The peopling of floral scrolls with living creatures is a decorative device which enjoyed unrivalled popularity throughout the whole history of Imperial art and in almost every country of the Empire. Its full cultivation and flowering were achieved in the Roman age; but its roots, like those of nearly every Roman art-motif, are in the late classical Greek and Hellenistic worlds. These roots were varied and complex. The primitive notion of spirits indwelling in trees and plants, and at a later stage personified in visible shape, may have played a part; some of the constituent elements can certainly be traced back to religious symbolism; and more immediate was the influence of the naturalistic trend of fourth-century art, which favoured the idea of rendering birds, insects, and small beasts in their native setting. In Hellenistic and Imperial times, as these elements mingled and the motif became more widespread, fancy came gradually to outweigh fact; and a delight in incongruity for its own sake found ready expression in the peopling of vine- and acanthus-rinceaux with mythological and genre scenes and figures, framed in the foliage or poised on slender stems, or with human figures and such solid quadrupeds as dogs, bulls, horses, bears, panthers, and lions, careering through the leafy whorls or springing from the hearts of flowers.

Lydus quotes a Capito three times, a Fonteius four times, Capito and Fonteius together twice, and adds that they were contemporaries of Varro and Sallust. He calls Capito a ἱερεύς, makes him an interpreter of the Etruscan Tages, and attributes a few oracles to him as well as some calendary matter. He calls Fonteius a ῾Ρωμαῖος, refers to his oracles, his interpretation of the Etruscan discipline and his work περὶ ἀγαλμάτων, and gives his ‘Tonitruale’ in Greek translation. Modern scholars identify Capito with the antiquarian Sinnius or the jurist Ateius Capito, and they assign Fonteius to the imperial period but leave him unidentified. There is nothing in the texts to support these suggestions. My conjecture is that we have to deal with one man, C. Fonteius Capito, who was ‘ad unguem factus homo, Antoni, non ut magis alter, amicus’ (Hor. Sat. 1, 5, 32 f.). Born c. 80, he became pontifex apparently after 44, when, with the help of Antony, Lepidus was elected pontifex maximus. He soon received another office from Antony—according to Groag that of a tribune—and went with him to the East. These facts emerge from an (unpublished) inscription from Cos, of c. 39, which calls him a ἱερεύς and records that, on his request, the rule of Antony was accepted on Cos.

The importance of the Roman army in the social history of the Empire needs no stressing. The financial and, therefore, the social position of the soldier cannot be assessed without an estimate of the pay that he received during service and of the lumpsums or grants of land that he secured on retirement. The conclusions of A. von Domaszewski on this question have hitherto been generally accepted; but, as they have recently been sharply challenged and certainly rest too frequently on ingenious combinations which do not give the certainty claimed for them, it may be convenient to assemble and analyse the evidence once more, even if the conclusions suggested prove to agree in the main with those of Domaszewski. I shall seek to distinguish certainty from probability, and both from mere conjecture.

The two Latin inscriptions described below were both re-composed during the years 1946–7 in the course of the work of systematisation carried out in the ruins of Lepcis Magna. In the case of the earlier inscription, of Augustus, a small number of inscribed blocks belonging to it had been uncovered during the Italian excavations of 1930–4; but the greater part of the inscription was covered by soil and only brought to light when the fallen west perimeter wall of the Market was re-erected in 1947. It has the distinction of being the earliest dated Latin inscription hitherto found at Lepcis.

The second inscription, of Vespasian and Titus, has no such structural context. Most of its blocks were built into the upper courses of the towers of the main surviving Byzantine Gate of the city; but others were found scattered at street-level in the vicinity of this gate, and had obviously fallen from it. The blocks are inscribed on both faces with parts of two identical texts, and, as explained below, the nature of the bi-frontal inscription suggests that it formerly belonged to a monumental arch, spanning one of the main streets of the city: but the position of this arch cannot be established with any certainty. Certain fragments of this inscription have been published previously; but a relatively complete text is presented here for the first time.

