Describing Caesar's feast in Alexandria, Lucan comments on the folly of the Egyptians in displaying their riches to him, an armed guest already waging civil war, when even the more virtuous and austere Roman generals of antiquity - Fabricius, Curius and Cincinnatus - would be tempted to take such wealth in triumph for their country.

Deianira complains that Hercules, as a slave of Omphale, did not refrain from telling to the Lydian queen his famous labours; among them, the Hydra:

quaeque redundabat fecundo vulnere serpens

fertilis et damnis dives ab ipsa suis (Ov. Her. 9.95f.)

‘It will be admitted that redundabat, which usually means to “overflow”’ can only be applied to the Hydra by a very strong metaphor; but it is not only a strong one, it is quite unexampled: so A. Palmer in The Academy 49 (1896), 160. But his own emendation rebellabat has not convinced anybody – rightly, as I think. Still, his remark was not otiose: redundabat refers to the ‘growing again’ of Hydra's heads, that seem to ‘spring’ from the wounds like water's spurts; it is not impossible, but nevertheless it is strange. Why does Ovid use this metaphor ‘a fluctibus desumpta’ (Burman) here? I suggest that there is a precise reason: Ovid does not mention the name of the serpens, but of course his reader knows it and also knows why the Hydra is so called: πò τν ὐδτων.

M. J. Dewar argues that in Georg. 1.511–4 Virgil may have been drawing a disquieting parallel between Orestes, evoked through an imitation of Aeschylus (‘Choeph.’ 1021–5), and Octavian, present a few lines above (498ff.).

Pausanias probably supports this suggestion; he shows that the link Octavian-Orestes existed quite early and in a sense favourable to Octavian, even though it may soon have been used in a negative sense by anti-Caesarian propaganda on account of the dark side of the myth. In front of the temple of Hera in Argos there was still visible in the second century a statue representing Orestes, but identified by the inscription as Augustus. Certainly this parallel Augustus-Orestes was not proposed-and preserved - with polemical purpose in a famous sanctuary and in the Augustan age. Given the resemblance between history and myth and the moral weight of the famous myth itself, it is unlikely that we have to do with the mere re-use of any old statue.

It is not enough, says Quintilian (7.10.16), to assemble the various parts of a speech. The orator must arrange his points in the natural and logical order for his purposes, and he must unify the different sections so skilfully that no join will show (‘ne commissura perluceat’), producing a single body instead of assorted limbs. If we define ascommissura (or transitus) the rhetorical device which welds together different themes or chapters with an associative link in word or thought (sometimes matching like with like, more frequently depending on antithesis), Tacitus already had this lesson by heart when he wrote the Germania. That he exploited the same technique in his major works, concentrating on his transitions as hard as would Macaulay on his, appears not to have been noticed, and certainly has not received systematic study.

In De alimentorum facultatibus, Book I, Ch. 1, Galen begins his discussion of the powers of foodstuffs by a rough sketch of the opinions of earlier physicians on this subject. He says that according to some of them these powers are only known (γνσθαι) on the basis of experience (πεῖρα), according to others on the basis of a combination of experience and reasoning (λογισμς), whereas a third group gave priority of importance to reasoning (202.4–6 Helmreich). Galen proceeds to say that there is considerable disagreement between these physicians on the topic in question and that, consequently, an unbiased testing of their opinions is necessary; this testing should operate by means of argumentation (πδειξις). Now there are two different starting-points for argumentation, i.e. perception and ‘distinct thinking’ (ἢ γρ ξ αἰσθσεως ἢ κ νοσεως ναργος). Then the text printed by Helmreich runs as follows: κα μς ναγκαῖν στιν ἢ θατρῳ τοτων ἢ μφοτροις χρσασθαι πρς τν το προκειμνου σκμματος εὕρεσιν: ‘It is necessary for us, too, to use either one of these (i.e. either perception or thinking) or both of them in order to find the question under discussion.’

