¹ SGDI 5635 and Michel 666 (in part). The copies both of Pococke and of Guérin are extremely faulty, but between them a reasonable text can be established as far as that is ever possible with a list of proper names. Collitz and Bechtel make three alterations: 1. 1 ()ει(ν)ήρεω for Σιδηρέω[ς] Boeckh; ΣΕΙΔΗΡΕΩ, Pococke, ΕΙΔΗΓ, Guérin. l. 5 Πό(ρ)κεω for Ποίκεω: ‘Der hergestellte Name ist für Maroneia zu belegen und lässt sich aus griechischem Sprachmaterial deuten; vgl. Ἄντυξ, in Mylasa (BCH XII, 33, no. 14 (2)).’ But Ποίκης can be established for Teos, Strabo xiv 633 where F reads Ποίκης, x Πύκνης alii Ποίκνης; Tzschucke, followed by most modern texts, alters, surely wrongly, to Ἄποικος to agree with Paus. VII, iii, 6. Ποίκης will be a hypocoristic form of Ἄποικος. L. 28 Ε(Ι)καδίου for Ἑκαδίου after CIG 3089 l. 6. The first and last of these corrections seem unnecessary, the second clearly wrong, Michel dates the inscription to the second century B.C., but adds a query; since it is no longer extant and the only copies are so bad, I cannot see how we are to arrive at a date except on internal evidence, which will not allow even such modified precision.

² Article ‘Teos’ in PW V, 539 ff. This is the most recent and best work on the subject.

³ The same man occurs in CIG 3082 and 3083, cf. Le Bas-Wadd. 108 and Rogers, , AJA IX, 422 sqqGoogle Scholar.

⁴ . Ti. Claudius Zenodotus is not Φιλαίης but κυδωνίδης, and his brother Philisteus has apparently no family name after the πύργος name as we should expect, though it is possible that the stone is broken at this point. These two inscriptions show that what I have called the ‘family names’ in these inscriptions do not give the name of the man's actual father, but that of the γένος.

⁵ It might, however, be worth consideration whether, since the expression occurs on inscriptions referring to two brothers and on no other known Teian inscription, it may not have some connexion with the relationship rather than with the individual position of the two men.

⁶ For the symmories of Teos cf. BCH IV, 175, no. 35, CIG 3065; references in PW IVa, 1165–6.

⁷ La Polis Grecque, 137–8.

⁸ ‘Les “Pyrgoi” de Téos,’ Rev. Arch. XXVIII (1928), 185–208Google Scholar, where a fuller account of earlier views and references to the relevant literature will be found.

⁹ ῥύμη = ‘street’ according to L and S, but it more probably means a quarter of the town; cf. the word ἄμφοδον in the Stratonicean inscription referred to above and REG XXXVIII, 122. A parallel to ῥυμάρχης is ἀμφοδάρχης, Philo, Belopoeica xciii, 8Google Scholar (Droysen Heerwesen, p. 262); other references in L and S.

¹⁰ Béquignon does not actually use the word annual, but since he thinks that ἔτος is to be supplied with ἄναρχον I suppose he must assume that the periods of service were of a year each.

¹¹ The fact that a board like the ten στρατηγοί at Athens was elected on a tribal basis provides no parallel; an important point is that individuals could be, and often were re-elected.

¹² From SIG 344 = Welles, Royal Correspondence in Hellenistic Period, no. 3/4 we know that at some time between 306 and 302 (?304/3 Welles, p. 25) Antigonus proposed to effect a synoecism of Teos and Lebedos which suggests that in his opinion Teos could with advantage be increased in size.

¹³ SGDI 5636 l. 4 . I do not see the point of Ruge's remark that, in the fragmentary state of the inscription, it is possible that the name of the first tower may have been lost; there is no room for it anywhere and the beginning of the text is sufficiently well preserved. ( etc.). In any case, however, the second tower has no name.

¹⁴ Ed. Meyer, , Gesch. des Altertums (1937), III, 282Google Scholar ( = 1893 II, 307) ‘das Gebiet von Teos zerfällt in “Türme”, d.h. offenbar Adelsburgen, die den Namen einzelner Personen tragen.’ Wilamowitz, , Sitzber., Berlin 1906, p. 63Google Scholar, n. 4 ‘Die πύργοι sind natürlich villae, Landhäuser des grund-besitzenden Adels.’

