When ancient historians wanted to take cognizance of what their predecessors had written, they seem usually, though not always, to have preferred an anonymous reference to one by name. To this practice the most noteworthy exception is Polybius. In his Histories he mentions about a score of previous writers, including philosophers, orators and historians, a large number of whom are the object of quite bitter self-righteous polemic. This harsh trait in Polybius' character appears slightly odd, when one considers his generally sanguine and optimistic view of the world he lived in and of the advantages which he himself enjoyed over his predecessors.

Of several scenes of the Orpheus myth which had been depicted by earlier artists, only one found its way into the repertory of Roman mosaicists; Orpheus sitting in the wilderness, holding the wild animals spellbound to the music of his lyre. H. Stern has recently compiled a catalogue of forty-seven of these mosaics, and to this number must now be added at least eight more, namely at Edessa, Tarsus, Trento, Trier, Thysdrus, Lepcis Magna, Ptolemais (the subject of this paper), and Tobruk (see below, p. 17), making a grand total of fifty-five, of which forty-five are in the Western Provinces; Roman Britain alone accounts for nine.

The tribunate of the plebs, according to a statement that Marcus Cicero puts into the mouth of his brother Quintus, was an office born in sedition and destined to create sedition (‘in seditione et ad seditionem nata,’ Leg. III, 19). There were two major periods of sedition. The first, the time of strife between patricians and plebeians, lasted from the birth of the tribunate in the early fifth century to the Lex Hortensia of 287 B.C. The second is usually dated from the tribunate of Tiberius Gracchus in 133 to the dictatorship of Caesar, a time when the tribunate was repeatedly an instrument of revolution, now accompanied by violence. The general view is that in the century-and-a-half between these two periods the tribunes, except in the time of Gaius Flaminius, were in accord with the Senate and were indeed agents of senatorial rule.

My object in this paper is to examine the implications of a contradiction in Augustan poetry which seems clearest in the Odes of Horace. He exhorts his emperor to curb the present extravagances of vice and restore marriage to its ancient dignity, and praises him for having done so. But this high moral tone ill fits his claims to be an erotic poet: content to love marriage as a bachelor, he addresses many odes to girls not burdened, nor likely to be, by the chains of wedlock. Indeed erotic poets have seldom found their inspiration in the stable affections of lawful marriage. In spite of the disparate freedom (even greater then than now) allowed to the male, it nevertheless seems reasonable to confront Horace first with his praises of matrimony, then with his more licentious outpourings, and require an explanation.

There are many periods of history—and unfortunately not always the least important—that survive, for us, entirely or mainly in one version. Ever since the development of modern critical historiography it has been recognized that it is the historian's duty to test that version, and highly skilled methods in the use of evidence have been worked out in order to enable him to do so. We all know some of the outstanding results in our field: the age of Augustus, once seen through the eyes of court literature; the ages of Tiberius and Claudius, once known only through the resentment of those who had suffered under tyranny; one could name many other periods that have gained a new reality in the last few generations and that now stand out in three dimensions.

During the fifty years that have elapsed since the great cycle of Dionysiac paintings was unearthed in the villa named after them at Pompeii, the search for their understanding has made reassuring progress. This search will never lead us out of ultimate ignorance about the deepest meaning which such visual imagery had within the context of an ancient mystery religion and we may resign ourselves to awareness of our own limitations in this respect. The best we can hope for is that we may learn to understand the surface meaning of images, whose full significance was only made clear to the initiated, to the degree to which, in antiquity, an uninitiated person could understand them. The last two published extensive discussions of this cycle of murals show clearly, however, that we are still far from having achieved even that degree of understanding. The impasse that has been reached can be overcome only by introducing new evidence and it is my hope that some progress can be made by doing so in the following pages.

The Roman revolution, which transformed an oligarchic Republic into the Principate of Augustus, had its origin, as Sallust (BJ, 41–2) saw, partly in the misery of the poor, in a social crisis, whose origins I cannot discuss here; it began with the Gracchi and with agrarian reform, and agrarian reform remained a leitmotiv in the turbulent century that followed. I need only mention the laws or bills of Lucius Philippus, Saturninus, Sextus Titius and the younger Drusus, the settlement of the Sullan veterans, the proposals of Plotius, Rullus and Flavius, the agitation of Catiline, and the land allotments of Caesar, the Triumvirs and of Augustus himself. Modern accounts tend to obscure or even deny the unity of this theme throughout the period. It is true that in the earlier phase reformers were more concerned to find remedies for social distress as such, and in the later to provide homes for veterans. But the Gracchan settlers and the veterans had two things in common: they were mostly countrymen, and they desired to obtain a secure livelihood by owning their own land. According to Appian (BC 1, 27), whose testimony we have no right to reject, the work done by the Gracchi was not lasting (cf. n. 1). Hence, the distress they had tried to alleviate persisted or revived; the governing class remained indifferent. Unorganized and unarmed, the followers of the Gracchi could save neither their leaders nor their own interests; men of the same class, with arms in their hands, were the essential instruments for bringing down the Republic.

Comparable to the legates who govern the consular provinces of Caesar, and standing at parity or sometimes above them, the high equestrian dignitaries, by their careers and promotions, serve to explain how a whole system worked. They also illuminate a process. Enormous benefit accrues to the study of social history. The men themselves can seldom be grasped as personalities, given the dearth of literary evidence, nor do they often impinge on great events.

