^{page 27 note 1} e.g. Frisk, , Eranos xli (1943), 63.Google Scholar

^{page 27 note 2} La Casa degli Indeuropei,98.

^{page 27 note 3} As in early English cruck buildings, Walton, J. cf., Antiquity 22 (1948), 179 ff.CrossRefGoogle Scholar

^{page 27 note 4} e.g. Chapman. The use of σκηναί for the Homeric encampment is at least as old as Pausanias (10. 25. 3), who does seem to mean tents of some sort, though it is possible that the σκηναί of, for instance, Xen. An. 3.5. 7 are some solider type of building.

^{page 27 note 5} The Argive Heraeum model: AE, 1931, i ff., and Markman, cf., Studies Presented to D. M. Robinson, i. 259 ff.Google Scholar

The Perachora models: Payne, Perachora, 34 ff.

The Samos models: (a) apsidal, AM lxxiv (1959), 18Google Scholar; (b) oval, AM lv (1930), 16 ff.Google Scholar

The Smyrna houses: JHS lxxii (1952), 104Google Scholar; BSA liii (1958–1959), 10 ff. and pi. 74Google Scholar; Akurgal, Die Kunst Anatoliens, 9 and 12 f.

^{page 27 note 6} AA lxxix (1964), 2, p. 184.Google Scholar

^{page 28 note 1} On Mycenaean palace roofs, Blegen, cf.AJA xlix (1945), 35 ff.Google Scholar where other references are given. I hope to discuss this vexed question elsewhere.

^{page 28 note 2} Eranos, xli (1943), 59 ff.

^{page 28 note 3} Scanned κλīσĭον or κλīσιôν (cf. Aỉγυ-πτίους, Od. 4. 83). This scansion is, however, only remotely possible.

^{page 28 note 4} e.g. at Mycenae (Wace, Mycenae, 79) and particularly often at Geometric sites, e.g. Emporio (Boardman, Chios: Greek Emporio, 37, where other examples are cited).

^{page 28 note 5} Loc. cit. 63.

^{page 28 note 6} Rooms 8–9–11–14, Pendlebury, , BSA xxxviii (1937–1938), 77 f.Google Scholar

^{page 28 note 7} Akurgal, , Die Kunst Anatoliens, 13.Google Scholar

^{page 28 note 8} κλισίη does have a few formulaic epithets: εṽτνκτος (three times Il., once Od.), εṽπηκτος (twice Il.), and κατηρεøής (once Il.), but these are scarce and relatively simple. Details of its decoration and furnishing are never given, as they are for example of Alcinous' and Odysseus' houses, nor are parts like storerooms and stairs ever mentioned (always excepting the passage in Il. 24 which will be discussed below).

^{page 28 note 9} For the very few exceptions to these remarks, see below.

^{page 29 note 1} e.g. Il. 24. 452, Od. 14. 5 f. As I hope to show elsewhere, the αύλή of the Homeric palace was of this type too, and quite different from the internal court of Mycenaean palaces.

^{page 29 note 2} e.g. Il. 4. 433 (sheep), Od. 15. 556 (pigs). Oddly enough, cattle are never mentioned in an αύλή, but this may be just coincidence.

^{page 29 note 3} Thus Palmer's statement, that it was a ‘secondary enclosure … in the centre of the (palace) αύλή’, is quite erroneous ( Trans. Phil. Soc., 1948, 96).

^{page 29 note 4} σταθμός in this sense appears to be found on the Mycenaean tablets, where ta-to-mo occurs twice, in a heading, with lists of sheep. Chadwick argues strongly for its identification with σταθμός {JHS lxxxv (1965), 189Google Scholar), though Palmer doubts it (Gnomon, xxix (1957), 569Google Scholar).

^{page 29 note 5} Od. 13. 407 ff.

^{page 29 note 6} Od. 14. 5. There is no reason, contra Stanford ad loc, why this should not mean a porch or vestibule as usual.

^{page 29 note 7} Cf. the seat in front of a Geometric house at Zagora (ADelt, 1960, 248 f.).

^{page 29 note 8} Chamoux, , REG lxv (1952), 282 f.Google Scholar

^{page 29 note 9} Loc. cit. 281 ff.

^{page 29 note 10} Od. 14. 13 f.

^{page 29 note 11} Od. 14. 16 and 532 f.

^{page 29 note 12} Stanford, note on Od. 14. 34.

^{page 29 note 13} Od. 16. 155 fr.

^{page 29 note 14} Od. 16.41.

^{page 29 note 15} e.g. at Karphi, , BSA xxxviii (1937–1938), 67Google Scholar; at Emporio, Boardman, Chios: Greek Emporio, 37.

^{page 30 note 1} Od. 14. 395 and 16. 78.

^{page 30 note 2} Od. 14. 381.

^{page 30 note 3} Od. 16. 165 and 17. 521.

^{page 30 note 4} Smyrna: Akurgal, Die Kunst Anatoliens, 9 ff.; Zagora: Ergon, 1967, 75 ff.

^{page 30 note 5} Eumaeus had moved into a new κλισίη while Odysseus was away (Od. 14. 7 ff.), and Odysseus expresses no surprise or particular interest on learning this (Od. 13. 407 f.).

^{page 30 note 6} WD 604 and 607.

^{page 30 note 7} Perhaps Eumaeus would have transferred his attention to agriculture if Odysseus had remained in Ithaca and given him an ο̑ỉκος, a κλ̑ηρος, and a wife (Od. 14. 64). This would have been a change for the better: an ο̑ỉκος is a better class of dwelling than a κλισίη, and moreover carries with it connotations of a ‘household’.

^{page 30 note 8} Other, clearer signs of a difference between ôỉκος and δόμος are discussed elsewhere (JHS xc (1970), 117–20Google Scholar).

^{page 31 note 1} With the possible exception of οἴκονδε, Il. 23. 856, which might, however, mean ‘to his home’ here. In any case, this word is very generalized in meaning.

^{page 31 note 2} Il. 16. 231.

^{page 31 note 3} Il. 19. 212: front door, either of main room or of porch, as above.

^{page 31 note 4} e.g. Il. 19. 243 ff., where from Agamemnon's hut are brought seven tripods, twenty cauldrons, twelve horses, and eight women. We need not take the horses, at least, too literally; they must have been kept in the αὐλή.

^{page 31 note 5} Il. 24. 125.

^{page 31 note 6} Od. 14. 419 ff.

^{page 31 note 7} Line 596.

^{page 31 note 8} For example, most of book 9 (over 600 lines in fact) takes place in Agamemnon's and Achilles' κλισίαι.

^{page 31 note 9} This paper is based on parts of a Ph.D. thesis in the University of London, written with the aid of a scholarship from the New Zealand University Grants Committee. I am grateful to my supervisor, Professor T. B. L. Webster, for invaluable advice and encouragement, and also to Professor Agathe Thornton and Professor H. A. Murray for useful suggestions and help.