¹⁵ For extent cf. Strabo, XIV 644: Welles, op. cit. no. 3 (p. 20) l. 98; Livy XXXVII, 284. On the question o fertility cf. Athen. IV 160a (barley); SGDI 5633 (sheep and cattle). SIG 37, 38 (Tod GHI no. 23) shows that in ca. 470 Teos had to import corn and had some difficulty in doing so (A 6/7); but this was surely due to the abnormal circumstances of the time: in ca. 306–302, as we see from Welles no. 3, 94–101, Teos and Lebedos appear as normally exporting grain. The passage is from the first letter of Antigonus about the synoecism and runs as follows:

¹⁶ For full list, with parallels, see Appendix A.

¹⁷ Wilamowitz, op. cit., 74–5.

¹⁸ Paus. VII, iii, 4 (sc. Erythrae) . The fact of a mixture of races in all the Ionian cities is well enough established, Erythrae is merely a particularly good example. Teos in the tradition is fairly pure, except for the rather suspect Minyans and the undefined Ionians.

¹⁹ Hdt IV, 159, 161, 186.

²⁰ Cf. (i) Δάδας, founder of Themissos in Caria, Steph. Byz. s.v. Θεμισσός, (ii) τὰ Δάδασα, a fortress in Cappadocia, near Comana, Dio Cass. XXXVI, xii, 2, (iii) τὰ Δαδάστανα, a town inBithynia, Amm. Marc. XXV, x, 12, (iv) Δαδόκερτα, a fortress in Greater Armenia, Steph. Byz. s.v. (comparison with, e.g., Τιγρανόκερτα shows that Δαδο- is a personal prefix). For the form and its Asiatic connexions cf. Kretschmer, , Einleitung in die Gesch. der gr. Sprache, p. 337Google Scholar.

²¹ The omission of the father's name is strange and Ruge, op. cit., p. 555, thinks it an argument against these people being high officials, which seems to be justified. On my view they will be fairly important people and would be expected to have a father's name, but if I am right this is a strange inscription in any case and the important things are the names of the πύργος and of the γένος.

²² L. 21 is a doubtful case.

²³ Since it is agreed that this list records a series not in time but in space, we want a spatial expression rather than a temporal one to restore with ἄναρχον and ἄναρχα where they occur: the obvious word on any view is χωρίον. This is the ordinary word for a plot of land, for instance in cadastration; cf. Inschriften von Magnesia, no. 122 passim. In OGIS 225 l. 1 ( = Welles, , Royal Correspondence, no. 18, 1. 6Google Scholar) it means an estate or fief in the sense in which I interpret πύργος, being used interchangeably for βάρις, on which I shall speak later. In Modern Greek χωρίον means a village.

²⁴ I must express my gratitude to Dr. Onions and Mr. C. E. Bazell for assistance on the philological side of this paper.

²⁵ Schuchardt, , Sitzber., Berlin 1935, p. 186Google Scholar calls it ‘gewiss eine erstaunliche Sache,’ but claims to have known it all along or at least suspected it.

²⁶ Cf. Kretschmer, op. cit., 113 and Boisacq s.v.

²⁷ It also occurs as the name of a fort in Pieria, on Pangaeum, Hdt. VII, 112, and as a place-name in Crete, Plin. N.H. IV 59, cf. Plut., Lyc. 31Google Scholar.

²⁸ Stesichorus fr. 28 ; Aesch., PV 956 ; Eur., , Phoen. 1098, 1176Google Scholar, where it refers to the citadel of Thebes.

