by those living about the place. Three villagers of Pannus were therefore called upon to point out where the road had formerly run. Their names are given in the hyparch's delimitation, evidently to legalize the transaction in case of dispute. No doubt the eastern boundary thus re-established was fixed by boundary stelae, though this is not stated by the hyparch. Presumably there was no plot of the royal domain in this locality, since it would have been referred to in determining the eastern boundary line had such a plot existed in the local land registry office.

An analogous description of boundaries has been preserved in the case of a long-continued boundary dispute between the Cretan towns of Itanus and Hierapytna. Several descriptions of the boundaries of the city-state territory of Itanus, as fixed in previous decisions in this case, were introduced as evidence before the arbitration board of citizens of Magnesia in 139 B.C. The most complete of these reads as follows: "It was decided by the people of Itanus and Praesiae to make peace for all time on the basis of the territory which each now holds, the boundaries of which are these: As the Sedamus flows to Karymae, to the ridge of the hill and beyond to the crest, and straight along following the crest, thence in a straight direction to Dorthannae to the pond, and as the road runs, southward of the road leading through Atron, thence to Mallus in a straight line to the sea."5 A map was also introduced in evidence upon which it was "clear at a glance" that the temple of Zeus lay outside the disputed territory.

The Aristodicides document consists of three letters of Antiochus I Soter (281-261 B.C.) to the satrap of the Hellespontine satrapy, named Meleager, headed by a covering letter of Meleager to the people of Ilium. The king sent an order to the satrap, stating that

1 Ibid., 11. 43-45.

2 W. Dittenberger, Sylloge Inscriptionum Graecarum (2d ed., Leipzig, 1900), No. 929.

In this case called Tepiopioμós, ibid., 1. 57.

▲ Ibid., n. 7.

Ibid., 1. 62 ff.

* Ibid., 1. 71, καθότι καὶ διὰ τῶν ἐπιδεικνυμένων ἡμῖν χωρογραφίων εὐσύνοπτον ἦν. 70.G.I. 221.

he had given, in full ownership,' a large estate in the Troad to Aristodicides of Assos. He commands the satrap to arrange that 2,000 plethra of land bordering upon the towns of Gergitha or Scepsis be assigned to Aristodicides and that it be "bounded off and assigned" to the territory of the Ilians or Scepsians. Again it is apparent that the poσopioμós, the detailed description of the boundaries, is available neither at the central land registry of the king nor at the satrapal registry of Meleager. The actual work of delimitation must be referred to the local registry.

3

Meager though it is, the evidence shows that the local registry offices of the royal domain could produce written descriptions of boundaries (poσopioμoi) of the areal units by villages, and the city-state registration bureaus equally detailed descriptions of their territories. The Mnesimachus document found in 1910 at Sardes by Buckler and Robinson and published by them with an excellent commentary, shows that the successors of Alexander were giving out large and small portions of the old Persian royal domain in Asia Minor under heritable leases, either to soldiers or in return for other types of distinguished services. The villages and allotments granted to Mnesimachus paid a fixed rental in money into the financial chest of the subdivisions of the satrapy in which each lay. In the time of the Mnesimachus document, fixed by the editors in the period 306-301 B.C., these subdivisions were called chiliarchies. The grant to Mnesimachus fixes for us a land classification in the Seleucid kingdom which must have corresponded to the gift land (yn v Swpea) of the land under grant (yî év áþéσe) in Ptolemaic Egypt." The payment of rental by Mnesimachus, and later by the temple organization which took over the land from him, proves that the title to the land still rested with the king.

5

1O.G.I. 221, 1. 20, dedwkaμev as opposed to kжikexwpηkaμev, "we have granted to." Cf. douvat, 1. 27, and see Haussoullier, op. cit., 31, n. 2.

2 Ibid., 11. 21-25.

3 In the American Journal of Archaeology, XVI (1912), 11–82.

4 Ibid., p. 55.

Ibid., pp. 22-23.

• The editors of the document (p. 69) suggest that these chiliarchies may be the same satrapal subdivisions later known as hyparchies.

7 Wilcken, Papyruskunde, I 1, p. 284.

The land granted to Mnesimachus, though regarded by him as a single estate' and mortgaged in its entirety, consisted of noncontiguous areas. The grant was made under two types of areal units, by villages, and by much smaller units called "allotments" (kλîpoɩ).3 The view of the editors of the Mnesimachus document that the villages granted to him were "probably not laid off by metes and bounds" is certainly incorrect. Just as in the case of the village of Pannus sold to Laodice," the boundaries of each of the village units granted to Mnesimachus must have been sought out in the local registry and the actual local transfer of the land to Mnesimachus made with the greatest exactness as to boundary details. Without such provision the local land registry would have been in a continual state of confusion and a clear title to the property impossible to maintain.

The central land register of the royal domain of the Seleucid kings was constructed necessarily upon broader lines, with an apparent absence of details of the delimitation of the properties. The units on the royal, or central, register were villages and allotments, Kλnpo. The allotments were presumably of definite areas at the outset of the Seleucid régime, as was probably the case also under the early Ptolemies in Egypt. But in Egypt the size of the cleruch grants, even in the late second century B.C., had become quite variable in the same locality and the actual areas of the holdings of 1 AJA, p. 12, col. I, ll. 2, 4, olkov.