Not only was Gellius' preface received in the fifteenth century at the end of his work instead of the beginning, but it arrived almost or wholly without the Greek, which had to be patched up by guesswork; between siluarum and quidam early editors read ‘ille κηρ⋯ον, alius κ⋯ρας ⋯μαλӨε⋯ας’, the first two names in the similar passage, Plin. N.H. pr. 24. Salmasius, in the preface to his Plinianae exercitationes, printed a text ‘ex vestigiis antiquae scripturae optimi exemplaris [sc. MS P = Paris, BN lat. 5765] partim etiam coniecturis nostris correctiorem’; following κ⋯ρας he gave, in the right place but with the wrong accent, ‘alius Κ⋯ρια’. But when eleven years later he came to annotate Simplicius' commentary on Epictetus' Ἐγχειρ⋯διον, alerted by Simplicius' statement (taken from Arrian's own epistle dedicatory) συν⋯ταξεν ⋯ Ἀρριαν⋯ς, τ⋯ καιριὼτατα κα⋯ ⋯ναγκαι⋯τατα ⋯ν øιλοσοø⋯ᾳ κα⋯ κινητικώτατα τ⋯ν ψυχ⋯ν ⋯πιλεξ⋯μενος ⋯κ τ⋯ν Ἐπικτ⋯του λ⋯γων, he remarked: ‘Quidam et inscripsere libros suos olim τ⋯ κα⋯ρια, quod maxime ad rem quam tractabant pertinentia eo opere persequebantur’, citing Gellius with ‘alius κα⋯ρια and commenting ‘Ita enim ex veteri codice ibi scribendum est, non ut vulgo editur, κ⋯ριον [sic]'. Nevertheless, editors preferred his first thoughts to his second; Hertz, in his separate edition of Gellius' preface (Progr. Breslau, summer 1877) and in his editio maior (Berlin, 1883–5), gives three parallels:

Plin. N.H. pr. 24, ‘Κηρ⋯ον inscripsere quod uolebant intellegi fauom’, where the Latin translation guarantees the reading;

Clem. Alex. 6.1.2.1 (pp. 422–3 Stählin-Früchtel-Treu) ⋯ν μ⋯ν οὖν τῷ λειμ⋯νι τ⋯ ἄνӨη ποικ⋯λως ⋯νӨοȗντα κ⋯ν τ⋯ παραδε⋯σ⋯ [‘orchard’] ⋯ τ⋯ν ⋯κροδρ⋯ων øυτε⋯α οὐ κατ⋯ εἶδος ἕκαστον κεχώρισται τ⋯ν ⋯λλογεν⋯ν (ᾗ κα⋯ Λειμ⋯ν⋯ς τινες κα⋯ Ἑλικ⋯νας κα⋯ Κηρ⋯α κα⋯ Π⋯πλους συναγωγ⋯ς øιλομαӨεȋς ποικ⋯λως ⋯ξανӨισ⋯μενοι συνεγρ⋯ψαντο), where again the sense requires the honeycomb;

Philost. VS 565 ⋯πιστολα⋯ δ⋯ πλεȋσται Ἡρώδου κα⋯ διαλ⋯ξει κα⋯ ⋯øημερ⋯δες ⋯γχειρ⋯δι⋯ τε κα⋯ καίρια τ⋯ν ⋯ρχα⋯αν πολυμ⋯Өειαν ⋯ν βραχεȋ ⋯πηνӨισμ⋯να, where hertz emends κα⋯ρια to κηρ⋯α.

In the preface to his Nodes Atticae, Gellius explains his choice of title:

quoniam longinquis per hiemem noctibus in agro, sicuti dixi, terrae Atticae commentationes hasce ludere ac facere exorsi sumus, idcirco eas inscripsimus Noctium esse Atticarum (praef. 4)

He then proceeds to enumerate other titles used for miscellaneous works similar to his own, both Greek and Latin, which, he claims, are far more refined and witty than his title (§§4–9). Attractive as Gellius' explanation may be, it raises some serious difficulties2 and should not be taken at face value, especially since it seems to establish a novel type of relation between title and work. None of the titles in Gellius' list seems to have been based on the circumstances of the inception of the work, nor indeed does any other extant title prior to the publication of the Nodes Atticae. There is no reason to deny Gellius the credit for inventing a novel principle of titling, yet titles based on the circumstances of composition rather than on the content of the work fail to perform a primary function of titles, namely an initial direction of the expectations of prospective readers.

