal II (AKA 186 l. 16, 209 l. 16. 220 l. 17 and 345 l. 132, and Iraq 14 (1952), 33 l. 24Google Scholar), and also once by Tukulti-Ninurta I (King, L. W., Records of the Reign of Tukulli-Ninib I, 90–91 ll. 12–13Google Scholar) and in a letter addressed to Esarhaddon (RCAE No. 628 l. 14).

²¹ Luckenbill, op. cit., 129 ll.50–51 and 131 ll. 62–63.

²² Iraq 7 (1940), 90 ll. 7–9Google ScholarPubMed.

²³ Luckenbill, op. cit., III l. 69. This text also gives the height of the platform of Sennacherib's Southwest Palace on Küyünjik as 190 tipku (ibid. 106 l. 6), indicating that although it stood slightly lower than that of the ekal māšarti on Nebi Yūnus, it was still on a level with if not higher than the city wall at this point. The various accounts of this building differ in the height of the platform. The earliest puts it at 170 tipku (ibid. 96 l. 78) and the two latest at 190 (ibid. 106 l. 6 and 119 l. 18), whilst in a fourth version written in the intervening period Sennacherib records that he first made it 160 tipku high, but then raised it by a further 20 to 180 (ibid. 100 ll. 53–54). There may, however, only be an inconsistency in these versions of 10 tipku, the first being written before the subsequent increase in the height of the platform, and thus the 170 in this and the 190 in the latest two accounts correspond with the 160 and 180 of the fourth.

²⁴ Little can be added to Baumgartner's observations on the usage and meaning of the Akkadian words for foundations, uššū, išdu and tem(m)en(n)u (ZA 36 (1925), 220 and 236–253Google Scholar; see also Orientalia 35 (1966), 234–239Google Scholar); in short, although it is quite probable that there is some difference in meaning, this is not apparent in the texts, even where these terms appear side by side. There is thus no evidence to support the suggestion made in the Chicago Dictionary that išdu is used of the ‘damp course’ (CAD 7, 235–236Google Scholar), nor that tem(m)en(n)u refers to the foundation trench as put forward by Falkenstein as one of the meanings of the Sumerian temen (Orientalia 35 (1966), 236–239Google Scholar). Likewise there appears to be neither any linguistic nor archaeological grounds to justify Sidney Smith's equation of uššū with the actual foundations and išdu with the lower part of the wall resting thereon (Essays Presented to J, H. Hertz, 385-396). He bases this proposal on a Sumerian text of Ur-Baba of Lagash (VAB 1, 60–61Google Scholar) and two temples excavated at Ur, the Ur III Gig-par-ku (AJ 6 (1926), 367–368Google Scholar) and the Late Babylonian Harbour Temple (UE IX, 39–40Google Scholar). These were built with their foundations as a separate entity, in plan identical to the superstructure but probably considered, for cultic reasons, a buiding complete in itself, buried and reserved for the gods. On these foundations stood the superstructure Using Ur-Baba's text Smith identifies the foundations proper as uš, which he equates with the Akkadian uššū: but for the lower part of the walls of the super-structure there is no distinctive Sumerian word, and these he identifies as išdu. It is doubtful, however, that uš actually corresponds to uššū (Orientalia 35 (1966), 229Google Scholar), and there is also no evidence in the Akkadian texts for such a difference in meaning. Furthermore no example of this building technique, either in a temple or any other type of structure, has been found in Assyria, where foundations tend to be but the lower part of the wall, rarely being of different construction.

