I have not read the Thebais more than three times, nor ever with intent care and interest; and although in putting these notes together I have consulted a large number of editions—Bernartius, Tiliobroga, Geuartius, Cruceus, Gronouius, Barthius, Veenhusen, Beraldus (Delphin), ed. Bipontina, Lemaire (with Amar), Queck, O. Mueller (books I-VI), Kohlmann, Wilkins, Garrod, Klotz, and the translations of Marolles, Nisard, and Mozley (Loeb)—it may well be that profitable matter has escaped me and that some of my comments have been made before.

As a preliminary to an estimate of the work of Kleisthenes, I have sought in this paper to constate certain facts about the nature of the evidence.

The last Latin colony in Italy named by Livy is Aquileia; but Velleius Paterculus in a well-known passage says that Luca received a colony in 177 B.C., and follows his usual practice of not stating whether it was of the Latin or of the citizen type. Livy does not know of a colony at Luca, but does mention a citizen colony planted at Luna in 177 b.c.

Everyone knows the statement of Servius that Virgil was compelled by Augustus to alter the second half of the Fourth Georgic after the fall of Gallus, and that he substituted the story of Aristaeus for the laudes Galli. This statement, often doubted by older generations, has had such a remarkable success in recent years that anyone who ventures to impugn it must feel that he is pleading with a halter round his neck before a one-sided jury. It is notable, however, that these jurors, though one-sided, are by no means united, save in their determination to uphold the credit of Servius. One group will assure us that the episode of Aristaeus shows obvious signs not only of haste but of immaturity, and they draw a pathetic picture of the grief-stricken Virgil, ordered to bring out a second edition minus the praises of his friend Gallus, and half-heartedly throwing together some bits of juvenile stuff which he happened to have handy. Others assure us that the tale of Aristaeus shows us a more mature Virgil, nearer to the Virgil of the Aeneid. But the Servians really cannot have it both ways. They must make up their minds in the one direction or in the other if they are to be taken seriously. They make it doubly difficult for us to take them seriously when they try to tell us what Servius meant—but of that more anon. The present paper is an attempt by a converted Servian to examine frankly the arguments which have been advanced in support of Servius. Most of these may be found in Skutsch, Aus Vergils Frühzeit, pp. 140–147. One is very reluctant to use disparaging words about an eminent scholar who has served the cause of learning in so many ways; but it can scarcely be denied that the work just mentioned, in spite of its ingenuity and enthusiasm, shows a lamentable lack of three great essentials—a judicial temper, accurate statement of facts, and cogent reasoning. There is also at times a regrettable vagueness of expression, suggesting that the author is concealing inward doubts or shrinking from thought. All these deficiencies are to some extent present in the section with which this paper is concerned.

In C.Q. XXVI, pp. 35 sqq., Mr. H. D. Broadhead comments unfavourably on my essay, ‘Prose-Rhythm and the Comparative Method’ (C.Q. XXIV, pp. 164 sqq.; XXV, pp. 12 sqq.). I wish my reply to be explanatory rather than controversial. In a few places Mr. Broadhead has mistaken my wording, and he has, I fear, a poor opinion of my aesthetics. But those are personal matters; I will try in this article to defend my position generally, illustrating my remarks on the classical languages with some English analogies, and meeting, I hope, in the course of the argument the more important of Mr. Broadhead's objections.

A new edition of the Homeric Hymns is in preparation by Mr. W. R. Halliday, Principal of King's College, London, and myself. In the meantime there are some passages in the Hymn to Hermes which call for longer treatment than would be natural in an edition. Some of these notes are suggested by the substantial and valuable edition of Professor L. Radermacher (Wien, 1931).

Professor D'Arcy W. Thompson has recently revived a conjecture of Lauth on Geoponica, XIII, 8, 1, which runs as follows: ὅϕεις οὔκ ἔσονται ⋯ν χωρ⋯ῳ ⋯⋯ν ⋯ʊ⋯νθιονἢ ⋯ρτεμ⋯σιον ἢ ⋯βρ⋯τονον περ⋯ τ⋯ν ἔπα$$υλιν ϕυτε⋯σῃς. τοὺς δ⋯ ⋯ντας ⋯λ⋯σεις ⋯⋯ν (directions for a fumigation follow). The conjecture is that ⋯ντας is the Egyptian hontasu, ‘lizard.’ That this would make sense is obvious; but the usage of the Geop. itself, to say nothing of other authors, indicates that the word is simply what it appears to be, namely the pres. part, of εἰμ⋯. Thus we find in X, 46 τ⋯ ς⋯κα σκώληκας ¿ὔ π¿ι⋯σει⋯⋯ν μ⋯λλων φυτε⋯ειν ⋯μπ⋯ξῃς σκ⋯λλῃ τ⋯ν κλ⋯δ¿ν, τ¿ὺς δ⋯ ⋯ντας ⋯ναιρεῖς ⋯⋯ν κ.τ.λ. Here the explanation ‘lizards’ would hardly be reasonable; still more cogent, however, is XII, 8, 2–3, ¿ὔκ ἔσονται … τ⋯ς δ⋯ οὔσας κ⋯μπας διαϕθερῖς, where no such conjecture is possible. Not dissimilar is XII, 19, 9, τ⋯ς τε γ⋯ρ οὔσας ψ⋯λλας ϕθε⋯ρεικα⋯ ἄλλας οὔκ ⋯ᾷ γεν⋯σθαι. All these extracts are taken from authors of the second and third centuries A.D., though we have no means of knowing how closely their actual wording is followed; it is therefore in point to cite a passage of Galen, the opening paragraph of the tractate περ⋯ αἱρ⋯σεων Medicine, he says, has for its aim health, and a physician must know how to bring about health if absent (μ⋯ παρο⋯σαν) and preserve it if present (παροȗσαν). καλεῖται δ⋯, he continues, τ⋯ μ⋯ν ⋯ργαζ⋯μενα τ⋯ν μ⋯ οὖσαν ὑγ⋯ειατν τε κα⋯ βοηθ⋯ματα, τ⋯ δ⋯ φυλ⋯ττοντα τ⋯ν οὖσαν ὑγιειν⋯ διαιτ⋯ματα. In other words, ὢν can be used in much the same sense as παρών. This is no novelty of Roman times, for we may compare, for instance, Soph. Ant. 1109, οἵ ' ὅντες οἵ τ'⋯π⋯ντες, and El, 305 τ⋯ς οὔσας … κα⋯ τ⋯ς ⋯πο⋯σας ⋯λπ⋯δας.