Hence

(A.D. 892) is 'stratêgos of Cephallenia and Longobardia '.1 Gay has rightly concluded that it is not till after this year that Longobardia became a separate theme. But, on the other hand, there is no evidence that the separation was made before A.D. 900. Hence no inference can be drawn from the omission of Longobardia as to the date of the list.

The fact that the list includes the themes of Strymon and of Samos cannot be held to date it; for though the creation of these themes is often ascribed to Leo, this is by no means certain. The case of Thessalonica is a warning. Gelzer attributes the theme of Thessalonica to the Neuordnung of Leo VI (op. cit. 130); but this theme appears in the Taktikon of Michael III.3 The themes of Strymon and Samos do not appear in that document, but they may have been formed before the accession of Leo VI. The evidence, however, already adduced seems sufficient to date the source of the first list of Philotheos to the reign of Leo.

The lists of precedence in Sections II and III (cod. Lips.) agree with list 1 of Sect. I in omitting the hetaeriarch, but there are some variations in order. (a) In Section III the Drungarios of the Fleet follows, instead of preceding, the Logothete of the Course, and (b) the Logothete of the Flocks precedes, instead of following, the Protospathar of the Basilikoi (the latter does not occur in Section II); (c) in Section II the Comes Stabuli precedes ¿ èк проσάπov tŵv Oeμárov, but Section III agrees here with the lists of Section I. The variations are common to both MSS.

Another point of difference to be noticed between Section I and Sections II, III, is the treatment of the Magistri. In Section II we have αἱ δὲ λοιπαὶ πᾶσαι τῆς δευτέρας ὑπάρχουσι τάξεως οἷον ὁ μάγιστρος, ὁ μáyloτpos, and in Section III (ad init.) simply 8 μáylσтpos. In both cases we might expect οἱ μάγιστροι.

We may turn to the evidence of the Jerusalem MS. collated by Uspenski. (1) In this MS. in the lists of precedence, both in Section II and in Section III, we find the Hetaeriarch (μέγας ἑταιριάρχης) immediately after the Drungarios of the Watch. The fact that he occurs in both lists shows that the omission in the Leipzig MS. is not accidental. (2) The Stratêgos of Longobardia appears after the Stratêgos of Sicily in Section II. He is not mentioned in any of the lists in the Leipzig MS. On the other hand, the stratêgos of Nikopolis is omitted in the Jerusalem MS.; but this may be a mere scribe's error

1 Chron. Vulturnense (Muratori, R. I. S. i. 2. 413).

2 L'Italie méridionale, 171-4.

3 Uspenski, 115.

Phil. 713, 728.

(there are several other omissions in H which are clearly accidental). (3) Instead of åvlúñaros naτpíkios the Jerusalem MS. has throughout simply ἀνθύπατος. (It also has in most cases σπαθάριοι instead of σnalaρоkavdidáтo, but probably this is merely a mistake of the scribe.) (4) In Section II where the Leipzig MS. has ó μáyισтpos å μáyιoтPOS the Jerusalem MS. has & μáyloтpos; but this may be due to parablepsia. (5) The precedence of the protospatharioi of the Chrysotriklinos is said in L to have been established máλa (Section III, p. 732), but in H it is attributed to Leo VI.

The probable inference seems to be that the Jerusalem fragment belonged to a slovenly copy of a later recension of Philotheos than that which is represented by the Leipzig text, which was copied from the original. The editor, whether Philotheos himself or another, brought the treatise up to date by inserting the Stratêgos of Langobardia, and repaired the error of omitting the Hetaeriarch. The discrepancies between Section II and Section III seem to be due to the circumstance that Philotheos was using old lists of different dates and he did not succeed in eliminating all the inconsistencies.1

(4) Scope of the following investigation. General comparison of the Constantinian with the later Byzantine System.