A short note can also be added on one usage of a fourth word, dannatu, translated by the Chicago Dictionary under subheading 3.b) as ‘bottom of the foundation trench’ (CAD 3, 90Google Scholar). In such contexts, however, it is most commonly used of the solid mudbrick of a decayed building (see ZA 36 (1925), 38–40Google Scholar. and AHw 160), and occurs in texts which describe the rebuilding of a structure. This necessitated the removal of debris and other ruined material, e.g, an-ḫu-su ú-ni-kìr a-šar-šá ú-me-si dan-na-sa ak-šú-da ‘I removed its ruined parts, I cleared its site, I reached its solid brickwork’ (Andrae, W., Die Festungswerke von Assur (WVDOG 23), 166, ll. 9–10Google Scholar). That is the decayed and fallen mudbrick was cleared away until the builders reached that part of the structure which had remained sound, at which point they could begin their restoration. In two inscriptions of Tukulti-Ninurta I dannatu is also used of the solid natural earth or bedrock. In one he describes the ‘New Palace’ at Assur (Preusser, C., Die Paläste in Assur (WVDOG 66), 30–31Google Scholar): qa-qa-ra-te ma-da-te lu-ú-me-is-si 80 mu-šá-ri zi-iq-pa [ú]-še-la-a a-na šu-pa-li dan-na-su ki-ṣir šadīⁱ lu ak-šud ‘I cleared much ground, I went down vertically 80 mušarū (and) below I reached its solid ground on the bedrock’ (Weidner, op. cit., 12 ll. 74–79; see also ibid. 5 ll. 45-51 and 10 ll. 22–26). That is, not wishing to build his palace on insecure disturbed ground surface, formerly the site of private houses, he completely cleared this area down to firm ground, in this case bedrock. Similarly when digging a moat round the walls of Assur, he penetrated down into the dannatu, the bedrock: ḫi-ri-şa rabā^a a-na li-me-it dīri lu aḫ-ri dan-na-su ki-şir šadīⁱ i-na aqullāt ^meš(at)erī lu-pi-ṣi-id 20 mu-šá-ri a-na šu-pa-lu mē ^mešna-aq-be lu ak-šud ‘I cut a great ditch around the wall. I dug into its solid ground, the bedrock, with copper pickaxes, I reached 20 mušarū below the water-table’ (ibid. 32 ll. 7–8).

²⁵ Borger, op. cit., 60 ll. 51–53.

²⁶ Iraq 20 (1958), 106–108, 21 (1959), 98–129, 23 (1961), 1–14, 24 (1962), 1–25, and 25 (1963), 6–37Google Scholar, and Mallowan, M. E. L., Nimrud and its Remains, 369–470Google Scholar.

²⁷ The general architectural formulae found in such buildings are set out by Loud, in RA 33 (1936), 153–160Google Scholar and repeated in Loud, G. and Altman, C. B., Khorsabad II, The citadel and the town (OIP XL), 10–13Google Scholar.

²⁸ Loud and Altman, op. cit., 75–78. See also Iraq 25 (1963), 36–37Google ScholarPubMed and Mallowan, op. cit., 456.

²⁹ See Loud and Altman, op. cit., pl. 68 and Mallowan, op. cit., 371–373.

³⁰ Luckenbill, op. cit., 130 ll. 70–71, 131 l. 58 and 132 l. 67. On bābānū see the dictionaries and Babyloniaca 2 (1908), 168–176Google Scholar.

³¹ Luckenbill, op. cit., 133 l. 82.

³² Ibid. 130 ll. 65–70; and a shorter version on 132 ll. 66–67.

³³ Iraq 21 (1959), 113Google ScholarPubMed and Mallowan, op. cit., 424–426.

³⁴ Luckenbill, op. cit., 133 ll. 83–85.

³⁵ Borger, op. cit., 62 ll. 32–34.

³⁶ Luckenbill, op. cit., 129 ll. 53–56 and 131–132 ll. 64–65.

³⁷ As recognised by Wiseman, (Iraq 14 (1952), 3–6)Google Scholar and hinted at in the Chicago Dictionary (CAD 4, 54 §3′Google Scholar) but ignored by Soden, von (AHw 191–192)Google Scholar. Thus in these texts ‘palace’ can be rendered in Akkadian either by ekallu in the singular or collectively in the plural form, e.g. Borger, op. cit., 62 l. 35 and 61 l. 3 respectively.

³⁸ AfO 18 (1957–1958), 351–353 ll. 52–89Google Scholar; see also KAH II, 66 ll. 27–42 and 67 ll. 3–14Google Scholar.

³⁹ WVDOG 66, 30–31Google Scholar. Tukulti-Ninurta I names the ‘New Palace’ é.lugal.umun.kur.kur.ra (Weidner, op. cit., 10 l. 30, 12 l. 79 and 39 §32 l. 3), while Tiglath-pileser I refers to that which contained the bīt šaḫruri, bīt labuni and ekal ^iṣkakkēme ^meš as é.gal.lugal. šár.ra.kur.kur.ra (AfO 18, 353 l. 78Google Scholar), that is probably one and the same building. Wiseman, on the other hand, has suggested that these three structures formed part of the Anu-Adad Temple at Assur (CAH rev. ed. II, Ch. XXXI 23), but Tiglath-pileser states that in the construction of this temple he used some cedar wood, and with what remained he decorated the bīt šaḫuri (AfO 18, 352 ll. 59–62Google Scholar). On these three terms see ibid. 354–355 and 358–359, and on bīt šaḫuri also JAOS 39 (1919), 71Google Scholar, ZA 40 (1931), 1–5Google Scholar, and Syria 21 (1940), 6–8 and 160–161Google Scholar.

