the purchaser (the document is a sale) had moved to Syene after the date at which it was written 1).

Lastly, it may be remarked that sailors are fairly common in these papyri. Patermuthius and John were sailors before joining the army, and several other persons are described as ναύτης. One κυβερνήτης occurs.

4. Clergy.

After the soldiers, the most numerously represented class in these papyri, so far at least as witnesses are concerned, are the clergy. There is, however, not much of interest to be noted in this respect. The majority of the clergy are διάκονοι. Several πρεσβύτεροι occur; they are variously described as πρεσβύτερος ἐκκλησίας Συήνης, πρεσβύτερος τῆς τοῦ Θεοῦ ἁγίας ἐκκλησίας Συήνης, and πρεσβύτερος Συήνης. John, μονάζων, and John, αυλιζόμενος ἐν τῷ μοναστηρίῳ Πανιπάνης, perhaps one and the same person, have already been mentioned (above, p. 166). A nun (uorapi) also occurs. In Inv. 1792 one witness, in a very illiterate signature. appears to describe himself as arazrootir. which. if correctly read, is presumably to be corrected to dræpróótng2). Finally, in Inv. 1800 occurs an ἀρχιδιάκονος τῆς ἁγίας Μαρίας Συήνης.

The large proportion of clergy who occur as witnesses in this and still more in other collections of late Byzantine papyri is doubtless to be explained by the great increase of illiteracy, or at least of the inability to write Greek, so that only among the clergy could many persons able to subscribe their names be met with. That even in this class there were many who could not write Greek is clear from many pieces of evidence, notably from the case of Abraham, Bishop of Hermonthis, in P. Lond. 77. In the present collection none of the ecclesiastics who occur as principal parties (John, poréşor. John, avióueros, Tsone, uore) is able to write.

5. Scribes.

For the most part the scribes of these documents do not add to their names any descriptive epithet. That neither this fact nor the absence of the usual cortos or Agios proves anything as to their profession is shown by the fact that the same scribe will sometimes omit and sometimes insert his title 3). From a comparison of the various

1) The earlier part being lost, the date and the descriptions of the parties are missing, but the document must have been found at Assuan, as it was with the other Syene papyri.

2) Prof. Heisenberg doubts this, saying, das ist sprachlich sehr schwierig." 3) That the variation is not due to difference of date, a scribe once a civilian having entered the army later, is shown by the case of Mark son of Apa Dius, who in A. D. 577 (Inv. 1801) signs as orgatióing úgiduoi Evýrys, but in documents of a later date omits the description.

scribal signatures it appears that professional notaries played comparatively little part at Syene, and that the majority of the deeds, even where the parties concerned were civilians, were written by soldiers. The following is a complete list of the scribes whose names are preserved in the British Museum collection. Descriptions preserved in any one document, when the same scribe is responsible for more than one, are added to the name. The numbers after some of the names give the number of documents written by the scribe; and in every case where he is also the writer of documents at M(unich) I have added the number in that collection also. Abbreviations are extended.

Αβραάμ Δίου.

Φλαούιος Ἀβραὰμ Παμήτ, κεντυρίων λεγίωνος Συήνης.

Αλλάμων Πέτρου, ἀπὸ βικαρίων (2, Μ. 1).

Φλαούιος Δίος Βασιλείδου, ἀπὸ ἀριθμοῦ Συήνης (23)*).

Διος Παπνονθίου.

Θεοδόσιος Απολλωνίου, νομικὸς Βαβυλώνος?)?).

Θεόφιλος, ἐλάχιστος διάκονος).

Φλαούιος Λάζαρος Πέτρου, ἀδιούτως ἀριθμοῦ Συήνης.

Μάρκος Απα Δίου, στρατιώτης ἀριθμοῦ Συήνης (54), Μ. 2)5).

Among these scribes Dius son of Papnuthius enjoys an evil preeminence by virtue of the excessive badness of his Greek. Dius son of Basilides, in both the documents assigned to him, writes a very illegible hand (he is certainly the worst writer of them all), and commits many grammatical blunders. Allamon son of Peter is also very uncertain in his constructions. The most accomplished penman is Theophilus the deacon, whose writing is easy and elegant, and whose Greek is moreover comparatively correct. The general impression derived from these documents is, however, that in this outpost of the Empire the Greek language was rapidly decaying. 1) This scribe is somewhat doubtful. In Inv. 1788 the scribal signature, if it was ever added, which is doubtful, is lost; but the body of the document seems to be in the same hand as the subscription, which is by Dius. In Inv. 1789 the scribe was Dius son of The badly written patronymic can at need be read as B[co]thɛidov, and the hand of the document, though at first sight unlike that of Inv. 1788, seems, on a more detailed examination, to have a good deal of similarity to it.

