SUA CLASSICAL REVIEW are The cd pee ka A I, cit af, the fal’ double Pceadeasthl VIEW in May, July, feet November, and THE CLASSICAL QUARTERLY EDITED BY F. W. HALL, M.A., St. John’s College, Oxford. R. HACKFORTH, M.A., Sidney Sussex College, Cambridge. Board of Management ; Prof. J. F. DOBSON, M.A., Chairman. -—y -* Prof. A. C. CLARK, Litt.D., F.B.A., Hon. Treasurer. L. WHIBLEY, M.A. Hon. Secretary. Prof, R. 8. CONWAY, Litt.D., P.B.A, Prof. H. WILLIAMSON, M.A. surer, and Secretary, repre- Representing the Oxford Philological Society. senting the Council @f S34 | >. @ ROBERTSON, Of A: Classical Assoctation. Representing the Cambridge Philological Society. | With the Chairman, Trea- E. A. BARBER, M.A. With the co-operation of Prof. Wm. GArpneR HALE, University of Chicag Prof. MAurRIcE Hutton, Toronto; Prof. T. G. Tucker, Litt.D., Pec ay Vol. XXII. APRIL, 1928 CONTENTS HORACE AND THE Morar FuncrTion oF Poetry. . . . . . . . J. Tate MISCELLANEA. . . . : Piette’. ot au yees wea . « T. W. Atten Tue Eouires Ittustres . ... . ERR toe SEW wie aw ig cite 2 Pears THE SPONDEION SCALE . . . . « ss. ve . Re P. WInNINGTON-INGRAM Tue Perrptus Maris EryTrHrRazki ...... +» M. P. CHar_LeswortH Ilapa:rteioBar='AGereiv . 2. 1 ew ees te os . Gan MM, Bomaiac Tue Suorter Giosses or Pracipvs. . . . .... i. j. W.. Presse Ciretgacnvs alin Omens 05 os ee ee EO. Pewee NIcANDER’s SIGNATURE, QuesTIons WITHOUT ANSWERS . oe Bp eOeEe Notes oN Festus AND TERENCE . . . . ee W. M, Linpsay SUMMARIES OF PERIODICALS . Lonpon: JOHN MURRAY, ALBEMARLE STREET, W. New Yorx: G. E. STECHERT. & CO., 31-33, EAST torn STREET, Price for single numbers, 4/- net, or 4/2 post free. Yearly Subscription 16/- post free ; U.S. A. $4, Combined Yearly Subscription for the CLASSICAL QUARTE RLY and the CLASSICAL REVIEW, 25/- post free; U.S.A. $6.25. THE CLASSICAL ASSOCIATION THE objects of the Classical Association are to promote the development and maintain the well-being of classical studies, and in particular (a) to impress upon public opinion the claim of such studies to an eminent place in the national scheme of education; (b) to improve the practice of classical teaching; (c) to encourage investigation and call attention to new discoveries; (d) to create opportunities for intercourse among lovers of classical learning. Membership of the Association is open to men and women alike. The annual subscription is 5s. (life composition, £3 15s.), and there is an entrance fee of 5s. (not charged to Libraries), Members receive a copy of the annual Proceedings of the Association and, on payment of 2s. 6d., of The Year's Work in Classical Studies (both post free). They may also obtain the Classical Review and Classical Quarterly at reduced prices, provided that the subscriptions be paid before January 31st in each year. Subscriptions sent in later than that date must be at the rates offered to the general public. Inquiries and applications for membership should be addressed to the Hon. Treasurer, Mr. H. F. Hose, Dulwich College, S.E. 21; or to either of the Hon. Secretaries, Dr. E. Norman Gardiner, 186, Woodstock Road, Oxford, and Miss E. C. Gedge, Crayford Rectory, Kent; or to the Hon. Secretary of any of the district Branches—viz., Mr. F. H. Sandbach, The University, Manchester; Dr. G. A. Auden, Education Office, Margaret Street, Birmingham ; Mr. R. B. Onians, The University, Liverpool; Professor Granger, University College, Nottingham; Mr. H. A. Pitman, 4, Prince’s Buildings, Bristol; Mr. Basil Anderton, The Public Library, New Bridge Street, Newcastle-upon-Tyne; Miss E. J. Vaughan, Westfield College, Hampstead; Mr. L. J. D. Richardson, Fernshaw, Hilary Road, Llanishen, Cardiff; Mr. E. E. Bibby, The University, Leeds; Mrs. Jackson, The Cottage, Cordwell Valley, Holmesfield, Nr. Sheffield; Mr. E. D. T. Jenkins, University College, Aberystwyth; Miss M. E. Lees, University College, Bangor; Miss M. K. Wakefield, Maidstone Girls’ Grammar School, Kent; Mr. S. Price, 33, Walton Well Road, Oxford; Mr. J. M. Gaussen, 16, Lewes Crescent, Brighton; Mr. G. W. Dyson, University College, Southampton; Mr. F. Fletcher, University College, Exeter; Mr. C. W. Baty, 49, Waterloo Road, Bedford; and Mr. J. E. Scott, Gonville and Caius College, Cambridge. Che Society for the Promotion of Roman Studies 50 BEDFORD SQUARE, LONDON, W.C. 1 President: PRINCIPAL STUART JONES, D.Litt., LL.D., F.B.A. The Society embraces the history, art, and archaeology of Rome, Italy, and the Roman Empire, down to about A.D. 700. In connexion with the Hellenic Society, the Roman Society maintains a joint library of Greek and Roman art, archaeology, and history, and a collection of lantern slides and photographs. Members are entitled to borrow books and slides, and these can be sent through the post. Communications respecting books and slides should be made to the Librarian at 50 Bedford Square. Afternoon meetings for the reading and discussion of papers are held at Burlington House, Piccadilly, W.1. Notices of these are sent to all members. Annual Subscription one guinea. Life membership composition for persons over fifty years of age ten guineas. For others fifteen guineas. There is at present nq entrance fee. : Persons desirous of joining the Society are asked to communicate with the Secretary, Miss M. V. Taylor, at the Haverfield Library, Ashmolean Museum, Oxford. THE JOURNAL OF ROMAN STUDIES (11 x 74 inches) is published by the Society in half-yearly parts, and sent post free toall members. Vols. I.-XVII. include papers by English, American, Danish, Russian, French, and Italian scholars, and, amongst others: ANDERSON, J. G. C. Festival of Men Askaénos. ASHBY, THOMAS. Roman Malta. ASHBY, THOMAS, and R. A. L. FELL. The Via Flaminia. ATKINSON, D. The Governors of Britain from Claudius to Diocletian. BAGNANI, GILBERT. The Subterranean Basilica at Porta aggiore. BAYNgss, N.H. Three Notes on the Reforms of Diocletian. Bury, J.B. The Notitia Dignitatum. CALDER, W. M. Roman Inscriptions in Asia Minor. CHEESMAN, G. L. The Family of the Caristanii. COLLINGWOOD, R. G. Hadrian's Wall: A History of the Problem. DAvtigs, Prof. G. A. T. Topography and the Trajan Column. DELLA Coote, M. Publius Paquinus Proculus. DESSAU, Prof. A Roman Senator under Domitian and Trajan. Charges. HarpDy, E. G. Notable Iudicia Populi on Capital HAVERFIELD, F. Roman London. HAVERFIELD, F., and STUART-JONES, H. Romano- British Art. Jones, H. STUART. A Roman Law concerning Piracy. Claudius and the Jewish Question. MACDONALD, G. The Corbridge Gold Coins. The Agricolan Occupation of North Britain. MACKAIL, J. W. Virgil and Roman Studies. Prerce, 8S. R. The Mausoleum of Hadrian. RAMSAY, Sir W. M. Studies in the Roman Province alatia. REINACH, SALOMON. A Bronze Statuette of Zeus. RUSHFORTH. G. MCN. Magister Gregorius de Mira- bilibus Urbis Romae. oe STRONG, Mrs.S.A. The Archaeological Exhibition of 1911 at Rome. ; VAN BuREN, A. W. The Ara Pacis Augusti. WARDE FowLer, W. The original meaning of the word ‘Sacer.’ * Mundus Patet.’ WHEELER, R. E. MORTIMER. Roman Colchester. The Roman Town Walls of Arles. Vol. XVII. is in active preparation. ‘ § HO incline functi traditi * passe Plato of lite provic ‘reviv Strabc I source we tak freque | tional to illus remar variou of cou! discrec I am literat poetry tions. from rt I Plato t Stoics Chrysi which It ise school: 1 Hove 2 In ar of the A belief (X doctrines 3 Philo (Berlin, NO. I THE CLASSICAL QUARTERLY APRIL, 1928. HORACE AND THE MORAL FUNCTION OF POETRY. THE modern admirers of Horace who take him seriously as a moralist are inclined to attribute an undue degree of originality to his views on the moral function of poetry. The conception of the poet as teacher was, of course, the traditional Greek view. But Professor A. Y. Campbell! thinks—in spite of ‘passages from Strabo and Plutarch ’—that this conception ‘after the days of Plato and Aristophanes lapsed as completely as did the production of the sort of literature that had justified it.’ Strabo and Plutarch, he asserts, merely provide evidence that the older Greek view revived; the forces inspiring this ‘revival’ were ‘not Greek, but Roman.’ ‘ The Greeks got it from the Romans, Strabo from the spirit of the Augustan age.’ In view of what Strabo and Plutarch themselves tell us concerning their sources, this theory is sufficiently strange.” But it will appear even stranger if we take into account the work of C. Jensen on Philodemus? to which I shall frequently refer in the course of this article. My object, then, is, in the first place, to show briefly that the Greek tradi- tional view of the function of poetry did not die out as is alleged, and, secondly, to illustrate from Horace himself the true position, namely, that Horace in his remarks on the moral function of poetry revived nothing, but simply reflected various well-known aspects of the still flourishing Greek conception. I do not, of course, wish to deprive Horace of whatever credit (or, as some will have it, discredit) is due to him for appropriating this Greek attitude towards poetry. I am concerned merely to mitigate the statement that ‘ Horace’s views on literature are derived primarily from himself.’* His account of the function of poetry may (though I greatly doubt it) represent deep-rooted personal convic- tions. But certainly his views are not original; nor do they mark a departure from recent Greek criticism. I. In the first place, poetry, especially Homer, continued for centuries after Plato to play its part in the education of the young.’ Of this educative use the Stoics—as formerly Protagoras*—wholly approved. Zeno, Cleanthes and Chrysippus all contributed something, as Plutarch’ tells us, to the devices which the schoolmaster adopted for extracting moral lessons from the poets. It is evident that we may accept Strabo’s statement that the poets were read in schools ‘ for the sake not of mere amusement but of moral correction.’® 1 Hovace (1924), p. 29. # Campbell, op. cit., p. 68. 2 In any case, how could we regard ‘the spirit 5 Strabo I. 2, 3; Heraclitus, Quaest. Hom, 1: of the Augustan age’ as the source of Strabo’s Dion Chrys. XI, etc. How the schoolmaster belief (X. 3, 23) that sound physico-theological used the poets to enforce morality may be seen in doctrines had been handed down by the poets ? Plutarch, De Audiendis Poetis. 3 Philodemus aber die Gedichte, funftes Buch 6 Plato, Prot. 326. (Berlin, 1923). 7 Op. cit. 33-34. 8 Loe, cit. NO. II. VOL. XXII. E 66 J. TATE It was the wisdom of the poets which justified this use of their works. Just as Protagoras in Plato’ had affirmed that the early poets were sophists in disguise (who concealed their doctrines because the time was not ripe for disclosure), so the Stoics (and others) declared poetry ‘a kind of first philo- sophy.”* They asserted, too, that the wise man alone is poet.* Cleanthes even held the view that poetry is superior to philosophic discourse as a vehicle for expressing the highest truths of philosophy.* He carried his belief into practice by writing several poems, including the famous Hymn to Zeus. On this traditional belief in the philosophic wisdom of the poets the Stoics —and above all, Chrysippus°—placed such extreme reliance that by their etymologies and allegories they practically turned the poets into Stoic philosophers.® : But it was not only the professed Stoics who retained this didactic theory of poetry. The Stoic Strabo has transcribed for us the views of the non-Stoic Polybius, who argued against Eratosthenes’ (the alleged champion of ‘ delight ’ as the exclusive aim of the poet), and held that the general opinion was correct, namely, that the Homeric poems were a philosophic treatise.2 Similarly Plutarch is openly indebted to the Stoics in his essay ‘How the Young Man should read Poetry.’® Though opposed to their allegorical interpretations, which he regarded as impious” and absurd," he is at one with the Stoic and traditional view of the function of poetry. It is clear, then, that the assumption made by Aristophanes and his audience persisted continuously down to the time of Horace and later: poetry is to be valued chiefly for its instructiveness.“ But further confirmation is afforded by Philodemus, On Poetry,“ where considerable light is shed on the views of Ariston of Chios (a moderate Stoic who was one of Eratosthenes’ teachers), the Alexandrian Neoptolemus of Parium,“ and other grammarians. Ariston, for example, in rejecting the ‘ physical’ interpretations of the myths, while contending that the poet must combine good thoughts with a good style, is evidently a precursor of Plutarch, whose attitude, as we have seen, is very similar. Again, Neoptolemus held that the function of poetry is twofold: (x1) to charm the spirit of his hearers (the #vyaywyia of Eratosthenes), and Piutarchi Commentationum De Aud. Poet., of which the concluding paragraphs may be consulted in 1 Prot. 317b. 2 Strabo I. 1, ro. 8 Strabo I. 2, 3, from which it follows, as Strabo later (I. 2, 8) insists, that the good poet must be a virtuous man. 4 See Philodemus quoted by Zeller, Siéoics, p. 323, 0.1. 5 See von Arnim, Stoic. Vet. Frr. Il.. p. 255. 6 Cf. Cicero, De Nat. Deorum I. 15, 41. 7 Strabo, Joc. cit. E. held that all poets aim at puxaywyla not ddacxadla, so that the poet must not be judged primarily by his knowledge. Cf. Aristotle, Poetics 1460b. 8 Strabo I. 2, 17. 9 He is, of course, more largely indebted than he acknowledges. See Schlemm, De Fontibus F, M. Padelford, Essays on the Study and Use of Poetry by Plutarch, etc., pp. 121-123. 10 Amat. 757¢. 11 De Aud, Poet. 318. 12 Frogs 1008 sqq. 13 See Jensen, of. cit. 14 That even the Alexandrians were far from holding any such theory as ‘art for art’s sake’ might be illustrated by the fact that even Eratos- thenes imitated the didactic vein of Hesiod. On his Hermes see Couat, La Poésie Alexandrine, Pp. 465 sqq. 18 Jensen, op. cit., pp. 131 and 135. (2) to | thenes? Strabo! subserv bitter p that he delight. No said to as Jens Ac Horace embodi He Greek c tolemus The pre municat Butcher! own ide concern impress which d Th into tw celst Ra instruct 1 Op. ci Tw Terely axovovras | 2 Of co charm is was perha upon this the poet’s 3 Who : acts like a of knowle nit, 4 Jenser 5 Op. ci ir works. phists in ripe for st philo- thes even thicle for ) practice he Stoics by their to Stoic c theory on-Stoic ‘delight ’ ; correct, Similarly ing Man etations, stoic and and his : poetry ation is 1 on the sthenes’ marians. e myths, d style,” , Is very twofold : es), and , of which msulted in and Use of » far from urt’s sake’ en Eratos- ssiod. On lexandrine, HORACE AND THE MORAL FUNCTION OF POETRY 67 (2) to give them good and profitable teaching.” He thus conflates Eratos- thenes? with the more distinctively Stoic view. Of course, the Stoics—witness Strabo*—did not deny the delightfulness of poetry. Rather they made it subserve the moral purpose by rendering instruction attractive; it sugared the bitter pill of precept and precedent. With regard to Homer, Neoptolemus held that he generally (though not always) succeeded in combining instruction with delight.‘ Now it was from Neoptolemus of Parium that Horace in the Ars Poetica is said to have taken ‘ praecepta eminentissima.’ This statement of Porphyrion, as Jensen°® has established, is no exaggeration. Il. Accordingly I propose to set forth in the second place the chief passages in Horace which recall the Greek traditional views on the function of poetry as embodied both in Neoptolemus and in other Greek sources. Non satis est pulchra esse poemata; dulcia sunto et quocunque volent animum auditoris agunto. A.P. 99-100. Here the principle of poetic charm is laid down in accordance with the Greek critics; 1. 100 paraphrases the term Wuyaywryia of Eratosthenes, Neop- tolemus and others.® It is not enough for a poem to satisfy the canons of art.’ The proper criterion is its effect upon the minds of those to whom it is com- municated. This standpoint is characteristically Greek. Even Aristotle—as Butcher® points out—did not hold that a poem attains its end by ‘ realizing its own idea,’ but in the impression made upon the mind of the hearer. This concern for the audience, and, in consequence, for the ethical nature of the impression which they are to receive, is the root of all the ethical preoccupations which distinguish Greek literary criticism.® The poet, then, must excite the interest of his hearers. But the latter fall into two classes: the first consists of those who merely desire pleasure—the celst Ramnes of A.P. 342; while the second consists of those who demand instruction also—the centuriae seniorum of A.P. 341. Accordingly A.P. 343-4. Omne tulit punctum qui miscuit utile dulci, lectorem delectando pariterque monendo. 6 Cf. Timocles (Meineke, Com, Frr. II. 800) apparently following Aristotle: ~uxaywyndeis. 7 For this sense of pulchra, cf. Epist. Il. 1, 72. 8 Aristotle’s Theory of Poetry, third edition, p. 208. 1 Op. cit., excursus I., especially p. 123: deity T@® Terelw woinrn peta THS Wuxarywylas Tov rovs dxovovras woedeiv kal xpnommoroyeir. 2 Of course the demand that the poet should charm is much older than Eratosthenes, who was perhaps merely the first to insist exclusively upon this (in the popular view) minor aspect of the poet’s office. 3 Who says (I. 2, 8) that the pleasure of poetry acts like a love-philtre in stimulating the desire of knowledge. Cf. Plutarch, De Aud. Poet., ad tnit, 4 Jensen, op. cit., p. 113. 5 Op. cit., p. 108. 9 Even the author of the wep? tous is far from judging poetry by a purely literary standard. E.g., he censures the Theomachy though acknow- ledging the sublimity of its imagery (IX. 6-8)— presumably on the ground that an impious passage, however awe-inspiring, will not affect the normally moral man with the rapture which is the end of poetic imagery. 68 J. TATE The poet who pleases both parties by combining instruction with delight has satisfied every test. Some poets aim merely at giving pleasure (delectare— A.P. 333), others at the improvement of their readers (prodesse—Ibid.). It is the practice of the most successful poets ‘simul et iucunda et idonea dicere vitae’ (A.P. 334). Idonea dicere vitae is a translation of Neoptolemus’ xXpnotmoroyeiv (to give useful precepts), as prodesse is of mdedeiv.1 Horace has clearly adopted the statement of Neoptolemus. It should be noted that if we are to say that on the Horatian view the true poet ought to instruct, we should remember that it is the ‘ought’ of expediency, not of morality. He will give moral counsel because the audience, or an important section of the audience, desires it, and it is therefore the best policy.” How, then, is the poet to be equipped for his work of instruction? The Greek traditional answer would have been that he is extremely wise, perhaps omniscient,® and that his wisdom and knowledge are in some sense the result of divine inspiration.‘ It is on the poet’s wisdom that Horace, in harmony with the Stoics (and their predecessors), lays most emphasis. Not only must the poet have his matter well thought out beforehand,’ he must have studied philosophy, especi- ally ethics: A.P. 309 sqq. Scribendi recte sapere est et principium et fons: rem tibi Socraticae poterunt ostendere chartae, etc. Line 309 is obviously an echo of the Stoic dictum already cited, which numbered the poetic art among the accomplishments of the ideal wise man. ‘Sapere’ is wisdom in general (as Oreili takes it), and especially moral wisdom, as Il. 312 sqq. clearly show.® The poet who has studied ‘Socraticae chartae’ will be able to supply his characters with appropriate words and actions (l. 316). ‘Socraticae chartae’ is certainly a wide enough term to include Stoicism. But a consideration of |. 309 and of ll. 311 sqq. makes it probable that the term means exclusively or, at least, pre-eminently the Stoic philo- sophy, to which Horace himself had given a certain amount of study.’ More- over, Stoicism claimed to be, and was called, Socratic. For example, Heraclitus® clearly means Stoicism by Ywxpatixy copia; and Persius® speaks of the ‘ Socratic bosom’ of his Stoic teacher Cornutus. Even in przhistoric times (of which Horace seems to take an Epicurean, if not Darwinian, view), ‘sapientia’ was the distinguishing mark of the poet. See A.P. 391 sqq. His wisdom in those days found indeed a more practical expression : 5 A.P. 311, where provisam rem =mpovootpeva of Neoptolemus (Jensen, of. cit., p. 115). 6 For this sense of sapere cf. Epist. I. 2, 40. 7 A. Y. Campbell more or less agrees (p. 250): e.g. Horace ‘ himself used the works of the later schools much more than Plato’s ; and the follow- 1 See supra, p. 67, n. I. 2 Cf. A. Y. Campbell, op. cit., p. 64, who dis- cusses the Horatian view of ‘the supreme duty of the poet.’ I cannot find anything in Horace about this ‘ duty.’ 3 Cf. Plato, Rep. 598de ; Xen. Symp. IV. 6-7, etc, ing lines’ (i.e. 312 sqq.) ‘ suggest Stoicism.’ * Cf. Pindar, fr. 150 (Bergk); Plato, A pol. 22b, Laws 682, etc. 8 Quaest. Hom. 79. 9 V. 37. » The m moved givers < tion as mixed both hi (1030 si Sil as appl praise. fari pos which | Ariston Ac moral selectec the pra his poe probab! improv He moralit poetry | set befc praise ¢ allegori testimo functio! An Homer. state by very co 1 With (1. 5). 2 Supra * Cf. S 70, et at. ; 4 Cf. x suppose ( gone min ight has slectare— ). Itis a dicere tolemus’ race has at if we e should will give udience, 1? The perhaps le result ics (and have his » especi- atc. |, which se man. wisdom, shartae’ actions include yrobable c philo- More- xample, » speaks urean, if le poet. ractical povoov peva, ). 