⁴⁰ Lines 73, 77 and 87. Although the bīt labuni was likewise partly of pistachio wood, it is simply referred to as bīt ^iṣbuṭni (l. 5 8), possibly due to its lesser importance.

⁴¹ AKA 146 ll. 14–16. For the date of this monument see AfO 12 (1937–1939), 377Google Scholar and JSS 4 (1959), 204–215Google Scholar.

⁴² AKA 186 ll. 18–19 and 220 l. 18; Iraq 14 (1952), 33 ll. 25–26Google Scholar; II R 67 l. 67; Lie, A. G., The Inscriptions of Sargon II, 76 ll. 13–14Google Scholar; Winckler, H., Die Keilschrifttexte Sargons I, 166Google Scholar ll. 18–19 and 170 l. 13, and II, pl. 43 obv. ll. 19–22; ZDMG 72 (1918), 182 l. 35Google Scholar; Luckenbill, op. cit., 96 l. 79, 100 l. 56, 106 ll. 14–18, 119 ll. 20–21, 129 ll. 53–56 and 131–132 ll. 64–65; and Borger, op. cit., 61 ll. 9–10.

⁴³ CAD 4, 380Google ScholarPubMed. Campbell Thompson suggests ‘willow’ (DAB 289–291), whereas for the stone ^abanušū he gives ‘diorite, dolerite’ (DAC 163), which favours the more widely accepted equation of ^iṣušū with ebony. Unless otherwise noted, the identification of the various types of wood, metals and stones in this article is taken from Campbell Thompson's DAB and DAC.

⁴⁴ Campbell Thompson equated the Akkadian taskarinnu with the Syriac ’eškār’ā, and thus translated it ‘boxwood’ (DAB 348; see also WO 1 (1950), 368–371Google Scholar and JNES 26 (1967), 269–270Google Scholar). Alternatively Wiseman has suggested ‘walnut’ on the grounds that box is not a suitable building material, whereas walnut is and grows in considerable quantities in Assyria (Iraq 17 (1955), 3–4Google Scholar). There is, however, no etymological evidence for this and, as kindly pointed out to me by Professor Saggs, box can grow to a height of some thirty feet or more, and it is possible that in their campaigns the Assyrians passed through virgin forests where it was to be found in such a state. Furthermore ^iṣtaskarinnu may not have been used in the construction of the building, but in its decoration, for which box would be both effective and adequate. This, however, still leaves unanswered the question as to which Akkadian word refers to walnut, and since this wood must have been widely used, the identification of iṣ^taskarinnu must remain open until new evidence is brought to light.

⁴⁵ BSOAS 19 (1956), 317–320Google Scholar.

⁴⁶ DAB 300. Von Soden simply describes it as a building timber from Syria, (AHw 196)Google Scholar, and the Chicago Dictionary as a precious wood (CAD 4, 75–76Google Scholar).

⁴⁷ Luckenbill, op. cit., 106 ll. 14–20. Other materials also found in such lists are ^iṣbuṭnu ‘terebinth’, ^iṣtarp'u ‘tamarisk’ ^isdaprānu ‘juniper’ and ^iṣmeḫru ‘poplar(?)’ (DAB 267–268), ‘type of fir’ (AHw 641).

⁴⁸ Iraq 20 (1958), 110 and Mallowan, op. cit., 293–294Google ScholarPubMed.

⁴⁹ AHw 61, CAO 1/2, 193–194Google Scholar and Afo 14 (1941–1944), 70–72Google Scholar. This type of figure was used for both column bases (Luckenbill, op. cit., 110 ll. 31–35, 123 ll. 33–34 and here) and as colossi flanking doorways (ibid, 110 l. 23 and 123 l. 31, and Borger, op. cit., 61 ll. 15 and 18). No such colossi have yet been discovered, but column bases of this type have been found at Nimrud, in the Southwest Palace (Layard, A. H., Nineveh and its Remains, I, 376Google Scholar, and Barnett, R. D. and Falkner, M., The Sculptures of Tiglaathpileser III, 23 and pls. CVIII–CXIGoogle Scholar.