The following pages are not a complete commentary on Philotheos. The investigation is confined to the determination of the functions of the officials, and to the origin of the offices and of the orders of rank. I have not entered upon the subject of the fees (σvvýbeta) paid for dignities and offices, and the Imperial bounties (evσeßlai, áñrokóμßia, dopa) to which the dignitaries were entitled. The latter and main part of the book of Philotheos-Section IV-is important for my purpose, as it throws light on many difficulties which arise out of the earlier part; but a commentary on it belongs not to this inquiry, but to a treatise on the court ceremonies.

From Philotheos we derive no information as to the civil government of the provinces, except so far as finance is concerned. The provincial judges are not mentioned. We hear nothing of oi ἀνθύπατοι καὶ ἔπαρχοι τῶν θεμάτων or οἱ πραίτορες τῶν θεμάτων who appear in the Takt. Usp. (118, 119). A large question of considerable

1 In Phil. 78811 we meet the катеnáv∞ of Paphlagonia. In the time of Philotheos, and since the early years of Michael III, the governor of Paphl. had been a σrparnyós (Phil. 713,, Takt. Usp. 113). Under Theophilus he had been a Katepano (De adm. imp. 178,), and perhaps Theophilus raised the dignity of the theme. It looks as if Philotheos were here using a document dating from more than sixty years back.

difficulty, touching the position and the districts of these officials, and their relations to the Stratêgoi, is involved, and I have not been able to discuss it in the present investigation.

A few remarks may be made here as to the general character of the organization of the ninth century as contrasted with the older system which it superseded.

If we compare the scheme of administration which was founded by Diocletian, and completed by his successors, and which remained intact, except in details, till the beginning of the seventh century, with the later Byzantine system, we find that while there is no break in continuity, and the changes seem to have been gradual, the result of these changes is the substitution of a new principle.

The older system has been described as a divine hierarchy. Gibbon designates its principle as 'a severe subordination in rank and office'.1 There was a comparatively small number of great ministers and commanders-in-chief who were directly responsible to the Emperor alone. All the other administrators were ranged under these in a system of graded subordination. In the Notitia Dignitatum of the East we can count twenty-two high offices,2 to some of which all the rest were in subordinate relations.

3

In the ninth century it is quite different. There is no hierarchy of this kind, so far as office is concerned. The number of independent officials responsible only to the Emperor is enormously larger. Instead of twenty-two it is about sixty. And these numbers do not fully express the magnitude of the change. For in the fifth and sixth centuries the territory ruled from Constantinople was far more extensive than in the ninth. It included Syria and Egypt and extended to the Danube. Long before the ninth century, Syria and Egypt and a great portion of the Balkan peninsula were lost.

This change was brought about in two ways. (1) The whole provincial administration was reorganized. The provincial territory was divided into a number of military districts, or Themes, and the governor of each theme, who was primarily a military commander, had also a certain civil jurisdiction. He was independent, subject only to the Emperor. He was not under the orders of any Master of Soldiers or Praetorian Prefect. In fact the Masters of Soldiers and the Praetorian Prefects disappeared. (2) The great central

1 Decline and Fall, c. xvii, p. 169, in Bury, new ed. vol. ii (1909).

In the reckoning I omit the castrensis, and include the Proconsul Asiae, who was not under the vicarius Asianae or the Praef. Praet. Orientis.

The hierarchy of rank remains and has been developed into a more elaborate scale.

ministries of the Master of Offices, the Count of the Sacred Largesses, and the Count of the Private Estate, each of which consisted of many different departments, and had an extensive range of functions, were broken up into a large number of offices with restricted competence. These changes were not brought about at a stroke, by a single deliberative act of administrative reform. They came about by a gradual series of modifications, but they all tended in the same direction, to substitute the principle of co-ordination for that of subordination, and to multiply supreme offices instead of placing immense powers in the hands of a few. We cannot point to any single emperor as the Diocletian of the new system. It is probable that Leo the Isaurian did much to normalize it, but it was in the seventh century under the Heraclian dynasty that the older system had broken down and been irrevocably abandoned, and the chief principles of the newer had been introduced. Even in the sixth century we can discern some foreshadowings of the change.