2, 40. } (p. 250): the later 1e follow- sm,’ HORACE AND THE MORAL FUNCTION OF POETRY 69 A.P. 396. Fuit haec sapientia quondam publica privatis secernere, sacra profanis, etc. The myth that Orpheus charmed tigers and lions, the myth that Amphion moved stocks and stones, are allegorical hints that primitive poets were law- givers and founders of cities and of civilization. The view of the poet’s func- tion as moral and instructive is thus confirmed by, and indeed, unconsciously mixed up with, the belief that such had been the office actually performed by both historic and prehistoric bards. The parallel with Aristophanes’ Frogs (1030 sqq.) is very striking. Since sapientia is the distinguishing feature of the successful poet, ‘ sapiens’ as applied to any particular poet is little more than a conventional term of praise. Hence we find Tibullus (in Efist. I. 4, 9) described as ‘ qui sapere et fari possit quae sentiat.’' Ethical wisdom and a good style—these qualities which Horace here ascribes to the poet are the same as were demanded by Ariston of Chios. Horace has supported the opinion that the proper function of poetry is moral improvement by a reference to prehistoric ‘facts’ which have been selected, or perhaps fabricated, to suit the theory. But he can also point to the practice of Homer, chief of poets, who, as the Greeks believed, composed his poems with a view to instructing and edifying his hearers. Indeed it was probably because they held this view of Homer that they set up moral improvement as the proper aim of poets in general. Hence Horace (Epist. I. 2, 1-4) declares Homer a better teacher of morality than Chrysippus and Crantor; just as Cleanthes? had preferred poetry to philosophic discourse. Moreover, in the person of Odysseus, Homer set before us an ‘utile exemplar’ of the efficacy of virtue and wisdom. This praise of Odysseus is also in the Stoic manner.* Combined with the mildly allegorical treatment* of the Odyssey in the following lines, it is further testimony to the influence of Stoicism on the Horatian view of the poetic function. Another quality of Homer is selected for mention in A.P. 401 sqq.: Post hos insignis Homerus Tyrtaeusque mares animos in Martia bella versibus exacuit. Homer, like Tyrtaeus, displayed the services which the poet can render to the state by stirring the warlike spirit of his hearers. Such praise of Homer was very common among the Greeks. Aristophanes® declared that Homer won all —whether in the treatise of Heraclitus or else- where. It should also be pointed out that the statement (ibid., n. 3) that Heraclitus’ source was the Roman Varro—though it has Zeller’s 1 With the sense of sapere cf. sapiente bonoque (1. 5). 2 Supra, p. 66, n. 4. 3 Cf. Strabo I. 2, 4; Heraclitus, Quaest. Hom. 70, et at.; Seneca, De Const. Sap. II. 1. 4 Cf. Xen. Mem. 1.3, 7. There is no need to suppose (cf. Campbell, p. 276) that Horace had gone minutely into the study of Stoic allegories countenance (Stoics, p. 336)—must be regarded as exploded. Consult, e.g., C. Reinhardt, De Graecorum Theologia, 5 Frogs 1034-1039. 70 J. TATE his glory by teaching the discipline and doughty deeds of war, and that many men, such as the hero Lamachus, owed their martial prowess to their familiarity with his poems. In Plato’s Jon’ the rhapsode claims to be the best of generals because he surpasses all men in knowledge of Homer. In Xenophon,? Niceratus claims to have learned generalship and many other arts from Homer. That the arts which Homer knew included strategy, was a view to be reckoned with even in the second century of the present era; this is clear from the rhetorical treatise of Hermogenes,* who praises Homer as the best of poets and rhetors, but adds regretfully that perhaps he is not the best of generals or statesmen, although he produced a most excellent imitation of such in accordance with the best models. Horace, it may be observed, has dropped the somewhat exaggerated claim that Homer taught the science of warfare. He contends merely that the poet rouses military ardour. Here he agrees with Isocrates,* who says that Homer was honoured because he praised those who fought against the barbarians,° and because his verses were likely to arouse in his hearers the desire to perform similar exploits themselves.® But the argument for the wisdom of the poets is clinched by the fact that poetry does actually perform an educative function. Epist. II. 1, 126-131, may be taken as representative of the theory and practice of Greek education, and of Roman education in so far as it was in this respect influenced by Greek. The poet instructs both by precept and by example: ‘ pectus praeceptis format amicis ’ and ‘ orientia tempora notis instruit exemplis.’ ’ From the practice of the Greeks comes also the other argument which is adduced by the same passage to prove that the poet is ‘ useful to the state’ (l. 124). The poet knows how to obtain the favour of the gods by song (ll. 132-138). He too affords consolation to the needy and troubled. Somewhat similar is the statement in A.P. 405: ‘ludusque repertus et lon- gorum operum finis ’"—poetry affords amusement, and to the weary relaxation.’° In the fourth book of the Odes Horace dwells on yet another proof of the utility of poetry, which again has been derived from the Greeks." If there were no poet to sing the deeds of brave men, what hope would virtue have of 1 541. (Rep. 607a). Cf. also Pindar, Pyth. I. 12. 2 Symp. IV. 6-7. ® Line 131: imopem solatuy et aegrum. Cf. 3 Walz, Rhetoves Graeci III., p. 375. Hesiod, Theogony 98 sqq. ; Aristotle as quoted by 4 Panegyr. 159. Diog. Laert. V. i. 19; Cicero, Pro Archia 16. 5 This un-Homeric view of the Trojans as Euripides (Medea 190 sqq.) denies that poets barbarians appears also in Horace (Epist.I.2,7, have yet learned to do this, ‘Graecia barbariae lento collisa duello ’). 10 Cf, Cicero, Pro Archia 12: ‘suppeditat .. . 6 There is no sign that Horace, any more than _ubi animus reficiatur.’ Aristophanes or Isocrates, found any difficulty in 11 Odes IV. &, 13 sqq., especially Il. 20 sqq. the praise of Homer as ‘ excellent for recruiting purposes’ (cf. Campbell, op. cit., p. 60). 7 Cf. Plato, Prot, 326; Phaedrus 245a:... Tovs émvyvyvouevous wacdever. 8 The reference is simply to the Greek choral And 9, 25 sqq. (‘vixere fortes ante Agamemnona,’ lyric. Cf. Plato who obeys tradition so farasto _ etc.). retain ‘hymns to the gods’ in his ideal state neque si chartae sileant quod bene feceris, mercedem tuleris, etc. lasting | be igno: Th with w traditio it was tendenc praised of ‘div compar 4 It has represe and th betwee acquire Greek « contras Ci them 1 Democ the Gr it is cl with tl in rega the div logical ration: explici a strai than A confou same t 1 Cf, 22-57; | VII. 16. 2 Fryv. 3 Pha 4 See at many miliarity generals Niceratus r. That ned with hetorical | rhetors, atesmen, with the ed claim the poet t Homer barians,° perform act that 31, may ion, and ? Greek. ; format vhich is > state’ oy song yubled.?® et lon- ation.?° of the f there have of 2. um. Cf. uoted by ychia 16. at poets itat . qq. s, mnona,’ HORACE AND THE MORAL FUNCTION OF POETRY 71 lasting fame? Noble characters would lack this incentive, and posterity would be ignorant of their illustrious examples.’ There now remains Horace’s treatment of inspiration. In all the passages with which we have so far dealt Horace has been in accord with Greek tradition, ancient as well as recent. But here we have proof positive that it was from recent criticism that he derived his views. The older Greek tendency had been to regard the best poetry as divinely inspired. Democritus? praised the poetry which is written under a divine impulse. Plato* had spoken of ‘divine madness’ and ‘madness of the Muses.’ With such utterances compare— A.P. 295 sqq. Ingenium misera quia fortunatius arte credit et excludit sanos Helicone poetas Democritus. It has never been pointed out that Democritus is here unconsciously mis- represented. The old distinction between the poetry which is divinely inspired and the poetry which is not had long since been replaced by the distinction between that which is the result of natural ability and that which springs from acquired skill. Horace has adopted the antithesis, common in the more recent Greek critics,* between Svvayis (natura, ingenium) and téxvn (studium, ars). The contrast appears again in— A.P. 408 sqq. Natura fieret laudabile carmen an arte quaesitum est; ego nec studium sine divite vena nec rude quid prosit video ingenium. Cicero® similarly misinterpreted Democritus and Plato in attributing to them the view that madness is a necessary element in the poet’s character. Democritus and Plato had spoken of divine madness. Cicero and Horace, and the Greek critics whom they follow, have removed the epithet ‘divine.’ Thus it is clear that Horace, though he speaks of himself on occasion, in accordance with the literary convention, as priest of the Muses,® is a thorough ‘ rationalist’ in regard to those goddesses. He is at one with the later Greeks in turning the divine enthusiasm of which Plato and others had written into a psycho- logical aberration. Among the Greeks it was perhaps with Aristotle that this rationalistic treatment of poetic inspiration began. For it was he who first explicitly distinguished between the poet with a natural gift and the poet with a strain of madness.’ But Horace is probably following a much later authority than Aristotle, for in him this distinction does not appear; rather, the two are confounded, the natural gift and the strain of madness being regarded as the same thing. 1 Cf, Cicero, Pro Archia 14; Theocritus XVI. 5 De Divinat. I. 37, 80. Cf. Pro Archia 18, 22-57; Pindar, Ol, X. 91, Pyth. III. 114, Isthm. where ‘ divine afflatus’ is to the Roman evidently VII, 16. a very bold metaphor. 2 Fry. 18 and 21 (Diels). 6 Odes III. 1, 3. Cf. Epist. 11. 1, 133. 3 Phaedr. 2454, etc. ? Poetics 1455a, 4 See Jensen, ep. cit., p. 123. 72 HORACE AND THE MORAL FUNCTION OF POETRY It is interesting to observe that Horace is manifestly out of sympathy with any view which lays stress on the ‘ fine frenzy’ of the poet.’ But more to our present purpose is the bearing of this attitude on the moral function of poetry. The view of inspiration as évvayis makes it impossible to claim for poetry any kind of divine authority in doctrine. The poets are no longer regarded as divinely inspired sources of ethical truth.” A few other instances*® might be given from Horace to illustrate his indebtedness to Greek utterances on poetry. But enough has, I hope, been written to substantiate my conclusion. On the moral function of poetry we find in Horace simply his account of the conceptions of recent Greek critics (based on the older traditional views), conceptions which, since they were so largely adopted by the Stoics, come to us mingled with a good deal of more or less popular Stoicism. It is evident that Horace in his remarks on the poetic function has followed his own advice to the poet;* he has thumbed well the exemplaria Graeca. J. TATE. QUEEN’S UNIVERSITY, BELFasrT. 1 Cf. Sikes, Roman Poetry, p. 62. In the alli- ance of genius and art, Horace ‘certainly seems to regard Art as the predominant partner.’ Horace’s attitude may in part be due to Roman hostility towards anything in the nature of enthusiasm. This is still clearer if we compare Pseudo-Longinus (XIII.) who, although ap- parently as great a rationalist as Horace in his Homer and (in a lesser degree) of Hesiod as in some sense a divine revelation. Cf. Julian, Or. IV., p. 149¢ (Homer was 6eéAnr70s) ; Proclus, In Plat. Remp. 405 (riv ‘Opjpov rounrixhy ri ard Gedy wpunuévnv). But such miracles were defi- nitely relegated to the remote past, e.g. Dion Chrysostomus (XXXVI. 34) is careful to make his remarks on divine inspiration apply only to the treatment of inspiration, makes much ofit inthe very ancient poets. sense of high emotional power. 3 For a point which Horace probably owes to 2 Cf. Plato, Ion. 534e, Laws 682,719, etc, Of the Homeromastiges see my note on Epist. I. 19,6 course the philosophers who revered the early in C.R., December, 1927, p. 218. poetic traditions could still regard the works of 4 A.P. 268, ment work run < ador sami Rza shov RY athy with ore to our of poetry. oetry any rarded as strate his ope, been of poetry ek critics hey were d deal of remarks * he has TATE. esiod as in Cf. Julian, ); Proclus, chy Thy ard were defi- eg. Dion o make his only to the ly owes to bist. I. 19,6 _-p6Es — MISCELLANEA. I. ONOMACRITUS AND HESIOD. THE new volume of Oxyrhynchus Papyri (No. XVII., 1927) provides a frag- ment of Hesiod (No. 2075) which is interesting in more than one respect. The work is the Catalogi and the subject the apotheosis of Heracles. Lines 16-23 run as follows: —vuv § dn Geos eats kaxwv 8 eEnrv0e TavTwr —fwe & evOa rep adror orXAvpTria Swpar’ exor[ Tes —alavaros xat aynpos exwov xarXd{[to]dupov nBylv —rratda dios weyarowo Kai npns XpuvcoTredsd| ov —rov mpi pev p’ nXOnpe Gea ANevKwrevos np[ —ex Te Dewy paxapwv [ex] re Ovntwv avOpw|mrev —vuv § ndn mredirnxe tie Se pwv e€oxov adrA[ wv —ala[vatlov pera y’ avrov epicOeven x[polpr|cjwvr[a. We have therefore, for the first time, a specimen of a Hesiodic text adorned with onyeia. It was well known that the Alexandrines applied the same methods to Hesiod as to Homer, and the evidence is collected by Rzach in Pauly VIII. 1226, but hitherto no MS., uncial or minuscule, had shown the actual signs. Next, it is not difficult to perceive the reason for the obelization of these eight lines. As the editor remarks, they partially coincide with » 601-4: tov dé per’ eioevonca Binv ‘Hpaxrnetnv ¥ te \ 7. £ / ay —eldwrov* autos bé¢ pet aOavaracr Oeoict —réprerat év Oarins nai yet Kadriogupov "HBnv —raida Atos peyaroto kal” Hons ypucorredinou. The obeli on these lines are preserved in the lost minuscule MS. ‘J’ according to the collation of Heinsius,’ and the reason for the athetesis is given in the scholia: a0erotvtas xai Néyovtas ’'Ovopuaxpirou elvat, rodrov [604] brs ’Ovopaxpirov éureroqobat pacw nOérnra: 5é. It is an easy inference that the eight lines in Pap. 2075 have the same origin, and are to be added to the relics of this Orphic poet and oracle-forger? collected in Kinkel, E.G.F., pp. 240, 241. There the fourth fragment deals with Heracles, and ends ’Ovopa- kpitos dnow év trois éreot. Further, the section on Heracles in the Theogony 943-4, 950-5 lies under the same suspicion, since the scholiast says aJerotvtas Heinsius’ statement. 2 Herod. VII. 6. Aynpos and regiAnxe may be put down to the style of sixth-century epos. 1 MSS. of the Odyssey in which signs occur are H1, H3, L4, L8, M4, R4, R8, Us, U6 (bd), Ug, V4. There is, therefore, no reason to doubt 74 T. W. ALLEN épeEns atixo. évvéa, though to which lines he refers is not clear. (See the note in Rzach ed. major.) From these data consequences follow : 1. The Hesiodic text was less well defended than the Homeric; it was weaker in the proportion of 8 to 3. Onomacritus gained admission for three of his lines to the Odyssey, eight to the Catalogi, besides possibly the Heraclean portion of the passage in the Theogony. } 2. The Alexandrians had the flair to detect these amplifications, but not the power to removethem. I commend this circumstance to the consideration of those who hold that the current vulgate was established by them. 3. The Alexandrians were aware of the composite nature of their text, how it had been augmented by Onomacritus and others, mostly now anonymous. They were unaware of the basic fact that the poems had been put together, and the Trojan war itself invented, in the century of Onomacritus, and why not by Onomacritus himself? Wise and prudent they overlooked this ; it was revealed to the babes of the Fatherland. 4. A terminus post quem is provided for the fifteenth Homeric Hymn. Il. A PTOLEMAIC ILIAD. In Mr. Milne’s Catalogue of the Literary Papyri in the British Museum, 1927, long expected and welcome, there is a fragment (No. 251) of Book M of the Iliad of the second century B.c., containing parts of lines 128-140, 176-191, 249-263. There are alterations and additions, as is common in Ptolemaic Homers. It would be interesting to know if these fragments are part of the same book as the Flinders Petrie fragment of the same century, which contains some lines of A. Other Ptolemaic pieces of the third century have been grouped together. A few suggestions, based on Prendergast, may be made upon the lacunas. 129. |xomottnv for wodvtoitnv. Meaningless, since we have od[umoirns in 182. 138. |wa.a[. Perhaps stadyevov (as P 96) 8’ adaparra. mean a rearrangement of the names in 139, 140. This would 178. Read Naor apyevor Se Kat ws Kpatep|n TEp avayKn. 179a. Read ws tpwecow apnye tatnp Zevs| «nde § ayauous. 183a. Read A 502 i.e. xopanv bd erepoto Sia Kpotadoio| tepycer. 188a, instead of 189, emmopayov Bare Sovpt Kata Kpalrepny vomvr{v. The papyrus gives vpCuiwny|v; vousvn may conceivably be a real form; but Crénert’s note (if ‘ Cr.’= Crénert) that it is connected with the place ‘Tpyivy is extraordinary. The ‘rho’ is presumably a correction of Z absorbed into the text. In 188b mAevyov is as good as mvevpove. 1 Similarly Megara, which beat in vain at the the Hesiodic Catalogi fr, 96. 8. door of the Homeric Catalogue is encamped in ee " 7 —_—— > ge 5 a TI —___———. —_—_~, Gece. - - 2 T movec one— down whose a few untou V Paris Introd reduce does r refere! facsim scribe gr. I. almos: It is a bear < its gr evider as the NV two oO correc oppos: the fa hands beyon this p out to Q e note it was three aclean it not ration 1927; f the )-IQI, maic f the itains been unas. ‘OLTNS vould nf v. real acted 1ably is as — = ——_ i eS es nS cece eataags ee ae a gy : —- ~ = . MISCELLANEA 75 190a. Perhaps opvis ws medio Suwxwv opveov ad[ro from N 64, because there is no other case of a\Xo in their position. 1gla. jAvoel de yura. 250a. arr erlapvywpev pnd apBarropeOa epyor or the like in the tone of B 435. 254. nde x[ cove cOus Kovenv pepev «.t.r. The marginale |y I presume added vnov. That to the line above (].v] must also be a correction, perhaps @veAXav for OverAAny or vice versa. 255. Perhaps Oupov alredOivv0ev tpweoor Se xvdos orate. IlI. Varicano Greco I. This MS., for many years lost owing to the action of Cardinal Mai, who moved it from 796 to what he thought a better place, was being read by some- one—whose lot I, occupied among the horrible copies in which Homer has come down to us, envied—in May and June last. This must have been Mr. Post, whose paper on it in the last C.Q. interests me very much. I venture to make a few palaeographical remarks on it, leaving the philological argument untouched. Vat. gy. I. is not in the hand of the British Museum Lucian and the Paris Clement (a. 914), though Sir Edward Maunde Thompson says so in his Introduction, p. 222. Sir Edward had nothing before him but the unsatisfactory reduced facsimile in Cavalieri-Lietzmann. In its natural size Vat. gr. I. does not recall these two MSS. (with which I have long been familiar), and on reference to the full-sized facsimile of Lucian and Clement and the diminished facsimile of Plato I find nothing to contradict this impression. Hence the scribe Baanes and the implication of date (about A.D. 915) disappear. Vat. gr. 1. is merely a rather cursive hand of S. IX-X, and, as Mr. Post says, of almost (I should say quite) equal age with Paris 1807 and Clarke 39 (a. 895). It is an accident that the dated ninth-century minuscules are all stiff. They bear contemporary cursive hands in their margins. Moreover, Parts 1807 and its group cannot be dated with any greater precision, and therefore on the evidence of hand Paris 1807 may as well have been copied from Vat. gr. I. as the other way about. Mr. Post, however, couples the two Platos by a closer tie. He says that two omissions in Paris 1807 and one in Vat. gr. I. are supplied by the same corrector ‘as the minute identity of the handwriting proves.’ It is difficult to oppose such an emphatic statement, but when I look at the two passages in the facsimile of Paris 1807 (f. 201r, and 215r) and magnify all the marginal hands on the facsimile page of Vat. gr. I. with my eye I can see nothing beyond a resemblance of type. Marginalia, especially the small marginalia of this period, can rarely be identified, as Sir Edward Thompson himself pointed out to me once—the smaller the liker. On the other hand, I agree with Mr. Post that this minuscule marginal 76 MISCELLANEA hand in Vat. gr. I. is much older than was supposed. It is, in fact, contem- porary. So, naturally, is the small semi-uncial scholia hand. Even the minuscule hand which quotes the ‘ patriarch’s book’ and the ‘ antigrapha ’— on which the interest of Vat. gr. I. depends—is not late, and there is no reason to put it into the eleventh century. Between 800 and 1000 there must have been a great many Platos in the world; it seems unnecessary to force a direct connexion between the casual survivors. T. W. ALLEN. contem- ven the rapha ’"— ere is no Os in the he casual \LLEN. Tn he -_ Renee THE ‘EQUITES ILLUSTRES.’ THE question as to who were the class of Equites called ‘ Illustres,’ or in Greek émidaveis—as to whether, indeed, such a class existed apart from the ordinary ‘ Equites Romani ’—is a very puzzling one, and no agreement has yet been reached on the matter. The phrase has been used in all periods of Roman literature except the very earliest, but the disputed use is only that referring to the period from Augustus to M. Aurelius. It is used by Republican writers like Cicero,’ and in reference to Republican times by Livy,” but in these cases can be explained as either a quite general usage for ‘distinguished knights,’ or a transference of an idea belonging to Imperial times.* For the period of M. Aurelius and afterwards we have the definite classification of the Equites made by that Emperor and Verus.* But on the use of the phrase for the period between these two there is much disagreement. A. S. Wilkins, writing in Smith’s Dictionary of Antiquities,® denies the existence of a separate class known as ‘ Equites Illustres,’ but most scholars accept it, as we shall see. To prove that a separate class or category did exist, it seems sufficient to point to Tacitus’ use of the phrase ‘ ex illustribus,’® and to a similar use of trav émipavav by Dio Cassius.’ Both these imply a class. Moreover, Tacitus would hardly say that Senators and ‘ Equites Romani Illustres’ were forbidden to enter Egypt without the permission of the Emperor unless the phrase had some definite connotation.