⁵⁰ This word is variously read in Akkadian as ašnān, pindū and ezennū (DAC 163–164 and CAD 1/2, 451–452 and 4, 427Google ScholarPubMed), and likewise its meaning is as obscure. Campbell Thompson suggested that it was used of a feldspathic-pyroxcnic rock, which is basically a basalt and augite stone to be found in Iraq and which forms the matrix of garnet, thus possibly explaining the talismanic qualities of an. še.tir (DAC 163–164). Alternatively since the Assyrians usually only employed stones of the limestone variety for building purposes and as Sennacherib compares the appearance of an.še.tir stone to that of cucumber seeds (Luckenbill, op. cit., 132 ll. 72–74), it is possible that this term refers to a limestone with a high fossil content.

⁵¹ naburru (see ZA 36 (1925), 226–227Google Scholar, and Porada, E., Essays in the History of Architecture, presented to R. Wittkower, 10–11Google Scholar) is used in a few instances in the phrase ištu uššīšu adi naburrišu in place of gabdibbu, obviously referring to some upper feature of the building (Luckenbill, op. cit., 130 l. 72, and Borger, op. cit., 4 l. 17; 21 l. 22, p. 22 Episode 26, c:E l. 22, and p. 88 l. 10), and twice by Sennacherib in other contexts. One is in the passage quoted above in which he describes the ekal ^abanpīli u ^iṣerini on Nebi Yūnus, and the other is in his account of the town wall of Nineveh: 40 libitti i-na na-a[l-ba-ni-ia] rabīⁱ ú-kab-bir-šú a-na e-liš a-di šap-la-[ti …] na-bur-ri-šú a-na 39 libitti ú-tir-[ma] i-na 3 uš 20 ti-ip-ki libitti ša pān zu-ú-[uq-ti (?)] mu-sir-šá e-la-niš a-di pa-aš-ki-šú ri-ši-šu ul-li-ma ‘I made it 40 bricks wide by my great brick mould. From below to above [ Άhellip;] I added its naburru to 39 bricks, and by 200 courses I raised the brickwork of the front of the ridge(?) of its enclosure(?) upwards as far as its pašku, its top’ (Iraq 7 (1940), 90 ll. 4–9Google Scholar). Recent excavations have shown that Sennacherib built this wall in two stages. On the outer faΆccedil;ade, the lower stage was faced with stone and surmounted by stepped crenella-tions of the same material. These enclosed a paved causeway, doubtless for manoeuvring soldiery. Above this towered the main bulk of the wall (Sumer 23 (1967), 77–78 and pls. V–VIIGoogle Scholar). Thus Sennacherib describes how he set the naburru ‘crenellations’ on the first stage, 39 bricks high, while the main part of the wall was five times this height. Similarly in his Hittite style wing on Nebi Yūnus the outlines of the ^fdlamassatu statues set on their stone plinths reminded him of such stepped crenellations.

⁵² Luckenbill, op. cit., 132–133 ll. 75–82.

⁵³ See principally ZA 45 (1939), 108–168Google Scholar, Orientalia II (1942), 251–261Google Scholar, and ZDMG 108 (1958), 66–73Google Scholar.

⁵⁴ II R 67, l. 68. It is found in two earlier texts, in a Mari letter (ARM I, 26 l. 10′Google Scholar) and in a Middle Assyrian ritual text from Assur in which it appears to have been a cult structure in the Temple of Ashur (KAV 42 14–16Google Scholar).

⁵⁵ Lie, op. cit., 76–78 ll. 17–3, Winckler, op. cit., I 166 ll. 20–21, and II pl. 42 rev. ll. 5–8 and pl. 43 obv. ll. 23–24, and ZDMG 72 (1918), 182 ll. 36–38Google Scholar.

⁵⁶ Luckenbill, op. cit., 97 ll. 82–84, 106 ll. 20–22 and 119 l. 22.

⁵⁷ V R 10, col. X ll. 101–102.

⁵⁸ See CAD 6, 184–185Google ScholarPubMed where it is wrongly stated that bīt ḫilāni not only refers to the portico, but also to ‘a room or section of a palace provided with a portico’.