²⁹ On this see Ramsay, , Hist. Geog. of Asia Minor, 286Google Scholar, Cities and Bishoprics, I, part ii, p. 419Google Scholar; Rostowzew, Römisches Kolonat, 253 sq.Google Scholar, Anatolian Studies presented to Sir William Ramsay, 374 n. 1. These comments of Ramsay and Rostowzew were the starting point for this paper, but I think they give too great a political importance to Eumenes' action. Ramsay, (Cities and Bishoprics, 420)Google Scholar says: ‘Eumenes regarded the territorial aristocracy as the supporters of King Antigonus, and tried to strengthen his cause by enlisting the sympathy of the lower classes … Eumenes and the Attalid kings allied themselves with the people; and apparently the great nobility was weakened or destroyed by them.’ Rostowzew goes so far as to describe as a ‘Kampf der hellenistischen Herrscher gegen die feudale Struktur Kleinasiens’ what was surely in essence merely the action of a condottiere temporarily at a loss for funds.

³⁰ Cf. Ramsay, , Hist. Geog., p. 357Google Scholar (Acta SS. Sergii Bacchi, 7th Oct., 842 sq., Anal. Bolland. XIV, 385)Google Scholar. This place, called Tetrapyrgium, was near the Euphrates.

³¹ Ramsay, op. cit., p. 286.

³² Polybius XXXI, xviii, 11, Strabo XVII, iii, 22. Now the most famous of all.

³³ Pap. Ox. II, 243, 1. 15; Preisigke, , Hermes LIV, 423 sq., Ed. Meyer, Google Scholarid., LV, 100 sq.; A. Alt. id., LV, 334 sq.; Hasebroek id., LVII, 621 sq. The mipyos in the vineyard in the parable of the Wicked Husbandmen (S. Mark XII, i) is more probably to be taken as a watch tower, with Alt, than in Preisigke's meaning of Wirtschaftsgebäude.

³⁴ Schuchhardt, , Sitzber., Berlin 1935, p. 186Google Scholar considers that Kretschmer (op. cit.) has proved that the Greeks distinguished between πύργος and τύρσις, the former standing for the ‘nordische Volksburg,’ of which he thinks Mycenae and Tiryns are examples, and the latter for pre-Indogermanic towers indigenous to the Mediterranean area, which he conceives as similar to these of Sardinia and Malta. There is no evidence of any sort for this and it is no more likely than the Germanic origin of the word.

³⁵ This well-known passage gives a vivid picture of the kind of feudal life to which I am referring. For remains of such towers in Mysia cf. Schuchhardt, ‘Ursprung und Wanderung des Wohnturms’ (Sitzber., Berlin 1929, pp. 448–9)Google Scholar.

³⁶ It is also used by Hippocrates, , de Articulis XLIII, 27Google Scholar ( = Foesius 808) . This is the only place in Greek where it is used for something ordinary and not out of the way. The de Articulis whether by Hippocrates or not, is an Ionian work of the fifth century and so does not upset the contention that Asia Minor is the only place where τύρσις could be used of a common object.

³⁷ It was popular at Alexandria, where it was seized on by writers in search of an ‘elegant variation’ for πόλις or ἀκρόπολις, cf. Lycophron 717, 834, 1209, 1273; Nicander, Alex. 2Google Scholar; Ps. Orpheus, Argonautica 153Google Scholar; Anth. Plan. 279 l. 2; SEG VIII 497 l. 7 (I am indebted to Mr. M. N. Tod for the last reference. It comes from a poem on the tomb of a native of Apamea, but its occurrence is more likely due to the love of literary ornament on the part of the writer which is evident throughout the poem than to Anatolian reminiscences of the subject of the epitaph).

³⁸ This last statement is denied by Kretschmer, , Glotta XXII, 111, n. 1Google Scholar ‘Im Etruskischen ist τύρσις nicht nachgewiesen.’ It is doubtful whether our knowledge of Etruscan is extensive enough for such a negative judgement; and it is in any case likely that Dionysius knew more about the Etruscan language than we do. It is perhaps more important that the port of Caere was called by the Greek name Πύργοι; on the other hand archaeology shows that it was almost a Greek port, cf. Blakeway, , BSA XXXIII, 170 sqq.Google Scholar; JRS XXV, 129 sqqGoogle Scholar.

³⁹ The word also occurs as a place-name in Thrace, Μόσσυνος Athen. VIII 345c, and in Macedonia in Byzantine times, Μοσυνόπολις.

