¹ The material is collected in Pauly-Wissowa, s.v. Iulius 386. We have two literary sources outside the historians, (a) Oracula Sibyllina XIII, 21–80, the only connected narrative for the period: retrospective prophecies, well-informed but conscientiously oblique and stylised. See Pauly-Wissowa, s.v. Sibyllinische Orakel 2158–2162; Olmstead (below). (b) [Aristides] XXXV Keil, εἰς.βασιλέα Groag suggested that the ruler addressed was Philip; the case has been revived most recently by Swift, L. J., Greek Roman and Byzantine Studies 7 (1966), 267 ff.Google Scholar It may be so. In any case the speech illuminates background more than events, and much that is said must be merely tralatician (n. 60 below). On the period in general, Olmstead, , Class. Phil. 37 (1942), 241–262, 398–420CrossRefGoogle Scholar. Recent literature: Walser and Pekary, Die Krise d. röm. Reiches 21 f.

² Decline and Fall, ch. X (on Decius and his successors).

³ For example, we dispose of about 200 census-returns from an original total of about 30 million (Hombert and Préaux, Recherches sur le recensement dans l'Égypte romaine 40).

⁴ .

⁵ κατὰ τὰ κελευσθέντα (in POxy. 78, 14 κατ᾿ ἐνκέλευσιν): not a special order for an individual case, but a general instruction like that which initiated the fourteen-yearly census for the whole country. One such order survives: the edict of Mettius Rufus, providing for a registration of property (Mitteis, Chrest. 192).

⁶ Minor corrections in Berichtigungsliste 1, 314, confirmed by Samuel, , JJP 13 (1961), 51Google Scholar. Comment: Taubenschlag, , Opera Minora II, 190 f.Google Scholar

⁷ κατ᾿ ἄνδρα βιβλίον.

⁸ The brothers in PLeit. 16 have obvious motives of self-interest: less property, less tax. In POxy. 78 Sarapas acts against his own advantage, .

⁹ The addressee is the strategus Aurelius Dius, alias Pertinax, who is attested in office on 28.1.245 and 17.7.246.

¹⁰ Many of the readings are difficult. Dr. Mariangela Vandoni has kindly sent me a photograph of the original. In 18–19 I suggest .[, cf. PLeit. 16, 10. (At 8 the ṗapyrus has Πανάρους genitive. After 10, perhaps a lacuna.)

¹¹ The editor's text of lines 13–15 runs: (from εὖδος) makes an unlikely phrase. Better to punctuate after καθολικοῖς; ειδεσι (εἰ δὲ σι-?) belongs with the destroyed line following.

¹² Later, indeed, there were two rationales functioning together: but, as the evidence stands, not before the fourth century. See Lallemand, J., L'administration civile de l'Égypte 80 ff.Google Scholar

¹³ This curious text is to be published elsewhere.

¹⁴ The proclamation ends with the note Ὀξυρυγχίτου, which implies that copies were sent to nomes other than the Oxyrhynchite—very probably to all the nomes in Egypt. But the schedule which follows carries no such proof of universality, and might reflect only the local conditions of the Oxyrhynchite. See note 23.

¹⁵ The schedule is not necessarily complete at the end: POxy. 2664, introduction.

¹⁶ Rostovtzeff comments on this aspect of Philip's reign: Social and Economic History of the Roman Empire ² 489 and 746, n. 57.

¹⁷ POxy. 78 belongs with the ‘general’ registrations of property—that is, those which cite official κελευσθέντα a (see Harmon, , Yale Classical Studies 4 (1934), 182 ff.Google Scholar). We know of no such registration since 146 (Chron. d'Égypte 22 (1947), 123CrossRefGoogle Scholar); the next was to be in 298 and after (Lallemand, L'administration 174 f.). Marcellus and Salutaris seem to have revived a long disused practice.