During the sixteenth century Galen's De constitutione artis medicae (i.224–304 Kühn) enjoyed a great success: in about fifty years it received four different Latin translations and three commentaries. Certainly this is also true of other medical classical texts, but such success is surprising for a treatise which did not have a wide circulation either in the Middle Ages or in the seventeenth century and later. In fact it is preserved in its entirety in only one Greek manuscript (Florence, Laur. plut. 74.3 = L of the twelfth or thirteenth century, with later corrections = L) and in a Latin translation by Niccolò of Reggio, who worked mainly for King Robert I in Naples in the first half of the fourteenth century. Furthermore, in his edition of 1679 René Chartier made a mistake, which the humanistic editors of the Greek Galen had avoided. The last part of the De const, art. med. itself enjoyed a considerable fortuna as an independent tract on prognosis in the Greek and Latin manuscript tradition. The editors of the Aldine and the Basle editions knew such an excerptum, at least in the manuscript Par. gr. 2165 (= P) of the sixteenth century, and rightly decided not to print it. Chartier found it in the manuscript Par. gr. 2269 of the fifteenth or sixteenth century, and published it in the wrong belief that it was a new treatise of Galen's (vol. ii. 170–95 = viii.891–5). He was followed by Carl Gottlob Kühn in his edition of 1821, who printed the De const, art. med. in the first volume (289–304) and the De praesagitura in vol. xix.497–511. The error was not publicly detected until Kalbfleisch in 1896.

In three articles published in Blackwood's Magazine (November 1849, April and May 1850), one Wilson, under the nom de guerre of Christopher North, propounded the view that Shakespeare's Othello operates on a double time scheme. The represented time in Cyprus (Acts II to V) is some thirty-three hours, lasting from about 4 p.m. on Saturday till the early hours of Monday morning. If we take this time scheme at face value, there has been no opportunity for Desdemona and Cassio to commit adultery: Iago's insinuations and Othello's suspicions are manifestly absurd. However, another time scheme is in operation as well. By its clock, the protagonists have been in Cyprus for more than a week. For example, we find Bianca, a local courtesan, complaining that it has been ‘seven days and nights, / Eight score eight hours’ (III.iv.171–2) since she last saw Cassio, who by the first time scheme had arrived in Cyprus only the day before.

Tall tales abound in Ctesias' Indica, as scholars have not hesitated to emphasize, heaping ridicule on the author's enthusiasm for the fantastic and on his apparent lack of regard for the truth. However, by no means everything in the work is absurd or wrong, and marvels too are no surprise. After all, as a resident of the Persian court for a number of years at the end of the fifth century B.C., Ctesias had seen items from India which would have been truly remarkable to Greeks of his time. He had seen, for example, elephants, which few Greeks before Alexander's Asian campaigns had done, and, it should be added, much of what he says about these animals is quite correct. The following pages discuss what he relates of the bird which he calls the β⋯ττακκος, the parrot, or rather what Photius in a not entirely problem-free section of his summary of the work preserves of the original description. As with Ctesias' account of the elephant, this is the first Greek description, so far as we know.

pausanias Preserves what we know about the Little and the Great Daidala, religious celebrations which took place in Plataia from the classical into the Roman period (Paus. 9.3.1–9). To his account can be added a fragment from Plutarch's work (Eusebius, Praeparatio Evangelii 3.1.6 = Plutarch fr. 157), and a brief mention in Menander Rhetor (Peri Epideiktikon, ed. L. Spengel, Rhetores Graeci iii, p. 367.7). At the celebration of the Little Daidala, which occurred about every six years or so (Paus. 9.3.3), the Plataians made an image from the trunk of an oak tree (Paus. 9.3.4); they called the image a daidalon, because ‘the men of old’ called the wooden images, the xoana, daidala (Paus. 9.3.2). Every sixty years, the Plataians celebrated the Great Daidala, to which other Boiotian states sent representations (Paus. 9.3.5–6, Men. Rhet. iii, p. 367.7). At this Great Daidala, all the images which had been made at the Little Daidala were gathered together and burnt (Paus. 9.3.8). The process by which this was done was to allocate by lot to each of the important Boiotian towns one of the daidala, and to distribute the rest amongst the lesser Boiotian states, who would pool their resources so as to be able to participate in the ceremony (Paus. 9.3.6). Each large city, or a group of smaller cities, was thus responsible for one daidalon.

