BRIEF MENTION.

In his excellent Essai sur la rhétorique grecque M. Navarre makes a strong plea for the rehabilitation of rhetoric, which, according to his report, is fallen into unmerited neglect in Modern France (A. J. P. XXI 472). A similar plea was made some years ago by M. d' Haussonville in a critique of George Sand quoted in my Essays and Studies (p. 50): Les préceptes de rhétorique, he says, qui ont cours dans nos écoles ne diffèrent pas de ceux que la jeunesse studieuse recueillait autrefois sous les portiques d'Athènes et de Rome. Celui qui fait un jeu de ces préceptes et qui ne sait discerner l'éternelle vérité cachée sous leurs formules arides pourra peut-être surprendre un succès d'un jour; mais il s'exposera à voir couler tôt ou tard sa réputation fragile, comme un édifice dont l'architecte aura embelli la façade sans en asseoir la base d'après les lois de l'équilibre géométrique. No one can be surprised at these pleas, these warnings, who knows how much the sanity of French prose is due to rhetorical studies. With the new era the old charm will disappear. On the one hand, we shall have, nay, we have, utter carelessness, on the other, sublimated symbolism. And if rhetoric is neglected in France, what shall we say of the scant attention it has received in Anglo-Saxon countries? In the classical domain Mr. Sandys has done some good work, and it is to me a welcome sign of the times that Mr. ROBERTS has attracted so much attention and gained so much reputation by his admirable editions of the nepì vous (A. J. P. XX 228), and of the Three Literary Letters of Dionysius of Halicarnassus, to which he has now added Demetrius On Style (Cambridge University Press). To be sure, the author of the epi vous, the so-called Longinus, has never lost his hold on the modern world. He is a critic of exceptional delicacy of feeling and breadth of sympathy; and while the work may be rooted in antique rhetoric, its foliage and its flowers seem to belong to our world, and we can hardly realize that we are separated from him by the whole width of the Christian centuries. Dionysios, on the other hand, is no sooner set up than he is bowled over. Blass, who has done so much to bring him back into notice, has actually suffered by reason of his championship; and almost everybody that deals with him seems to be afraid of being suspected of spiritual affinity with the schoolmasterly critic. Usener classes him among the magistelli and Norden vilipends him, so that I am glad to see how stoutly Mr. ROBERTS stands by him with Mr. Saintsbury as an pedpos in time of need. For my part, Dionysios

accompanies me during three out of the four years of my cycle, and, while he is not exactly the man of my counsel, he is assuredly indispensable for any serious study of Attic oratory. If he is narrowminded in his judgment of Plato and Thukydides, that very narrowness shows us how potent rhetoric was in every sphere of antique literature; and in the matter of Thukydides, I am free to confess that I would rather consider the great historian a perverse genius, as Dionysios has done, than look upon him as a Laokoon, struggling with the twin serpents of diction and syntax, which had not yet been tamed to the docility of Aesculapian snakes (A. J. P. XIV 397, Shorey, Tr. A. P. A. XXIV, 1893, p. 82.)

As for Demetrios, nothing could be more timely than the revival of his admirable manual; for though Mr. ROBERTS is kind enough to say that rhetoric is not neglected in the United States of America, the cultivation of it must be very recent. It is not so very long since James Russell Lowell said: 'If I have attained to any clearness of style, I think it is partly due to my having had to lecture twenty years as a professor at Harvard. It was always present to my consciousness that whatever I said must be understood at once by my hearers or never. Out of this, I, almost without knowing it, formulated the rule that every sentence must be clear in itself and never too long to be carried, without risk of losing its balance, on a single breath of the speaker'. In other words, he formulated a rule that he might have learned in advance from Cicero,' from Dionysios,' who gives the veûμa redelov avdpós as the measure of the period. No wonder that one hails with satisfaction the prospect of a new edition of the De Compositione by so competent a hand as Mr. ROBERTS, if indeed we may construe his suggestion as a promise. So, too, if Oliver Wendell Holmes, that other great light of American literature, had read Dionysios, he would not have written in so tentative a fashion of 'the pneumatic or rather pneumonic character of rhythm'. Homo mensura is written all over our tables of weights and measures; the lungs are the windgauge of style. But Lowell and Holmes are classics and Apollo saved them. But Apollo will not always save. If there is an elementary rule in what the ancients called composition, it is that prose rhythm and poetical rhythm must be kept apart. All classical scholars know the rule. All French stylists insist on it. Everybody remembers how absurd Buffon made himself with his quatrain on the horse; though perhaps everyone that quotes 'un poète mort jeune à qui l'homme survit' does not remember the gentle rebuke that De Musset gave Ste.-Beuve. Tu l'as bien dit, ami, mais tu l'as trop bien dit'. Dryden's prose is admirable, but, as Mr.