⁵⁹ Also found in Luckenbill, op. cit., 97 l. 82 and 119 l. 22. Four derivations have been proposed for appāti: (i) appu ‘nose’, that is a structure projecting out from the main building (ZA 45 (1939), 134–135Google Scholar and Orientalia 11 (1942), 254Google Scholar). (ii) aptu ‘window’, that is a building with windows (ZA 45, 135Google Scholar, Orientalia II, 254Google Scholar and CAD 6, 184Google Scholar). In support of this is the equation of ḫilāni with the Hebrew ballōn, but alternatively Hittite derivations have also been proposed for this word (ZA 45, 134–135 and 140Google Scholar, and AfO 9 (1933–1934), 127Google Scholar). (iii) appatu ‘upper surface, top’, that is a building with an upper storey (Orientalia 11, 254Google Scholar); this word is used elsewhere of the top of a stick or the rim of a pot (AHw 59 and CAD 7, 236 and 239Google Scholar), but there is no evidence that it can also refer to an upper part of a building. Nor is there any evidence that the Assyrian porticos were thus equipped, although this may have been the case in the Syrian prototypes. And finally (iv) appannu ‘a building term (portico ?)’ (AHw 59–60). This word is found in the Nuzi texts, probably a Hurrian loan-word. There is no evidence that it refers to a portico, but such a structure probably did exist in the Stratum II palace at Nuzi (Starr, R. F. S., Nuzi I, 127Google Scholar) and also in the House of Shilwi-teshub at the same site (ibid., 340), and thus this may be the more preferable though by no means proven derivation of bīt appāti.

⁶⁰ ur.maḫ/urmaḫḫu ‘lion statue’ is used of portal statuary here and in three other texts: in the Broken Obelisk (AKA 147 l. 17), by Tiglath-pileser III on the Central Palace at Nimrud (II R 67 l. 79), and by Esarhaddon on the Nebi Yūnus ekal māterti (Borger, op. cit., 61 l. 17; see below).

⁶¹ uggallu, which is only found in connection with the bīt ḫilāni, may either be a Sumerian loan-word' úg.gal ‘great lion’, or be made up of úg ‘lion and, an Akkadian word gallu. In a letter sent to Sargon reporting on the progress of the building of Khorsabad (RCAE No. 452) reference is made to the bīt ḫilāni there, including: gul-la-a-te [ Άhellip;] šá šap-la tim-me ‘the gullāte [Άhellip;] which (fit) under the columns' (obv. ll. 5–6), i.e. the column bases (see CAD 5, 128Google Scholar, AHw 297 and Salonen, A., Die Türen des alten Mesopotamien, 92Google Scholar). gallu, the second dement of úg-gallu, may therefore be a variant of gullatu, the term meaning ‘column base in the form of a lion’.

⁶² The phrase ^iṣ(a)dappi kulūl bābāni is used both in the descriptions of the bīt ḫilāni, as here, and of the columns used to support the lintels of openings between rooms, e.g. ^iṣtim-me erī rabūti ^mešiṣtim-me ^iṣerini ṣīruti ^mešiṣa-dap-pi ku-lul bābāni ^meš-šin e-mid ‘I set the boards of the kulūlu of its doorways (upon) great columns of copper (and) tall cedar columns' (Borger, op. cit., 61–62 ll. 22–23). In similar contexts ḫittu ‘lintel, architrave’ is also used (e.g. Luckenbill, op. cit., 110 l. 40), and thus kulūlu may either be a synonym or refer to another, closely related part of the doorway, as for instance von Soden's ‘BekrΆouml;nung von Toren’ (AHw 505).

⁶³ Lie, op. cit., 76–78 ll. 17–3.

⁶⁴ ZA 45 (1939), 143–168Google Scholar, and Iraq 14 (1952), 120–131Google Scholar.