® Assuming, then, that the ‘ Equites Illustres’ were a special section of the Equites, what was their status, and why were they so called ? To this question scholars have replied in various ways, and their answers may be classified for convenience under three heads: I. The Traditional View.—The ‘ Equites Illustres’ were the same as the ‘ Laticlavii ’"—i.e. Knights of Senatorial rank preparing to enter on a Senatorial career. This view was first put forward by J. Lipsius, and was for a long time the accepted one. Belot, for instance, in his Histoire des Chevaliers Romatns,° and Marquardt in Historia Equitum Romanorum,” both accept it. It is also the view of Cagnat," Macleane,!* and Cornish.** The main support for the 1 E.g. Verr. II. 3. 24; III. 24. 60: Philipp. 7 LIX. ro. 8 Ann. II. 59. 4. XIII. 13, etc. ® Vol. II., p. 404 sq. (Paris, 1873). 2 E.g. XXIX. 34. 17; XXX. 18. 15, etc. 10 Chap. IV. 3 Cf. Ramsay, Manual of Roman Antiquities, s.v. 11 In Daremberg et Saglio, s.v. Equites. 4 Cf. Friedlander, Sitteng. I., p. 292 (Leipzig, 12 On Juvenal X. 95. 19I0). 13 Concise Dict. of Greek and Roman Antiquities, 5 S.v. ‘ Equites.’ 6 Ann. IV. 58. 1. S.v. 78 H. HILL theory rests on two passages of Statius,’ one of Josephus,? and one of Isidorus;* and the best exposition of it is to be found in the work of Belot. II. Mommsen attacked this theory and propounded a new one.* This was that the ‘ Equites Illustres’ were those holding high Equestrian office, such as the four great ‘ Praefecturae,’ and that the title was synonymous with ‘ Splendidi,’ ‘ Primores,’ etc. His view receives support from Kiibler® and from Bury.° III. ‘ Maecenas’ Theory.—More recently a third theory has found favour with scholars. It is that the ‘ Equites Illustres’ were those Knights who, like Maecenas, were qualified by birth and wealth to enter the Senatorial Order but preferred to remain simple Knights. This is the view of Madvig,’ Sandys,® Ramsay,’ Furneaux,’ and Spooner," though there are some slight variations in each case. The most recent writer on the subject, Dr. Arthur Stein,’* comes to a purely negative conclusion. He rejects the first view, suspends judgment on the theory of Mommsen, and does not even mention the third. Of the three theories none seems to me to be unassailable, and all alike suffer from one fundamental weakness. The main arguments in each case are founded on passages of Greek or Roman writers in which the phrase ‘ Equites Illustres ’ (or tmwzeis émipaveis) does not occur! Thus the supporters of the Traditional Theory appeal to Statius IV. 5 and V. 2. 12, and Mommsen to Tacitus, Annals XVI. 17, and Agricola 4, none of which passages contains the phrase ; while Madvig and his followers refer to Maecenas, who is nowhere called ‘ Eques Illustris.’ Let us examine each theory in turn and test its value: I. The Traditional Theory need not detain us long, because Mommsen has already disposed of it.* The references given by Belot prove nothing. They merely concern the Senatorial class of Equites elsewhere called ‘ Laticlavii,’ but in no way connect these with the ‘ Equites Illustres.’ Even the basic assumption of the theory will not bear investigation, since none of the men called by this title can be proved clearly to have belonged to the Senatorial Order, and some, such as e.g. Ti. Iulius Alexander, a Jew, were plainly not eligible for that rank.™* II. Mommsen, while taking others to task for irrelevancy, himself fell into the same error, for he classes expressions like ‘ Primores equestris gradus,’ etc., with ‘ Equites Splendidi’ and ‘ Equites Illustres.’ It is well known, of course, that ‘ Splendidus ’ was a general epithet very commonly applied to Knights,” and it seems clear that the other expressions Mommsen quotes are either 1 IV. 5, and V. 2. 12. 10 Tac, Ann., Vol. I., Introd., p. 103 (1896 ed.). 2 Ant, XIX. 1. 3. 3 Orig, IX. 4. 12. 11 Note on Tac, Hist. 1. 4 (1891 ed.). 4 Staats. ILI.1 563 (1888 ed.) ; French Tr. VI.,? 12 Der voimische Ritterstand (Munich, 1927), p. 173. pp. 101 sqq. 5 In Pauly-Wiss., s.v. Equites. 13 Loc. cit., and cp. Pauly-Wissowa, s.v. 6 Student's Roman Empire, p. 43. 14 Tac, Ann. XV. 28. 4. ? Verfass. und Verwalt. d. r.S.1., p. 170. 15 See Cicero, e.g. passim. 8 Companion to Latin Studies. 9 OP. cit. gene Mor Agn en bl grav agai! Mor must entit Knig some of as Had: and | Alex: to ar ‘ Pri as ‘S$ those * Spl posit of Ti were more but, ‘Tra ‘Lat Thus some An € biog calle ‘Eq men eo wo = Sar 1 one of selot. This was 2, such as plendidi,’ id favour who, like al Order Sandys,® ariations nes to a udgment all alike case are Equites s of the msen to ains the nowhere isen has They ticlavii,’ 1e basic he men natorial nly not fell into is,’ etc., course, ights, - either 1896 ed.). 1, 1927), THE ‘EQUITES ILLUSTRES'’ 79 general or very different from the phrase under discussion. In particular, Mommsen quotes Annals XVI. 17—‘ Equites dignitate senatoria ’—and Agricola 4, which concerns ‘ Procuratores Augusti,’ who cannot be included en bloc as holding the highest Equestrian office. But apart from this lack of clarity in its exposition, there are several grave objections to Mommsen’s theory. In the first place, the very rarity of the phrase ‘ Equites Illustres’ is against it. If we regard the most prominent Equites, with Dio Cassius* and Mommsen himself, as those holding the four great Equestrian Prefectures, we must presume that, if this theory be correct, a very large number of men were entitled to be called ‘Illustres.’ One would expect, for instance, that important Knights like Cornelius Gallus, Sejanus, Tigellinus, Macro, Burrus, etc., would somewhere have this title. But this is not so. In order to test Mommsen’s theory, the writer has collected the names of as many as possible of these ‘ Praefecti’ from the time of Augustus to that of Hadrian, and examined a very large number of references to them in Latin and Greek authors and Inscriptions. In no single case, except that of Tiberius Alexander (with whom I shall deal later), was the epithet ‘ Illustris’ applied to any of them. The most prominent Equites are usually referred to collectively as ‘Primores Equitum,’? ‘ Principes Equestris Ordinis,’* etc., and individually as ‘ Splendidissimus,’ or simply ‘ Splendidus.’ Inscriptions make it clear that those who held the highest Equestrian offices were generally called ‘ Splendidissimi.’* In addition to this negative evidence against the theory, there is the positive fact that none of the men called ‘ Illustres,’ with the exception again of Tiberius Alexander, seems to have held high office. For the most part they were rather obscure, and several times we are not even given their names. III. The ‘ Maecenas’ Theory, though strongly supported, is, if anything, more vulnerable than the last. It is open to several very serious criticisms ; but, first, it may be well to point out how far it is from the theory I have called ‘Traditional.’ Lipsius said that ‘Equites Illustres’ was synonymous with ‘ Laticlavii.’ Maecenas was, as Velleius tells us, ‘angusti clavi fine contentus.’® Thus the two views are diametrically opposed. If this theory were true, we should expect to find the epithet ‘ Illustris ’ somewhere applied either to Maecenas himself or to someone like Maecenas. An examination. of passages referring to Maecenas quoted in dictionaries of biography, editor’s commentaries, etc., has convinced me that he is nowhere called ‘Illustris’ or émidavjs, but merely ‘Eques Romanus,’ ‘ Eques,’ or ‘Equestris Ordinis,’ etc., and in Greek iamevs or avnp immevs, etc. Several men definitely compared with Maecenas or resembling him in their careers are 1 LIT. 24. 32; and cp. Dio Cass. LVI. 42; LVII. 11, etc. * See e.g. C.1.L. V. 875 al. 5 Vell, Pat. 11. 88. 2 (reading of Heinsius). 2 Tac. Hist. I. 4; Suet. Aug. 100, etc. 3 Vell. Pat. 11, 127. 3; Suet. Aug. 66; Juv. IV. 80 H. HILL mentioned by classical writers. Such are Sallustius Crispus,! Annaeus Mela,* L. Vestinus,® and Minicius Macrinus,* to name only a few. Other Knights again, such as Septimius Severus,° Geminius,® Samius,’ and the poet Persius,® men of enormous wealth, must have been eligible for the Senate if they had wished to enter it. All these, then, according to this theory, were entitled to be called ‘Illustres.’ Yet none of them receives the title in any extant writings. Moreover, none of the men mentioned as ‘ Equites Illustres’ would fit into this category, since most appear, as has been said before, to be men of little influence; and one at least—Tiberius Alexander—held the high office of ‘ Praefectus Aegypti.’ Finally, it is extremely unlikely that a title, and that an honourable one, would be conferred on men for shirking their responsibilities. And if the qualifications of an ‘ Eques Illustris’ were so purely negative, it would surely be extremely difficult to define the category. Who, then, were the ‘ Equites Illustres’? The only way to answer the question is to examine the passages in which the phrase is used, and try to draw some conclusions from them. They are very few, and the following is, to the best of my knowledge, a complete list :° Tacitus, Annals— II. 59: Senators and ‘ Equites Romani Illustres’ forbidden by Augustus to enter Egypt without permission. IV. 58: With Tiberius to Capua went ‘Eques Romanus praeter Seianum ex illustribus Curtius Atticus.’ (Note the order of the words, which seems to exclude Sejanus from the ‘ Illustres.’) IV. 68: Titius Sabinus, ‘ illustris eques Romanus,’ dragged off to prison ‘ob amicitiam Germanici.’ VI. 18: Pompeia Macrina, a descendant of Theophanes of Mytilene, exiled. She was the daughter of an ‘illustris eques,’ whose name was apparently Pompeius Macer. (See Strabo XIII. 2. 3, p. 617.) XI. 4: ‘Equites Romani illustres quibus Petra cognomentum’ brought to trial on false charges—really for conniving at meetings of Mnester and Poppaea Sabina. XI. 35: ‘Illustres equites Romani,’ involved in intrigues of Silius and Messalina, meet death bravely. (N.B.—It is doubtful whether the names which follow are those of the ‘ Equites Illustres,’ but if they are they give us no assistance.) XV. 28: Tiberius Alexander (later ‘ Praefectus Aegypti’ and a supporter of Vespasian. See Tac. Hist.), sent as ‘minister bello’ to Corbulo in his negotiations in the East. He is called ‘illustris eques Romanus.’ The phrase is not found elsewhere in Tacitus, nor in any other Roman author writing in or of this period. 6 Tac. Ann, VI. 14. 1. 8 Suet. Vit. Pers. ® This list is, I find, identical with that given by A. Stein, op. cit., p. tor, note 4. 1 Tac. Ann. III. 30. 2 Ib. Id. XVI. 17. 7 Id, 1b. XI. 5. 2. 3 Id, Hist. 1V. 53 ; and Spooner, ad loc. 4 Plin. Ep. I. 14. 5. 5 Stat. Sev. IV. 5. unde will most state alluc hous of G was of hi Illus: ment The is th obvic in tk ‘ Min Emp the t high | to see the fr they | of Ec oppo! quite appli comitr I in cet 1 Stra us Mela,? ’ Knights Persius,® they had ntitled to writings. would fit » be men igh office able one, nd if the ld surely in which They are complete Augustus Seianum is, which 0 prison M ytilene, se name 617.) brought Mnester ilius and ather the t if they jupporter Corbulo is eques r Roman ib. XI. 5. 2. 1) that given THE ‘EQUITES ILLUSTRES'’ 81 Dio Cassius (im meds émedavys)— XLV. 10: A Republican reference, probably quite general, but may be a transference. LIV. 2: xai immeis xai yuvaixes érmipaveis prevented from appearing in the arena. (The epithet may be applied only to yuvaixes.) LVII. 14: Drusus flogged imzevs émigpavys, and was given the nickname of ‘ Castor ’—no doubt a reference to the patron god of the Equites and a play on the word ‘ castigare.’ LIX. 10: Knight trav émipavavy (unnamed) compelled to fight in the arena by Gaius for having insulted Agrippina. I have not noticed any other use of the phrase referring to the period under discussion. It is not used in inscriptions until very much later. From this list it can be seen how comparatively rare the phrase is. It will be noticed, too, that it is used only by Tacitus and Dio Cassius, the two most scrupulously accurate and detailed historians of our period. A careful scrutiny of the above extracts reveals one feature either clearly stated in or deducible from practically all of them. This is that the men alluded to by this title were closely connected with the Emperor himself or his household—that is, they all belonged to the Imperial ‘ Court.’ Curtius Atticus was a friend and ‘comes’ of Tiberius, and Titius Sabinus of Germanicus. We know nothing of Pompeius Macer himself, but his father was a personal friend of Tiberius,’ and Augustus entrusted to him the revision of his library. This argues a Court upbringing for his son. The ‘ Equites Illustres,’ called Petra (of whom nothing else is known), and those others mentioned in Amnals XI. 35, were concerned in important Court intrigues. The case of Tiberius Alexander is equally clear. This passage of the Annals is the only one in which he receives the title of ‘ Eques Illustris,’ and it obviously refers to his earlier years, since we find him as ‘ Praefectus Aegypti ’ in the Year of the Four Emperors.® In our passage he is acting as the ‘ Minister’ of Nero in the East—that is, as the personal representative of the Emperor. This again argues close connexion with the Court, and no doubt the title ‘ Eques IIllustris’ is used because the youth has not yet risen to any high Equestrian office. Is it, by the way, too great a stretch of the imagination to see in the subsequent promotion of Tiberius the hand of Poppaea Sabina, the friend of Judaism ? The remainder of the extracts quoted do not give the names of the Knights they refer to, but all except one have connexion with the Court. The exclusion of Equites from Egypt has more point if we apply it to those Knights who had opportunities of intriguing against the Emperor. Dio Cassius LVII. 14 is quite indecisive ; but in LIX. 10 it is very likely that the phrase ray émipavav applies to a courtier, since such a man would have more opportunities of coming into contact with Agrippina. I would suggest, then, that ‘ Eques Illustris’ was an informal title given in ceremonial language to courtiers of Equestrian rank. That is to say, any 1 Strabo XIII. 2. 3, p. 617. 2 Suet. Caes. 56. 3 See Tac. Hist., and Josephus, saepe. F 82 THE ‘ EQUITES ILLUSTRES’ Knight taken into the inner circle of the Emperor’s friends and holding no State office was unofficially described as ‘ Illustris.’ This interpretation, I submit, not only fits the facts as we have them, but also disposes of the difficulties which we have seen prevented the acceptance of the other theories I have enumerated. It explains the rarity of the expres- sion, since the title belonged rather to the language of the Court than to literary language. It also explains why people like Sejanus and Burrus, intimate friends of the Emperor, did not receive the title. They were already designated as ‘ Praefecti,’ etc. This theory also receives support from two other sources. The Younger Pliny tells us that the friendship of the Emperor might excuse a man from holding State offices. This will explain why the ‘ Equites Illustres’ mentioned were not men of great prominence in public life. Again, we are told that when M. Aurelius and Verus reorganized the Equestrian Order, they used the word ‘TIllustris’ to describe those Knights who were of high rank without holding any official position. This was only a rather wider application of the word. One question may be asked in considering this theory. Why were Maecenas and Proculeius called és ta wpa@ta i710 tod ’Avyovotou Tipwpevos? and holding no official position, not called ‘ Illustres’? The answer is that they were friends of Augustus in the early years of his Principate when no Court existed. Apropos of this use of the word ‘ Illustris’ (or émepav7s), it is interesting to note that the words have often, even when used with words other than ‘ Eques,’ some connexion with the Imperial Court. Thus the phrases ‘illustris femina ’ or ‘ illustris vir,’ though they have, of course, a general application too, might often be translated ‘ lady (or gentleman) of the Court.’* The same is true also of éaidavys.© It may be that these epithets were used informally in referring to courtiers, but our information is too scanty and the words too common in other usages for any definite assertion to be made to this effect. As to the duties and privileges of these ‘ Equites Illustres’ it is impossible with the few facts at our disposal to draw any safe conclusions. Their importance must have varied considerably in different reigns. We know, for instance, that the Emperor often made use of his friends of both Orders to advise him on State or personal matters; and we have had one case of an ‘Eques Illustris’ as the Emperor’s personal agent. They may have been members of the various Imperial ‘ Consilia,’ though here again we have no definite information. But whatever their insignificance in public life, we cannot doubt that they often played an important part behind the scenes. H. HIv. UNIVERSITY COLLEGE, SWANSEA, 1 Ep. IV. 24. 2 Cf. Friedlander, of. cit. XV. 37. 6, al. ; Suet. Tid. 45; id, Calig. 36, etc. 3 Dio Cass. LIV. 3. 6 Dio Cass. LV. 10; LX. 32; LXI. 20; + See e.g. Tac. Ann. III. 49. 2; XI. 36. 3; LXII. 24, al. thes« figur unsk pass: toxe use. piece whic conn pass espe scale of tl §§ 1 TOU | dvateo yeveo TO pb TOTE TO Kd CavT TOV apo TOUTE (111) elvat (112) (113) Tépou To be 1 O1 Aristic interv: quarte some ¢ howev iolding no them, but .cceptance he expres- t than to friends of gnated as > Younger man from nentioned nized the e Knights was only Vhy were “Lp@pevot,® er is that when no resting to | Eques,’ s femina ’ oo, might ; true also | referring ymmon in m possible s. Their know, for Irders to ase of an ave been have no life, we nes. HILL. lig. 36, etc. LXI, 20; THE SPONDEION SCALE. PSEUDO-PLUTARCH DE MUSICA, 1134F-II35B AND II37B-D. Our information about the early stages .of Greek music is so slight that these references of the Pseudo-Plutarch to a scale employed by the legendary figure Olympus take on an immense value for us. The dialogue itself is an unskilful patchwork, but the author’s sources are often good. These particular passages are almost certainly both derived with small alteration from Aris- toxenus, in whose time the traditional music ascribed to Olympus was still in use. For the elucidation of the scale’s history and structure we have three? pieces of evidence to help us. There is the first passage mentioned above, which deals with the discovery of the enharmonic genus by Olympus, which is connected, at first sight obscurely, with the Spondeion scale; and the second passage, which discusses the scalar limitations of the orovdedfwv tpozos, with especial reference to an elementary polyphony. There are also the ancient scales quoted by Aristides Quintilianus (ed. Meibom, p. 21). Our best starting-point will be Plutarch’s account (1134F) of the origin of the enharmonic, which runs as follows (in Weil and Reinach’s edition, §§ 104-117): (104) “Orvytros 82, as ’Apsotokevos now, vrodkapBavetar UO TOV povotKav Tov évappoviov yévous evpetns yeyevicOar* (105) Ta yap mpd éxeivov tavta Sidtova Kai ypwyatixna Hv. (106) brovoodc. 5é tHv eipeow ToravTny Twa yevéoOat. (107) avaatpedopevoy tov "OdvyTov év r@ diaTov@ Kai diaBsBalovta TO péros TodraKis ert THy Sidtovoy TapvTdtny, TOTE wey ATO THS Tapapéons ToTé 5é aro THS wéons, Kal TapaBaivovta thy Siadtovov Atyavoy, (108) catapabeiv TO KaAXOS TOU HOoUS, Kal oUTW TO ex THS avaroyias cUVETTHKOS GVETHMA Bavpmda- gavta Kal arodeEduevov, év TovT@ troeiv érri tod Awpiov rovov. (109) ovTEe yap tav tov Siatovou idiwy ote Tay TOU ypwpaTros amTecOa, AdXr’ ovde THY TIS dppovias. (110) elvas 8 a’t@ Ta Tpa@Ta TaY évappoviwy To.adTa* TiWéacte yap TOUTwY TMpaTov TO orovdeciov, év @ ovdepia Tav Siatpécewy TO idiov eudaiver, (111) ef pH TIs eis TOY GUYTOVwWTEpOY oTOVdeLacLOY BréTwWV avTO TovTO SudTovoY elvar amresnaon: Shrov 8 dre nal wevdos nal experes Onoes 6 TowodTo TiBeis: (112) Weddos perv, Sti Siécer EXaTTOV éoTe TOvoU TOD TrEpl TOV HryEemova KELpévoOU ° (113) éxperes Se, Ott, nal ef Tis ev TH TOD Toviaiou Suvaper Tein TO TOD TUYTOVO- tépov orrovderac mod idiov, cvpBaivor dv dvo é€Rs TiWec Oat Sirova, 76 pev aovvOeror, 7d 58 civOerov. (114) Ta pev ody TpaTa Tov évappoviwy ToLadTa* (115) TO yap 1 Or four, if we include the reference of for it was the passage of Aristoxenus, that Aristides Quintilianus (p. 28, Meibom) to an Plutarch himself makes use of. Aristides prob- interval grovdeacuds, consisting of a riseof three- ably was in possession of books of Aristoxenus quarters of a tone. This must certainly have which are lost to us, and very possibly culled some connexion with the Spondeion scale. It is, from one of them his list of ancient scales (cf. however, possible that Aristides’ only authority J. F. Mountford in C.Q., Vol, XVIL.). 84 R. P. WINNINGTON-INGRAM év tais pécats évapuovioy TruKvov, © viv xXpavrat, ov Soxet Tod wonrod elvas. padiov & dori cuvideiv, dav tis d4pyate@s Twos avAODYTOS Gkovan* acivGeToOY yap Bovrerat elvar nai Td év tais pécais Hysroviovs (116) dorepoy dé To HpsToOvioy SinpéOn ev te trois Avdiots nai év trois Dpvyios. (117) paiverae 8 “Odvpmtros avéncas povotkhy TO ayévyntov Tt Kal ayvoovpevoy bd Ta EuTpocbey cicayayeiy, nal apynyos yéverOar ris “EXAnuixns Kai Kardis povotnis. The text of this passage is fairly sound, and, although we may have doubts about the exact order of the sentences, the meaning is fairly clear. On the grounds of the last sentence Aristoxenus or Plutarch has been accused of muddle-headedness. This is quite unfair. It is true that Olympus according to this account did not introduce the quarter-tones of the enharmonic genus, but it is equally true that these quarter-tones were not the only characteristic of the enharmonic; the omission (hypothetical) of the G and the resulting Sirovos Avyavos were an essential preliminary to the splitting of the semitone, which may originally have been in the nature of an ornament. We need not, of course, believe any actual musician to have omitted a note in the diatonic series. Olympus, as we have seen, was a legendary figure; the enharmonic was no accidental discovery. The tradition amounts to this—that Greeks at some point before accurate musical history begins came in contact with a type of music which used the dirovos acvvOeros, and adopted it and the quarter-tones, which were an elaboration of it. Plutarch calls it ‘ Hellenic and beautiful.’ On any theory it is curious that music, regarded as the invention of Olympus and subsequent to the diatonic, should be called ‘ Hellenic.’ It is accounted for by the later popularity of this style, and particularly by Aristoxenus’ glorification of it. The gapped scale, then, which the enharmonic implies, was a real contrast to a completer series of diatonic notes which we can assume to have been pre-existent in Greece, and ‘Olympus’ really did enrich Greek music with a new element. The legend of the manner of his discovery is a picturesque fiction emphasizing that contrast; we need not pursue its details so far as to ask in what exact scale, mode, or key Olympus was playing when he made his great ‘ discovery.’ We are told, however, that he constructed a Dorian scale é« rijs dvadoyias. This can only mean that the later Greeks were aware of an old scale of this sort: EF ABC (E). ‘Such,’ continues our author, ‘ was the character of the first enharmonic compositions of Olympus, for they count as the first of them the Spondeion, in which none of the (tetrachordal) divisions manifests its peculiarities.’ The otovoeiov, then, bore some relation, possibly of identity, to the earliest enharmonic music.’ Further, it contained no notes characteristic of any of the genera—that is, it had neither the diatonic lichanos and Sitrovos civberos, nor, at least as yet, the wuxvov. Weil and Reinach (§ 110, n.) quote a number of passages to prove that it was a Dorian scale. This seems very probable, in which case it must have been very similar to the scale written out a few lines 1 Probably it was the name given to the wise knownasraAdpa. We know that Phrygian Dorian varieties of the early enharmonic, other- and Lydian varieties also existed. See p. go n. above There name looki (cvv0 accor the o that 1 this | sugge authc cano! an ol ap. N dtécet EF A passa hepta stren; than of so eithes to thi diets oTrove by a absur Pyth: mista possil surel} purel Pluta \ ment: limite It is < if we and vu Thre the 7 1 Th it is ra 2 Gr the sa ov elvas. feTov ‘yap 1) [Lt TOVLOV Orupros rayayety, e doubts 1as been Ilympus armonic he only >» G and splitting nament. . note in Ire; the is—that contact and the nic and vention lic.” It arly by irmonic hich we lly did of his ed not lympus or, that hat the rmonic ion, in ' The sarliest any of vOeros, umber ble, in v lines hrygian >. go n. THE SPONDEION SCALE 85 above. Further information, however, is given about it in the Plutarch text. There was, he says, in it an element that might be described as diatonic— namely, ‘the higher spondeiasmus’; this, however, was an erroneous way of looking at it, and implied an unmelodious succession of intervals—a ditone (cvvOerov) followed by a ditone (dcvvOerov). This is of course impossible according to the principles of Aristoxenus ; and it is carefully explained that the ozovderacuds interval is not a tone, but three-quarters of a tone, smaller, that is, than the disjunctive tone which precedes it. We get, then, a scale of this sort:!' EF AB C*. The look of it not being liked by editors, it is suggested that Aristoxenus, who can almost certainly be accounted the actual author of the sentences in question (the argument proceeds according to the canons laid down in his Harmonics, Bk. III.), has made a mistake, interpreting an old orrovdevac os of three semitones (the Pythagorean Sdieous, see Philolaus, ap. Nic. 9; Adrastus, ap. Theo., p. 55, Hiller) as though it were of three modern d:éoecs, or quarter-tones.2, When it was seen that this alteration gave a scale EF AB D(E) not only closely resembling the tpdzos ozrovderafwv of the later passage in the dialogue, as usually interpreted, but also the supposed defective heptachords of Terpander and Philolaus, belief in this hypothesis was much strengthened. We shall see that there is no more reason to believe in the one than in the other. Apart from that, however, we should not accuse Aristoxenus of so grave a mistake without good reason. This theory leads us to assume either that the interval known as omovderacpos changed from a tone and a half to three-quarters of a tone concurrently with the change of usage of the word Sieous, which is absurd ; or that there never was an interval ovovdevacpos in the arrovéetov scale, but merely one of three semitones, which Aristoxenus mistook by a verbal mistake for the later omovderacyuos that he knew, which is also absurd; or else that the only ozovderacpes there ever was was of three Pythagorean d:écecs, and that Aristides (p. 28, Meibom) merely perpetuated the mistake of Aristoxenus and described a never existing interval, which is just possible, but unlikely. Aristoxenus in the Plutarch passage is describing, surely, a scale he knew, not one he had laboriously reconstructed from a purely verbal tradition. We must then believe in the scale as given by Plutarch in default of some valid reason for rejecting it. We come next to the tpozros o7rovderdfwv. Various notes of this scale are mentioned (Plut., § 172, etc.) in an attempt to prove that the ancients had limited their art by omitting from their scales notes they might have employed. It is a curious way of putting it, but not so erroneous as is sometimes suggested if we believe that the complete diatonic series was known comparatively early, and used concurrently with less complete scales of possibly non-Hellenic origin. Three notes in particular, according to this passage, were not employed in the tpdmos orovderdfwv—namely, tpitn, vntn, and ‘ tpitn cuvnppévov.’ (The 1 The symbol * after a note here indicates that (D) EF AB) C$D. But he assumes, further, it is raised by a quarter of a tone. that Aristoxenus was completely misinformed 2 Greif also (Rév. des Et, G., 1910) makes about the scale, and imported the disjunctive the same assumptions to produce the scale tone into his account. 86 R. P. WINNINGTON-INGRAM erroneous text of Weil and Reinach will provide a convenient starting-point for our argument.) However, though they were not employed in the melody, they were employed in the accompaniment in the following manner: tpirn consonantly with tapurarn ; vyrn dissonantly with mapavyrn and consonantly with péon and tapapéon; tpirn cuvnppévwv dissonantly with wapavyrn and Axavos (again adopting for the moment the readings of Weil and Reinach). According to their interpretation we now have a scale with the following scheme : péros - - EFGAB D Cc 6 fs Kpovats - - This is naturally brought into relation with the supposed defective scale of Philolaus, with its tpinusroviov dovvOerov, and with the ozovdeiov of the earlier Plutarch passage, emended to correct the supposed mistake of Aristoxenus and endowed with a diatonic lichanos. (For the moment we can disregard the notes of the accompaniment ; the rest of the scale existed before they were employed, which cannot have been earlier than the time of Archilochus.) We have seen that there is no good reason to believe in this ‘howler’ of Aristoxenus. There is no more reason for supposing that the oovdeiov ever had a diatonic lichanos. In the earlier passage we are told that the ozrovdefopr of Olympus was an enharmonic of the early type which had no quarter-tones— such were the first enharmonics; later the semitone was divided. This is the implication. It is, then, natural to suppose that the ozovéeiov remained enharmonic, and that Aristoxenus (if Aristoxenus it is) in referring to its notes refers to the notes of the enharmonic genus, not the diatonic. This is the natural assumption in the absence of specific instructions to the contrary. The notes of the melody then become EE*F ABC; those of the accom- paniment, B* and E (‘ tpirn cvvnupuévwv’ shall be left for separate and later consideration). This interpretation, besides being logical, has certain striking advantages and confirmations. In the first place, the scale it gives corresponds very closely with the scale of Olympus’ enharmonic, as deduced from the Plutarch account, differing from it only in the split semitone of the lower tetrachord. It also corresponds with the early ozovéeiov discussed in the same passage, except that there is no suggestion that the B-C interval was widened to a three-quarter tone. But above all it simplifies the account here given of the polyphony of the ozrovéeafwv tpoTos. Let us take the three accompanying notes one by one. First, we are told (172) that the ancients abstained from the note rpirn, the omission of which contributed to the characteristic effect of the style. In the accompaniment, however, it was used consonantly with trapurdrn. Of course, the tpirn in any genus would always form the consonance of the fifth with the wapumdrn. There is no need here to mark a lacuna with Weil and Reinach, as there is no reason why Plutarch should necessarily have given more than one example; further, on our interpretation there is no other note with \ or dis splitti: There was n that t sumab a forti implie semitc S not fc additi used i to it. this \v practi we in discor polyp’ for it Greek conso unnec as we I must Philo! Philo Script termi whict B, nc but g origir and t Tapa Tapa, the te Philo to act a sca 1 St reserv ng-point melody, Pr: TpiT sonantly mT and teinach), ollowing scale of e earlier stoxenus yard the ley were lochus.) vler’ of lov ever rovoOetoy tones— ; is the mained 'S notes is the ntrary. accom- d later ntages S very utarch chord. issage, l toa of the inying TpiTn, . In . OF : fifth il and given " note THE SPONDEION SCALE 87 with which the pecoruxvov, tpirn, would form a likely interval, consonant or dissonant. Apart from the question of polyphony, the absence of the splitting of the semitone becomes most interesting when we turn to § 115. There we are told that the splitting of the semitone in the tetrachord pécov was not the work of Olympus, implying that this was the first step; further, that the old-fashioned style of flute-playing in the day of Aristoxenus (pre- sumably) tended to leave even the semitone in the tetrachord pécwy undivided— a fortiori, that in the tetrachord Ssefevypévwv. Here in section 172 we have implied a Spondeion scale,} in which only the lower tetrachord has its semitone divided ! Sections 175-177 deal with vyrn. All accounts agree that this note did not form part of the early melodic compass, and most of them ascribe its addition to Terpander. It is not surprising, therefore, to find that it was not used in the Spondeion melodies, when even the diatonic scale had not attained to it. When the ‘ Dorian vyrn’2 was added, it was not inserted ruthlessly in this well-established style of music, but used when polyphony came into practice in two ways—dissonantly with apavyjrn, consonantly with péon. If we interpret mapavyrn as the enharmonic wapavytn (=C) we abolish that discordant D-E interval, which has always made the orthodox theory of polyphony, if not slightly ridiculous, at least incomprehensible, and substitute for it an interval, the major third, just such as we should have expected the Greeks to employ if once they had wandered beyond those they recognized as consonant. The concordant use with uéon (E-A) needs no comment. It is unnecessary to add with Weil and Reinach xal mpos tapapéonv from 179, as we shall see when we examine the following sections. Before proceeding to the last of the accompanying notes, however, we must digress to discuss shortly the defective scales ascribed to Terpander and Philolaus, which can be closely related to the Spondeion. We will deal with Philolaus first. Nicomachus, in the ninth chapter of his Enchetridion (Mustct Scriptores Graeci, Teubner), quotes from him to illustrate the old Pythagorean terminology. He is describing the well-known harmonic framework (dppovia), which can be represented by the notes E-A-B-E. Only he calls the note B, not wapapéon, but tpitn. Howisthis? Nicomachus is at pains to explain, but gets himself into a terrible muddle in doing so. He has to imagine an original heptachord EFGAB DE, different from that he assumes in Chapter XI., and that the disjunctive tone (rod dsafevyvivros révov) was inserted between mapavntn (D) and tpirn (B)! Tpirn thereupon suffered a change of name to mapayéon. Except for the last statement this is pure invention to account for the terminology of Philolaus, and took this form because Nicomachus imagined Philolaus was dealing with a normal diatonic scale. We in our turn now have to account for the Pythagorean’s use of tpirn. It is true it can only apply to a scale in which a single note separated it from v#jrn. Is the series more likely 1 Strictly the term ‘Spondeion’ should be This is its development, the crovded{wr rpdros. reserved for the scale without quarter-tones. 2 Section 270. 88 R. P. WINNINGTON-INGRAM to have been B-D-E or B-C-E? In view of what we have seen of the old style (the apyasxol rporoe of Aristoxenus, p. 23), we can have no doubt that Philolaus was using the nomenclature of a scale in which vjrn (E) had been added to the old Spondeion without the division of the semitone B-C. B-C-E is not the tetrachord of later music, but the trichord of the archaic style (cf. Plutarch, § 171). We may now turn to the Aristotelian problems for an account of the work of Terpander. Four problems of the nineteenth book deal with the hepta- chord, the octachord, and the position of uéon in them. Problem 47 takes what we may describe as the orthodox view—that the heptachord was merely the octachord without the disjunctive tone, and that its uwéon was the lowest note of its upper tetrachord. Problem 44 implies the same point of view. Problems 7 and 32, however, in answering their respective questions, refer to an omission of the note rpitn, and the second of them associates this definitely with Terpander. Gevaert* was probably right in regarding these references as interpolations ; he was entirely wrong, however, in ascribing them to followers of _Nicomachus and interpreting them diatonically. They must have been written by someone who knew the old Spondeion scale, and that in it the note tpitn, the weodrrucvoy, was not regularly employed, who knew also the tradition that Terpander had added the vjtn. He therefore conceived that the innovator, finding a scale EE* F ABB*C (a kind of heptachord), had suppressed the TpiTn (B*) in order to add the vytn (E). ~ As an historical procedure this is of course improbable. It was not the addition of vyrn that caused the omission of tpitn; but the Plutarch passage is borne out in an interesting manner, and we can even believe that vjrn came to be used with the old Spondeion before the splitting of the upper semitone. Gevaert objected that the scale in question must be diatonic, not enharmonic, because otherwise the hypothetical omission of vqTn would leave a scale bounded not by a seventh, but by a sixth. But it is just such a scale that the interpolator contemplates, similar in compass to the old Spondeion and to the Syntonolydian of Aristides Quintilianus. Nor need we be distressed by the variation in the use of tpirn between Philolaus and the Problems. Before the semitone was divided and the pycnon came into being, the nomenclature preserved by Philolaus may well have been used; the intro- duction of the wecoruxvoy obviously renders a change necessary. Nicomachus was thus right on at least one point. We can now return to a consideration of the polyphony of the ozovierafwv tporos. A third suppressed note is there dealt with. The MSS. are unanimous in reading cuvnupévwv vyrn (or some minor variant). Weil and Reinach, thanks to their interpretation of the nomenclature as diatonic, felt themselves com- pelled to emend vy#rn to rpirn, thus landing themselves in hopeless difficulties. On their theory, of course, vjtn curvnupévor is the same note as the rapavyrn, already mentioned as a melodic note; also, it is not dissonant with the lichanos (D-G), as by § 179 it should be. But what is the rpirn cuvnppéevov 1 Problémes Musicaux @ Arvistote, Gevaert et Voligraf. doing in but not as acco! of what argued | Phrygia oun pLpe a modul Phrygia turn to Dorian, DEF A term Tra which v It was Mexaves to § 176¢ is needl Tapape ynTH ar sixth (I have be just the of the Tapavn manusc It is h became by an eee as the worth \ hesitate employ St scale. were él interva tone. would in the | we ha OTrOVvOE scale. of the old loubt that had been . B-C-E haic style the work he hepta- 47 takes is merely he lowest of view. refer to lefinitely rences as followers ive been the note tradition inovator, ssed the this is of mission ner, and n before juestion mission But it ipass to s. Nor aus and ) being, > intro- machus eration nous in thanks $; com- sulties. aVNTN, th the L[LEV@V THE SPONDEION SCALE 89 doing in this scale at all? The Aeolian used it and perhaps the Mixolydian, but not the Dorian, to which this scale is related. Secondly, it is made to act as accompaniment to a note (mapavjrn) higher than itself in contravention of what must have been a universal law (Ar. Prob. 19, 12). Thirdly, it is argued in the text that the early knowledge of this note was proved by the Phrygian melodies (ra Dpvyia). Now what possible part can the note rtpirn cuvnppévwv have had in a Phrygian scale? Weil and Reinach’s suggestion of a modulation into the Aeolian is very weak. Yet it is this very mention of the Phrygian that gives us the clue and bids us restore the MS. reading. If we turn to the Phrygian of Aristides we see immediately that it gives us the Dorian, with the substitution of D for E; purged of quarter-tones, it is DEF ABCD. It is this upper D (called here vijtn cvvnppévwv because the term tapavyrn is already occupied by the enharmonic note of that description) which was used at a later stage in the accompaniment of the old oovéeiov. It was so used, to follow the MSS., against wapavyntrn and rapapéon and Myavds dissonantly. Weil and Reinach wish to transfer cal mpos tapapéonv to § 176, as on their theory it makes a hideous discord. On ours, however, it is needless ; vitn ocuvvnupévwv makes the interval of a minor third (D-B) with Twapapéon, and this is as probable and reasonable as the major third between yytn and wapavyrn (§ 176). Acxyavdés also provides the interval of a major sixth (D-F); again perfectly reasonable, and the type of interval which might have been expécted. I[lapapéon is not so obliging; D-C is a tone, which is just the interval which has been so fortunately eliminated from the discussion of the vyrn SueLevypevwrv. It is perhaps a little arbitrary to reject the mpos Tapavntny altogether, and as we have fortunately been able to confirm the manuscript reading in the rest of the passage it is a pity to abandon it here. It is however conceivable that the wapapéonr originally stood alone, then became accidentally duplicated, and one of the two was altered to mapavyrnv by an ignorant copyist. This emendation has no ingenuity to commend it. If it is not accepted it will leave us precisely where we were before as far as the polyphony is concerned. It will only be accepted if it is considered worth while to eliminate the discord of the second. Probably no one would hesitate to do so if he were not already accustomed to the idea of its employment. Still unexplained is the three-quarter tone interval in the old Spondeion scale. No hint is given irl the second Plutarch passage that abnormal intervals were employed ; but the first passage is explicit enough, and implies that the interval between wapapyéon and mapavynrn (enharmonic) was a three-quarter tone. If the lower tetrachord had a normal semitonal mu«vov (E-F), this would mean dissonance between F and C. Stranger phenomena are known in the history of scales, but this does not seem very like Greek music. Surely we have here the explanation of the debated expression 0 ouvrovwrepos orovdetacpos. The three-quarter tone interval occurred in both parts of the scale. Of course it was only the upper interval which, from its vicinity to the go R. P. WINNINGTON-INGRAM disjunctive tone, could create an illusion of the diatonic.1 The misconception arose as follows. If F-E is three-quarters of a tone (}4), then the interval A-F is 44, which is almost identical with the interval 33, made up by the disjunctive tone and the upper o7rovdevacuds, differing in fact only by y3s- The uninstructed, assuming A-F to be a ditone, assumed further that C*-A was the same. Light is thrown upon the subject from two other sources. In the first place there is to be compared the hemiolic chromatic of Aristoxenus, the small intervals of which together make a three-quarter tone. This shows that he recognised a mux«vovr slightly larger than the enharmonic semitone, but smaller than the (minor) tone of the normal chromatic. The fact that he calls it chromatic, whereas the Spondeion was regarded as enharmonic, need not trouble us. Aristoxenus was always anxious to preserve the purity of the lowest type of enharmonic thanks to his connexion with virtuoso flute-playing. In any case, such a distinction as this was arbitrary and academic. Still more interesting is the duadov diatonic of Ptolemy (Harm. I. 16), of which the lowest interval is }}—i.e. roughly a three-quarter tone—the complete tetrachord being {7XioX42. It does not occur in the lyre scales he describes in Harm. II. 16. In fact, the account suggests that Ptolemy had invented it himself. It may, however, be a more or less conscious reflection of the Spondeion, which may not have become completely obsolete by then. If so, the original intervals of Olympus’ scale were E}}? F?; A$ Bi; C. We are told that Olympus in his discovery of the scale made frequent use of the tritone B-F, which probably means that this interval was much in use in this style. It is certainly more likely to have been the undecimal tritone 4, than the harsher $$ or the strict tritone 77§. It is not so easy to deal with the amplification of the scale through polyphony. The splitting of the three- quarter tones produced an interval of a fifth between the two pecoroxva; this is clear enough on any showing. Nyrn (E) was added, making an interval of a fifth with wéon and of 3 with mapavyrn, less harmonious unfortunately than the major third earlier suggested in this connexion. Nyt cuvnppévwr also was added, making dissonant intervals with Avyavds and tmapapécn. It is therefore extremely difficult to decide how the interval C-E (7s) was divided by the note D. If we take Ptolemy’s tetrachord (}{ xis Xs), we have a reasonable interval between D-B, i.e. £, a minor third, but an interval of $3 between D-F. If, on the other hand, we make our tetrachord }3 x 3° x i, with the larger tone in the middle, according to the canon which Ptolemy laid 1 The lower D, the irepuyrdrn, was not in use only occurs in this passage and ‘in the earlier in this scale, as the continuation of the second passage about the Spondeion. They must only Plutarch passage informs us in the following refer to a special type of early melody, and in words: ob &’ d&yvoay dwelyovro év trois Awplos sections 183-185 Plutarch is still discussing the rob rerpaxdépdov rotrov (rav imdrwr). Now the rpéros omovdecdgwy and cognate scales. Indeed Dorian scale of Aristides possesses the note D. this must be so, as deixovro can have no subject But there is no need to suppose that ra Adpa but of wddax, which is made more explicit in necessarily refers to all Dorian melodies. This § 181 (’OAdpmov re xal rdv dxodovOncdvrwvr éxelvy). expression, together with ra Ppiyia and ra Avéia, down (2 now $¢. gating. have all between the E a preserve was imp in tune. in their have or oTrovoEete diatonic suppose scale, w melodic to demo vindicat 1 The p if we sup was a mil of a fifth b onception e interval ip by the r3s- The C*-A was . the first the small s that he it smaller = calls it need not y of the -playing. till more hich the trachord ‘ribes in rented it 1. of the If so, We are > of the e in this 4+ than with the e three- va; this erval of ely than ov also It is divided have a il of 33 3, with ny laid ie earlier nust only y, and in ssing the Indeed o subject ‘plicit in vy éxelvy). THE SPONDEION SCALE gI down (and broke), the interval D-F becomes 8, but that between D-B is now $$. This problem, however, is not peculiar to the tuning we are investi- gating. It is a natural law governing scales that a diatonic octave cannot have all its fifths and fourths perfect. Our own diatonic C-C has a false fifth between D-A. The normal Greek tuning has the same false interval, only in the E octave it becomes a false fourth between A-D. If then the Greeks preserved their normal structure of two tetrachords joined by a major tone, it was impossible for all the harmonies deduced from the Plutarch passage to be in tune. We need not here inquire whether they tolerated one harsh interval in their polyphony or in what way they solved this tuning problem.’ We have only to admit that the harmonies to which the three-quarter tone orrovoecacpuos leads us are rather more difficult than those given by the normal diatonic tuning. If we believe them to have been too difficult, we must suppose that the }{ interval disappeared in the accompanied version of the scale, which preserved its original character only in the limitation of its melodic notes. All of which does not affect the main purpose of this paper— to demonstrate the enharmonic character of the orrovderafwv tpdros, and so to vindicate the manuscript readings of the Plutarch passage which describes it. R. P. WINNINGTON-INGRAM. 1 The problem is in point of fact soluble only by dividing the two wvu«vd slightly differently. if we suppose that the disjunctive tone (B-A) As this was flute music, and the whole wu«véy was a minor tone (3,9), and that the consonance was produced from a single hole of the pipe, this of a fifth between zpirn and rapurdrn was obtained may not have presented much difficulty. SOME NOTES ON THE PERIPLUS MARIS ERYTHRAE!I,)’ THE document commonly called the Periplus Maris Erythrae has within recent years attracted a good deal of attention from scholars interested in Roman policy in the South-East. Though many of the statements it contains are definite enough, the uncertainty that prevails as to the date of its composi- tion and to some other details renders the interpretation of these statements somewhat difficult, and has opened the way for rather novel views. The ball was set rolling by Kornemann who in 1921 extended the limits of the Roman Empire beyond the Red Sea, declared that the phrase év@ecpos Bactreds (as applied to an Arab chief) must mean a soctus et amicus, and (like Voltaire in Gibbon) generously presented the Empire with the island of Socotra. He was followed by W. Schur who was engaged in sketching out an ‘ Orientpolitik’ for Nero of a large and grandiose nature: he accepted a great deal of Kornemann’s essay, approved his rendering of &0eapos Baotdevs, and though he rejected the notion of a Roman garrison on Socotra he added some ideas of his own, as, for instance, that direct sea-trade with India only began under Nero; he also agreed with those scholars who placed the discovery of monsoon voyages by Hippalus as far back as the late Ptolemaic period. Rostovtseff, in his magnifi- cent book, seemed inclined to follow Schur, believing in the occupation of Aden (and possibly Socotra) and in a Roman garrison on Syagrus. Dessau, on the whole, has paid little heed to these fancies. While acknowledging that there was a certain amount of Roman inter- ference in the Red Sea and beyond (as will be seen later), I find it hard to credit that it was as thorough-going as Kornemann and Schur would assume. Before building up any theory of Roman policy a man must have some data to go upon, and my purpose in this paper is the comparatively modest one of 1 It would be cumbersome to give in full the rather large bibliography that has grown up round the subject, but the following are all important : Text.—A new edition of the Periplus by Hjalmar Frisk (Gothenburg, 1927) supersedes the older one of Fabricius (Leipzig, 1883) and others. LITERATURE. — W. Vincent: The Commerce and Navigation of the Ancients in the Indian Ocean. Vol. II. London, 1807. W. Schoff: The Periplus of the Erythvaean Sea. (Translation and Commentary.) Philadelphia, 1912. W. Otto: Article ‘Hippalus’ in Pauly-Wissowa, VIII. 1660, 1661. J. Tkac: Article ‘Saba’ in Pauly-Wissowa, I. (Zweite Reihe), 1298 sqq. R. Sewell: ‘Roman Coins found in India.’ Journal of the Royal Asiatic Society, XXXVI. Pp. 591 sqq- 1904. E. Kornemann : ‘ Die historischen Nachrichten des Periplus maris Erythraei tiber Arabien.’ Janus. Pp. 558qq. tI921. W. Schur: ‘Die Orientpolitik des Kaisers Nero.’ Kilio, Beiheft XV. 1923. H. G. Rawlinson: Intercourse between India and the Western World. (Second edition.) Pp. ror sqq. 1926. Some treatment of the subject-matter will also be found in Dessau: Geschichte der Rémischen Kaiserzeit, I:, pp. 379 sqq. and in Rostovtseff : The Social and Economic History of the Roman Empire, pp. 91 sqq. trying tc matter o the date the ques date of 1 been ma to them with the The simple n the reig1 is its lat posed be from a Petraea India ar and the resumpt: writer sz an’ aut, trade-rel Herrma second Nabatae Thi of the I1 easily id means ¢ us on fir kings, al and 70. reigning the thro cannot — reigns Cc A.D. 40 1 Korne 2A, H Zeitschrift 1913, PP. | 3 Dussa Safad, pp. : an imagir Clermont- Glase 'S 1as within erested in t contains composi- tatements The ball e Roman stXevs (as oltaire in He was litik’ for nemann’s ected the Own, as, - he also yages by magnifi- 1 of Aden 1, on the in inter- hard to assume. > data to t one of Vissowa, I, in India.’ XVI. Pp. achrichten Arabien.’ s Kaisers India and Pp. ror r will also Rémischen stovtseff : he Roman M. P. CHARLESWORTH 93 trying to ascertain what we can regard as fairly certain and what not in the matter of dates and facts. I therefore propose to limit myself to four points: the date of the composition of the Periplus, the date of the florust of Hippalus, the question of the nationality of the éxarovrapyns at Leuce Come, and the date of the expedition against the town of Arabia Eudaemon. All these have been matters of controversy, and until some agreement is reached with regard to them it is useless to discuss Roman policy. In the following sections I deal with these four questions. The dating of the composition of the Periflus should be a comparatively simple matter, yet it has been strangely obscured. It is obviously later than the reign of Tiberius, as the reference to avtoxpdropes (§ 23) shows; but what is its latest limit? Here Kornemann allows himself no doubt:! it was com- posed between the years 87 and 105 of this era. The first terminus he takes from a paper by A. Herrmann,” the second from the annexation of Arabia Petraea by Trajan. Herrmann declares that about the year 87 traffic between India and China was resumed owing to conquests made by Chinese generals and the consequent reopening of trade-routes to traffic: as evidence for this resumption he cites § 64 of Periflus. We turn to that passage and find the writer says, ‘eis 5¢ trav @iva tavtny ove Eat evyepas atredOeiv:> orraviws yap an’ avTns twes ov ToAdXol Epyovra:’ this is strange proof of a resumption of trade-relations, and the passage would suggest the very opposite of what Herrmann wishes to prove: it is obviously prior to such a resumption. The second terminus is, however, incontestable since the Periplus speaks of a Nabataean king, Malichas or Malchus (§ 19), and there were none after 105. Thus the passage on which Herrmann relies betrays him, and the names of the Indian princelets given in the Periplus are unidentifiable, or rather too easily identifiable with anyone, to be of any use. We might despair of any means of dating our document were it not that the mention of Malichas puts us on firm ground. We now know fairly definitely the reigns of the Nabataean kings, and Dussaud and Macler® fix the reign of the last Malichas between 39 and 70. He cannot have succeeded earlier than 37 when Aretas IV was still reigning, nor can he have reigned himself beyond 71 when Rabel III came to the throne, and the Periplus must be placed between the years 40 and 70, and cannot possibly be later. (It is interesting to note that Glaser* places the reigns of two Arabian princes mentioned, Charibael and Eleazus, between A.D. 40-70 and 25-65 respectively). Further precision is unnecessary pp. 37-8. I have said nothing about Zoscales, Nambanus, and the rest. We cannot be sure of Zeitschrift dey Gesell. fiiv Evdkunde 2u Berlin, their identification, and even if they are rightly 1913, Pp. 553 $qq. identified, South Indian chronology (and Abys- 3 Dussaud et Macler, Voyage archéologiqgue au sinian) is none too securely established for us to Safa, pp. 171 sqq. It is true they want to foist use to control the Periplus, Hence I must an imaginary Malichas IV on Nabat, but M._ regretfully dissent from the views of Schoff in Clermont-Ganneau will have none of it. J.R.A.S., 1917, pp. 827 sqq. 4 Glaser, Die Abessiniey in Avabien und Afrika, 1 Kornemann in Janus, L., 1921, pp. 55 $94q- 2A. Herrmann: ‘Ein alter Seeverkebr.' 94 M. P. CHARLESWORTH (though I should be inclined to place it A.D. 50-65 if pressed), and any hypothesis based on a Domitianic date, or on one earlier than Claudius, falls to the ground. This too, I am glad to note, is the conclusion of Tkac. IT. Hippalus’ title to fame is that he was the first European to observe the periodicity of the monsoons in the Indian Ocean and to use this knowledge for sailing directly across the ocean instead of hugging the coasts of Arabia and Carmania. Until recently he was assigned a floruit in the middle of the first century of our era, and Tka¢ in his enormous article upon ‘Saba’ (in Pauly- Wissowa) is prepared to agree. I consider this a little too late, but with the date upheld within the last twenty years by some scholars (notably Otto and Kornemann!?) I find myself entirely at variance: they have sought on various grounds to place him late in the Ptolemaic age, roughly circa 70 B.c. It must be confessed candidly at the outset that we do not know as much about Hippalus as is usually implied: authorities declare that he is mentioned in the Periplus, by the Elder Pliny, and by Ptolemy.” This, though true, is misleading, for the Periplus mentions him as a man, an experienced sea-captain, while Pliny imagines Hippalus to be the name either of a wind or of a headland, and Ptolemy regards it as the name of a sea. In order to find a date for him Kornemann uses the famous passage wherein Pliny describes the sea-route to India (N.H. VI. 96-107). Here Pliny recounts the course followed by Alexander’s admirals, who kept to the coast: afterwards came the use of the monsoon (ventus hippalus), which was followed a generation later (secuta aetas) by a quicker route to Sigerus, and this was the course for some time till in Pliny’s own day a shorter cut was discovered to the ports of Muziris and Neacyndon. Kornemann assumes that Pliny derives all his information about this route (save the last item) from Juba, who was writing in the Augustan age, and that ergo the florwit of Hippalus falls con- siderably before that age. He finds this notion confirmed by some inscriptions of the late Ptolemaic period of 62 and 51 B.c., which speak of a otparnyos éri ths “Ivduens nai ‘Epv@pas Oaracons (inscriptions upon which Otto also relies). This seems unconvincing: the inscriptions prove that there was such an officer, and presumably he had general control over the Red Sea ports and trade with them and beyond the Straits of Bab-el-Mandeb; he was probably in charge of the ¢vAaxes tod ’ApaBiov pvyod of whom Strabo has a story.* Obviously Egyptian merchants sailed beyond the Straits to visit the Ptolemaic colonies in East Africa or Socotra; some few reached India by the long coast- ing voyage. But that men were sailing direct across the ocean, or that the atpatnyos had jurisdiction far into the "Ivésaj Oaddoon cannot be proved from these inscriptions. And to assume that Pliny’s account depends on Juba is 1 Otto in Pauly-Wissowa, Hippalus. Korne- mann, /.¢c., pp. 55 $qq. 2 Peviplus,§ 57. Pliny, N.H. VI. 100, 104, 172. Ptolemy IV. 7, 12. 3 Dittenberger, 0.G.J.S. I,, Nos. 186 and 190. * Strabo, p. 98. SO] also conj voyage dc Befo: difficulty silence of lived and and knew does. TI What Str is the se Ptolemie: vessels dé pax Rom vessels a word abc had visite Indian tr to assum was no H This Pliny anc if we con that, thor reign, the certain p German the acces Thes by the te! section : chapters Whereve the Red 20-15 B.C Here he gives | statemen sets out then ‘eal naurgatio what he 1 Pliny, 2 Strabo, 3 Strabo, and any lius, falls serve the ledge for ‘abia and the first n Pauly- with the Itto and 2 various It must sh about >d in the sleading, in, while and, and passage re Pliny e coast: llowed a was the vered to y derives who was ills con- criptions Tparnyos itto also vas such orts and orobably a story.* tolemaic g coast- that the ed from Juba is ) and 190. SOME NOTES ON THE PERIPLUS MARIS ERYTHRAEI 95 also conjecture: all we can safely say is that Pliny’s account of Onesicritus’ voyage does, for he expressly says so. Before giving my own interpretation I should like to point out another difficulty which Kornemann and Otto ought to face, and that is the complete silence of Strabo on this point. On their own hypothesis, if Hippalus had lived and made his discovery about 70 B.c., Strabo, who had travelled in Egypt and knew much about Indian trade, ought to have mentioned him: he never does. This may be an argumentum ex stlentio, but it seems to me important. What Strabo does emphasize in his remarks about India and voyages thither is the setback that trade experienced under the weak rule of the later Ptolemies,? whereas Hippalus’ discovery should have given it a fillip. ‘Few vessels dared,’ says Strabo, ‘ peer round the Straits in Ptolemaic days,’ but the pax Romana had brought greater feeling of security and ‘as many as 120 vessels a year now set sail for India and the Aethiopian headlands.’* Not a word about direct voyaging, about monsoons, or about Hippalus, yet Strabo had visited Myos Hormos, had picked up information there, had talked about Indian trade, and was interested in the blowing of winds generally.* It is fair to assume that he does not mention Hippalus, for the simple reason that there was no Hippalus to mention as yet. This has cleared the ground somewhat, and we may date Hippalus before Pliny and the Pertplus, and after Strabo; but greater precision can be attained if we consider what ‘after Strabo’ can mean. It has often been pointed out that, though portions of the Geographica refer definitely to events in Tiberius’ reign, the work as a whole appears to have been completed under Augustus and certain portions revised later: this accounts for such notices as those of the German triumph of Germanicus, the installation of Zeno as king of Armenia, the accession of Ptolemy of Mauretania, and so on. These revision passages can usually be detected (as Pais has pointed out) by the tell-tale word véwors.° This word does not occur in Strabo’s Egyptian section : he appears to have left Egypt by 20 B.c., and there is nothing in his chapters on Egypt and Arabia which refers to events later than that year. Wherever he completed his work he had no later information upon Egypt and the Red Sea, and so his account of that region is not really later than 20-15 B.C. at most. Here Pliny comes to our aid with a controlling limit. In Book VI. 96 sqq. he gives the stages in the development of voyaging to India, and his detailed statement is borne out by a more general one in the Pertplus. What Pliny sets out to do is to retail first Onesicritus’ account (as related by Juba), and then ‘eam nauigationem quae his annis comperta seruatur hodie.’ This nauigatio underwent subsequent modifications, and Pliny gives them, but what he emphasizes is that whereas Onesicritus made a coastal voyage the 1 Pliny, N.H. VI. 96. 2 Strabo, pp. 118, 798. 3 Strabo, p. 118. 4 Strabo, p. 29. For Myos Hormos and Indian trade see pp. 118, 686, 798, etc. 5 Pais, Ancient Italy, pp. 379 sqq. 96 M. P. CHARLESWORTH nauigaito was by monsoon. The Periplus gives a shorter but similar account, again emphasizing the difference between coastal and monsoon voyages. Here are the two passages: N.H. VI. 100 sqq.— Stc (coasting) Alexandri classis nauigauit. Perip. 57— Tovrov 5€ dAov rdv cipnpévov mepimrrovy amd Kayns xai ris "Evdaipovos ’ApaBias of pty piKxpoTréepors wrolorts teptxorAmilovres ExXeor, postea ab Syagro ... Patalen fauonio quem hippalum ibi uocant, peti certis- simum uidebatur. Secuta aetas propiorem cursum tutiorem- que indicauit si ab eodem promunturio Sigerum peteret ... . « . donec compendia inuenit mercator ... nauigant uento hippalo ad primum em- porium Indiae Muzirim, ... et Nea- cyndon. apwros S€ “Immados xuBepvitns . . . Tov dia weAdyous éLevpe mA . . . "Ad’ od péxpe Kal viv ot pev €ts Atpupexny mdéovtes Eri mirciov tpaxnXi- e 4 > 4 (ovres ot 5€ eis Bapvyafay .. . ov wAciov H Tpeis Huépas avTéxovot Kai... mwaparAéover TOs eipnyéevous KdArovs. In these two contrasted passages we have an account of the same phenomenon from two points of view—that of Pliny, who appears to have had such oral information as reached him in Rome (nunc primum certa notitia patescente), and that of the writer of Periplus, who knew the voyage by experi- ence. Both lay emphasis on the fact that early voyagers had to sail along the coast; both associate the monsoon voyage with Hippalus. Then comes a slight difference: the author -of the Periflus says that those who wish to gain Barygaza or the North Indian ports take one route, those who are making for Limyrike (South India) have to throw the ship’s head more off the wind, tpaynrivew (§ 57). Pliny mentions the Northern ports (Patale) as being the object of Hippalus’ voyage, then says that a generation later men learnt to reach Sigerus, which is half-way down [it is identified with the Meli-zigara of the Periplus (§ 53) and the Meli-zegyris of Ptolemy]; finally came the compendia and the voyage to Muziris and Neacyndon (Nelcynda), ports of Limyrike (Pertp. § 53). Thus whereas both Pliny and Periflus recognize the North Indian ports as being the first to be reached by monsoon voyages, the Periplus says in general terms that a man must tpayndilew to reach the southern ones, while Pliny distinguishes two stages: obviously as time passed men grew more experienced and daring and ventured finally to make the long direct voyage. Now if we assume that Pliny was writing shortly after a.pD. 70 we have to find a date for Hippalus’ discovery, which is not much before Pliny’s time (his annis comperta), and in addition allow for a generation (secuta aetas), a period of time longer than a generation (diu), and a recent discovery just before Pliny wrote. This seems to me to leave only one possibility: the increase in Egypto-Indian trade after 30 B.c., mentioned by Strabo, stimulated merchants and captains to free themselves from Arab intermediaries and cut short the SOM long coas Egypt) H the ocean found the forty yea Nelcynda. discovery am inclin trust this The « of discuss soldiers, a was not: that the 2: (like Dr. \ does not r That able doub (p. 779); A first disinc contingent Nabataean sentative. Nabat, ans Axumite k Arabia (or limit of his LaBaiwv x right to stz as mapaXn returned, | under Tib stationed | parallel oc was allowe undertook important : was a coho conditions 1 For the re 2 For this places in the 0.G.1.S. 199. account, voyages. lauigauit, 1 fauonio sti certis- tutiorem- munturio Cate . os num em- et Nea- 1e same lave had a notitia y experi- long the a slight to gain king for ie wind, sing the sarnt to igara of mpendia Amyrike > North Periplus rn ones, ‘Ww more yage. have to ime (his eriod of e Pliny ease in rchants iort the SOME NOTES ON THE PERIPLUS MARIS ERYTHRAEI = 97 long coasting voyage. Shortly after 20 B.c. (that is after Strabo had left Egypt) Hippalus made his great discovery, sailing by monsoon direct across the ocean to the ports of the Indus delta: a generation later (A.D. 10-15) it was found the monsoons could be utilized to reach more southerly ports, and about forty years later (A.D. 50-55) men began to voyage direct to Muziris and Nelcynda. (Although Pliny dates these stages rather precisely the process of discovery must have been gradual and not per saltum.) Thus on the whole I am inclined to place the discovery of Hippalus between 20 and Io B.c. and trust this will be found to agree with all the data available. ITT. The éxarovrapyns at Leuce Come (Perip. § 19) has occasioned a good deal of discussion and dogmatic statement. Was this official, with his force of soldiers, a Roman ora native? Vincent assumes he was: Fabricius is sure he was not: Schoff is doubtful: Rostovtseff was sure of his being Roman and that the 25 p.c. tax was levied in order to drive trade to the Egyptian ports, but (like Dr. Whewell) he has changed his mind since writing that article: Dessau does not regard him as Roman. That the kingdom of Nabat was a vassal kingdom there can be no reason- able doubt: Strabo expressly says ‘viv dé «axeivos ‘Pwpaiors eiosy bnKdor’ (p. 779) ; Augustus arbitrated as overlord between Herod and Aretas, and was at first disinclined to confirm Aretas in his kingship. In return Aretas supplied a contingent of troops, as later Malichas did for the Jewish war, and we find Nabataean kings sending gifts and golden crowns to the princeps or his repre- sentative... Leuce Come on the Red Sea was the southernmost point of Nabat, and so the extreme limit of Roman control, and that explains why an Axumite king of the first century, recounting his tale of victories in Africa and Arabia (on the so-called Monumentum Adulitanum), names as the northern limit of his conquest in Arabia Leuce Come—‘ azo re Aevetjs Koyns éws trav LaBaiwy ywpas érodhéunoa.’