CORN DOLE ARCHIVE

were complete for the whole city and that the city was divided for this purpose into twelve areas. Since the number of quarters named in the papyri is more than twice that, see Rink, Strassen- und Viertelnamen, p. 52, some at least of the twelve areas will have included more than one quarter, and it will have been only for convenience that they were designated by the name of a single quarter.

It has already been established that an Oxyrhynchite phyle tribe consisted of amphoda quarters (e.g. Mertens, Les services, pp. 15, 24, 129), so that the question naturally arises whether these twelve areas correspond to the tribes, as might be suggested by the title of the phylarch in 1119 13 "to the phylarch of the quarters about to perform liturgy," and of the systates successor to the phylarch, in PSI 1108, 5–6 and 2715 5–6, "systates of the tribe of the Gymnasium Promenade and other quarters," cf. 1116 5–6; P. Flor. 39 4. That they do correspond to the tribes seems to be confirmed by 2930, which is a piece from the beginning of a list of recipients, compiled by a person described as "a former phylarch of the tribe of the Thoeris Promenade and the Lycians (Parembole?)." This evidence, together with the separate responsibility of the phylarchs for compiling the registers of corn dole recipients in this archive, strongly inclines me to think that these areas are equivalent to tribes and that there were twelve tribes at this period.

The Oxyrhynchite tribal system is imperfectly understood, but we know from the document published in TAPA xcix (1968), pp. 259–63, and from three others to be published along with a reprint of it in a forthcoming volume, that it began in A.D. 206/7 with six tribes, each of which provided public servants for a year in rotation till A.D. 229/30. In the next three years A.D. 230/1–232/3 the tribes acted in pairs. A little later we find a nine-year rotation period in effect, at least between A.D. 244/5 and A.D. 253/4 (1119). Nothing more is known except that there were at least three tribes in the reign of Aurelian (1413 12, 13) and that a tribal system continued as late as A.D. 396 (P. Flor. 39). If PSI 86 is from Oxyrhynchus there were at least sixteen tribes c. A.D. 367–75.

It is rather uncertain how much support the evidence in this archive provides for the view that the Roman corn dole was organized on the basis of the Roman tribes, see 2928 introd. and below, pp. 14 seq.

The phylarch in Oxyrhynchus

This official has been well investigated by Professor Mertens, Les services, pp. 16–30. At the time of his work, however, there were no dated mentions of the phylarch between A.D. 254 and the appearance of his successor the systates (A.D. 287, Mertens, p. 19). These papyri, together with P. Wis. 2 (Aurelian; see BASP iv 34, CE xliv (1969), pp. 134–8) and 2764 (Probus), fill that gap to some extent.

One additional point of importance emerges. The phylarch was thought to function only in years in which his tribe had the duty of providing public servants (Mertens, pp. 22–4). From this archive it is clear that even in years in which his tribe was not serving the phylarch was responsible for keeping records, at least of the corn dole. In