καὶ

2 Ibid., pp. 12-13, κῶμαι αἶδε (αἱ) καλοῦνται τοβαλμουρα . . . . καὶ ἄλλαι κῶμαι ἢ καλεῖται τάνδου καὶ κομβδιλιπια . καὶ ἄλλη κώμη Περιασοστρατα ἄλλη κώμη ἐν ̓Αττούδδοις ἡ καλεῖται Ἴλου κώμη.

Ibid., ll. 6, 8.

4 Ibid., p. 54.

5 O.G.I. 225, 11. 40-49.

* AJA, XVI, 12-13. τοβαλμουρα κώμη

....

κῶμαι τανδου καὶ κομβδιλιπια . καὶ ἄλλη κώμη Περιασοστρατα; O.G.I. p. 225, 1. 2, Πάννου κώμη ; O.G.I. 221, 11. 28-29, τὴν Πέτραν καὶ τῆς χώρας τῆς Πετρίδος έργασίμου πλέθρα χίλια πεντακόσια; O.G.I. 262, 1. 6, κώμην τὴν Βαιτοκαικηνήν, granted by one of the kings named Antiochus to the temple of Zeus Baetocaece.

7 Mnesimachus document, AJA, XCI, 12–13. The kλpot about Smyrna granted to certain mercenaries colonized in Palaemagnesia are of two types, ordinary allotments and "cavalry" allotments, κλýpoνs iππɩкοùs in O.G.I. 221, II. 42-43.

8 AJA, XVI, 74.

P. Teb. I, Appendix I, p. 548.

the "hundred arourae" and the "thirty arourae" cleruchs fail in many cases to agree with the titles. So also in Western Asia, variation in the size of the allotments, as private ownership developed, must have come rapidly. In 125-26 A.D. Hadrian wrote to the proconsul of Asia stating that he should take as a standard of measurement for the kleroi a mean between the largest and the smallest holdings in that vicinity, if it was impossible to find out the size of the allotments as established by the kings (the Seleucids).2

The indications that the central land registry of the Seleucids was without detailed descriptions of its village and allotment units and unable to keep up to date its information upon transfers from the royal domain are as follows:

1. The king, in all the available cases, must order the satrap to have the delimiting done locally.3

2. In the case of the sale to Laodice the central register is unable to indicate what Tórо (perhaps meaning single plots upon which peasants are living outside the village) are on the estate called "the village of Pannus."4

3. The village of Petra, conferred upon Aristodicides of Assos, had already been granted to one Athenaeus and the grant had not, as yet, been recorded at the central registry.5

The central register did, however, show, in the case of the village of Baetocaece, that it had once been held by an official named Demetrius, but was again available for assignment. For the royal domains about Babylon the local, or detailed, register would, of

1 P. Teb. I, Appendix I, p. 547. Cf. P. Teb. I, 27, 11. 7–8; 30, 1. 17 (34 arourae); 31, 1. 13 (34gg arourae); 32, note to line 18 (24 arourae); 54 (10 arourae).

2 The ingenious attempt of the editors of the Mnesimachus document (AJA, XVI, 73-75) to determine the actual size of the allotments granted to Mnesimachus leads to no acceptable result. The computation is based upon three factors, none of which is applicable with certainty to the situation about Sardes in the time of Mnesimachus. These are a 4 per cent return on land values (this must have been quite variable); an unconvincing comparison between the size of plots of city land about Magnesia with the kleroi about Sardes; and an unproved assumption that these city plots of Magnesia were approximately of the same productive quality as the allotments granted to Mnesimachus.

3O.G.I. 225, 1. 30; 221, II. 21–23; 262, 1. 8, "according to the existing boundaries." ▲ Ibid., 225, 11. 3-4.

5 Ibid., 221, 11. 50-54.

O.G.I. 262, 11. 6–7.

course, be in Babylon itself and immediately available to the officials of the king. This explains why, in a grant made to his wife Laodice by Seleucus II in 233-232 B.C.,2 the size of the grants (15, 12, 30 measures) is recorded.

There was a land register maintained for the Lydian satrapy at Sardes, at a bureau called the Baoiλikal Ypapai. This may have been the name of the land-registry office or of a general record office of the satrapy, of which the land registry was merely one department. In the sale to Laodice the satrap Metrophanes receives an order from the king to record the sale in this Sardian bureau. It is a justified assumption that similar land registers for the satrapies were established in the other provinces of the Seleucid kingdom.

There is no evidence regarding the extent of the details kept upon the satrapal registers. I have shown, however, that the actual descriptions of the boundaries were obtained through the hyparch, an official of the subdivision of the satrapy. One has, therefore, a strong impression that the satrapal registry office was not supplied with details as to the land units of villages and allotments. CORNELL UNIVERSITY

ITHACA, N.Y.

1 Haussoullier, op. cit., 18. The cuneiform tablet recording this gift is in New York.

2 Rostovtseff in Klio I, 299, n. 1.

3 Rev. de Phil., XXV, 18.

4O.G.I. 225, 11. 23-24, καὶ τὴν ὠνὴν ἀναγράψαι εἰς τὰς βασιλικὰς γραφὰς τὰς ἐν Σάρδεσιν.

O.G.I. 238, n. 2. See the fuller discussion by Haussoullier, op. cit., 22-26.