B. DIGNITIES (αἱ διὰ βραβείων ἀξίαι).

In the sixth century, apart from the exceptional titles of Caesar, nobilissimus, and curopalates, there were a number of dignities, unattached to office, which could be conferred by the Emperor. The highest of these was the Patriciate (introduced by Constantine), which was confined by a law of Zeno to men who had been consuls or prefects, but was opened by Justinian (Nov. 80) to all men of illustrious rank. There were also the titular offices of the consulship, the prefecture, and the stratêlasia (magisterium militum). The acting administrative officials were distinguished as in actu positi or μпрактоι 1 from the titular officials (ǎπρакто), who were of two kinds, (1) illustres vacantes, and (2) illustres honorarii.2 The vacantes not only bore the title but wore the cingulum, the insigne of office; the honorarii had the title but not the cingulum. But in all cases the dignity was conferred by codicilli. In the case of most offices, the titular dignity was probably conferred only on those who had once held the office, but the consulship, the prefecture, and the stratêlasia were regularly conferred on others than officials. The

1 In later texts we generally find the forms μπратоs and äпparos, e. g. Cer. 239, κἂν στρατηγὸς ἔμπρατος κἄν τε ἄπρατος. Cp. περὶ ταξ. 50219 ἐν ταῖς ἐμπράτοις προελεύσεσιν. In Cer. 798 we find a curious third term μεσόπρατος. From this passage it would appear that μrparos was specially used of the Stratêgos, and μeσóñparos ñaτpiktos was applied to Patricians who held official posts in the capital (ὁ ἐμπολιτικὸς ὀφφικιάλιος).

2 C. I. 12. 8. 2. Cp. Mommsen, Eph. Epig. v. 129.

comitiva, which was in principle an order of the same kind, had been appropriated with its three grades to particular offices, to which it belonged as a matter of course.

In the course of the seventh and eighth centuries, the number of these orders, or titular offices, was largely increased, and they were conferred by investiture with insignia. There were several schools of officers in the palace, who had various duties connected with the Imperial service: silentiarii, vestitores, mandatores, candidati, stratores, spatharii. All these titles came to be used as ranks of honour, and were conferred upon all the more important civil and military officials according to their degree. The chief of the school of spatharioi was entitled the protospatharios, and this term was adopted to designate a higher rank than spatharios-the rank next to Patrician itself. Between the spatharioi and protospatharioi was interpolated a new class of spatharokandidatoi. To the hypatoi (consuls) was added a new and higher class of disypatoi (bis consules).

The protospatharioi were probably not instituted as an order before the end of the seventh century. In the seventh century, the Patricians and Hypatoi were the two most eminent ranks, and the añоeπάрxwv (ex Praefectis) and σrparŋλáraɩ were still very high dignitaries. In the course of the next two centuries these orders were rearranged and multiplied. The Patricians were divided into two ranks: the ordinary Patricians (TерíßλеTто), who retained as their insigne (ßpaßeîov) the ivory tablets, and those to whom the dignity of Proconsul was added (åv¤úñatoɩ kaì ñaтρíkɩɩ) who had purple tablets. More important and interesting is the creation of a new and higher rank, that of μáуtorρol. This innovation was obviously connected with the abolition of the office of magister officiorum. At first it was intended that there should be only one magister (as there was only one curopalates); very soon we find more than one, but throughout the ninth century the dignity was sparingly conferred.

In this place it will be convenient to add a note on the use of the terms ἄπρατος, λιτός, and παγανός which occur in Philotheos. ἄπρατος (vacans), to which reference has already been made, is used of persons who bear the titles of offices of which they do not actually perform the duties (e.g. otpaτnyol, ảoŋkpîτaɩ, &c., see Phil. 71011, 737 39 597). λιτός is applied to persons who have orders (dignities διὰ βραβείων), but are not ministers or officials; Phil. 72915 of Auroì åvðúñatol, ib.22 λιτῶν πατρικίων (where there is question of an office being conferred on such), 73015 mayavós1 seems to be a less technical term, and to 1 The nearest equivalent of wayavós is 'ordinary'. Cp. Cer. 5482 Яμépav π. ordinary day (not a s ecial feast), 234, kvρiakην π. ordinary Sunday, 367 iло

1