¹⁸ The state based its records on the various kinds of ἀπογραφαί. The last known returns in the different groups are dated as follows: donkeys, 119/20; property (general), 146 (note 17); camels, 216/7; sheep and goats, 238 (PPhil. 8, introduction); uninundated land, 240 (Chron. d'Égypte 38 (1963), 120 f.Google Scholar); property (regular), 256; census, 257/8 (but see Hombert and Préaux, Recherches sur le recensement 52 f.). This material, taken from Aegyptus 15 (1935) 133–7Google Scholar, except when another source is specified, may be outdated and misleading. So far as it goes, it illustrates the total breakdown of the state's information-service in the half-century after the death of Caracalla. When Diocletian set out to revive Egypt, he revived the information-service as well (Lallemand, L'administration 172–7).

¹⁹ Lallemand, L'administration 28–9.

²⁰ SB 7696: commentary JEA 21 (1935), 224 ff.Google Scholar

²¹ POxy. 2664, 6–8:

²² POxy. 2664, 13–14, note.

²³ P. Mertens, Les Services de l'État Civil 16 ff. Mertens emphasises that conditions may have been different outside Oxyrhynchus. Indeed, SB 7375 attests four phylarchs for an unnamed city in the reign of Alexander Severus; this city may, I suppose, have been one of the Greek πόλεις, where exceptional constitutions are expected; more probably, it was another nome-capital (Mertens 23 f. suggests Hermopolis). The schedule in POxy. 2664 testifies only to Oxyrhynchite conditions (note 14). On the other hand, the recently published PWis. Sijp. 2 does not touch our problem: Dr. Rea points out that the dating formula (ll. 45–6) belongs to Aurelian, not to Septimius Severus.

²⁴ For similar levies, see Wallace, , Taxation in Roman Egypt 168Google Scholar. Rostovtzeff, SEHRE ² 489, regards the ἐπιβλή as an increase in the rent of state-land.

²⁵ The text of O Bodl. I, P 138 is tacitly corrected in OBodl. III, index VIIIB, s.v. Φόρος.

²⁶ Wallace, Taxation 334, 352. A connection had already been seen between these exactions and the expenses of the millenary.

²⁷ BIF AO 47 (1948), 192 f.Google Scholar

²⁸ On the reversal of 302, see Lallemand, , L'administration 206 f.Google Scholar

²⁹ PLond. 1157 R and V contains a miscellany of documents relating to Alabastrine in the Hermopolite nome (V, 1–11 = Wilcken, Chrest. 375, above p. 134). The recto is an account of money taxes for a year 6; this account mentions decaproti three times. Before col. 1 stands another document, inserted by a different hand : a return, from Alabastrine, for the census of 229/30. The account, therefore, dates from 226/7 at the latest (see Wilcken, Grundzüge 217). Hence the normal view that decaproti were introduced by Septimius Severus.

³⁰ PLeit. 16, introduction. Again, decaproti play an important part in the sale of derelict land in Wilcken, Chrest. 375.

³¹ For example, PSI 1072, 21–4. τὸ ἀπότακτον is equivalent to φόρος ἀπότακτος (ἐκφόριον ἀπότακτον), e.g. at PSI 739, 15; 1072, 23.

³² So PBerl. Leihg. 5, 5–6 (an account of transport charges), εἰσὶν ἀποτάκτου κβ (ἔτουσ) πυροῦ (ἀρτάβαι) ᾿Βσιη: ‘die auf das Dorf entfallende, von einer oberen Behörde festgestellte Gesamtsumme’ (Kalén). PCairo Isid. 42, 6 (receipt for payments of delegatio), κερκ(εσούχων): ‘If ἀπο(τάκτου) is correctly resolved, it refers to the total assessment imposed on Kerkesoucha for delegatio’ (Boak and Youtie). Other examples: SPP XXII, 183, 35–7; PStrass. 66, 8; PAmh. 119, 8; PFay. 39, 15; OMich. 752. PMil. Vogl. 179, 3 ff. might be interesting, when elucidated: The use of ἀποτάσσειν as ‘earmark’ seems not to be relevant, though attested in literature and inscriptions (Plut., Mor. 201 A; Robert, , Hellenica IX, 14–18Google Scholar and Nouvelles Inscriptions de Sardes 16 f.).