⁶⁵ At Khorsabad the portico leading into Room 15 of Palace F and that into Room S of Ashurbanipal's North Palace at Nineveh have been identified as bīt ḫilāni (Orientalia 11 (1942), 257Google Scholar; Meissner, B. and Opitz, D., Studien zum Bit Ḫilâni im Nord palast Assurbanaplis zu Ninive (Abhandlungen der Preuss. Akad. Wiss. 1939)Google Scholar, and Iraq 14 (1952), 125)Google Scholar; but these do not correspond to the descriptions of this structure, and instead are to be grouped with those doorways of which the lintel was supported on pillars, as referred to by Sennacherib (Luckenbill, op. cit., no ll. 36–40 and 123 ll. 35–36) and Esarhaddon (Borger, op. cit., 61–62 ll. 22–23), and in the temple texts of Ashurbanipal (Piepkorn, op. cit., 28 col. I l. 18, LAAA 20 (1933), 81 l. 29Google Scholar, and Smith, S. A., Die Keilschrifttexte Ashurbanipals, 19 l. 12Google Scholar; it is also possible, however, that in these texts Ashurbanipal refers to the ‘sacred trees’ which stood by the entrance to the shrine, as found at KhorsabadΆmdash;Place, V., Ninive et l'Assyrie I, 120–121Google Scholar, Loud, G., Khorsabad I, Excavations in the palace and at a city gate (OIP XXXVIII), 97Google Scholar, and Loud and Altman, op. cit., 61).

⁶⁶ Layard, op. cit., I, 376 and Plan 2.

⁶⁷ Lie, op. cit., 78 l. 4, and Luckenbill, op. cit., 97 l. 86, no l. 42 and 123 l. 37.

⁶⁸ Naumann, R., Architektur Kleinasiens, 83–104Google Scholar, and Lloyd, S., Proceedings of the British Academy 49 (1963), 167–173Google Scholar.

⁶⁹ Sumer 12 (1956), 321Google Scholar. 33, and Borger, op. cit., 63 l. 48.

⁷⁰ Sumer 12 (1956), 32 ll. 36–41, and Borger, op. cit., 63 ll. 52–54 and 61–62 ll. 22–23Google Scholar.

⁷¹ Borger, op. cit., §27 A. This is dated to the limmu of Atarilu, i.e. 673 B.C., whereas the first text is that of Banbā, i.e. 676 B.C.

⁷² Borger, op. cit., 61 ll. 9–10.

⁷³ Sumer 12 (1956), 30 ll. 18–32, and Borger, op. cit., 61 ll. 5–8Google Scholar.

⁷⁴ RCAE No. 22 rev. ll. 5–8.

⁷⁵ RCAE No. 84 rev. ll. 2–6.

⁷⁶ CAD 2, 274–275Google Scholar and AHw 131–132; for further references to bītānu sec Borger, op. cit., 62–63, and JNES 24 (1965), 328–330Google Scholar.

⁷⁷ Sumer 12 (1956), 30 ll. 18–19, and Borger, op. cit., 61 l. 5Google Scholar.

⁷⁸ Salonen, A., Die Hausgeräte der allen Mesopotamier I, 278Google Scholar.

⁷⁹ JNES 24 (1965), 328–333Google Scholar.

⁸⁰ bītānu is found only once in association with the crown prince; Sennacherib built one for his eldest son in Assur (Luckenbill, op. cit., 152 No. XV l. 3). He also dedicated the rebuilt bītānu in the Temple of Ashur on behalf of a younger son (ibid. 150 No. X l. 3, and 151 No. XI l. 2). In neither case is there any indication that it referred to a specific structure and not generally to the inner part of the building.

Oppenheim also suggests that the phrase é.gal.tur.ra ‘small palace’, like bītānu ‘small house’, referred to the crown prince's palace, it being used twice by Esarhaddon in this connection (Borger, op. cit., 69 §30 l. 10, and 71 §43 l.22). It is,however, simply to be read as ekalla ṣiḫra^ra and is commonly found in accounts of palaces that are being rebuilt and enlarged, e.g. Sennacherib on his palace on Küyünjik: ekalla ṣiḫra^ra ša a-tu a-na si-ḫir-ti-ša aq-qur-ma ‘I completely demolished that small palace’ (Luckenbill, op. cit., 99 l. 48). Furthermore it is to be noted that in his account of the rebuilding of the North Palace on Küyünjik, formerly the bīt rēdūti, that is the official residence of the heir apparent, Ashurbanipal refers to it neither as é.gal.tur.ra nor bītānu (V R 10, col. X ll. 51–108).

⁸¹ P. E. Botta and E. Flandin, Monument de Ninive II, pls. 148–150, and V, 53–56 and 164–166, and Place, op. cit. I, 149–151, II, 6–7 and 36–42, and III, pl. 37 bis. Koldewey and Parrot have identified this building as a bīt ḫilāni (von Luschan, F., Ausgrabungen in Sendschirli II, 188Google Scholar, and Parrot, A., Nineveh and Babylon, pl. 10 B on p. 8Google Scholar).