2) This person, if the extension of Ba is correct, belongs of course to a different category from the rest.

3) The word diάzovog is lost, but can be certainly restored from another document, in which this scribe subscribes for an illiterate person.

4) This includes one papyrus (Inv. 1799 Mon. 109) of which the beginning

is in the British Museum, the conclusion at Munich.

5) To these may be added Picovios Aßqaàμ Movaciov, the scribe of Mon. 106; for the beginning of that document is in the Brit. Mus. (Inv. 1803a); see above, p. 164, note 2. Being called Flavius, he too was no doubt a soldier.

Klio, Beiträge zur alten Geschichte XIII 2.

12

6. Classes of Documents.

A complete description of the contents of the papyri cannot here be given, but the nature of some of them may be briefly indicated.

It has already been remarked (above, p. 163) that the three documents from the neighbourhood of Thebes all relate to loans. Inv. 1794 and 1805 are actual loans1), in the latter case some articles of jewellery being pledged as security. Inv. 1793 recto is apparently a receipt for the full price of an ear-ring deposited as security for a debt; the borrower had evidently found that she was unable to redeem her pledge, and being in want of money she surrendered it to her creditor in return for the difference between its full value and the amount of her debt. Of the documents written at Syene four are loans.

Two documents are concerned with the inheritance of Jacob, which plays so large a part in Wenger's Vorb. (pp. 12-17). Inv. 1790 having already been published by the Palaeographical Society, it is unnecessary to say anything further concerning it here 2). The other document, Inv. 1792, which is dated in either 584 or 585 (see above, p. 165), is connected with the disputes between brother and sister concerning the inheritance. The loss of a few letters from the beginning of each line and the fact that the papyrus is badly rubbed in places render portions of the document at present unintelligible, but the general sense is clear. John acknowledges to his sister Kako his abandonment of all claims on the zorrorqucion лQayuάτor, but reserves from this surrender certain property, including τῶν ξύλων τοῦ πλοίου ... ἱσταρίου (cf. Vorb. p. 13).

In Inv. 1791 a certain Jacob son of Apa Dius, a sailor, becomes surety to Patermuthius for the appearance of τὸν Ψανὸ καὶ Σανμπού at an arbitration to be made by Μρκο (sic) τῷ λογιωτάτῳ δικστην (sic). This is probably the Mark 6702a6rizós who in P. Mon. 103 (Vorb. p. 16f.) acts as arbitrator between John, the brother-in-law of Patermuthius, and his mother Tapia. As by a fortunate accident some fragments from the beginning of this document are preserved among the London papyri, the arbitration can be dated in July, 583. It seems likely, therefore, that the date of Inv. 1791, which is given by the indiction only, is to be taken as 586 (see above, p. 165). Whether the dispute with which it is concerned had anything to do with the inheritance of Jacob it is impossible to say.

An interesting document is Inv. 1796, an agreement between Patermuthius and his wife Kako concerning their property, perhaps due

1) 1794 is perhaps rather a bond to pay interest on a loan than the loan itself.

2) There were a few misreadings in the transcript given by the Pal. Soc.; see the corrigenda in part 10.

to a desire to avoid such disputes concerning the inheritance as they had themselves been parties to in the case of that of Jacob. After a verbose preamble in which they state that their motive in concluding the agreement is to prevent the survivor being in any way defrauded by their heirs. (μήποτε κτλ. εὑρεθείη τὸν ζῶντα ὑστερούμενον παρὰ τῶν ἀμφοῖν ἡμῶν τέκνων), they declare, νοοῦντες φρονοῦντες λογισμοὺς ἔχοντες ἔρρωμένας τὰς διανοίας τὰς φρένας ἀπαθεῖς τὰς αἰσθήσεις ὑγιεῖς ἐπὶ ποδῶν βαδίζοντες ἐπ ̓ ἀγορᾶς παρερχόμενοι εἰς ταύτην τὴν ἔγγραφον ὁμολογίαν), that on the death of either of them the survivor shall have full and entire possession of all the property of the deceased, whether acquired by inheritance, purchase, or personal exertion, and that on the death of that survivor their common property shall be divided in equal shares among their children, no child being favoured at the expense of the others. They then provide that if any child attempts to set aside this agreement he shall be disinherited and pay a fine of 12 solidi. The nature of the document suggests that it may have been drawn up fairly soon after marriage as a kind of marriage settlement; but the parties must, as already said (above, p. 165), have been married at least as early as 582, whereas the date of this document is A. D. 583-4 (above, p. 165).