³³ The broken phrase in PMil. Vogl. 97, 9 f. might refer to taxation: ἐκ τοῦ προτεθέντος κ[ατ᾿] ἄνδρα βιβλίου ἀποτάκτο[υ. ἀπότακτον does not appear in the parallel clauses POxy. 78, 18 ff.; PLeit. 16, 10; P5/249, 8ff.

³⁴ Against Mertens, Les Services de l'État Civil 25, n. 141, who suggests that ἀπότακτον designates the minimum financial qualification for nominees to the office. POxy. 1562, 11, from the same reign as POxy. 1409, has the phrase πρὸ τῆς τοῦ ἀποτάκτου συστάσεως—unfortunately in a mutilated context.

³⁵ ‘Diese feststehende Sache’ (Preisigke, , Wörterbuch I, 196Google Scholar); ‘die betreffende Bestimmung’ (Preisigke, Fachwörterbuch 30). Grenfell and Hunt doubtfully suggested ‘the appointed office’, POxy. 1187, 13–15, note.

³⁶ Compare: σύστασις ‘delegation of a tax among the individuals of a community’ (PGoodsp. 12, 1, 15; SPP XX, 88, 9 ff.); ἀσύστατος ‘without tax assigned’ (PLond. III, 1249, 5–7, p. 227); συστάτης ‘nominator of liturgists’, one who assigns individuals to the offices (Mertens, Les Services de l'État Civil 32–5).

³⁷ p. 136 above.

³⁸ POxy. 2664, 5–6, illustrates what σύστασις might mean in such a case : the emperors order the reduction of liturgies, Marcellus and Salutaris use their own judgment (ὅπως ἐπεκρείναμεν) in some of the details. I have even wondered whether the συστησάμενοι POxy. 1187 might be identified with Marcellus and Salutaris themselves.

³⁹ Lines 102–3. ‘Diese neue Steuerfestsetzung’, Wörterbuch iv, 264.

⁴⁰ Each new régime was naturally eager to hear of the abuses of the displaced rival: the motif is clear in POxy. 2664, 6–9 (already, and more explicitly, in the edict of Ti. Julius Alexander, OGIS 669, 4 f. and 14). Christian sources credit Decius with a real personal hostility to Philip and his works (Class. Phil. 37 (1942), 398Google Scholar).

⁴¹ This is not to say that ἀπόταικτον is peculiar to Philip's reign. POxy. 1409 and 1562 (Probus) may refer to a later imposition: see note 76.

⁴² POxy. 2664, 2, note.

⁴³ Lallemand, L'administration 80 ff.; POxy. 2664, 1, note.

⁴⁴ POxy. 2664, 1, note.

⁴⁵ For the literature, see PMerton 11, p. 152.

⁴⁶ Line 5 is printed: : the acting prefect was (it seems) only the acting holder of his other office: an extraordinary and therefore suspect titulary. Dr. Anna Swiderek has been kind enough to send me a photograph of the original. From the photograph I hazard a reading which restores a normal formula: looks very probable; —already conjectured by Dr. John Rea—is consistent with the traces). This does little towards determining the date: the iuridicus functioned as prefect in 176 and in 225 and perhaps in 215/6, and at other dates before and after these.—In line 7 read ].. κύριε for ]…. ρισ. In line 24 Dr. Rea has read ]τῦβι for π]αῦνι (deleting the supralinear stroke in the line below).

⁴⁷ ‘Wende des 2 und 3 Jahrh.’, according to the editors.

⁴⁸ Preisigke, , Wörterbuch III, 379Google Scholar.