⁸² Sumer 12 (1956), 32 ll. 54–56, and Borger, op. cit., 62 ll. 30–31Google Scholar.

⁸³ MDP 30 (1947), 1#x2013;119Google Scholar, and see also Iranica Antiqua 5 (1965), 98–99Google Scholar.

⁸⁴ Luckenbill, op. cit., 129 ll. 58–60 and 132 ll. 69–70. In both palace and tctnple texts ^iṣerinu ‘cedar’ is the usual roofing material, and it is only for Sennacherib's palace on Küyünjik and that of Sargon at Khorsabad that ^iṣšurmēnu ‘cypress’ was also used (ibid., 106 ll. 23–26 and 119 ll. 22–23, and Winckler, op. cit., I, 166 l. 21 and II, pl. 42 rev. l. 8 and pl. 43 rev. l. 5). In all cases these references are apparently to a flat roof and in no text is there any indication of vaulting, although this may also have been used.

⁸⁵ Luckenbill, op. cit., 129 ll. 60–62 and 132 l. 71. Other types of wood given in the Late Assyrian palace texts as the material for doors arc: ^iṣerinu cedar, ^iṣdaprānu juniper, ^{iṣmusukānnu} sissoo, ^iṣtaskarinnu box/walnut(?), ^iṣušū ebony, ^iṣburāšu juniper, ^iṣsindā oak, ^iṣašuḫu fir. Although ivory is not mentioned in any case, a door discovered in Fort Shalmaneser at Nimrud did contain elements of this material (Iraq 25 (1963), 26–27Google Scholar and Mallowan, op. cit., 451), and it is found in the Late Babylonian texts (e.g. VAB 4, 138 col. IX l. 9Google Scholar). Other materials used for the metal bands (mēsiru, see Salonen, , Die Türen des alten Mesopotamien, 73Google Scholar) are siparru bronze, kaspu silver, zaḫalū a silver alloy (?) (CAD 21, 12–13Google Scholar; but DAC 60 ‘gold leaf(?), overlay(?) or perhaps elcctrum’), and ṣāriru a type of gold. Bands of ḫurāṣu gold are also found in Esarhaddon's temple texts (Borger, op. cit., 5 vi 12, 23 l. 6 and 87 l. 23). In one of his accounts of the bīt akštu at Assur Sennacherib gives a long description of the scenes with which he had these bands decorated (Luckenbill, op. cit., 140–141 obv. l. 5 to rev. l. 2).

⁸⁶ Luckenbill, op. cit., 129–130 ll. 62–65 and 132 ll. 72–75. See also below.

⁸⁷ Sumer 12 (1956), 30 ll. 22–32Google Scholar.

⁸⁸ Borger, op. cit., 61 ll. 7–8.

⁸⁹ Sumer 12 (1956), 32 ll. 36–53Google Scholar.

⁹⁰ Borger, op. cit., 61–62 ll. 12–29.

⁹¹ CAD 1/1, 286–287Google Scholar, AHw 31 and ZA 37 (1927), 218–219 n. 2Google Scholar.

⁹² AKA 147 ll. 17–18.

⁹³ II R 67 l. 79, Central Palace at Nimrud.

⁹⁴ Luckenbill, op. cit., 109–110 ll. 20 and 22-23 and 123 ll. 30–31, on the Southwest Palace on Küyünjik; and ibid., 129 l. 64 and 132 l. 75, on the ekal kutalli/māšarti on Nebi Yūnus.

⁹⁵ Sumer 12 (1956), 30 l. 27, and Borger, op. cit., 62 l. 41Google Scholar.

⁹⁶ Iraq 17 (1955), 134 l. 8Google ScholarPubMed.

⁹⁷ KAH I, 14 obv. l. 22 and 15 obv. l. 21Google Scholar.

⁹⁸ Lie, op. cit., 78 l. 3.

⁹⁹ Luckenbill, op. cit., 97 l. 85.

¹⁰⁰ Ibid., 106–107 Ho 32–36 and 120 ll. 25–27.

¹⁰¹ Baghdader Mitteillungen 3 (1964), 148–156Google Scholar.

¹⁰² CAD I/I, 286–287Google Scholar, and AHw, 31 and 532.

