It is a well-known fact that of the three governors of Britain who were responsible for the Flavian advance of the northern frontier two had had previous military experience in the area in which they were now called upon to operate. Petilius Cerialis was in command of the IXth legion during the critical times of Boudicca's revolt (Tac. Ann. XIV. 32). And although in that capacity he succeeded in losing the larger part of his forces, his subsequent appointment by Vespasian to the command of the army despatched to reconquer the Rhine frontier from the revolted Civilis suggests that his military record was more distinguished than Tacitus would imply. Agricola has still better claims to be considered as a British specialist. In 61 as tribunus militum he was serving with distinction on the staff of Suetonius Paulinus in N. Wales (Tac. Agr. 5), and in 70 he was sent as legatus legionis to the XXth legion at Chester, in which capacity he served under Cerialis during the initial stages of the northward advance (Tac. Agr. 8). Of Julius Frontinus, the successor of Cerialis and predecessor of Agricola, we are less well informed; and it is the purpose of this note to show that in all probability he too had had previous military experience in Britain.

The language of Greek Tragedy can be considered as a whole by virtue of the characteristics which distinguish it from that of other branches of Greek literature, and the resemblance between the three tragedians in this respect is more noticeable than the differences. Still, if we compare Aeschylus and Euripides it is impossible not to feel a marked change of tone, in λ⋯ξις as in δι⋯νοια and ἤθη. As in E. the familiar legends are frequently set in a more everyday atmosphere and the characters cast in a less heroic mould, it is natural that the tone of the language should be lowered, partly by the frequent use of distinctively prosaic expressions and partly by the introduction of what appear to be colloquialisms. This change of tone in language was at once noted by Aristophanes and is referred to by Aristotle as an innovation of E. On the other hand E.'s style presents a certain anomaly, since while deliberately securing a closer approximation to the language of prose and ordinary conversation he also shows a poetic and archaizing tendency in the use of Aeschylean and Homeric words and forms not found in Sophocles. This may be due to a reluctance to depart too far from the poetic tradition of Greek tragedy, and possibly to a scholar's interest in the language of Aeschylus and Homer, in a minor degree a foreshadowing of the learned archaism of the Alexandrian poets.

These suggestions for the betterment and elucidation of the text of the De Beneficiis are additional to those already published in the Classical Quarterly in January, 1934. They are based on a conviction much deepened since that time that Buck1 is right when he says: N (i.e. codex Nazarianus) allein, und zwar ohne seine Ueberarbeitungen von späteren Händen, darf die Grundlage des Textes von de beneficiis bilden. Préchac3, the latest critical editor in this field, substantially confirms Buck's sweeping conclusion by an independent survey of the evidence. The readings designated in the Teubner text3 as N2 and N3 are themselves conjectures, not readings drawn from independent sources of testimony, and as conjectures they must be judged. Gertz4 had long before discerned the truth even if he seemed somewhat cavalier in his attitude toward other manuscripts than N.

Text, interpretation and metre present a tangled problem in this threnody, and the solutions of editors differ widely. The chief function of detailed metrical study in such corrupt passages of lyric is to weight the scales in favour of—or more often against—certain methods of handling the text. The positive results of this present attempt to apply metrical criteria are necessarily modest and tentative; negatively they are, I think, sometimes decisive.

In the fifth book of Livy there is a passage of which hitherto no good explanation has been forthcoming:

Romae interim multiplex seditio erat, cuius leniendae causa coloniam in Volscos, quo tria milia civium Romanorum scriberentur, deducendam censuerant; triumvirique ad id creati terna iugera et septunces viritim diviserant (Livy v, 24. 4: 395 B.C.). (Livy then goes on to describe the viritane division of the ager Veientanus.)

The method of approach to detailed criticism of prose-writers and poets adopted by Dionysius is in a large measure comparative. The procedure of comparison is threefold: firstly, the bringing together of passages from authors to elicit points of resemblance or of difference between their styles secondly, the assumption of the existence of common critical standards for all works of art, whether literature, painting, or sculpture thirdly, the use of metaphor and simile to illustrate matters of criticism which need the assistance of some visual or mental image to make clear the exact meaning of the point of criticism. The metaphorical vocabulary is the richer for the greater preoccupation of Dionysius with aesthetic appreciation than with philosophical analysis; and any seeming differences in technical vocabulary are compensated by the freedom with which his vocabulary in general draws on the Greek of his own and earlier times. The following list gives some of the more interesting metaphorical expressions used in the rhetorical works, with notes on their meaning and, in some cases, with illustrations of the use of the same or similar terms in other Greek critics and in the volumes of scholia.

In attempting to date the Shield several complementary methods are possible. Roughly these may be classed as literary, historical and archaeological. The literary method indicates that the Shield comes late in the Hesiodic corpus: in particular the use of the F is careful. The historical method suggests a preciser upper limit. Wilamowitz believed that the point of lines 393–401, which give the season in which the combat between Herakles and Kyknos took place, can only be that a commemorative festival was held—at the temple of Apollo at Pagasae, since the scene is laid in Apollo's sanctuary. Mazon continues that the reputation of the temple dates from the end of the Sacred War and the establishment at Pherae of a dynasty with close connections with Delphi (which Kyknos was in the habit of defrauding): he therefore sets the terminus post quern at about 590. This argument is attractive and possible. Several scholars have attempted to secure a lower limit from the last sentence in Hypothesis A: κα⋯ Στησ⋯χορος δ⋯ φησιν ' Hσ7iota;⋯δου εἶναι τ⋯ πο⋯ημα. But as Wilamowitz observes, we do not know whether the poem in which this mention came (plausibly a Kyknos) was a genuine work of Stesichoros. Certainly it need not make the Shield much earlier than the Stesichorean poem, since our ‘Hesiod’ wrote his epyllion for incorporation in the Ehoiai which was attributed to the genuine Hesiod, and an insertion into such a work could have been very quickly accepted. Mazon's lower limit of 560 is then on insecure grounds. Thirdly there is the possibility that analysis of the decoration of Herakles' shield might give useful comparisons to the conclusions of archaeologists on the chronology of Greek art. It is this method that I shall attempt.