Shortly after his arrival at Carthage, while he is waiting for Dido to meet him, Aeneas finds that the walls of her temple are adorned with pictures of the Trojan War. Sunt hie etiam sua praemia laudi, he cries to Achates, sunt lacrimae rerum et mentem mortalia tangunt. The description of the pictures which follows is a remarkable example of Virgil's ability to use a traditional device in such a way as to strengthen and illuminate the main themes of his poem. It is my object here first to reinterpret one of the scenes which has been misunderstood, and then to discuss how Virgil has chosen and arranged his episodes so that the description of a picture gallery becomes a part of an epic poem.

Ancient and modern scholars are so unanimous in their condemnation of Varro as a writer, that a study of his ‘style’ may seem to be valueless. Cicero paid ready tribute to his great contemporary's learning, but studiously forbore to say anything about his writing, a fact which was observed by Augustine, who admitted Varro's inferiority in this respect. Quintilian, in a guarded way, makes the same criticism; for him Varro is ‘plus scientiae collaturus quam eloquentiae’. In recent times Norden has castigated the De, Lingua Latina as exhibiting the worst Latin style of any prose work, and his opinion of the Res Rusticae is not much higher.

Editors and commentators ancient and modern have not responded very well to the challenge of the two riddles which round off the contest between Damoetas and Menalcas at the end of Eclogue 3 (104–7). The first is generally regarded as impossibly difficult; the second as impossibly easy. Critics take refuge in quoting Servius' despairing statement: sciendum aenigmata haec sicuti pleraque carere aperta solutione. Yet it is most unlikely that Virgil would introduce insoluble or meaningless riddles into the Bucolica. If there is no solution the lines become pointless, even tasteless; and these are not epithets to be applied lightly to Virgil's poetry.

In a recent article, Dr. A. F. Norman has attributed to Eunapius the authorship of a fragment in Suidas (Adler A 2094 s.v. ), which clearly relates to the siege of Maiozamalcha. His arguments are cogent and must, I think, be accepted. Some slight additional support for the attribution is provided by the fact that it contains the adverb of which, as Vollebregt pointed out, Eunapius was particularly fond. Norman compares this fragment with the relevant passages in Ammianus Marcellinus (24. 4. 23) and Zosimus (3. 22. 4) and points out that if the attribution is correct, ‘now, and for the first time, we have a reference to the same incident of this campaign from the narratives of Ammianus, Eunapius, and Zosimus’.

The lettering of this inscription begins at the very top of the block, just below the straight edge, and stops half-way down the block, the lower part being smoothed but uninscribed. As the inscription is not set centrally on the block, it is probably the continuation of an inscription which ran on a block once superimposed upon it. Doubtful letters are those which are marked by the dot underneath; and W. Peek reported in Ath. Mitt. lxvi (19,41), 200 n. 2, that he had seen the top of the v which Vollgraff suggested in the restorationThe inscription is dated to the seventh century B.C. by epigraphists. The stone itself was seen in 1729 by Fourmont and rediscovered in 1928 by Vollgraff; it was in the wall of a Venetian tower on the Larissa or acropolis of Argos. No one knows whence the Venetians had taken it. Therefore Vollgraff's suggestion that it had originally been set up in porta arcis regiae is no more than an attractive speculation.

The methods of assessing a writer's spirit vary in usefulness according to his genre. If he is a satirist much may often be learned through an examination of his names. This is certainly true of Horace, and one might have thought that in recent years a fair amount of attention would have been paid to this aspect of his work. Yet to the best of my knowledge no special study has been published in the present century. Certain points have been well noted by scholars like Vogel, Becher, and Marouzeau, and a few editions contain summaries of the material. The last detailed discussion, however, was that of Cartault, and one must admit that it was not wholly unbiased. So it seems reasonable to review the evidence again, making use of the work done by Marx, Cichorius, Münzer, and others. We do not have to inquire about all the characters in the Sermones; only satirical references need be considered, and even here there is room for selection, because some of the figures are so obscure that nothing helpful can be said about them.

