¹ I was enabled to do this work by a Rome Scholarship at the British School at Rome and a Research Scholarship at Trinity College, Cambridge, supplemented by a grant of £25 from the Rouse Ball Research Fund. A further grant of £50 from the Rouse Ball Research Fund has helped to meet the cost of publication. Mr J. B. Ward Perkins and Professor J. M. C. Toynbee suggested and encouraged this research, and I was also helped by discussions with Mr R. M. Cook. Mr C. N. Johns, Controller of Antiquities in Libya, put his house at my disposal during my first visit. Mr R. G. Goodchild, his successor, put me up on subsequent visits, and greatly assisted me with advice based on his deep local and general knowledge. The Antiquities Department, whose archives contain rich photographic records, gave me every assistance. Abdul Hamid shared with me much of the exertions of map-making, and Mr Sheppard Frere very kindly consented to prepare the maps for publication. Mr Alan Rowe has kindly allowed me to refer to his work which was as yet unpublished at the time of writing. To these people and institutions I wish to record my sincere thanks.

² SEG ix 2, where there is recorded the wholesale distribution of corn by Cyrene to Greece in 330–326 B.C., when it was ravaged by famine.

³ Horn, fig. 31, where a reconstruction of S 185 is illustrated.

⁴ Pietrogrande, A. L., Africa Italiana, iii, 1930 pp. 107–40Google Scholar.

⁵ SEG ix 193, 194.

⁶ The best discussion of them is to be found in Chamoux, pp. 293–300, with bibliography.

⁷ Norton, pp. 157 ff.

⁸ Ferri, Silvio, ‘Il Santuario di Budrasc’, Notiziario Archeologico, iii, 1922, pp. 95–9Google Scholar.

⁹ Abbreviations given in brackets after works mentioned in this section are used for reference purposes elsewhere.

^9a Often, but incorrectly, known as the Ras el Hilal tombs. They lie on the middle plateau, beside the road from Lamluda to Ras el Hilal.

¹⁰ Burton Brown, p. 149. The dating cannot be far wrong.

¹¹ Cf. Robinson, under the numbers 22, 23, 26 (a) and (b), 50. The correct reading of 50 is perhaps Ἀριοτοτέλευς not -εις. Fig. 16 shows a recessed panel in the bottom of a plinth.

¹² Cf. Oliverio, G., Documenti Antichi dell' Africa Italiana, Cirenaica II i 1933, No. 124Google Scholar; and hence SEG ix 236.

¹³ Oliverio, op. cit., says in a footnote to p. 116 that for reasons beyond his control transcripts are published instead of photographs of certain inscriptions. This is done at Plates L, LI. These include inscriptions Nos. 115–18, 120, 124. Nos. 115–18, all from bases from N 171, are accurately cut and regularly spaced, although this is not plain from Plate LI, figs. 71–4. The lettering of No. 120 is indeed rough, but the reading and transcript differ widely from the truth. The correct reading is:

Over this has been carved by a later hand:

There is red paint in the lettering. Below is

πλήρης

It is manifest that O. had not seen the original or he could not have published the text he did; he relied on the transcript furnished by an inept assistant. Hence one accepts the reading of No. 124 (Pl. L, fig. 79) with some reserve, especially since it is most unusual for the numeral signifying the age of the interred not to be preceded by L ( = ἐτῶν).

¹⁴ E.g. N 236.

¹⁵ Fouilles de Delphes, ii, Bousquet, Jean, Le Trésor de Cyrène, Paris, 1953Google Scholar. The upper part of the building is dated (p. 69) to 335–330 B.C.

¹⁶ The photographs are 1602 E 632, 1603 F 777, 1691 F 815.

¹⁷ W. B. Dinsmoor, Architecture of Ancient Greece, 1950, p. 155.

¹⁸ This monument is described by Rowe, , Bulletin of the John Rylands Library, xxxvi, 1954, p. 493Google Scholar as ‘… a huge masonry tomb, shaped like an Egyptian mastaba (a rectangular structure with sides sloping towards a common centre and with flat top), dating from between the fifth and early fourth centuries B.C.’ The ‘sloping sides’ are in fact rather weathered steps; the ‘flat top’ is missing, but was certainly a ridge roof; there is no detectable Egyptian influence of any kind; the material R. found includes a model black-glaze kantharos and krater, and an alabaster alabastron, all closely paralleled by finds made by Breccia at Sciatbi, near Alexandria, in tombs which do not antedate that city; finally, although this monument is large, there are some hundreds of others of like dimensions round Cyrene, and some considerably larger within sight of this very tomb.

¹⁹ Pacho, Plates XLV, XLVI; Smith & Porcher, frontispiece. Pacho's is much the more detailed illustration.

²⁰ Breccia, E., La Necropoli di Sciatbi, Cairo, 1912Google Scholar. See, e.g., Tav. XVII, 18. The altar, if it is an altar, illustrated in Tav. XVIII, 19, has interesting affinities with built tombs at Cyrene. But these inconsiderable monuments cannot have influenced the architecture of the necropolis at Cyrene; any influence there was flowed rather in the other direction. Tav. XXXI, 34a shows that Cyreneans were buried at Sciatbi.

²¹ Burton Brown, p. 148 ff.

²² Weld-Blundell, pp. 132 ff. But the form of this capital does not seem to be due to a rock-fault, as Prof. Studniczka there suggests.

²³ Robinson, No. 51.

²⁴ Oliverio, , Africa Italiana, ii, 1929, p. 122Google Scholar, says ‘… un tempio di Ecate, che pare non sia proprio l'attuale, fu costruito insieme col portico dorico-ionico … nel 107 d.c., … Distrutti ed incendiati ambedue nel 117 dai Giudei, fu ricostruito solo il tempio …’. An examination of the structure of this portico shows that it in fact antedated the temple and ran considerably farther west. Its Greek date is attested by the recessed panels cut in the riser ofits step and by the columns, which are fluted. Doric limestone columns were fluted in the Greek period; the forum, Augustan in date, has columns fluted except for the bottom part; and later buildings of the Imperial period have unfluted limestone columns with bases of the type already described. Even the Temple of Apollo, which was wrecked in the Jewish revolt, was rebuilt with unfluted columns. The rebuilt peristyle of the Temple of Zeus is an exception. But it may not have been badly damaged, and the enormous size of the blocks forbade replacing more than what had been destroyed. In contrast, Roman columns of marble are often fluted.

²⁵ Rumpf, A., JHS, lxvii, 1947, p. 12Google Scholar. This dating appears to rest on the fact that the women's flesh is not shaded. This is borne out by Beechey's illustrations, some of which R. reproduces, but not by Pacho's (Plate LIV). It is difficult to tell from the photographs which of the two was correct.

²⁶ Comparetti, Domenico, Annuario della R. Scuola Archeologica di Atene, vol. i, 1914, pp. 161–7Google Scholar. C.'s text is established from a photograph taken by the Missione Halbherr, and requires revision.