1 E. g. De Oratore III 182.

De Compositione c. 23, p. 171 (R.)

Pater says, he will fall into verse, whereas Wordsworth never does, according to the same authority, and Dickens's bad blank verse is a stock illustration. And yet, despite this simple rule, a much-admired American writer of our day has actually reeled off hundreds, if not thousands of more or less perfect decasyllabic verses under the delusion that he was translating Homer into rhythmical prose. And sometimes I cannot help thinking that it would be better if our classical scholars themselves had read something more of Quintilian than the first chapter of the Tenth Book, to which most of them seem to confine their attention. Perhaps we might have been spared the theory that every Latin sentence is constructed on the principle of a diminuendo toot; perhaps we might have been spared the revolt against Quintilian's doctrine (Inst. Or. IX 4, 26) that makes the verb the file-closer. In this whole range of studies, we must be content still to acknowledge the ancients as our masters. There is no appeal from the ear.

Mr. ROBERTS's Introduction gives us a sketch of the Greek study of style. To one who knows his Rhetores Graeci, his Gerber, his Blass, his Norden, the sketch seems rather too sketchy, but Mr. ROBERTS dedicates his book 'Iuventuti lucide scribendi ac venuste studiosae', and his xeipaywyía is meant for those who have yet to learn that there is such a thing as a serious study of style. And even those who are more or less familiar with the subject will be delighted with the skill and the grace of the exposition. As a translator, Mr. ROBERTS has gained high repute by his previous renderings but in the matter of translation I am prone to be a devil's advocate; and as I have not compared Mr. ROBERTS's version with the original word for word, I am not prepared to say that he has solved all the problems that the translation of any Greek work on rhetoric brings up. The task, as I know from experience, is one of great difficulty. The ordinary dictionaries are of little help, and the special vocabularies often leave one in the lurch. The latest guide is not always the most trustworthy, and old Ernesti is frequently better than more recent Volkmann. This lack of lexical resource lends especial value to the Glossary which Mr. ROBERTS has appended and which has enabled him to reduce the volume of his notes. An important feature of these notes is the number of illustrations drawn from English literature; for nothing is better calculated to carry conviction to the mind of the student than just such cogent exemplifications of the universality of rhetorical canons. For these illustrations Mr. ROBERTS ACknowledges his obligations to Mr. WAY, who has won such renown by his poetical rendering of Euripides; and it is to Mr. WAY that Mr. ROBERTS owes also the versions of the poetical

citations and a number of valuable suggestions in the translation of the text.1

This notice has already run beyond the bounds of Brief Mention, and yet I have said little about the original text, about the 'goldene Schrift des Demetrios „epì épμnveías', as Wilamowitz justly calls it. As Milton puts Phalereus' next to Plato and Aristotle, I am content to leave him at the bidding of so excellent a judge in such excellent company. The book might be called a Rhetorical Testament, and a certain Biblical authority has been given to it by the short paragraphs into which it has been cut up by the old editor Victorius; and I am not surprised that Mr. ROBERTS recalls, by way of contrast, a like performance attributed to Robert Estienne. The canons of Demetrios are

sharp, clear, sensible. The illustrative passages are really illustrative and have saved for us many gems from the lost literature. In one of his commandments Demetrios warns against the multiplication of metaphor lest 'we find ourselves writing dithyrambic poetry instead of prose'. This is what we always tell other people. This is what we think very fine in Paul Louis Courier's Dieu, délivre nous du malin et du langage figuré'. But for all that, Demetrios is not averse to figurative language. The rhetoricians were teachers, and being teachers, knew the value of metaphor and simile, which lend wings to the seeds of doctrine, and which plant them in the field of the ear, if I may use a figure that underlies the purgatas aures of that 'crabbed coxcomb', Persius.