* The question then arises, did Rome assume the right to station her troops and to send an officer mapagvAaxijs yap and also as mapadnmrns of customs-duties in client-states? No certain answer can be returned, but I think that some parallels may indicate that she did. Thus under Tiberius we find that in the client-kingdom of Thrace there was stationed a Roman resident magistrate with troops under him.’ A closer parallel occurs in the district of the Alpes Cotttae or regnum Cott, where Cottius was allowed to remain as ruler, and where he built roads across the Alps and undertook the policing of them. Through this kingdom ran one of the most important roads linking up Italy with Gaul, and we happen to know that there was a cohors stationed in it to provide for security of travellers. Similar conditions appear to have prevailed in the Black Sea: whether Chersonese 3 Tacitus, Annals II. 67. 4 Ammianus XV. 10, and Suetonius, Tid. 37. 1 (cf. Jullian, Hist. de la Gaule, Vol. IV., pp. 288 sqq.). 1 For the references see my Trade- Routes, p. 254. 2 For this inscription, which Dittenberger places in the first half of the first century, see 0.G.1.S. 199. G 98 M. P. CHARLESWORTH was an ally or a free state she had to receive a Roman detachment under a centurion, and it has been suggested that the Roman post at Aj Todor was occupied as being the frontier between Chersonese and Bosporus:? all this garrisoning was part of the general pacification of the Black Sea of which Josephus speaks, and later on in the second century in allied territory on the east of the Black Sea Arrian speaks of detachments at such ports as Hyssu Limen, Apsarus, Phasis, and Dioscurias.” In general the Emperors were accustomed to arrange through their legates for the sending and stationing of centurions and a small force to towns which had commercial importance, as Pliny’s letter to Trajan about Juliopolis attests: a request for Juliopolis was refused because the town was not important or large enough.® (I suspect that some intention of this sort underlay the orders of Commodus to Quadi and Marcomanni only to meet once a month at a definite place ‘ éxatovradpxou tivds “Pwyaiou mapovtos,’ as narrated by Dio.)*4 It seems likely, then, that in uncivilized regions the Romans (whether dealing with client-kings or allied states) took the right of planting troops there for general security and also for helping to collect dues, and I think we may safely assume that the €xarovtdpyns at Leuce Come was a Roman centurion and the otpdrevya under him a small detachment: if so the word in the Periplus ‘amootréXXetae’ would correspond to the word mitteret in the Pliny passage. Whether the Romans took the 25 per cent. import tax as their own or merely handed it over to the king is hard to say. The rate is certainly high, as Rostovtseff observed, but it is only fair to point out that the king of | the Gebbanitae felt himself entitled in his realm to quartas partes of all goods brought through,° and that presumably the Romans, once in Arabia, made up their minds to do as the Arabians did. IV. We are now in a much better position for discussing and dating the subjugation or destruction of the port of Arabia Eudaemon, which corresponds in site to the modern Aden. Here the words of the Periplus are plain and precise (§ 26): it was formerly an important city (adds) offering good anchorage; in the old days when men did not dare (as Strabo too says of the Ptolemaic times) to make the direct voyage from Egypt to India, it was a great exchange mart: but a little before the writer’s time Caesar subdued it (viv 5é ov mpd ToAXOD TaV HyeTépwv ypovwy Kaicap avTny Katectpéyato). The meaning should be clear, yet the strangest interpretations have been placed on it by those who refuse to believe that Rome ever interfered so far South in the Red Sea, or who have an hypothesis to advance; Caesar has been changed #4 Dio Cassius LXXIII. 2 (Boissevain). 5 Rate levied by Gebbanite king: Pliny, N.H. XII. 68. In any case a rate as high as this was levied on some articles in Ptolemaic Egypt: see the Zeno Papyri, Nos. 73 and 75. 1 For these Black Sea garrisons see Minns, Scythians and Greeks: for Chersonese and Roman garrisons, pp. 523 sqq.; for Bosporus, p. 614. 2 Josephus, Bell. Iud. 11.16. Arrian, Periplus Euxini, 3, 9, 10, etc, 3 Pliny, Epistulae X. 77 and 78. SO] into Elis to confir Roman | writer of our time being by must me beginnin Kaicap. usually n to the C: divine h¢ Augustu: gerente C agnita, Ss about I expeditic about th occupati be dealt mapabanr by Rom: did not \ It m briefly regions. The received beyond t lord of E by the g! some eve enterpris been kep the Indiz tip of Ar about thi the town Sea at | centurio! Nabatae: 1 This de and Glaser under a odor was 1 all this of which ry on the as Hyssu eir legates ms which is attests: ortant or he orders mth at a Dio.)* (whether ops there < we may centurion rd in the ret in the x as their certainly ie king of all goods made up iting the ‘responds lain and ng good ys of the it was a ibdued it 0). The laced on th in the changed in). ig: Pliny, 1igh as this aic Egypt: SOME NOTES ON THE PERIPLUS MARIS ERYTHRAEI — 99 into Elisar and even Charibael!! But more recent opinion has swung round to confirm the reading: Dessau, Rostovtseff, and others have no doubt that a Roman fleet did deal with the city, the only question is ‘When’? To this the writer of the Periplus supplies a definite answer, since he says ‘a little before our times’; writing as he was between A.D. 50 and 65 (vide supra, p. 2), and being by then a man obviously of some age and experience, such a statement must mean that the expedition against Arabia Eudaemon took place near the beginning of our era, and this supposition is confirmed by his use of the word Kaicap. In the free-and-easy «own Greek of the first century that word usually means one of two things—it either refers to the reigning princeps or else to the Caesar par excellence, Augustus Caesar, a god himself and founder of the divine house of Caesars. The Kaicap mentioned cannot be Nero, but must be Augustus, and a notice in Pliny,? which speaks of the Arabicus sinus, 1 quo res gerente C. Caesare Augusti filio signa nauium ex Hispanstenstbus naufragus feruntur agnita, shows that a Roman fleet must have been operating in the Red Sea about 1 B.c. Thus Pliny and Periplus combine alike to suggest that a punitive expedition against the troublesome town of Arabia Eudaemon was sent out about the beginning of the Christian era: presumably the Arabs finding their occupation as middlemen gone began to indulge in piratical acts and so had to be dealt with, and as a result a once thriving town became the kp mapaSaraootos which we meet in the Periplus. That the town was garrisoned by Roman troops seems extremely unlikely: it had been taught a lesson and did not venture to interfere again. It may be as well to summarize the points I have tried to prove and show briefly how they would fit into the framework of Roman history in these regions. The feeble rule of the later Ptolemies had neglected the Red Sea: trade received no encouragement, piracy was rife, and few vessels dared to venture beyond the Straits. The measures that Augustus took when he became over- lord of Egypt resulted in a restoration of commerce, and men were emboldened by the greater security afforded them to sail in greater numbers to the East; some even reached India by the old coasting method. This increase fostered enterprise, and Hippalus discovered the periodicity of the monsoons, which had been kept secret by the Arabs, and used this discovery to voyage directly across the Indian Ocean to the ports of the Indus delta. The Arabs of the southern tip of Arabia found their business waning and gave trouble: in consequence, about the beginning of our era, a punitive expedition was sent against them and the town of Eudaemon sacked. In order, too, to control the trade of the Red Sea at both its northern ends Augustus ordered a small garrison under a centurion to be sent to the town of Leuce Come, in the client-kingdom of Nabataea, for the purpose of collecting the dues and customs there: possibly 1 This desperate effort was due to Schwanbeck, 2 Pliny, N.H. II. 168. and Glaser approved. 44.2601 4 100 SOME NOTES ON THE PERIPLUS MARIS ERYTHRAEI the high rate imposed may have been meant to drive merchants to the Egyptian ports which were more directly under Roman control. Greater knowledge of the monsoon voyages led to succeeding captains making bolder ventures and striking across in a more southerly direction for Sigerus and the harbours of the Ratnagiri district, and even a little later to the ports of the Malabar coast, Muziris and Nelcynda. But they had not dared to try further South, and this explains why the voyage of the unknown freedman of Annius Plocamus, who was blown by the winds directly to Ceylon, created such interest. By the end of Nero’s reign, however, direct trade with India was fully established even with the most southerly ports, and continued without a break for over a century; the vast increase in knowledge which this led to is faithfully mirrored in the work of Ptolemy. In such matters as these it is futile to hope for a definite proof or rigorous demonstration. All I hope to have shown is that the dates here suggested for the composition of Periplus (A.D. 40-70), the floruit of Hippalus (circa 10 B.c.), and the sack of Arabia Eudaemon (circa a.D. 1), are the most probable on the data we have at our command and involve the least difficulty; also that it is extremely likely that in the client-kingdom of Nabat a Roman centurion was sent during the trading season to collect the dues at Leuce Come. To do more was not my aim, but if these facts can be established some basis may have been laid for considering Roman policy in this region during the first century of our era. M. P. CHARLESWORTH. THI: (1gOI), a I thougt Homer (( correctiv in his FA 1924), 23 (1927), sl and beca discussio It n narrowly the use c critic’s te like circu are used usages al from SA of Arist: Mr. Alle proposed term, a0 As f scholia t and thos distribut over thc complete The sub unmista. 313, 9 subjects to offer ceased, < 1 Drerut speak incic B 558; bu connexion of my ow! For ‘ great EI to the Greater x bolder and the s of the ’ further Annius od such dia was thout a 2d to is igorous sted for (0 B.C.), on the at it is ion was To do is may he first RTH. IIAPAITEIS@AI = A@ETEIN ? THIS equation was set up by Mr. Allen in the Classical Review XV. 8-9 (1901), and remained, as far as I know, for over twenty years without effect.' I thought, therefore, when writing my External Evidence for Interpolation in Homer (Oxford: Clarendon Press, 1925) that the article had provided its own corrective and needed no criticism. About the same time, however, Mr. Allen in his Homer: The Origins and the Transmission (Oxford: Clarendon Press, 1924), 2359, reiterated his opinion; and later still Mr. Sheppard, 1b. XLI. 127-9 (1927), shows that he has been impressed by it. Under these circumstances, and because a good deal depends upon the interpretation of the word, a brief discussion now seems desirable. It must be noted at the start that the question is to be defined quite narrowly. The only stimulus to which Aristonicus and Didymus respond by the use of forms of a@ereiy is the presence of an obelus in the margin of some critic’s text, and our question is simply whether they also used mrapasteio@az in like circumstances. In modern parlance, to be sure ‘athetize’ and its cognates are used loosely for almost any unfavourable action by a critic, but such usages are not to the point here. For the underlying issue now is whether from $A on I’ 230 we are to understand: (I) that B 558 was not in the text of Aristarchus; or (2) stood in that text with an obelus in front of it. Mr. Allen believes in the second of these alternatives, and to uphold it proposed this equation, in which, unless we are to have an ambiguous middle term, a@ereiv must be interpreted rigorously in its technical sense. As parallels for this use of wapa:teio@a: Mr. Allen now cites ‘ the Odyssey scholia throughout.’ I know only one example from the Hypothesis (p. 6, 16) and those in Dindorf’s index. The latter are scattered from y to @, a curious distribution ; but as I have read the scholia on a-f, v-w, and glanced rapidly over those on u-u, without finding anything, I will assume that the list is complete. None of these examples gives Mr. Allen the slightest support. The subject is always a character of the poem, ordinarily not named, but unmistakable: Athena at y 366; Pisistratus at 6 199; Odysseus at ¢ 149 (07s), " 313, 9 102; and in the Hypothesis. It is obvious that a verb with one of these subjects does not mean d@ereiy or anything like it. The meaning, too, seems to offer no great difficulty. It is to ‘beg off,’ but the element ‘ entreaty’ has ceased, as in the English phrase,? to be dominant, so that the verb is used in 1 Drerup, Hom, Poet. 1. 327, did, to be sure, have been written; Mr. Allen gives 366 as the speak incidentally of Aristarchus’ athetesis of number of lines athetized by Aristarchus. B 558; but tbat was probably a slip. In this 2 A says: ‘Sorry, but I can’t go as I promised. connexion I should like to correct an oversight This, that or the other has happened.’ B of my own: p. 46 a misquotation of Drerup. reports: ‘A has begged off from going with us.’ For ‘great Alexandrian’ ‘ancient critics’ should Nobody misunderstands the situation. 102 GEORGE MELVILLE BOLLING the sense of ‘ getting out’ or ‘ getting rid’ of something by a variety of means, provided only that the method be more or less amiable and courteous. Its adjunct is either an infinitive or a noun of action to show what is avoided. This prepares us for the consideration of the passage that Mr. Allen regards as his best parallel. It is the scholium 6re wadsAXoyeiv mapyrnrac. GdXoTpior dpa oi émipepopevoar otixos eixoos Emtd on the beginning (A 365) of Achilles’ speech to his mother: ‘ Thou knowest; why tell all this to thee that knowest ?’ Taking wapyrnta: as equivalent to 70érnxe, Mr. Allen must wish for its subject the name of some critic, in this case Aristarchus. This parts company at once with all the Odyssean parallels, and seems to lead only to further difficulties. Mr. Allen unfortunately does not translate, but the meaning desired must be: The line is marked ‘ because Aristarchus athetized the retelling of the tale. The following twenty-seven lines are therefore spurious.’ I cannot acquiesce in this because, in the first place, a note of this type is impossible; a line is not marked simply because some other lines are athetized, but because it conflicts with such lines,? and can therefore be used to confirm the athetesis (cf. SA on B 668 or @ 56 as illustrations). Then I am afraid that I have translated with too much goodwill. To give the sense supposed some- thing like rovs wadiAXoyovvtas actixovs would be needed: ‘ Aristarchus has begged to be rid of (=athetized) the lines that retell the tale;’ the infinitive would rather mean: ‘ Aristarchus has begged off (got out) of retelling the tale,’ implying that the lines were not in his text where it is agreed they were. Finally, the logic is bad: lines are not spurious because Aristarchus athetized them; but Aristarchus athetized them because they were, in his opinion, spurious. All these difficulties vanish if we follow the Odyssean examples and make Achilles the subject. The line is marked ‘because Achilles has begged off (got out) of retelling the tale. The following twenty-seven lines are therefore spurious. But the usefulness of the note to support Mr. Allen’s interpretation of the scholium on I 230 has vanished also. One other example remains. In B 1 and K 1 Aristarchus read aAXou, but Zenodotus @AAo, while various unnamed scholars had other devices for bringing in the article. At the first passage Zenodotus’ reading is recorded (SABG) objectively, but the need of returning to the question seems to have disturbed Aristonicus. He wrote ($A) wapa:tnréov tovs ypadovtas, ay éott xai Znvddotos, ‘@AXoL pév,’ 4 Kal Tovs OTwaodv Bovropévous Sacivev, which I would translate :* ‘I must beg to be delivered from the people, Zenodotus 1 If ¢ 149 causes us to check, it is because the between British and American English. For scholiast feels 7rd dwrecOat Trav yyovdrwy as a cere- me, and for some others at least in this country, mony incumbent upon a suppliant, while we do it suggests an action quite mild and almost not immediately appreciate that feeling. apologetic, far from the ‘earnest diapproval ’ of 2 Or, as frequently, because it is the source of thedictionaries. Besides, to translate rapareiz- an interpolation, and so can confirm the athetesis @a by ‘earnestly disapprove’ is too much of the in another way. This is here irrelevant. style of translation for which Lewis and Short 3 I avoid ‘deprecate’ partly because in regard furnish the classic example: bibere ‘to arrive at to it there seems to be a difference of usage the region of the river.’ is one, V breathing feeling, is Monimos nuisance to be rid ‘Thank } We! ‘ because A 251 sc TIWeV Yypa ‘ oT? not next of aberetr all, there: had a na! same me famed li a cross | to make aberovpel abereirat Tov év 74 meant ; : presumpt Whi: ypapoper include . one exar phrase u been Ov had little escape t unless it placed a: Thi: these co! 1To as: quotation than the f to say ‘ the clearly car 2 They t that their 1 the papyt warranted. means, s. Its led. - Allen ONTNT AL. (A 365) to thee | for its »mpany further neaning ed the urious.’ type is ietized, sonfirm id that | some- ius has finitive e tale,’ 7 were. 1etized urious. - make red off srefore station », but es for orded . have yp €oTL which dotus 2. For ountry, almost val’ of pareic- of the | Short rive at ITIAPAITEIZ@AI = A@ETEIN ? 103 is one, who write @AAoz pév, and from those who want to use the rough breathing in any way whatsoever.’ The only thing to note, beside the feeling, is that the adjunct is a person, for which cf. Diog. L. 6. 82: The slave Monimos wishing a change of masters feigned insanity, and made a perfect nuisance of himself éws avrov o Seordtns wapynryncato ‘ until his master begged to be rid of him’; that is, was willing to take any price offered, and say ‘Thank you!’ besides. We now come to the crucial passage SA on I’ 230. The line is marked ‘because Idomeneus was posted next to Ajax Telamonios in accord with A 251 sqq.: mapairntéov dpa éxeivoy tov otiyoy tov ev T@ KaTadoyp wo TIW@Y ypapopevov ° ‘otnoe 8 adywv, iv’ ’AOnvaiwr iotavto dddayyes for the Athenians were not next to Ajax.’ So far mapa:teio@a: has never appeared as an equivalent of a0erety nor even as an approximate equivalent of it; the presumptions are all, therefore, that here it is not one either. On the contrary, it has always had a narrowly limited meaning; the presumption is that here, too, it has the same meaning. Accordingly I translate: ‘I must therefore beg off from the famed line that is written by some people in the Catalogue.’ If this were a cross reference to an athetesis, there were technical formulae in which to make it; cf., for instance, SA at @ 56: 4 &€ avadhopda mpos éxeiva Ta aberovpeva ‘ ToArEwv ex TOA wY éyyéotTraror avdpes’ (B 231), or ST at M 371: abereirat otv TO‘ Kai ce vo0ov rep éovta’ (@ 284). There was no reason why Tov év T@ KaTtaroyw aOerovpevoy should not have been written had that been meant; and the fact that technical language is not used raises still another presumption that we are dealing with a different situation. What this unusual situation is seems clear from the phrase rov iro twev ypadouevoy. In the technical language of Aristonicus’ trevés is never used to include Aristarchus; cf. Ludwich, A.H.T. I. 127-9, who can cite from $A but one example to the contrary, and that merely apparent. Consequently the phrase used is but a passive equivalent for what in the active voice could have been dv ovx éypadev ’Apiotapyos, a form of expression that the Aristarcheans had little occasion to use, and even then avoided.? I do not see how we can escape the conclusion® that B 558 was not in the text of Aristarchus, and unless it was in his text he could not have athetized it—that is, could not have placed an obelus in front of it. This solution is in harmony with two other facts, and while either of these could be explained away separately, the need of explaining away both is 1To assume without warrant a verbatim other lines it is more agreeable to say ‘which quotation of Aristarchus would alter no more others add.’ The note of Didymus on @ 73 is than the form of the argument. For an editor rather exceptional ‘ which is not in the editions to say ‘the line that is written by some people’ of Aristarchus’; probably he found that it was clearly carries the implication ‘but not by me.’ already in the copies used by some of his readers, 2 They use the text of Aristarchus and assume 3 I cannot agree with Mr. Allen that the use of that their readers had a text with the same lines; so common a word as 71s suffices to show that we the papyri show that the assumption was _ are dealing with the ‘ language of athetesis.’ warranted. When they deal occasionally with 104 GEORGE MELVILLE BOLLING an improbable combination. (1) When Strabo IX. 394 complains od rapa- déyovrat TovO’ oi xpetixoi, the natural understanding of the phraseology is that the verse had not gained admission to the texts of the critics. I may recall Plutarch’s use of é£e7dev with reference to I. 458-61, and to Athenaeus’ use of the same verb with reference to & 604-5, which are also based on the absence of these lines from the text of Aristarchus. In each case it has been claimed that the phraseology is loose, and that nothing more than an athetesis is under discussion. But in each case we can see more and more clearly that our MSS. are not in the condition to be expected had the lines been written and athetized by Aristarchus. (2) When Aristonicus uses trordocew of the insertion of a plus verse, the standard of comparison is ordinarily the Aristarchian text. Consequently, as Ludwich, A.H.T. II. 398343, pointed out, his explanation of the diple at A 273 as mpos Tods tbTordacovtas Tois "APnvaioss tov TeXapwvioy raises a presumption that B 558 was not in the text of Aristarchus. More cannot be claimed because the usage is not invariable, there being cases in which the standard of comparison is the original (A ristarcho tudtce) text. This method of interpreting the scholia! has been described as light- hearted. If it deserves the epithet, I may at least plead in extenuation a great eagerness to learn how more serious-mindedness can contribute to the elucidation of the passage. The result can, however, be tested in another way. It had long been claimed, most notably by Ludwich, that an athetesis had never led to the loss of a line. But Mr. Allen was the first to push the question further, and seek to determine exactly what the MSS. do with athetized lines. He reaches, Homer 305, the correct conclusion: ‘There are a few coincidences between Aristarchus’ atheteses and the reading of MSS. after his time, but there is no reason to suppose they are other than coincidences.’ I must strengthen the conclusion by attacking the examples cited. From his six coincidences must be subtracted: B 558, because its athetesis is the point being tested; H 353, because it is omitted by none” of our MSS.; I 695, because it is only an approximate coincidence; 2 556-7, because it too seems likely to be only approximate, and the statements about it show too much confusion.* Other examples cited, External Evidence 4, can replace these; but still the correct- ness of Mr. Allen’s_conclusion stands out: coincidences between omissions in 1 Mr. Sheppard seems to believe that I am the 3 Mr. Allen’s list is reprinted from Class. first to think the line was not in the text of Rev. XV. 244,. In one list we are told that Aristarchus. That is an injustice to Ludwich, (Q 556-7 are omitted, in the other that 2 556 (the Wilamowitz, Cauer, Bethe, van Leeuwen, and misprint 536 may be corrected) is omitted; in no doubt to others; the most I could claim both it is said that Aristarchus athetized Q 556. would be an effort to strengthen theirconclusion, In reality Aristarchus athetized 2 556-7, and 2 Quotations are a separate problem. If Dio what the MSS. do cannot be determined. The of Prusa (lv. 15) did not have the verse, he was intervening Oxford text adds to our perplexity probably quoting at second hand; cf. Class. by reporting no omissions here nor at I 694 nor Phil, XXII. 99-100 (1927), where Oldfather at ¥ 810. catches Epictetus at that trick. the M! extrem manus¢ TI of his. sion th If test: Cc the tes Aristar Phil. 2 of this no ma to con an ath include in Aris not to This ri: the MS cannot explan: purpos of Aris I vulgate congen B 558 | on A: it woul Aristot a verd. immed Mr. Al 1 In h hand, bt then spré with suc! 2 The reasonab XV.ga sions’ if 73. =T because because MSS. sh 3 The ov Tapa- gy is that nay recall us’ use of 2 absence n claimed ; is under that our 1 written w of the arily the nted out, A Onvaios : text of variable, vistarcho as light- uation a bute to ng been 1 to the ler, and reaches, between re is no hen the 2S must H 353, nly an ye only Other -orrect- ions in n Class. old that 556 (the tted; in d 2 556. 6-7, and d. The erplexity 694 nor IIAPAITEIS@AI = A@ETEIN ? 105 the MSS. and atheteses of Aristarchus are purely coincidences; they are extremely few in number, and involve only one, or at the most two, manuscripts. The case of B 558 is strikingly different. According to Ludwich fourteen of his MSS. omit,! and thirteen contain, the line. That points to the conclu- sion that the line is not one of those that Aristarchus athetized. If one is not ready to accept that conclusion he can apply the converse test: collect the cases in which there is, as at B 558, a marked divergence in the testimony of the MSS., and see what is the relation of such lines to Aristarchus. This Mr. Allen did not attempt; but in the Amer. Journ. Phil. XXXVII. 8-12 (1916) I have listed eighty-seven passages (not lines) of this sort, to which should be added a fortiori the passages that are found in no manuscript. The first thing that stands out is that it is impossible to to connect a single one of these passages with Aristarchus; there is not an athetesis, not a variant reading, ascribed to him. Furthermore, the list includes a very large proportion of the lines we happen to know were not in Aristarchus’ text: B? 168, 206, E gor, N 255, 731, ® 480, V 804, 2 558, not to mention N 808a, T 39a, and lines that have got into no manuscript. This raises another question: Why at A 196-7, ® 73, V 626, do we not find the MSS. in a similarly disturbed condition? For it is clear that this fact cannot upset the correlation noted, but itself is the thing that demands explanation. Into this I shall not here digress,*® it is enough for the present purpose that this test again points to the absence of B 558 from the edition of Aristarchus. I class the line, therefore, as a post-Aristarchean interpolation in the vulgate. That, of course, does not mean (cf. Ext. Ev. 15) that it and its congeners were first composed in post-Aristarchean days. On the contrary, B 558 is one of the lines for which the opposite can be proved. The scholium on A 273 shows that Aristarchus knew the tivés who wrote the line, and it would be no surprise to find this verse in a ‘wild’ Ptolemaic text. As for Aristotle (Rhet. I. 15) and Herodotus (V. 66) being acquainted with the line, a verdict ‘not proven’ seems the most appropriate. The anecdote is the immediate source of the interpolation; that it in turn drew on Hesiod (cf. Mr. Allen, Homer 237) is possible.* 1 In half of these it is entered by a second’ the conclusion I am about to draw. Mr. Allen hand, but that shows merely how the line was then spreading. Mr. Allen’s method of dealing with such evidence is not clear to me. 2 The evidence is frequently indirect, but reasonably satisfactory. Mr. Allen, Class, Rev, XV. 9 and 244, will recognize but three ‘ omis- sions’ in Aristarchus’ text, viz. E 808, II 613, 73. The first two must be set aside: II 613 because it was read in his second edition ; E 808 because the evidence is conflicting, while the MSS. show it to behave like any athetized line. 3’ The papyri supply the answer and confirm regards their evidence as casual, and it has usually been so treated. We should look upon them, on the contrary, as samples of a text the lines of which varied so little, that it is possible to predict what lines will, and what will not, be found in any vulgate papyrus. My predictions were first made in 1916 (Amer. Journ. Phil, XXXVII. 21) and seem (cf, Ext, Ev. 255-6) to have stood the testing of ten years satisfactorily. 4 Only let us not ascribe to Hesiod (cf. Mr. Allen, ib. 2353) the infelicity of making Ajax moor his ships {»’ ’A@nvalwy toravro pddaryyes. In 106 ITAPAITEIZ@AI = A@ETEIN ? The equation wapa:reic0a:=a0ereiv is therefore to be abandoned: (1) because it conflicts with the usage of the scholia ; (2) because the corollary that B 558 was athetized is contradicted by the behaviour of the MSS., which indicate (3) that the line was not in the edition of Aristarchus. GEORGE MELVILLE BOLLING. Oun1o STATE UNIVERSITY. its original setting (if there was one earlier than have described the action of one of the sons of the anecdote) the line need have referred neither Theseus in the rescue of his grandmother. to ships nor to Ajax. It could, for instance, TI Tu Festus’ detailed general similari and our WI! of the 1 article ‘ writer gloss. t being 1 argume had not to poin against Th of the : with by to the are thr no rela the Nc employ sentenc pretatic Tc if our c in batc were nt jumped and an but not no expl Tc his ret sugges’ In the a form (Fest. from I Dat., «¢ should a purp he sons of ler. THE ORIGIN OF THE SHORTER GLOSSES OF PLACIDVS. Tue relation of the glossae breviores of Placidus, or rather Pseudo-Placidus, to Festus’ epitome of Verrius has been the subject of much discussion ; but after Karl’s detailed examination in De Placids Glossis (Comm. phil. Ien. VII. 2), the view has generally prevailed that Ps.-Plac. is not dependent on Festus, and that any similarities between them are referable to the use of similar sources by Verrius and our compiler. When Goetz in C.G.L. I. accepted Karl’s arguments, it seemed that the question of the relation to Festus was closed. It has been opened again by Baehrens in an article ‘ Zu den Glossen des Placidus’ in Heymes 60, 396, in which the distinguished writer claims to overthrow the conclusions of Karl, and reverts to the theory that the gloss. brev. are a patchwork of items culled from previous commentators, Festus being the chief source, and Servius and Nonius additional contributors. His arguments are sometimes very plausible, and would doubtless convince anyone who had not examined for himself the whole of Ps.-Plac. It seems necessary, therefore, to point out once more that this theory as a whole has the balance of probability against it. There are many points of comparison between Festus and Ps.-Plac. In fact, of the 500 odd lemmas of the latter, nearly 200 in one form or another are also dealt with by Festus (i.e. in most instances by his epitomator Paulus). Karl’s objections to the theory that these items or some portion of them are derived from Festus are three in number: (1) The order of the Ps.-Plac. lemmas in question bears no relation to their order in Paulus or Festus ; (2) whereas Festus generally shows the Nominative or Infinitive in the lemma-words, Ps.-Plac. in many instances employs the case-form of Nouns or personal form of Verbs appropriate to the sentence in which they originally stood; (3) the material of the Ps.-Plac. inter- pretation differs in varying degree from that of Festus. To the first argument Baehrens makes noreference. This is surprising because, if our compiler is excerpting from Festus, one would expect the Festus items to occur in batches. Further, there is evidence (v. Lindsay, ].P. 34, 256) that the gloss. brev. were not originally in alphabetical order. We must therefore suppose that Ps.-Plac. jumped spasmodically from one part of Festus to another, choosing one item here and another there. For one who is merely excerpting such a procedure is possible but not probable. It is, at any rate, a cardinal omission on the part of B. to attempt no explanation of the surprising diversity of order in the supposed Festus items. To the second argument B. makes a reply which is even more surprising than his reticence with regard to the first. He resuscitates a fantastic view which was suggested in passing by Loewe (G/. N.) as one undeserving of serious consideration. In the items in question, Ps.-Plac., instead of using Festus’ lemma-word, substitutes a form which he found in one of Festus’ quotations. E.g. instead of Naema (Fest. 154, 19) Ps.-Plac. substituted Naentam, which he found in Festus’ quotation from Plaut. Tvuc. 213. Or sometimes, according to B., he wilfully substituted a Dat., etc., for a Nom., e.g. Farton (Ps.-Plac. 23, 2 and Fest. P. 78, 27). Why should he do so? B. suggests ‘um einen gelehrten Eindruck zu machen.’ If such a purpose on the part of a mere excerptor is credible, then he ought at least to have 108 J. W. PIRIE been consistent in his perversity: e.g. why does his ingenuity fail him in the case of Multt facere where Multi facit is at his disposal, or Nasiterna where oblique cases are quoted ? The third argument B, treats in detail, and shows (sometimes with much ingenuity) how certain details in Ps.-Plac.’s interpretations point conclusively to Festus as a source. There is one important objection, however, which may be brought against B.’s argument. He discusses about 55 items in detail, and on the strength of his conclusion drawn from these he attaches about 30 more to Festus. He takes credit for moderation in not claiming more, if we may infer that from p. 412, n. 3 (‘ Unsicheres ist absichtlich beiseite gelassen’). In reality it is by leaving out the uncertain items that he achieves a fictitious strength for his case. Besides his 80 odd items, there are about as many more where the same or a similar lemma is treated by Festus, but where the interpretation is either more pronouncedly different or altogether dissimilar. E.g. the lemma-word Ambulacns is used both by Ps.-Plac. (7, 11) and by Fest. P. (19, 21), but the interpretations are quite different. Similarly, Antigerio, Ad exodium (cf. F.P. 70, 26), Arse verse, Anate, Aeruscans, Adovans, Avilla, Acupedum (to choose examples from the A-chapter only), are all possible Festus-excerpts if we consider the lemma-word alone; the inter- pretations make the ascription highly improbable. In other words, the negative effect of these more divergent items ought not to be left out of account. If we pick and choose among the Festus similarities, we can make out a reasonable case for Festus-excerption ; but faith in our conclusion is severely shaken when we find that there are other items whose lemmas suggest origin from Festus, and that the difference between the first class and the second lies merely in a greater and a less degree of similarity to Festus in the interpretations. The closest similarities and occasional identities between Ps.-Plac. and Festus are to be found in a number of very short glosses like Apua: piscis minutus (Ps.-Plac. 7, 29; cf. F.P. 21, 5). Such glosses by themselves provide an unsafe ground for argument because the possible margin of variation is too small. If the Festus theory could be rendered highly probable from other and more decisive items, these would, of course, greatly strengthen the argument. They accommodate them- selves, however, just as readily to the theory of ‘common sources.’ Consequently, Ps.-Plac. 7, 24, Avtitus : artibus edoctus (cf. F.P. 19, 5, Artitus : bonis instructus avtibus) is really a neutral item, with a slight bias against B. because of the different participle. B. complains here, as in many other instances, that Philox. 23, 33, Artitus : wavtexvos Saisados is accepted as being of Festus origin, but not the Ps.-Placidean item. To this the answer is that Avtitus in Philox. is one of a large number of items which, taken in connexion with the express mention of Festus at 8, 21, prove beyond reasonable doubt that the epitomator is one source of this glossary. By itself it would prove nothing. B. says nothing of that curious double occurrence oi certain items which in so short a glossary is surprising. E.g. 6, 5 and 35 Actstum: 6, 7 Ad incitam and 7,12 Ad incitas: 6,20 Abludam and 6, 30 apluda: 6,11 Angvae and 7, 3 Ancras, as well as some others. Axcras : convalles aut arbovum intervalla, apart from the spelling of the lemma, is equivalent to the Paulus item (10, 22). B. founds an additional argument on the use of the Accusative, which in Festus he supposes to have been dependent on a verb like dicebant. If Ancras had not been present, he would in all probability have claimed Angrvae : intervalla arborum for Festus. As it is, however, he would presumably allot Angvae and one of each of the other pairs to some unknown authority other than Festus. In that case, we would seem to have an authority who is not Festus, but who, in some respects anyhow, is similar to him. Then, perhaps, we might dispense with Festus altogether, and saddle this other source with all the Festus similarities. However, the difficulty presented by these double occurrences disappe that a. Ad inctt B.’ fiunt is the firs derived part is : 13, with F. the firs believe Festus Sing. glossar to the < ‘ lixae,’ first. as a lat Or in Ps.- one (as in our cannot B. sup} order (14, 36 accept truncu: meanir For C Simila: square eviden In suppor commc malumn tenuem therefc that P slende: posses: assum B. arg deriva been 1 14, IO, is app comin; shown and it before h the case ique cases ith much usively to h may be nd on the to Festus. that from it is by his case. ame or a her more 1 mbulacnis ations are se, Anate, ter only), he inter- negative we pick - case for find that that the er and a d Festus > minutus n unsafe If the ve items, te them- quently, 5 avtibus) different 23; 333 not the a large estus at of this ch in so tam and ncvYas, as spelling ditional ve been d in all ever, he nknown ity who yerhaps, . all the irrences SHORTER GLOSSES OF PLACIDVS 109 disappears altogether if we suppose that Ps.-Plac. was transcribing marginalia, and that a later alphabetical arrangement of these brought items like Ad tncttam and Ad incitas into comparatively close proximity. B.’s explanation of 14, 37, Colurmis : ex cornu factis. nam et colurnt qui ex corulo fiunt is curious. The lemma-word is referred ‘ohne Zweifel’ to Verg. Geo. 2, 396: the first part of the interpretation is inapplicable to the Vergil passage, and is derived from Festus (F.P. 33, 15, Colurna hastilia ex corno arbore facta); the second part is added from a Vergilian commentator. The procedure seems too involved. 13, 24, Caculae : lixae aut servi militum. The noteworthy features on comparison with F.P. 39, 29 (Cacula servus militis), are the use of the Plural and the insertion (in the first place) of an alternative (and wrong) interpretation which we can scarcely believe Festus used. As regards the Plural, I would deny neither the possibility of Festus having had ‘Caculae’ nor that of Ps.-Plac. having substituted Plur. for Sing. The St. Gall glossary, which borrows this Festus item from the (full) Abolita glossary, actually shows ‘Caculae’ (IV. 214, 51). To my mind, the chief objection to the ascription of this item of Ps.-Plac. to Festus is simply that the interpretation ‘lixae,’ of which there is no trace in glossaries known to depend on Festus, comes first. B. thinks it a secondary addition. It is more natural to regard ‘servi militum’ as a later addition or correction to ‘ lixae.’ On p. 405 B. says, in effect, that where F.P. has several meanings we get these in Ps.-Plac., and where F.P. has only one explanation, and that, too, a secondary one (as in the case of fartor, Ps.-Plac. 23, 2 and F.P. 78, 27), we read only that one in our compiler. A very good point, if the evidence were more convincing. I cannot find another example parallel to fartor. Of the five examples by which B. supports the first part of his assertion, eccere and naentam show a reversal of the order of Festus’ interpretations (a small point, but it makes us pause): caespstibus (14, 36) only coincides with the double interpretation of Festus (F.P. 39, 6), if we accept ‘ramis’ as a reasonable paraphrase (which it is not) of ‘ frutex recisus et truncus,’ On the other hand, in the case of actus, where F.P. 16, 15 gives four meanings, Ps.-Plac. (6, 40) gives only one, and that not the first one of Festus. For Ciccum Ps.-Plac. 13, 23 gives two interpretations, F.P. 37, 12 only one. Similarly with Caculae mentioned above. We can, of course, make these items square by labelling the extra material as a later addition; but we must not trim the evidence to suit a theory. In connexion with Ciccum, B. uses an argument which would, if sufficiently supported, be decisive, namely the argument from common mistakes (i.e. mistakes common to Ps.-Plac. and Festus). Both Paulus and Ps.-Plac. confine Ciccum to the malum Punicum or granatum. Varro, on the other hand (L.L. VII. 91: membranam tenuem quae est ut in malo Punico discrimen), apparently takes a wider view. We are therefore invited to suppose that Festus made a mistake in epitomizing Verrius, and that Ps.