The eight inscriptions transcribed and discussed below concern, in the first instance, the early history of Cyrenaica as a Roman province; but since most of them certainly and all of them perhaps involve Pompey and Cn. Cornelius Lentulus Marcellinus, one of his legates in 67 B.C., they have a much wider significance in illustrating Pompey's policy and position in that year.

1 and 2. Two rectangular marble bases, each inscribed on one face with identical texts. Found at Cyrene, one in 1860, within the Temple of Apollo (now in the British Museum), the other in 1927, in front of the same Temple (left in situ).

Caligula was dead when Seneca wrote De Brevitate Vitae. The phrase ‘modo modo intra paucos illos dies quibus C. Caesar periit …’ might suggest a date in the 40's. But, in a letter, lamenting the briefness of life, Seneca remarks on the distant events to which an old man's memory can apply modo: ‘Et tamen hoc “modo” aetatis nostrae bona portio est.’ Lipsius' theory collapses. His successors have supported other dates, principally two: 49, first proposed by Hirschfeld in 1870; 62, suggested by Dessau in 1918.

H. Peter briefly stated the reasonable opinion that Trajan did not draft the rescripts himself or manage the details of law and administration which they involve, that being the work of the secretary Ab epistulis. This widely accepted view was established on a firmer basis by the investigations of A. Henneman into the ‘outer and inner style’ of the rescripts. His general conclusion was that Trajan settled the more important problems himself from material prepared by his secretariat, by giving a verbal decision which the secretary worked up into the drafts which we have. These were written in a uniform chancery style, but personal interventions and comments of Trajan are recognizable within the formal texts. Henneman used two main arguments to establish the secretarial manner.

Mommsen in his Ostgotische Studien (Ges. Schr. VI, 362 ff.) enunciated the theory that Odoacer and Theoderic were kings of their German followers, but ruled their Roman subjects as commissaries of the emperors, holding the office of magister militum with certain precisely defined additional powers. Stein (Bas-Empire II, 116 ff.) and Ensslin (Theoderich der Grosse) have considerably modified this theory, admitting that Theoderic acted as king of all his subjects. They nevertheless still maintain that he was at the same time magister militum, and that his powers were limited in certain respects by a formal concordat with the emperor. In my opinion Odoacer and Theoderic were kings pure and simple, in the same position as the other barbarian kings.

Eusebius, quoting Hegesippus, records that during Trajan's reign Symeon son of Cleopas, a cousin of Christ, was martyred in Judaea as a Christian ἐπὶ … Άττικοῦ τοῦ ὑπατικοῦ —which at first sight appears to mean ‘when Atticus was consular legate’. Atticus was clearly the legate of Judaea at the time. The suggestion that he was legate of the consular province of Syria can be disregarded, as there is no reason why an event in Judaea should be dated by the governor of another province. Two problems arise—the date of Atticus' command, and his rank as legate.

Under the late Republic (after Sulla) a province was normally administered by a proconsul (an ex-consul or an ex-praetor) and, in subordination to him, a quaestor. If a quaestor died in the province or left it before the arrival of a replacement, the proconsul might appoint one of his legati or perhaps even one of his cohors praetoria to be acting-quaestor—‘pro quaestore’, ‘ἀντιταμίας’; and this was strictly the title also of a quaestor who remained at his duty in the province after the year of his magistracy had come to an end (either because his tenure had been extended by the government at Rome, or because he was awaiting the arrival of a successor).

The western shore of the Dead Sea, the most remarkable natural feature in Palestine (fig. 4), is a strip of flat or very gently sloping land, mostly barren, its parched surface seamed with a wrinkled pattern of erosion channels. The strip represents in part the old bed of the lake, once larger than now, and in part the spill of detritus from the rocky cliffs which tower above it. These precipitous heights form an almost unbroken wall some 300 ft. high extending from Qumran to Sdom, being patterned in high relief only by the few wadis which bite through them and carry the rare but torrential winter rains. The cliffs themselves are but the floor of the plateau through which the Jordan valley has opened its deep and famous rift, reaching a depth of over a thousand feet below Mediterranean sea-level; and from the plateau itself rises the mountainous, pitiless desert of Judah, stretching to Hebron 21 miles westward and draining eastwards towards the rift-valley. Where the eastward face of the cliffs is cut by water-courses it is carved into promontories, but in the main these constitute an integral part of wide sweeps of hinterland and are not positions of strength. At one point only (fig. 5) do two wadis flow close together and then divide, to tear out deep ravines behind a promontory, and almost to cut it off on the landward side. The spot in question is Masàda (pl. XVI), a lozenge-shaped table-mountain, 730 yds. long and 215 yds. wide overall, lofty, isolated and to all appearance impregnable.

The inscription which we published in this Journal, XLVIII, 1958, 117–129 (hereafter referred to as ‘F – N’) from a photograph, has now been collated by Fraser (June, 1961), and we offer here some improved readings, and consider their consequences (I). We also consider (II) some points raised by others.

Destruction by gravel-working of the successive fort and fortlet at Llystyn, Caernarvonshire (JRS xlviii, 130; L, 210–1) is nearly complete. (a) In the retentura of the fort four further barracks lay each side of the via decumana. The NW and NE gates had double carriageways between flanking towers, like the SE gate, and there were internal angle- and interval-turrets of timber, (b) The fortlet (fig. 10), whose gate lay on the SW, was asymmetrically planned, with a quadrangle SE of the unmetalled axial street and parallel blocks to NW of it.