As I go to press I see that in the last number of the Classical Review (Feb. 1903), Dr. RUTHERFORD has made a savage assault on Mr. ROBERTS's edition. Dr. RUTHERFORD'S peculiar acerbity always stirs sympathy with the victim of his taws. How different the tone of Henri Weil's review of Dr. RUTHERFORD's Herondas in the Journal des Savants for Nov., 1891. It is quite as effective as Dr. RUTHERFORD's scolding and yet it is radiant with the charm of a iucunda senectus. But instead of imitating the mite ingenium of the dean of French Hellenists, the critic seems to have had his perfervid genius still further heated by the example of Roemer's review of his Scholia Aristophanica; and Mr. ROBERTS has suffered in consequence. Why, if I were disposed to make trouble, I myself might air one grievance I have against Mr. ROBERTS. In c. 213 Mr. ROBERTS translates the di' què ions of Ktesias by You were saved through me', and this is the very example that I selected in my S. C. G. § 163 for the rendering Thanks to'. But I recognize the fact that it is hard to get rid of the equation διά c. acc. = διά c. gen., nay, impossible to quell τὴν ἀμφίκρανον καὶ παλιμβλαστῆ Kúva (A. J. P. XI 372). Dem. 6, 45: di' ovç is not div, despite Sandys, and Lys. 26, 9: di' ovç is Thanks to whom'. Even in such Greek as the Life of Aesop, which Wilamowitz has actually used for his 'Lesebuch', to the horror of straitlaced Hellenists, the distinction is carefully observed. Cf. Vit. Aesop. c. 12 (p. 259 Eberh.), οὕτως ἔμεινεν ἐπὶ τῆς οἰκίας δι' Αἰσώπου ὡς καὶ δι' ἐκεῖνον ἀπῆρε

RADERMACHER'S Demetrius nepì épμnveías (Teubner) was published only a few months before ROBERTS's, and the English editor's work was virtually finished, so that he could make little use of his predecessor. But the Demetrios of Bangor and the Demetrios of Bonn are constructed on very different lines. ROBERTS writes for the general student; RADERMACHER, the able adjutant of USENER in his work on Dionysios, has the professional in view, and the two editions supplement each other. Both editors agree in rejecting the authorship of Demetrius Phalereus, which has few supporters nowadays, but ROBERTS gives a wider margin than RADERMACHER and takes in the first century before and the first century after Christ. RADERMACHER finds a terminus in the word oκapírns (c. 97), which occurs for the first and only time in Strabo, XVII, p. 817; and he thinks it impossible that Demetrios should be earlier than the first century after Christ. Demetrios, according to him, is untouched by the puristic spirit, and in language he is nearer to Plutarch than he is to Lucian. At my time of life I do not believe in impressionistic criticism, but I should not be grieved if someone shall make out a strong case for a somewhat earlier date. But to tell the truth, Demetrios's use of Aristotle pulls me one way, his admiration of Xenophon another.

RADERMACHER gives more scope to syntactical observations than does ROBERTS, and as I took RADERMACHER in hand after I read ROBERTS, I am naturally pleased to note here and there coincidences of judgment. So I was much surprised to find that ROBERTS was superstitious enough to retain, c. 269, the MS reading ei ovvapen. Any attentive reader must have noticed that Demetrios has a way of using +optative in the protasis and the future in the apodosis, and RADERMACHER has rightly put ei avvalein in the text. The optative was doomed, but it made a brave fight before it died, and though the ideal condition is very steady in classical times, the future indicative is occasionally found in apodosis from Homer, Il. 10, 222, down. Another el with subjunctive occurs in c. 76, οὐδὲν οὖν θαυμαστόν, εἰ . . . γένηται, where both editors follow the MS, but as θαυμαστόν

e normally takes the indicative, I should write without much hesitation γίνεται, though Roshdestwenski's γενήσεται is sufficiently plausible. In c. 5, ROBERTS follows Victorius and the other editors in reading οὐκ ἂν . . . γράψειεν instead of οὐκ âv・ ・ ・ ypáveɩ év, which RADERMACHER defends. Now, as I have said, S. C. G., § 432, the future indicative with av is theoretically a legitimate construction and is not to be excluded from later Greek, but in my judgment Demetrios is not late enough for that. As for the combination of aorist and perfect, on which RADERMACHER has a long note (p. 84) à propos of čoxev. . . kai