¹⁰³ Sargon also records that he had portal statues in the form of immeri šadī ‘mountain sheep’ (Lie, op. cit., 78 l. 3) and likewise Sennacherib at Küyünjik (Luckenbill, op. cit., 97 l. 85), but at neither site has any statue been discovered which can be identified as such.

¹⁰⁴ Borger, op. cit., 62 ll. 23–29, and Sumer 12 (1956), 32 ll. 42–48Google Scholar.

¹⁰⁵ Borger, op. cit., 61 l. 7. Orthostats are also referred to by Tiglath-pileser III for the Central Palace at Nimrud (II R 67 l. 81), and by Sargon and Sennacherib for their palaces at Khorsabad and Küyünjik (Lie, op. cit., 78 ll. 4–7, and Luckenbill, op. cit., 97 l. 86, no ll. 41–44 and 123 ll. 36–37).

¹⁰⁶ Luckenbill, op. cit., 107 ll. 40–44 and 120 ll. 28–32.

¹⁰⁷ AfO 19 (1959–1960), 141 l.15Google Scholar, and KAH I, 67 l. 9Google Scholar.

¹⁰⁸ Iraq 14 (1952), 33 l. 29Google ScholarPubMed, and AKA, 221 l. 20, 245 l. 15 and 247 l. 29.

¹⁰⁹ II R 67 l. 82.

¹¹⁰ AHw, 450 and Salonen, op. cit., 32, 76 and 78.

¹¹¹ KAH I, 15 obv. ll. 25–26, 42 l. 5′ and 71 l. 5 and Luckenbill, op. cit., 148 IV l. 4Google Scholar.

¹¹² S. A. Smith, op. cit., 13 ll. 29–30.

¹¹³ CAD 6, 208Google ScholarPubMed.

¹¹⁴ AfO 9 (1934), 41Google Scholar. See also ZA 36 (1925), 229Google Scholar and 45 (1939), 127, MIO I (1953), 88Google Scholar, and OLZ 53 (1958). 524Google Scholar.

¹¹⁵ Luckenbill, op. cit., 113 ll. 10–12. See also another version, Iraq 7 (1940), 90 l. 8, which is quoted in n. 51 aboveGoogle ScholarPubMed.

¹¹⁶ II R 67 l. 75.

¹¹⁷ Porada, loc. cit., 10. See also ZA 36 (1925), 227–229 and 45 (1939), 127Google ScholarPubMed.

¹¹⁸ Iraq 24 (1962), 8–9Google ScholarPubMed and pl. Vc, and Mallowan, op. cit., 462–463 and pl. 378.

¹¹⁹ Iraq 24 (1962), 9Google ScholarPubMed. Place found a similarly decorated crenellation on the parapet of the Khorsabad ziggurat (Place, op. cit. III, pl. 35, 7).

¹²⁰ Hence ‘Archivolte’ (Bezold, C., Babylonischassyrisches Glossar 214Google Scholar, ZA 45 (1939), 125–126Google Scholar, and Borger, op. cit., 62), but not Heidel's ‘arch’ (Sumer 12 (1956), 33Google Scholar).

¹²¹ Place, op. cit., I, 174 and III, pl. 14. George Rawlinson illustrates what appears to have been a similarly decorated doorway in the North Palace at Nineveh, based on one of Boutcher's drawings in the British Museum (Rawlinson, G., The Five Great Monarchies I, 335Google Scholar). Dr. R. D. Barnett, however, has kindly informed me that Rawlinson's reproduction appears to be an inaccurate copy, for that which is evidently the original drawing shows not an arch with two equal sides but a stone with an irregular loopshaped cavity, which he suggests may in fact have been a threshold. This drawing, moreover, is marked ‘Centre Palace Nimrud’.

¹²² Iraq 25 (1963), 38–47CrossRefGoogle ScholarPubMed, and Mallowan, op. cit., 455–455.

¹²³ Borger, op. cit., 71–73 §43–45.

¹²⁴ Mallowan, op. cit., 376 and 387, et passim.

¹²⁵ Layard, op. cit. I, 375–381, et passim, Gadd, op. cit., appendix 9, Iraq 14 (1952), 5Google ScholarPubMed, and Barnett and Falkner, op. cit., 20–30.

¹²⁶ Iraq 29 (1967), 42–45CrossRefGoogle ScholarPubMed, W. Nagel, Die neuassyrischen Reliefstile unter Sanherib und Assurbanaplu.