Scene V.viii of Terence's Phormio (894–989) brings to a climax the confrontation between the trickster Phormio and the two old men Demipho and Chremes. Phormio, exploiting his knowledge of Chremes' extra-marital affair in Lemnos, persuades Chremes to surrender any claim to thirty minae, extracted by false pretences, which have in fact been used to purchase a girl for Chremes' son Phaedria. Demipho urges resistance to this blackmail, suggesting that they have more chance of placating Chremes' wife Nausistrata if they themselves tell her about the Lemnian affair. Phormio reacts by moving towards Chremes' house and calling out Nausistrata by name. The old men try in vain to restrain him.

In 294 most modern scholars either accept rapidique or adopt Lachmann's rapideque. An exception is Romanes, who oddly favours rapidisque, which he takes with impetibus crebris, placing a comma after corripiunt. If rapidique is read, one has to assume that Lucretius is writing as though venti, not flamina, were the subject. There are parallels for this kind of grammatical irregularity (e.g. 1.190, 352, if the text is sound), but there is no need to assume an irregularity here, for, as E. J. Kenney has pointed out to me, the right reading is almost certainly rapidoque. rapidoque was favoured by Lambinus, but did not originate with him. He notes ‘ex libris scriptis alii habent, rapidoque rotanti, alii rapidique rotanti’, and Pius (1511) knew rapidoque, which is printed in the ed. Juntina (1512). rapido…turbine is strongly supported by 1.273 rapido…turbine and 6.668 rapidus…turbo, also by subito…turbine in 1.279, a line which, as we shall see, is to be closely compared with 1.294.

Evander promises Aeneas two hundred of his Arcadians for the war against the Italians, with as many cavalry under Pallas into the bargain; and puts his son under the Trojan leader's command:

hunc tibi praeterea, spes et solacia nostri,

Pallanta adiungam; sub te tolerare magistro

militiam et graue Martis opus, tua cernere facta

adsuescat, primis et te miretur ab annis.

Stephanie West suggested in a note in this journal (40 (1990), 280) that the presence of an anecdote in Lodovico Guicciardini's sixteenth-century L'Hore di Ricreatione furnishes a parallel for the fable alluded to by Horace, Ep. 1.2.42–3: ‘Rusticus exspectat dum defluat amnis: at ille / labitur et labetur in omne volubilis aevum.’ The parallels, and a third from nineteenth-century Sicily, allow her to imagine a tale, ‘part of Italian traditional lore’, already extant in Horace's time and presumably transmitted in rural regions down to Guicciardini and the Sicilian story-teller.

In the midst of his fulminations against Cleopatra, Propertius denounces her land of Egypt in the following ‘wholly admirable parenthesis:’

Noxia Alexandria, dolis aptissima tellus

Et totiens nostro Memphi cruenta malo,

Tres ubi Pompeio detraxit harena triumphos!

Toilet nulla dies hanc tibi, Roma, notam.

Issent Phlegraeo melius tibi funera campo

Vel tua si socero colla daturus eras.

The text in Annals 4.28.1 has exercised scholars' attention since the rediscovery of Tacitus in the Renaissance. The text of the Medicean manuscript for the central words reads: ‘vinctus peroranti filio praeparatur’. Two problems have been perceived: firstly that praeparatur lacks an expressed subject, although from the context it is perfectly clear that Serenus senior is meant; secondly, the meaning of praeparatur itself.