⁴⁹ Hornickel, Ehren- und Rangprädikate 17.

⁵⁰ Taubenschlag, , Opera Minora 1, 206 f.Google Scholar

⁵¹ After διατάγμᾳ[τος restore, e.g., κελεύοντος; the lacuna at the beginning of line 7 is of uncertain length. This reinterpretation of BGU 1578 raises legal difficulties, (a) What is the ἱερώτατον καθολικὸν διάταγμα It seems an unlikely designation for the provincial edict, to which petitioners do some-times refer (see Chalon, L'Édit de Ti. Julius Alexander 72–5). I suggest, dubiously, that the phrase translates edictum perpetuum. If so, the document deserves further investigation, (b) ἀχαριστία is the charge. I can find nothing similar in legal sources before CJ 8, 49 of A.D. 367 (later developments, Riccobono, , FIRA III, pp. 33 f.Google Scholar); the actio ingrati of the rhetorical schools (Seneca, Contr. 2, 5; Juv. 7, 169) had been entirely unreal (Seneca, Contr. 3, praef. 17).

⁵² Julianus appear s in Wilcken, Chrest. 171 and SB 4639. Julianus as rationalis: PMerton 11, p. 152.

⁵³ Préaux, , Chron. d'Egypte 38 (1963), 128 ff.Google Scholar He had already functioned as procurator and acting high-priest in 197 (PAchmim 8). The acting high-priest of 215 was procurator usiacus (BGU 362, VII, 24–6); but this is not evidence enough to assign Diognetus the same procuratorship (PGiss. 48, introduction).

⁵⁴ The parallelism is much overdrawn by Rostovtzeff, SEHRE ² 726.

⁵⁵ Rea, Parola del Passato (1967); cf. now BGU 2024.

⁵⁶ Notice the Correctores Aegypti: Ulpius Pasion in 257 (POxy. ined.); Claudius Theodorus in 258 (PIR II², 1041: διασημότατος); Claudius Firmus in 274 (PMerton. 1, p. 157 ff.: precedence over the prefect); Achilleus in 296 (ibid.; PCair. Isid. p. 250: jurisdiction like the prefect's). Similarly the two duces, whatever their precise office, who operated in Egypt under Gordian III : one διασημότατοςλαμπρότατος, the other perfectissimus (Gilliam, , Chron. d'Égypte 36 (1961), 386 ff.CrossRefGoogle Scholar; Lallemand, L'administration 67, n. 1).

⁵⁷ Miss Lallemand (L'administration 81–3) remarks that the functions of the rationalis must have overlapped those of the dioecetes. If the rationalis had no existence between Philip and Diocletian, this problem disappears. But the ramshackle Egyptian bureaucracy easily accommodated overlaps.

⁵⁸ On Philip's financial situation, see Syria 38 (1961), 275 ff.Google Scholar

⁵⁹ Gordian III had already reduced the weight of the aureus (Magie, Roman Rule in Asia Minor 700 f.).

⁶⁰ Zosimus 1, 20, 2 : τὰ μὲν κατὰ τὴν ἑῴαν ταῖς τῶν φόρων εἰσπράξεσι καὶ τῷ τὸν Πρίσκον, ἄρχειν τῶν ἐκεῖσε καθεσταμένον ἐθνῶν, ἀφόρητον ἅπασιν εἶναι βαρυνόμενα κ.τ.λ. [Aristides] gives a less gloomy picture of his emperor's tax-policy (16): οὐ τοῦ πλείονος ἐδεη οὐδ᾿ ἐζήτησεν οὐδὲ διὰ χρήματα κακὸς ἐγέυετο, ἀλλ᾿ ἀνῆκε καὶ ἐπεκούφισεν. But this may be mere commonplace; Swift refers to the locus in Spengel, Rhetores Graeci III, 375, 21–4. It has indeed been suggested that it was Philip who abolished the Egyptian poll-tax: the last of the few third-century references belongs (perhaps) to 243. But the whole question remains very obscure. For the evidence, see Bell, , JRS 37 (1947), 18–23Google Scholar.

