In the worship of the Roman Emperors there were at the beginning at least two elements that stood in sharp contrast to each other. The Greeks had been accustomed for some three centuries to the rule of hereditary Basileis; a basileus during his lifetime was often entitled θεός and honoured as such, and this title and honour sometimes continued after death. In Rome the attitude towards an Emperor was different: strictly speaking there was at first no such thing as an ‘Emperor,’ but merely a citizen entrusted with great powers; writers of the imperial age are never weary of pouring scorn upon ‘mores regii,’ on the evils of a court and of despotic rule. Neither by tradition nor by education were Romans favourably disposed towards worship of a living man as a god. True, they were accustomed to the notion that every man had a genius, and the genius was entitled to sacrifice on that man's birthday; even so, the genius was hardly 'a god,’ save in so far as men were liable to think of it as if it were that fundamentally different thing δαίμων.

The origin of the imperial civil service is generally attributed to Augustus, and its development to Claudius; but a precise understanding of what were the original elements due to Augustus and what the additions of Claudius is to seek, despite much sound piecemeal work on the activity of the latter. The problem has not been tackled as a whole. To speak of the ‘civil service’ of Augustus is perhaps to use a metaphor of dubious value. The term ‘civil service’ implies a regular and highly organised body of bureaucratic officials forming an administrative and executive personnel which, under the general guidance of the Government, or, in classical terminology, of the Magistrates, manages, often in minute detail, the affairs of the country concerned. Such systems were not unknown in antiquity, notably in Ptolemaic Egypt, but such was not the system of the Principate in general, even at the fullest development of the procuratorial service. Still less was such a system originated by Augustus, despite the retention in Egypt of Ptolemaic methods.

The relief shown in Plate IA was among those transferred to the Terme Museum from the Villa Ludovisi, where, about 1460, Cassiano Dal Pozzo saw it immured in the wall of a small building to the right of the entrance. Its original provenience is unknown. In his catalogue of the Villa Ludovisi sculpture Schreiber describes it as representing ‘a knight with two at tendants. … On the left a youth walks forward to the 1. He wears a sleeved tunic girt at the waist, and shoes; and carries on his right shoulder a short stave which has been broken off where the relief border is damaged. He is represented almost en face, and turns his head backwards towards the knight. With his left hand he leads by the bridle a richly harnessed horse saddled with a panther-skin. On the horse rides a young (beardless ?) man in short-sleeved tunic and cloak, the latter falling over his left forearm. He is laureate, and holds the horse's rein in his left hand, while the right is raised to the level of his head. … There follows a bearded man wearing tunic and shoes. In his left hand he holds the end of an object slung over his left shoulder (probably a sack …), and in his right hand he lifts up a hemispherical helmet (the left cheek-piece broken off), in the act of placing it on the knight's head.’

The silver plate of the late Roman and early Byzantine period is characterised by formal engraved ornament of an unusual and easily recognisable kind. The motives used in this ornament consisted at first of the quatre foil diaper, several forms of rosette, strips of tongue pattern, heart-shaped leaves, and groups of leaves variously arranged. In later examples the characteristic motive is a pattern of palmettes, each unit joined to the next by the smallest side leaf, which surrounds a central panel or in one case (no. 6 infra) ornaments the body of a round vessel. This last stage has been dealt with by Matzulewitch; but as the examples of this type of plate have never been studied as a whole, they are collected here.

The Preface to the Penitential of Gildas has hitherto been found in only one MS., preserved in the Bibliothèque Nationale in Paris ( = Lat. 3182). This MS. is a large parchment codex of the tenth—eleventh centuries, containing a collection of canons, decretals, penitential fragments, etc. taken from many sources, e.g. the Collectio Canonum Hibernensis, the Lex Salica, etc. After an extract on pp. 279–80 headed ‘Item synodus sapientia sic de decimis disputant’, the Gildasian preface follows on p. 280 with ‘Incipit prefacio Gildae de poenitentia’, concluding on p. 281 with ‘hue usque Gildas incipitnunc sinodus aquilonalis Britanniae’. Most editions of this penitential are reprints of F. Wasserschleben's text. Based on a collation by Dr. Knust, it is defective from the standpoint of readings and palaeographical information.

The following collation has utilised, apparently for the first time, a newly-discovered MS. of the penitential, now in the Bibliothèque Municipale at Cambrai ( = No. 625, ff. 52–53). If the catalogue date, viz. saec. ix, is correct, then this MS. is earlier than Paris Lat. 3182.

On pages 75–80 of the Harvard Studies and Notes in Philology and Literature, Vol. XX (1938), is an article (with facsimiles) by Mr. J. Milton French entitled ‘Milton's Annotated Copy of Gildas’. It deals with a copy, in the Widener Library (shelf number Br. 98.319 F), of H. Commelinus' Rerum Britannicarum … Scriptores Vetustiores ac Praecipui, Heidelberg (1587), given to the library about 1765 by Thomas Hollis. This book contains on pages 113–146 the ‘De Excidio et Conquestu Britanniae Epistola' of Gildas. A former owner has inserted thirteen brief marginal notes on pages 114–123, i.e. explanatory comments on various matters in Ch. 1–33 of the ‘De Excidio’ (numbered according to Mommsen's edition; Mon. Germ. Hist., Auct. Ant., XIII, Chron. Min. iii, Berlin 1894–8). The remainder of the book, containing Geoffrey's ‘Historiae Regum Britanniae’ (pp. 1–92), the ‘Historia Britannica’ in six books of Ponticus Virunnius (pp. 93–112), Bede's ‘Ecclesiastical History’ (pp. 147–280), a ‘De Gestis Anglorum Libri Tres’ (pp. 281–348), William of Newburgh's ‘Rerum Anglicarum Libri Quinque’ (pp. 353–496) and Froissart’s ‘Historiarum Epitome’ (pp. 497–568), has no marginalia.