The ancient coast road of Libya which runs close to the shore from Berenice (Benghasi) to Teuchira (Tocra) and Ptolemais (Tolmeita) later swings inland to reach Cyrene, and climbs to the second plateau of the Gebel Akhdar (some 500 metres above sea-level), which it reaches in the vicinity of Messa. From Messa eastwards the road can be traced without difficulty, in the form of a shallow cutting in the rock (probably of Greek origin), to Zavia Beda where the remains of the Roman town of Balagrae are still visible. On the Peutinger Map Balagrae is marked as 12 Roman miles from Cyrene, and this figure is accurate if one follows the course of the old road from Beda to the south gate of Cyrene. For its last three kilometres before reaching Cyrene, the ancient road is very clearly visible as a sunk track cut in the rock, with numerous wheel-ruts, and with ancient tombs flanking it. It is far from straight and makes several sharp bends, but there is no indication that Roman engineers attempted to eliminate these eccentricities when the road became one of the highways of the Empire.

Roman milestones previously found in the area of Cyrene have all come from the Cyrene—Apollonia road, and the Trajanic column which marked the first mile on that route was found in 1915 by Italian military engineers in the course of bringing back into use the long-abandoned Roman road, which descends the upper escarpment of the Gebel. The column is now in the Museum at Cyrene, but its exact original site can be identified from photographs taken at the time of its discovery: it stood a little east of the large circular tomb immediately below the gallery which now houses the greater part of the ancient sculpture from the Italian excavations.

East of Sabratha the coast of the Mediterranean Sea is formed by a line of soft sandstone cliffs. These, which are scarcely more than twenty feet high, are divided by low headlands into shallow bays. Below the cliffs, the foreshore is sometimes of harder rocks, flat or weathered into sharp spines and pinnacles; sometimes of fine sand, equally ideal for bathing or for beaching boats. Inland run low sand-dunes, barren now except for the desert shrubs on which Arab boys graze their goats. But this coast was fertile once, for here and there one may find scatters of pottery, Hellenistic, Arretine, and Roman coarse wares; or lines of square-cut well-laid masonry, which mark the sites of Roman baths, mausolea and villas. One such group of buildings forms the subject of this paper.

The site (Fig. 1) was discovered by Mr. J. B. Ward Perkins, the Director of the Sabratha Expedition (1948), who noticed foundations cut into the rocks of the foreshore, and short stretches of wall protruding from the sand dunes above the cliffs: clear evidence that a villa had once stood there. It was not possible to spare either Arab labour or supervising staff for a full-scale excavation of the site, but it seemed worth-while to make some record of the buildings. Mr. N. de l'E. Thomas and I decided, therefore, to devote what spare time we had to investigating the villa and, when its importance had been recognized, we were aided by four Arab workers.

Among the countless attempts by nineteenth-century architects to supply a reconstruction of some ancient or medieval building, none was more favourably received nor more widely reproduced than the general view of Old St. Peter's and the adjoining Vatican Palace ‘about the year MCCCCL’, drawn in 1891 by Henry William Brewer. Brewer is better known for his tentative reconstruction of the city of Paris in the first third of the sixteenth century; but like the rest of his imaginative reconstructions, his drawing of Old St. Peter's also is a notable performance, bearing witness to the preparation of many years; and it has since been accepted in all hand-books on early Christian art as an attractive, and seemingly very instructive, illustration. Nevertheless it will not bear close, critical examination. In the Jubilee year of 1450 neither the Sixtine Chapel nor the Belvedere were yet standing, nor had the angle projection of the Palatium Archipresbyteri yet received its two-storied archway.

Brewer was above all concerned to illustrate the grace of a grandiose, light-strewn quadriporticus. Its northern arcade, with its continuous sloping roof, rests against a long and extremely narrow south wing of the Vatican Palace. This fact alone is a clear indication that his drawing was based on nothing more precise or more reliable than the well-known plan of Tiberius Alpharanus, published in 1589–90, or on plans derived from it.