⁴⁰ Kretschmer, op. cit., 111 says that the derivation of Τυρσηνός from τύρσις ‘has long been rejected on morphological grounds,’ and prefers to derive it from Tyrra, a city in Lydia. He adds, however, ‘möglich ist aber, dass dieser Ortsname zu τύρσις gehört’ and quotes a form τύρσος from Hesychius, , Phot., p. 612Google Scholar, 13 to which, apparently, there are no morphological objections. The point is clearly a very fine one.

⁴¹ That Latin turris, Oscan tiurri, is a loan word seems to be accepted, but it is almost certain that it was not borrowed from Greek, for τύρσις is very rar e in Greek (outside Xenophon, who uses it to describe a foreign phenomenon) while πύργος is common. It is natural to assume it was borrowed from Etruscan and Dion. Hal. says the Etruscans did have the word; Kretschmer, since he rejects this, has to say that it was part of the language of the primitive inhabitants of both Italy and the Greek peninsula. But the connexion with Asia Minor is much better based. For this see further the letter of Attalus, brother of Eumenes II of Pergamum, , Ath. Mitt. XXIV (1899), 212–14Google Scholar no. 36 = Welles, Royal Correspondence, no. 47 where we have; l. 2 and III, p. 162, no. 325; . Apollo Tarsios or Tarseus occurs fairly frequently in Lydian inscriptions (cf. Kruse, PW s.v. ‘Tarseus’) and this may mean ‘of Tarsus,’ but Ταρσηνός or Ταρσηνή is more probably to be interpreted as meaning Tyrrhenian.

⁴² The existence of this word βᾶρις suggests a misreading or misunderstanding by Diodorus of his source in the passage quoted above (XI, xxxviii, 4). Referring to the estate given by Gelon to his wife where he was himself later buried, Diodorus says: . It is surely very odd in any language to refer to the weight of buildings and it seems likely that Diodorus' source used the word βᾶρις to describe them and that this was misunderstood. If that source was Ephorus, who is Diodorus' main source for Books XI–XV, he may well have been acquainted with the word from the country near his home in Cyme.

⁴³ Cf. also Jos., , Ant. Jud. I, iii, 6Google Scholar; X, xi, 7; XI, iv, 6; LXX, 2 Chron. XXVI, 19Google Scholar; Psalms, XLIV, 9Google Scholar; Dan., VIII, 2Google Scholar; Inschr. von Magnesia 122d 4–8, the cadastral survey quoted above where there are five βάρεις listed as χωρία.

In LXX, Psalm CXXI, 7 occurs the word πυργόβαρις, apparently ἁπ. λεγ. There is a city called Baris in Pisidia, Plin., NH V 147, and also one near Parium which Jones, (Cities of the Eastern Roman Provinces, 91)Google Scholar thinks may have grown out of Laodice's estate.

⁴⁴ op. cit., 444.

⁴⁵ But he was forced to mortgage them to the temple of Artemis in Sardis, to which misfortune we owe the record of Antigonus' bounty, cf. Sardis VII (1), 1–7Google Scholar.

⁴⁶ Il. M 312–14.

⁴⁷ I cannot refrain, in this connexion, from quoting a passage from Hasebroek which does not seem to have received the attention it deserves (Griechische Wirtschafts- und Gesellschaftsgeschichte bis zur Perserzeit, 120–1): ‘When in the sixth century Alyattes invaded the territory of Miletus his sole strategic aim (das Ziel seiner ganzen Kriegsführung) was, according to Herodotus (I, 17), to ruin her agriculture and horticulture, which therefore formed the basis of Miletus' prosperity.’ It is hardly necessary to point out that the Herodotus passage quoted proves precisely the opposite, that Miletus' prosperity was almost entirely unaffected by this campaign though it was carried out scientifically for a period of twelve continuous years.

⁴⁸ Ponticus, Heraclidesap. Athen XII 26Google Scholar.

⁴⁹ Cf. fr. 7:

.

⁵⁰ Hdt. V, 28, the arbitration of the Parians.

⁵¹ Fr. 12, 3–4.

⁵² Aristotle, , Politics IV, 1290bGoogle Scholar, cf. Wilamowitz, , Sappho und Simonides, 277Google Scholar.

⁵³ Fr. 3 ap. Athen. 526:

.