I am sure I was not the only student who, when he first made his way stumblingly through the opening lines of the Persae, called to mind the raising of the curtain on the first act of The Mikado. But Aeschylus tells us a good deal more in his opening lines than the identity of the chorus of elders. They are the men left behind when the Persian host went to Greece, and they have been appointed by King Xerxes, son of Darius, to guard the wealth and splendour of the royal palace and oversee the land.

What I want to show in this paper is that what seems so natural in the Persae is achieved throughout tragedy and old comedy through a conscious adherence, as far as the identity of characters is concerned, to a convention. It is a convention which arises from the nature of Greek drama as a literary form primarily designed for performance.

In Euripides’ Bacchae, first produced in the last five years of the fifth century, that is, after the tragedian’s death, there is a remarkable sequence of scenes which constitutes the turning point of the play. In the first of these scenes (787-861), making the last section of the great central epeisodion, the god Dionysus, masquerading as a mortal, tries to persuade Pentheus that in order to spy on the maenads on Mt. Cithaeron, he must put on woman’s dress, or rather, allow himself to be dressed in female garb by Dionysus himself (827). The elements of the ϑῆλνς στολή are detailed in advance at 830 ff. — long flowing locks, skirts down to the feet, a μίτρα for his head, and as the special equipment of the bacchant, a thyrsus and a dappled fawnskin. Pentheus strongly rejects the stranger’s suggestion, though fascinated by it, and when he goes indoors at the end of the scene he has not yet committed himself to the plan. However, it is clear to the audience from Dionysus’ speech to the chorus after Pentheus has left the stage that Pentheus, in his last attempt to exercise control and freedom of action, will not be able to withstand the power of the god. In order to fulfil the god’s plan of vengeance Pentheus must be driven out of his right mind, since only then will he accept the female disguise. When the two main characters return after the third stasimon all this has happened. Dionysus calls Pentheus out of the palace to display himself as a bacchant, and Pentheus looks to Dionysus for approval of his costume. As adjustments are made to it each element of the disguise is spotlighted — the curls and the μίτρα (928-9), the girdle and the long dress (936-7) and the thyrsus (941-2). In keeping with his costume Pentheus must learn to play the part of the bacchant, and Dionysus instructs him in the correct wielding of the thyrsus (943-4) before the two set out for Mt. Cithaeron.

Eduard Fraenkel’s celebrated remark, that ‘for Greek tragedy there exists also something like a grammar of dramatic technique,’ has recently been confirmed, explained and illustrated by a series of distinguished studies emanating from both sides of the Atlantic. These demonstrate clearly how recurrent incidents were habitually presented by the fifth-century Attic tragedians with similar structural patterns. The transmission and execution of orders, for example, became so conventionalized that a set of rules for what is permissible and not permissible in the dramatic treatment can readily be formulated. By the time that a playwright like Euripides began competing in the dramatic festivals towards the middle of the fifth century, there already existed a sufficient body of material in plays produced for the accepted and acceptable dramatic conventions to be equally familiar to dramatist, actor and spectator. Conventions, however, like other organisms, may develop; they are not necessarily static. An experimentalist such as Euripides could play on his audience’s expectations and experience by altering just one of the elements in the conventional model, and by a minimal shift in the pattern shock the audience and amend the convention. In this paper I should like first to exemplify the Euripidean method briefly with two instances from the Ion, where the chorus creates or sustains the challenge to the conventions, and then to examine at greater length Euripides’ use and exploitation of the dramatic conventions tied to off-stage cries.

However much we may disagree about the meaning and function of the Iris-Lyssa scene in Euripides’ Heracles (815-73), we can be certain of one thing: it was meant to be startling. We can find, it is true, a handful of examples of the appearance of a deity in mid-action in tragedy, and no single feature of the scene to be discussed is without parallel, but viewed as a whole it is unique among the extant plays and the plays for which we have fragmentary evidence. For example, the appearance of Athena in Rhesus 595 ff. and the probable appearance of Artemis and Apollo during the course of Sophocles’ Niobe (cf. frs. 441a, 442, 445 Radt), must have been striking, but in each case the supernatural visitation advances the plot along not unexpected lines. In our scene the intervention of Iris and Lyssa drives the second part of the play in a completely surprising direction. Again, it is remarkable that two speaking characters appear and then not to agree but to quarrel heatedly about Lyssa’s assignment. The only even remote parallel is the discussion between Kratos and Hephaistos in Prometheus where the deities speak and act on ground-level and are found not at a crucial turning-point in the middle of the play, but at its beginning, when the disagreement of the speakers is a curtain-raiser to the more significant discord which pervades the drama.

