Acknowledgements. This study was carried out as part of my thesis on Roman Crete to the Arab conquest. I should like to acknowledge encouragement and advice from Dr. H. W. Catling, Dr. St. Alexiou, Ephor of Crete, and Professor S. S. Frere.

¹ Numbers in brackets by the name of the site refer to the map, FIG. 1.

² Inscr. Cret. i. xxix. 1.

³ The analysis of the relationship between distance and economic farming is taken from Chisholm, M., Rural settlement and land use (London, 1968) 111 ff.Google Scholar

⁴ Acts 27: 8.

⁵ Sites of the Byzantine B./Venetian period are known— AR 1971–2, 23—so perhaps the desolation of the mountains should be attributed to Venetian or Turkish deforestation.

⁶ For the evidence on mineral extraction see Faure, P., ‘Les minerais dela Crète antique’, RA 1966, 45–68.Google Scholar

⁷ S. Spanake, Crete: a guide, 30:

στήν Ἄγια Βαρβάρα βρέχει

κὶ ὁ θεὸς δὲν τὁ κατέχει

⁸ These distances are very deceptive since they are ‘crow flight’ distances. When struggling across a very steep and muddy valley one finds that a map km. can take three-quarters of an hour.

⁹ A large villa was found there by Taramelli, A.. AJA vi (1902), 121 ff.Google Scholar

¹⁰ ‘Village’ is loosely used. Only one of these sites has been excavated, at Ayia Triadha. This seems to have been a large farm of the early period.

¹¹ Strabo x. 4. 11 (478), where it is called an ἐπίνειον while the Stadiasmus 323 has it as a λιμήν.

¹² However, of the two main sources, Strabo and the Stadiasmus, the first is definitely based on Hellenistic material and I believe that much of the latter is as well.

¹³ Inscr. Cret. iv. 285 concerns the annona in the late fourth century, which is evidence for a population which did not produce all its own food.

¹⁴ At some of these sites, e.g. Panayia (25), coins have been found (of Gordian III), but generally the sites are still occupied and modern refuse has made sherding impossible. However, at Ay. Photia (28) tombstones of the first to second centuries have been found. Also the post-Roman accumulation of silt may well conceal surface sherds one might normally expect.

¹⁵ Sites with cisterns, from west to east: Vagiona (20), Stavghies (24), Sternes (27), Ayia Photia (28).

¹⁶ As already suggested by Xanthoudides, S., ADelt. ii (1915) par. 24 ff.Google Scholar, where he reports the discovery of most of the sites.

¹⁷ The figures are from L. Allbaugh, Crete, a case study of an underdeveloped area (1953).

¹⁸ These have normally been seen as traders, e.g. in Hatzfeld's study Les Trafiquants dans l'orient hellénique (1919) 157. But although obviously true of some, Pompey raised halfa legion from Crete (Caesar, , BC iii. 4. iGoogle Scholar). There is also an inscription from Priansos, Inscr. Cret. i. xxiv. 5, of a tribunus militum of Leg. XXXI, probably a unit of Octavian's.

¹⁹ Silvae ii. 6. 67:

sive alma fidem messisque negasset

Cretaque Cyreneque

²⁰ Inscr. Cret. i. xxvi. 2, 3, where it is the proconsul act ing on the authority of Nero and the Senate.

²¹ Nero, Inscr. Cret. i. viii. 49; Domitian, , BCH xciii (1969) 841 ff.Google Scholar; Aurelius, M., Inscr. Cret. i. v. 333.Google Scholar

²² Chisholm, op. cit. 139 fig. 9d.

²³ Inscr. Cret. i. xxix. 1 (A.D. 120), for the shrine and κώμη.

²⁴ Cf. Strabo x. 4. 18 on the andreia, but compare x. 22 where he remarks on ‘present’ i.e. Augustan customs. Also R. F. Willets, Aristocratic society in ancient Crete, 1955.

²⁵ It has been suggested that this hill site represents the site of the Gortynian exiles in the third century B.C., the Αἰνάοι. Faure, P., KretChron. xix (1965) 222–3.Google Scholar

²⁶ Although earlier irrigation is known, Willets, op. cit. 214, it is generally more concerned with supplying the towns rather than the countryside with water. Note the cisterns in the agora areas at Lato and Dreros.

²⁷ Particularly in the form of small fifth to sixth century churches, e.g. in the Mallia and Frankokastelli plains. Indeed, the Mesara is unusual in the scarcity of its church sites.