The large harvest of editions and monographs devoted to Seneca's Phaedra in recent years have done full justice to the influence on Seneca's play of the Greek tragic precedents, even exploiting the Latin play as a quarry for their recovery and restoration; at the level of individual motifs of dialogue, Seneca's debt to Ovid's fourth letter of the Heroides has been confirmed and detailed. But concern with the adaptation of the myth, and the obvious verbal resemblances to Ovid, have combined to distract students from the influence of another Latin work—the fourth book of Virgil's Aeneid. The theme of Virgil's book, the quality of Virgil's portrayal of Dido and her passion, with its truly dramatic greatness (of which a distinguished editor has written ‘if Virgil had written nothing else … it would have established his right to stand beside the greatest of the Greek tragedians’), the acknowledged supremacy of Virgil's reputation as a poet in Seneca's generation, and Seneca's own fondness for quoting the Aeneid, are all strong arguments for expecting some reminiscence of Virgil's great queen in Seneca's delineation of Phaedra and her doomed passion for Hippolytus. Before Virgil only Catullus' Ariadne had approached the insight and sympathetic analysis which was achieved in Dido. After Virgil the lovesick heroines of Ovid's Heroides are rhetorically versatile but without the moral stature to give value to their sufferings, while their static portraits cannot offer the development of either action or emotion which is essential to drama. Unfortunately, because Ovid both in Heroides and Metamorphoses borrowed so much detail and imagery from Virgil, his dependence on the greater poet complicates and often frustrates attempts to distinguish the relationship of Senecan poetry to that of the two predecessors whom he admired.

Until the last years of the Republic, Roman governors might officially see their wives present in their provinces only in the form of honorific statues. The chaos of civil war was to bring to a number of senatorial husbands, as they mounted their campaigns in the eastern provinces, the solace of their wives' companionship in the flesh. First in time amongst these refugee wives was Caecilia Metella, who was present with Sulla at Athens in 86 B.c. Later, and in less auspicious circumstances, Cornelia was to travel to Lesbos and then to Egypt to be with the doomed Pompeius in 49–48 B.c. Occasionally, such loyal wives joined their husbands over die latter's objections. They might, it seems, be expected to remain in Italy in order to avoid danger or even proscription, and not least to defend the family interests so far as this was possible.

‘Translations (like wives) are seldom faithful if they are in the least attractive.’ In an age of translations (I say nothing of unfaithful wives), Roy Campbell's mot is a salutary reminder that attractiveness in a version may conceal gross infidelity to the content and meaning of the original; and that even if by some miracle a translator should succeed in reconciling the demands of readability and accuracy, his work is necessarily ‘at best an echo’ of the ancient text, or—to use Platonic language—a copy reproducing its model only imperfectly.

Optimates and Populares, Roundheads and Cavaliers, Whigs and Tories, Cowboys and Indians: the neat antitheses polarize views and enlist sympathies. Nor are they all misleading. What of ‘Julians and Claudians’? These labels were invented to distinguish ‘factions’ or ‘parties’ within the family of Augustus; the issue between them was the succession and the struggle continued until the end of the principate of Augustus' heir Tiberius. Thus comprehensible order is brought to the confused mass of information and opinion that the sources offer us. Moreover, the labels have a respectable pedigree in modern writing. They are constantly employed by F. B. Marsh in his authoritative work and survive as valid and useful in articles more recently published; where historians have trodden so boldly literary critics have not hesitated to follow. My purpose, at the cost of destroying neat and treasured theory, is to show that these terms are inapplicable to the two groups to which they have been applied, and therefore misleading. They are an impediment to serious inquiry into. complex and changing situations.

It is only in comparatively recent years that critics have come to realize the extent of humour that is to be found in Propertius. The view of Postgate (1881), ‘Occasionally he shows a vein of humour we should not have expected’, is still that of Butler and Barber in 1933: ‘It [iv. 8] is his one humorous poem, and surprisingly successful.’ But by 1966 Eckard Lefèvre could write a whole book on Propertius Ludibundus, who was in this respect, as in so much else, the inspirer of Ovid: ‘In accordance with the principles set out the present study is an attempt at a contribution to the understanding of Roman love-elegy, in that it seeks to establish the achievement of Propertius as the decisive step in the development of the humorous element in elegy before Ovid set that particular stamp on it’ (p. 20).

Both Herodotus and Aristotle record the first known contact between Athens and the Macedonian area. It occurs in the time of Peisistratus, specifically during his second (?) exile. Herodotus refers briefly to revenues which Peisistratus collected from the Strymon region, which helped him recover his power (i. 64. 1). Aristotle in more detail states: ‘Peisistratus joined with others in the settlement of Rhaecelus in the Thermaic Gulf area. From there he went to the Pangaion region from which he raised money …’ (A.P. 15. 2).

In spite of much sensible writing, by many scholars, on the real importance and significance of the literary convention, there remains among some students of the Classics a persistent folk-memory which asserts that speeches in the ancient historians are rhetorically decorative, deliberately deceptive, and historically inessential. That a speech can be one (or all) of these things is inarguable: but to assume that all speech in historical writing can be so dismissed is dangerous, because such assumption inhibits a full understanding of what a serious ancient historian may be trying to do with it. I should therefore like to look again at the convention, and at some of the ways in which it is used by the major Roman historians.

The prime dramatic character of Greek tragedy is agonistic. Its myths for the most part show men struggling toward some goal, in conflict with one another, or against some force of circumstance or destiny, which is often personified in a god. Tragedy is at the same time a dramatic form restricted severely by theatrical conditions. The number of its speaking actors is held to three. It avoids the staging of any physical action like an assault or battle, let alone catastrophe, or death—for whatever reasons of narrative tradition, aesthetic convention, or simple impracticability. Despite certain ritual or symbolic aids to representation such as music, dance, and gesture, it is in consequence a drama of extreme, sometimes exclusive, verbal concentration. Exposition, development, climax, resolution, action and reaction—all movement occurs in the narrow room of at most three stage persons at any one time debating to confirm or change their attitudes or intentions—often, with a single character so placed debating within himself, in monologue or soliloquy, or in relief or opposition to another voice. My subject here is this last mise-en-scène: the deliberate working up of the ordinary exchange between characters into the opposition of one character to one or two others in a formal debate.