Nothing can be said in detail about Jotapian's revolt. Bersanetti believed that it extended to Palmyra, but his inscription as it stands antedates the reign of Philip (Laureae Aquincenses 2 (1941), 265; Chilver, JRS 38 (1948), 166Google Scholar). Another eastern pretender, Uranius Antoninus, may begin his reign under Philip: Barbieri, L'Albo senatorio da Severo a Carino 405 f.

⁶¹ Johnson, JJP 4 (1950), 151 ff.Google Scholar

⁶² Braunert, Binnenwanderung 269 f.

⁶³ POxy. 1418; PErlangen 18. A new, and still unpublished, Oxyrhynchus document gives the same impression of crisis. The document is a proclamation of 19 March 246, issued on the authority of the iuridicus Aur. Tiberius; it orders the immediate registration of all stocks of grain in Oxyrhynchus and its nome, with the view to compulsory purchase by the state as a means of supplying the city and the δημόσιαι χρεῖαι—the necessities of taxation.

⁶⁴ SB 7696, 102 f. [Aristides] gives a general picture of misery.

⁶⁵ Bell, , JRS 37 (1947), 20Google Scholar.

⁶⁶ Compare Rémondon's remarks, Rev. Phil. 28 (1954), 203 ff.Google Scholar

⁶⁷ Or. Sib. XIII, 42–45. 46–9 look like a later addition: Scott, , CQ 10 (1916), 15 f.Google Scholar

⁶⁸ This is sometimes thought to explain the organization of the classis africana by Commodus (Rostovtzeff SEHRE ², 708, n. 4 gives the sources). At best, Egypt supplied only a third of the corn for the city of Rome (Josephus, , BJ 2, 385 f.Google Scholar). But even in the third century control of this supply was a powerful weapon (SHA, Gallieni 4, 1).

⁶⁹ Of course, the problem of finding liturgists was not confined to Egypt. Two of Philip's surviving rescripts serve to close loopholes (CJ 10, 39, 3; 10, 53, 3).

⁷⁰ The same oily and self-exculpatory tone distinguishes PFay. 20 (Alexander Severus remits the aurum coronarium) and PCair. Isid. 1 (Aristius Optatus announces Diocletian's new taxation system).

⁷¹ POxy. 1662; SB 7696, 102 f.

⁷² Class. Phil. 56 (1961), 3Google Scholar; Frend, Martyrdom and Persecution 403 f. But then few emperors were popular in Alexandria.

⁷³ Lewis, , JEA 29 (1943), 73Google Scholar, on Diocletian and his predecessors.

⁷⁴ CJ, 11, 58, 1.

⁷⁵ SHA, Aurelian 47.

⁷⁶ Rémondon, , Rev. Phil. 28 (1954), 199 ff.Google Scholar Considerable efforts were made, and results achieved. Observe the ἀπότακτον of POxy. 1409 and 1562: Probus' programme? SHA, Probus 9, 3–4 describes an increase in yield produced by the works which Probus, while still a general of Aurelian, carried out on the Nile. This detail, like the military ones in the same chapter, probably belongs to the prefect of 269/70, Tenagino Probus (Stein, , Klio 29 (1936), 237Google Scholar). Even so, the attempt coheres chronologically with the others.

⁷⁷ If we could be certain that Philip replaced sitologi with decaproti, and the amphodogrammateus with the phylarch, the parallelism would be extremely close.

⁷⁸ Who planned? The abatement of liturgies, POxy. 2664, is theoretically the work of the emperors. But elsewhere, to speak of ‘Philip's’ plans is merely convenient shorthand. Much may come from the emperor's brother Priscus, viceroy in the East. Priscus was familiar with Egyptian conditions: he had been iuridicus and acting prefect in 238/40.