The problems of this passage were concisely stated by M. Platnauer more than thirty years ago and his suggestions for their solution have been adopted and developed in A. S. F. Gow's magnum opus. Its authority—so the present writer suspects—is liable at this point to eclipse the meaning of the text

The Vatican palimpsest (A) omits the word enthymemate and leaves a lacuna, thus showing that the scribe found it written as Greek in his exemplar. Now A has been shown to belong to the fourth century, and therefore its authority must be greater than that of the other manuscripts (twelfth and thirteenth centuries) available for this part of Gellius. The same problem occurs in 12.2.14, non pro enthymemate aliquo, where Hertz (followed by Hosius) goes against the Greek of the manuscripts and adopts the reading of Carrio. Yet an examination of the use of the Greek dative in Cicero's Letters will show that it can be used fully as the equivalent of a Latin ablative, can be governed by prepositions taking the ablative in Latin, and can be qualified by a Latin adjective in the ablative.

People who trust modern indexes will suppose that the name of Sex. Clodius, the disreputable henchman of Publius, comes twice in the Ad Atticum letters, 14. 13. 6 and 14. 13 A. 2. The manuscripts give it as follows:

Sight, and its object light, appear to be universal metaphors in human language, both for intellectual apprehension or activity and its objects and also for the experience of aesthetic and moral values. The figure is applied equally to the course or end of a rational approach to knowledge, giving scarcely-felt imagery like ‘I see’, ‘look into’, etc., or to a pictorially described ‘illumination’ or ‘vision’ that lies beyond the range of reason. Some phrases are applicable in both senses; to ‘see the light’ may connote either logical grasp of a fact or religious conversion.

‘Auctoritas’ was naturally one of Cicero's favourite concepts. In the ideal republic power lay with the people, auctoritas with the Senate (‘Cum potestas in populo, auctoritas in senatu sit’, De leg. 3. 28). Alternatively, in a balanced state, potestas would lie with the magistrates, libertas with the people, but still auctoritas would be the property of the Senate, ‘in principum consitio’ [De rep. 2. 57).

Groundless assumption in scholarship is generally soon swept away. Seldom does an interpretation which has little to commend it survive as long as that which I here propose to refute.

Professor Kamerbeek in a recent article raises some interesting questions about the value of the manuscript of Sophocles Parisinus gr. 2712, universally known as A. Ever since the time of Brunck, who was the first to use this manuscript extensively, it has been considered second only to Laurentianus XXXII 9 (L) as a source of correct readings and as a witness of the old, i.e. pre-Byzantine, tradition of Sophocles. But the importance of A has been seriously challenged by Alexander Turyn, who has sought to show that in Ajax, Electra, and Oedipus Tyrannus—the three plays which constitute the Byzantine Triad of Sophocles—A's text is principally that of the recension of Manuel Moschopulus (fl. c. A.D. 1290)—or, more strictly, the version of that recension which includes scholia by Maximus Planudes and some slight variations in the poetic text.

The writer recently examined the two palimpsest folios of the manuscript Parisinus gr. 107B, otherwise known as the Codex Claromontanus, which contain most of the surviving passages of Euripides' Phaethon. Despite the damaging effects of the chemical reagents used in the nineteenth century the text is not wholly illegible and a collation was made where possible. A comparison of this with the standard texts published by Nauck, von Arnim, and Volmer revealed some puzzling discrepancies; it looks as if recent editors have not examined the manuscript, basing their work on their predecessors' false reports of the collation by Blass. Nauck writes (p. 599): ‘quoniam vero programmata academica in paucorum manus perveniunt, haud gravabor quae Blass legit hoc loco repetere’. von Arnim shows in his preface (p. 2) that he copied from Nauck, and Volmer says (p. 9): ‘J. de Arnim omnia tragoediae Euripideae fragmenta contulit in Supplemento Euripideo, quem librum prae aliis commentationibus in fabula restituenda adhibeo’. In future scholars who turn their attention to the Phaethon would do well to go back to the collation by Blass, but this too seemed faulty to me at certain points.