The collection includes a number of sales, which are of considerable interest in several ways; - for the formulae employed (cf. Vorb. section VI), for the light which they, with other Byzantine sales of houses, throw on the structure of the Graeco-Egyptian dwelling-house, and for the evidence they afford as to the topography of Syene. There are five of these sales, all of which are sales of house property. To them may be added Inv. 1790 (New Pal. Soc. VI, no. 128), which is a transfer of house property in consideration of the discharge of obligations which the vendor is unable to meet; and in the matter of evidence for the structure of houses we may also add Inv. 1801, a loan on the security of house property. I have not left myself room here for any discussion of the first two subjects, but one interesting point in connexion with the second may be noted. In Inv. 1802, in the specification of the property sold, occur the words άλλο (i. e. κέλλιον) δὲ ἀπηλιωτικὸν νεῖον εἰς λίβα εἰς τὸν πυλόνα καὶ τὸ ὑποπέσσιον) ἤτοι τχρηρε, where έχρησε is evidently a Coptic feminine noun synonymous with лолéбotov. Mr. Crum, in reply to a query, informed me that he could not trace any such Coptic word, but suggested that light might be thrown on it from Demotic sources. This suggestion proved

1) Cf. the formula in the „Schenkung auf den Todesfall", Zeitschr. d. Sav.Stiftung, XXXII, p. 326, and in the will in P. Cairo byz. 67 151. Prof. Heisenberg suggests a reference to Mitteis, Grundzüge, 246 (Elterliche Teilung c).

2) Cf. also P. Flor. 15, 17, where inoлέooiov is the right reading, Archiv 111, 532.

12*

fruitful; for Sir Herbert Thompson, whom I next consulted, has identified the word with a rare Demotic feminine word found, up to the present, in only one document. This is P. dem. Strassb. 1 (Spiegelberg, Dem. Pap. d. Strassb. Bibl. p. 18ff.), a conveyance of house property dated in the 9th year of Alexander. According to information kindly supplied me by Sir Herbert Thompson, the parts of the house are the following (I quote from his letter):

„hyt

hph

Copt. 2AEIT, elsewhere = usually лvior, but occasionally dúoα or лQоavior (Spiegelberg translates „Vorhof“).

a building, meaning unknown.

enclosure, courtyard (Spiegelb. „Hof“).

..

trt staircase, probably a large outside one descending from the upper storey to the courtyard (Spiegelb. Terrasse (?)"). usually a detached garden-house (Spiegelb. „Kiosk“). roupa, door of the house, as opposed to the лvλóv.“

syḥ

Spiegelberg translates hrr „Frauenabteilung“, „evidently because he identifies it with a well-known hieroglyphic word; but I doubt very much if he is justified in doing so on philological grounds; and I see no reason in the text for suggesting such a meaning" (H. T.).

The evidence of Inv. 1802 makes it clear that Spiegelberg's rendering must be given up; for there cannot, I think, be a doubt that zonos and hrr are to be identified. The лɛбoóc is shown by abundant evidence to have been part of, or in close connexion with, the πυλών; and the υποπέσσιον was obviously the space under the л866ós. Since then hrr goes along with hytлviv and trt = staircase or „Terrasse", the meaning лолéбotov suits it perfectly. That a rare Demotic word occurring in a papyrus of the reign of Alexander should thus be elucidated from a Greek document of the reign of Justin is a striking instance of the inter-connexion of papyrological studies.

πυλών

7. Topography of Syene.

As already remarked, some light is thrown by the contracts of sale on the topography of Syene. There is not sufficient material to give any detailed idea of the plan of the town perhaps the Munich texts will in this respect supplement the London collection, but the scanty data of the latter may here be set down.

Two houses are described as situated in the νοτινὸν μέρος τοῦ goovoiov1). It may be inferred from this that Syene was divided, like Hermopolis, into quarters, one of which was called to Poovptor; and on

1) Cf. Zeitschr. d. Sav.-Stift. XXXII, 327.