In this passage which contains the second part of the contest between the two poets, their choral songs (μέλη) are under attack. Euripides, who has just now undergone a crushing criticism in the ληκύϑιον-episode (1201-47), is, however, as yet full of confidence (cf. 1249 f. and 1261 f.). When considering the choral song which introduces this part (1251-60), one must bear in mind that it presents a double-sided mentality: a) on the one hand because of the economy of the play Aristophanes has as yet to leave the contest undecided and so to keep the public in suspense; b) on the other it is obvious that in the choral song, just as in the whole play, the poet is partial to Aeschylus. This mentality appears in a typical way in the song, of which the last part (1257-60) has been bracketed by most critics, for they object that the chorus is over-abundant here and repeats in the second part the same point which was mooted in the first part: namely that an excellent poet (Aeschylus) will be subjected to criticism and this fact makes the chorus ponder.

In enunciating what may be called the ‘five-act rule’ and the ‘three-actor rule’ Horace was prescribing for budding dramatists rather than describing any existing body of drama. Nevertheless the Menander discoveries of the past twenty-five years make it probable that both of these ‘rules’ were in practice observed by the writers of Greek New Comedy, in contrast to the writers of Roman comedy, who observed neither. That the Romans should have abandoned what appears to have been the Greek practice in these two respects is an interesting fact of theatrical history; more importantly it provides us with new avenues for exploring the old question of how much the Romans adapted their Greek originals and for assessing the original contributions of Plautus and Terence.

Astronomical themes, with a rather limited scope, were a minor but recurrent element in ancient poetry, and Horace makes considerable use of them in the Odes. Most of the examples involve risings and settings as guides to the time of year, and most are within the range of the phenomena mentioned in Hesiod’s Works and Days. But in Odes 3.29.17-20, for the special interest of Maecenas, Horace breaks new ground by drawing more topical material from the Julian calendar, to remind his friend that it is now July and too hot for working in Rome: the (evening) rising of Cepheus on the 9th, the (morning) risings of Procyon on the 15th and Regulus on the 29th, and the entry of the sun into Leo on the 17th. Such an array of calendaric lore is clearly designed to appeal to Maecenas, who must have had a particular interest in astronomical phenomena.

Six epigrams of Martial are relevant to the question of his interpretation of passer in Catullus 2 and 3. Five of these, 1.7 and 109; 4.14; 7.14 and 11.6 are mentioned by Howell in his discussion of the problem in A Commentary on Book One of the Epigrams of Martial (London 1980), 122; to these must be added 14.77 which, although it does not specifically mention passer, certainly refers to it in the words qualem, dilecta Catullo, Lesbia plorabat. One of these epigrams, 11.6, is cited by Giangrande to support his argument that passer in Catullus is an equivalent for mentula on the grounds that Martial also used passer with the same meaning. On this point Howell comments: ‘if one accepts the interpretation of the passer poems recently persuasively reargued by G. Giangrande. . . then one must reinterpret not only this epigram [1.7], but the other four . . . in which M. refers to Catullus’ passer,’ and then proceeds to interpret Martial accordingly. A reading of the epigrams in question, however, reveals that such blanket reinterpretation is not only unnecessary, but also does grave injustice to the subtle ambiguity of Martial’s epigrammatic style.

During the first half of this century the view came to be widely held by numismatists, and by many Roman historians too, that the types chosen for the Roman imperial coinage were to be interpreted as a means of influencing public opinion, of reconciling his subjects to the rule of the Princeps, and of explaining imperial policy to them; in short, as propaganda. Thus for C.H.V. Sutherland coins are, in essence, ‘organs of information’, while in the words of M. Grant, ‘Roman coinage . . . served a propagandist purpose far greater than has any other national coinage before or since . . . This was the means which the Roman government . . . used to insinuate into every home in the empire each changing nuance of imperial achievement and policy’.