Believing that there is still something to be said about the early history of clausulae in Latin prose, I set myself to trace the practice of the early orators, then that of the Rhetorica ad Herennium, accepting its conventional dating to 86–82 B.C., and lastly that of Cicero in De Inventione, assuming it to be roughly contemporary with the ad Herennium, and in his early speeches. But clausula-study itself, besides shedding light on the methods of composition used by the still unidentifiable Auctor ad Herennium, cast doubts on the conventional dating of his work. I gave a brief outline of my conclusions in C.R., N.S. vi (1956), 134–6.1 propose to attempt here a detailed justification of those relating to the date of the work, with a few modifications of some points.

Wishing to forestall an inquiry from Quinctius about the produce of his farm, Horace says that he will describe its forma et situs (1–4). What follows is not an impersonal description, but an account directed at Quinctius who is thought of as passing judgement on the farm (cf. laudes, 8; dicas, 11; iam si credis, 15). This involvement of Quinctius in the description must be extended to the protasis of the opening sentence of the description: continui montes si dissocientur opaca/valle…, temperiem laudes where the sense is something like ‘if you were to find yourself in a place where the mountains, which crowd close to one another, are parted…, you would praise its temperate climate.

The total surviving output of the Hesiodic school of Greek Epic is too little, in point of sheer quantity, to afford an insight into social life and outlook comparable with that offered by the Homeric poems. Worse still, it is too heterogeneous a collection to form a meaningful social document.

Are we to restore ‘sit on eggs’ after Hesychius or ‘cry ’, ‘lament’ after Nauck? In his recent supplement to the Loeb Aeschylus, Mr. Lloyd-Jones says that the latter ‘is far better suited to the context’, by which I am given to understand that he means that Aeschylus would have been very unlikely to employ the brooding metaphor in this passage. Admonished by Wilamowitz (Interpretationen, p. 57 n.) that ‘es unverzeihlich ist, das Bild der briitenden Henne zu vertreiben’, I fall back in some bewilderment on the fact that I personally like it very much and that the tomb-nest symbolism, though not absolutely presupposed by Cho. 247 ff. and 501(), would certainly have encouraged the poet to write these later passages. But there is a dearth of objective argument on this matter which I want, if possible, to supply.

The following suggestions for the interpretation of Parmenides and Melissus can be grouped for convenience about one problem. This is the problem whether, as Aristotle thought and as most commentators still assume, Parmenides wrote his poem in the broad tradition of Ionian and Italian cosmology. The details of Aristotle's interpretation have been challenged over and again, but those who agree with his general assumptions take comfort from some or all of the following major arguments. First, the cosmogony which formed the last part of Parmenides' poem is expressly claimed by the goddess who expounds it to have some measure of truth or reliability in its own right, and indeed the very greatest measure possible for such an attempt. Second, the earlier arguments of the goddess prepare the ground for such a cosmogony in two ways. For in the first place these arguments themselves start from assumptions derived from earlier cosmologists, and are concerned merely to work out the implications of this traditional material. And, in the second place, they end by establishing the existence of a spherical universe: the framework of the physical world can be secured by logic even if the subsequent introduction of sensible qualities or ‘powers’ into this world marks some decline in logical rigour.

Themistokles was ostracized in the late 470's, probably in spring 471 or 470; if we are to believe Thucydides, he did not write to Artaxerxes in Persia until 465 at the earliest. In some way or other his stay in Argos and visits to the rest of the Peloponnese, his wanderings in northern Greece, and his delay in Asia Minor must be extended to fill this gap of at least five years. There is evidence of a sort, there are arguments good and bad for the lengthening or shortening of any of these episodes, but none of this evidence or argument is conclusive. Between 470 and 465 no event in Themistokles' life can be securely dated; there is no fixed chronological pattern into which a reconstruction of the political history of Athens and the Peloponnese during these years must fit. Since the reconstruction which I attempt here is itself based on evidence which is far from adequate, plausibility is the most that can be claimed for it or for the chronological scheme which I infer from it.