-Plac. copied it. Truly, an assumed mistake on the part of Festus is a slender foundation on which to build! Similar mistakes of Festus (which, as we possess neither Verrius nor his epitomator, can be neither proved nor disproved) are assumed in the cases of flemina (p. 404) and capronas (p. 405). Again,in two instances B. argues a Festus connexion because of the use of expressions appropriate to a derivation which is not stated in Ps,-Plac., but was mentioned, or ought to have been mentioned, by Festus. 16, 7, Damium : sacrificium quod in operto fit, etc., and 14, 10, Calpar : vinum quod primum libatur e dulvo (cf. F.P. 60, 1 and 57,16). In operto is appropriate, B. implies, only where, as in Festus, a derivation from dapécror is coming, dolium only where, as he assumes for the original Verrius item, calpar is shown to mean dolium. In the first instance, the phrase 1m operto is perfectly natural, and it (or something equivalent) must have occurred in the definition of damium before anyone thought of forcing a derivation from dapécvov to suit it. In the second IIo J. W. PIRIE instance, one is again assuming a statement by Verrius and then its omission by Festus. Anyhow, a suggested derivation of so obvious a kind can quite well be referred to ‘common sources.’ By the way, why does not B. claim for Festus on the same argument 15, 2, Czcindela: scavabacorum genus quod volans lucet, where Ps.-Plac. might be regarded as suggesting by /ucet a derivation ‘a candela,’ which Festus expressly stated? Is it because Festus’ genus muscarum and Ps.-Plac.’s scavabacorum genus are too discrepant ? Karl, rejecting the Festus hypothesis, suggested two sources for these gloss. brev.: (a) A list of short (mainly one- or two-word) glosses such as were compiled by the ‘scriptores glossarum ’; (b) marginalia from MSS. which were of a somewhat longer character and sometimes included etymologies. To these Goetz (C.G.L. I. 62) added ‘ docti glossographi’ to account for items like Apluda, Angvae, Exanclare, etc., where the lemma-word shows Nom. or Infin, Karl’s division into two sections (v. De P.G., pp. 93 Sqq.) is not convincing, nor does he claim to carry it out systematically beyond the sections A and C. Indeed, to distinguish between the list of a ‘scriptor glossarum’ and the actual marginalia of MSS. seems unnecessary, if it is even possible. If some are marginalia, why cannot all be? And surely we need not postulate a ‘doctus glossographus’ to change an Acc. here and there into a Nom. Professor Lindsay’s view (J.P. 34, 257) that these items of Ps.-Plac. are derived directly from the marginalia of MSS. of Republican authors is at once more simple and more convincing. These marginalia, emanating ultimately from the hands of the earliest glossographers, would be repeated, doubtless word for word, by the editors of the second century. Verrius drew largely on such marginalia of the earlier period: a portion of them preserved by the archaistic enthusiasm of the second century A.D. has found a home in the glossary of Ps.-Plac. No other theory explains at once the similarities with Festus and the divergence as regards order of items and form of lemma-words. These marginalia occurring in the MSS. of a variety of authors—e.g. Livius, Lucilius, Titinius, Plautus, Terence, etc.—were doubtless written by many different hands. Not all of them would be derived from still older marginalia: fresh notes would doubtless be added from time to time, expressive of the reader’s own inter- pretation or culled from commentaries. There is no doubt that the writer of some of these glosses collected by Ps.-Plac. was using commentaries. The item Ingluviem : Cornutus ventrem, Plintus edacitatem shows, I believe, that a reader wrote in the margin of a MS. of a Republican author definitions which he found in Pliny (De Dubio Sermone ?) and in a work of Cornutus. The latter was a well-known Vergilian commentator, and so his ‘ ventrem’ might be a note on Geo. 3, 431. Pliny’s definition, however, does not suit the Verg. passage, and Cornutus did more than comment on Vergil. The Cornutus note may have come from a work the name of which occurs (in apparently corrupt form) in Charisius, p. 261 (ed. Barwick). In this work (‘liber tabularum castarum patris sui’) there was a note on ‘In mundo’ (Caecil. 276R), on which Charisius drew ; and it is suggestive that Ps,-Plac. 29, 5. In mundo: in expedito vel ad manum, <1tn> procinctu contains the same ‘in expedito’ which Charisius (but not Festus P.) employs. There is another Ps.-Plac. item which mentions Cornutus, 33, 10, Magmentum : alit pinguissimum extorum, alit secunda prosecta, Cornutus quicqutd mactatur, etc. The definiteness of the reference to Cornutus, compared with the vagueness of al# .. . ali, suggests to me that the last definition was added to the previous two by a reader who happened to find Cornutus’ note on the subject either in the work just mentioned or, perhaps, in his Vergilian com- mentary at Aen. 4, 57 (‘mactant ’). There are other items which suggest additions made by another hand either before or after Ps.-Plac. compiled his list. B. points to 21, 20, Experitus : non peritus sed ext wm pers studer There should expertu use hi means perstus case ¢ emerga refere! additic if the If son It is ‘dog inforn sapien utique wishir Simila of Col assulis it is [| differe definit who d in det before Ii in MS ‘Ps.-F its ite posses Expre to a |. word | fact, t and e margi: V (e.g. 3 note v lookin the O: ila suggest 2 On than B (p. 414 much n nission by te well be Festus on cet, where ela,’ which Ps.-Plac.’s oss. brev.: ed by the hat longer J. I. 62) sclave, etc., Oo sections rry it out en the list ssary, if it y we need here into -Plac. are nce more from the ‘ word, by lia of the m of the er theory s order of gy. Livius, ; different ‘esh notes wn inter- r of some The item ler wrote | in Pliny 2l1-known Pliny’s 1ore than the name ick). In 1 mundo’ lac. 29, 5. expedito ’ lac. item it secunda “ornutus, definition ” note on lan com- nd either on peritus SHORTER GLOSSES OF PLACIDVS IIL sed extra peritiam positus sicut expers extva partem. Experitus, as Festus tells us, meant impevitus. So, according to B., the original gloss was ‘non peritus,’ but a later student added a sed phrase, which destroyed the close connexion of non with peritus. There is, however, another possible explanation. Expersitus, to judge by its formation, should have had, and doubtless did have, a positive meaning = peritus (cf. experior and expertus). Someone used it with the unusual meaning of imperitus, and this anomalous use has been preserved for us by both Fest. and Ps.-Plac.: our item therefore means ‘ Experittus has here a rare sense: it does not mean (as one would expect) pevitus, but tmperitus: the ex- bears the sense of extra as in expers.’1 A more certain case of addition is to be found at 14, 18, Caudeam : tunceam quod iuncea[m] cauda emergat. dicunt et scirpum et tibim. The dtcunt clause, with its Hebrew word and its reference to ‘iuncus’ previously mentioned only in adjectival form, suggest an addition. So, too, we might label as an addition the second part of 14, 37 Colurnis, if the argument founded on the adversative use of nam in late Latin be right. If some additions have been made, there may be many more less open to detection. It is possible that 7, 13 Amtegenttos had as its marginal explanation originally ‘dogmate Pythagorae,’ and that a later investigator, finding another item of information about Pythagoras, added the irrelevant clause ‘qui negavit quemquam sapientiorem fuisse eo qui nomina rebus imposuit.’ Perhaps even the phrase ‘non utique plaustrum’ of 7, 2 Avceram was tacked on to an earlier gloss by a student wishing to contradict a definition which he found elsewhere (e.g. Fest. 14, 21). Similarly, where two different explanations of the same word are given, as in the case of Collativus : magnus, e conlatio factus, or Cocults : aeneis vasts ad coquendum vel assulis avidis, or Caespitibus : glebis terrae cum suo gramine, interdum alio loco ramis, while it is possible to suppose that these point to two occurrences of the same word in different passages with different marginalia, it is equally possible to regard the first definition as the original item, the alternative meanings being added later by someone who derived his information from a commentator. It would be unsafe to dogmatize in detail about the dim history of a glossary which perhaps existed for centuries before it was pitchforked (in part) into the ‘ lecture-notes of Placidus.’ It is, I believe, approximately true that these gloss. brev. existed as marginalia in MSS. of the second century A.D., and that a list of them was made by an unknown ‘Ps.-Plac.,’ a list which, in the course of time, had casual additions made to some of its items. As regards the date of this Ps.-Plac., the only definite information we possess is the terminus post quem provided by the reference to Cornutus and Pliny.? Expressions like nam and tsbim (mentioned above), in virtue of which he is assigned to a late date, are few in number, and may quite easily be additions; nor does the word appetentia, on which B., p. 400, founds an argument, afford us much help. In fact, there is no satisfactory evidence enabling us to label Ps.-Plac. a ‘ Spatling ’; and even if there were, we are more interested in the date of the material (i.e. the marginalia) than in that of the compiler. We need not attach much importance to the objection that some of the items (e.g. 36, 20 Oburuas) are too long to have been marginalia. The length of a marginal note would be limited only by the space available. We can reassure ourselves by looking at the length of some of the marginalia to the Paeans of Pindar in Vol. V. of the Oxyrhynchus Papyri. 1 I am indebted to Professor Lindsay for this these writers were resuscitating O. Latin words suggestion, and also for much kindly criticism. which they found in the glossaries of their period, 2 One would require much stronger evidence nd that their use of them in the particular cases than B. adduces for the use of Mart. Capella (Dat. or Abl. and Acc.) quoted by Ps,-Plac. is (p. 414) as well as of Apuleius (p. 409). It is merely accidental. much more likely that, e.g. in antiis and contumias, 112 SHORTER GLOSSES OF PLACIDVS Another possible objection is this: If Ps.-Plac. had the MSS. of a considerable number of Republican authors at his disposal, how comes it that he took so small a number of marginalia from each—a few items from each of Ennius, Lucilius, Naevius, Titinius, and so on? Further, when he took so little relatively from each, how did it happen that so many of the items he chose were exactly the words that had also been annotated by Verrius? A partial answer to the first of these questions is that the original glossary was certainly larger than that which we possess: how much larger, it is impossible to say; it may be that, as we have it, the glossary is only a shadow of its former self. Further, we need not imagine that these MSS. had each a great mass of marginalia. Cf. the fragment of Menander’s KéAaf in Oxyrhynchus Papyri, Vol. III., which in about too lines contains only one marginal note and vestiges of another. So that a play of Plautus might conceivably offer Ps,-Plac. less than a score of marginalia, of which many might escape his notice or be purposely omitted. As regards the coincidences with Festus, marginalia were doubtless for the most part traditional, i.e. the same words and phrases were annotated in successive editions. If Verrius, as is probable, made an extensive use of such marginalia, then the degree of coincidence between Festus and Ps.-Plac. ceases on reflection to surprise us. About two-thirds of Ps.-Plac. stands outside the Verrian material as we know it: if we admit the use of approximately similar sources and allow for the individual caprice, etc., of the compilers, an overlap of one-third in their choice of items is not really unreasonable on the doctrine of chances. J. W. Pirie. THE UNIVERSITY OF GLASGOW. In page 6: allow t! In printed followi they ar Now in the wo! was a V KE€pavvo} KEpavo § ‘Hymr a coroll Su Dr. Fa: the ext to éAar poets ai to cont the stor where, | the tree in the g (Miller speaks « It was éAaiauet pines al situatio! knowlec Greek Si Pp. 313; isiderable 0 small a Naevius, , how did | had also ns is that ow much is only a had each rhynchus note and Plac. less ourposely ss for the uccessive alia, then action to ‘ial as we r for the ir choice PIRIE. CALLIMACHVS AND OTHERS. In the new fragments of Callimachus’ Airca II., line 68, Oxyrhynchus Papyri XVI1., page 65, the context and (as Professor Hunt informs me) the traces of the letters allow the conjecture érov, so that the line could run peppvod por rreptyerow erov. In the British Museum’s fragment of the Scholia to Callimachus’ A/ra I., printed in Oxyrhynchus Papyri XVII., page 56, line 40, xep[.]vo, Professor Hunt, following Mr. Bell, hesitates to read xepavvo on account of the want of space; and they are probably right. The context refers to Zeus as lord of the thunderbolt. Now in Hymni Ptolemaici, in Collectanea Alexandrina, page 84, column vi., line 9, occur the words |ray epavoBinv yovov wxiv, where it always seemed that under this vox nihili was a word with the same meaning, and hence Professor Goodspeed corrected it to kepavvoBinv. But on comparing the two passages we may be right in thinking that xepavo Should stand in the Scholia, and that xepavoBinv should be restored to the ‘Hymn,’ but whether it is more than a mistake in spelling is not yet clear. As a corollary the word before it may have been, for example, vyepédovjra. Simias Rhodius.—On the ’AwoAdAwyv, line 7, in Collectanea Alexandrina, page 109, Dr. Farnell has pointed out to me that ¢Aaiasx, which Tzetzes, the only authority for the extract, gives, should be retained, and not altered, as has generally been done, to éAdraw, for line 2 comes almost verbally from Pindar, Pyth. X. 31, and both poets are speaking of a visit of Perseus to the Hyperboreans, Lines 7 and 8 appear to contain a reference to another passage in Pindar, Ol. III. 1 sqq., which contains the story of Herakles’ visit to the same land, the home of ‘the servants of Apollo,’ where, admiring the olive-trees, he asked and received the gift of one for planting in the treeless demesne of Olympia, that the sprays might be plucked for the victors in the games. The word vijaovs need present no difficulty, for Hekataeus of Abdera (Miller, F.H.G. I1., pp. 386, 387, fragments 2 and 5, taken from Diodorus II. 47) speaks of the island of the Hyperboreans, called ’EAigo.. in Stephanus of Byzantium. It was the mistaken popular etymology of “YzepBdpeo: which led critics to alter éAaiaws to éAdrawt, for they argued that olive-trees are not northern trees, which pines are. But, as Dr. Farnell has well said, they were a paradisiacal people, the situation of whose country was not defined, but moved or receded as the Greeks’ knowledge of geography widened. On the whole subject see Farnell, Cults of the Greek States, Vol. 1V., 99 sqq., and How and Wells’ Commentary on Herodotus, Vol. I., PP- 313, 314- J. U. Powe t. MoOwPwa* ARS N "I K "A N A iv O 2 OXFORD. It is unusual for Greek poets to sign their works. kai kev ‘Opnpeiowo xa eioéte Nexavdporo pynori éxow, and the Alexipharmaca xai x’ év@’ ipvorddowo Kai eioérs Nixdvdpoo pynjorw éxow. the body of his verses; in the Alexipharmaca at ll. 266-274: NICANDER’S SIGNATURE. Nicander ends the 7heriaca In addition, he embeds his name in vv b€ ai apweddeww EAtxaa evOpvrreo vipao cwo kai Bardevra wepi mropbeia Kodovcac. / 4 7 > A , 4 ai TE OV yupvwreiar evtpeheos vea TEppy agTnvov, Kapvoo AaxvdAoioo xddAvupa, , ‘4 4 \ , = > , eiarpav TOO odpxa wept oxvrAoo atov orale vrreréor kapvoo, Td Kacravio érpedev aia. e ‘ / , > 4 ‘ cia 5¢€ vapOnxoo vearnv e£aivuco vyddv, o Te IIpounGeiowo xAorjv avedé~ato pwpne. bv 5€ kai EprvAAo1 piro{woro réerndAa, though one can imagine the heat with which this conclusion would have been con- tested, if at Theriaca 345-353, enpapevor Kagierotv Exao TepikvoOeaT apXar Suoovynt vedrnta yéepac Topev Hpepiowt voaivwv > dn yap pa rupda Aniovrop’ Evirrov ppover* ov yap Tho ye Kaxodppading arédvnvto. wh been con- . LOBEL: QUESTIONS WITHOUT ANSWERS. THE appearance of P. Oxy. 2080 makes possible the solution of an old puzzle in the wepi ovyvrdgewr of Apollonius Dyscolus. Speaking of the inadmissibility of the definite article with aAAjAwv, Apollonius says (p. 60, Uhlig): xai rpopavée éori éx rH Tapa TAT. XpHrewo Kai TOU TvpTaperopevov Adyov. pera pyparor yap SiaBatixyy tporwrwv uidbeow onpaiver, aAAnAOVE Tpwcnte (Od. Tr 12) aAAnAove § EAHioav GAAnAwY dAccivovres BéXea orovdevra (Il, P 374). Lehrs observed that a dative was required in the second example, and proposed GAAjAow épicwow (Od. o 277). The truth is now seen to be aAAyjAow & éAvnoav (P. Oxy. 2080, 76). But this certainty raises up a problem of a different kind, to which no solution at present offers itself. Why has Apollonius here arranged his instances in the order—accusative, dative, genitive? The theoretical order of the cases of a Greek noun to him, as to his predecessors we do not know how far back, was nominative, genitive, dative, accusative, vocative, and it is implied, at any rate as far as the first two terms are concerned, by the very words which immediately precede those quoted above, dvédixrov 7d dAAjAwv cio GpOpov wapdGecw, for there being no nominative &AAnAo, he uses the next following case, dAAjAwv, to stand for all the cases in which the word occurs. And throughout the wepi avrwvvpiac we find him arranging his citations, whenever he is not prevented by the nature of what he has to say, in the theoretical order; so, for instance (p. 5, Schneider): ta dpOpa dvOvrdyera: raic GvTwvupiaw, MPOTAKTLKG pév * 0 yap HAGe (Il. A 12) tov & avrov AvxaBavroo (Od. & 161) Kat tae vydupoo vrvoe (Od. v 79) aXX’ apa kai tov “Hpn (Il. T 96), and, what is even more conclusive, the order in which the pronouns are displayed, that is, the skeleton on which the book as a whole is built, is: ¢yw, ov, i, éuov, coi, od, éuoi, coi, of, end, w€, €, Hpucio, ipeio, ofeio, Huav, tuav, Tpav, Hpiv, ipiv, opiow, nuac, indie, opie. Why, then, dAAyjAova, aAAjAow, aAAjAwv in exactly the reverse of the usual order? I cannot answer. The existence of this theoretical order of the cases leads to the recognition of another problem, to which I am as little able to give an answer as to the first. The writer of the tract repi oynydrwv, which comes to us under Herodian’s name, but is judged not to be his, gives three examples of the figure roAvmrwrov, which he describes as drav #ro Tac évopaviag 7 Ta dvépata cio Tacos Tig Trove peraBddAdovTer SiarcOdpeOa. rdv Adyov. He proceeds: wo mapa KXcoxdpe* | AnuooGevne vrerrn Pirirry ° AnpooGévove révno pév 6 Biow, peyadn 8’ 4 rappycia ° AnpooGéves roAA@v Sidopévwv ovdev odre tAROoe ovre KdAAoT G£iov éepavy ™podoviac ° Anpoobévny ’AréEavipor éfjrec 7rd dia ri wap’ avroia Aoyifer Oe - > , adikwo TE aréBaver, | & AnudoGevec. 116 QUESTIONS WITHOUT ANSWERS éore &¢ 7d rovovTov oXjpa Kal mapa Tit THY TonTwV, wo Tap’ ’"ApyiAdxy Kai "AvaxpéovTs 4 \ . > b , » de "A A | ~ xai mapa pev obv "ApytAdyp* . . . mapa d€ “Avaxpéovti eri tpidy KAevBotdAov pév eywy’ épéw KAevBotrAw: 8 eripaivopor KAevBovdAov S€ dtorKéw. Seeing that the first of these quotations contains all five cases and the third quotation three, in their correct theoretical order, it is not far fetched to suppose that the writer deliberately chose passages in which the cases occurred in that order. If that be so, we shall expect to find in the second quotation, that from Archilochus omitted for the moment above, not merely an example of the vyjya roAvrtwrov, but one in which the cases occur in their theoretical order. Yet the form in which the passage is ordinarily presented to the reader of Archilochus by his modern editors not only does not satisfy this condition, it does not even, as a rule, satisfy the condition that it shall be an example of roAvmrrwrov, and that though Dindorf (p. xxxili) pointed out what was logically required over a hundred years ago. Moreover, corrupt as they evidently are, the manuscripts (or some of them) do present each case in the expected place: vuv d€ Acddidroo pev dpxes Aewhidov Havn., Bod. } >» . AeaédiAoo Marc., Pariss. vivian Acwhirur 8 rdvra xeirasc Aewdire 5€ Havn., Marc., oa GKOVE (Aewdr)Aov Bodl. sscr. Acwidtroo 8 Pariss. But apart from remarking that AcwPiAws 8 rdvra xeirac (although Porson used it to correct Myrioyoo 5 rdvra moira: into Myridxy S& wdvra xeirat, Meineke, Comic. fr. 1V. 675) does not seem to be Greek for anything, I have no suggestion to make. E. Lose. OxFORD. Tr (Ancien ‘ glossoy is kept Di but was (P: (Pz (Fe The cor even in for us n Vevborus (OS In Verr (Fe And pro (Fe Sor by Ver original merely hopeless inform t Bu task, 7 Latina, of Virgil task for | "Avaxpéovts nd the third suppose that at order. If Archilochus \urrwrov, but in which the odern editors , Satisfy the ugh Dindorf ago. of them) do ‘OVE Porson used t, Meineke, ) suggestion t. LOBEL. ait EP FF re | FESTUS, DE VERB. SIGNIF. 284, 30. Tue alphabetical arrangement of dictionaries is a necessary evil. Elsewhere (Ancient Editions of Plautus, p. 13) I have pointed out that the lore of the early ‘glossographi’ on Plaut. Czst. 407: Diobolares schoenicolae miraculae is kept together in Varro’s De Lingua Latina VII. 64: Diobolares a binis obolis. Schoenicolae ab schoeno, nugatorio unguento. Mira- culae a miris, id est monstris, a quo Accius ait personas distortis oribus deformis ‘ miriones,’ but was separated in Verrius Flaccus’ dictionary : (Paul. Fest. 65, 8) Diobolares meretrices dicuntur quae duobus obolis ducuntur. (Paul. Fest. 110, 4) Miracula, quae nunc digna admiratione dicimus, antiqui in rebus turpibus utebantur. (Fest. 442, 7) Schoeniculas appellauit meretrices Plautus propter usum unguenti schoeni, quod est pessimi generis. The compiler of the Abolita Glossary took material from Festus, and this glossary—— even in the miserably shrunk and distorted form in which it has survived—preserves for us more or less recast extracts from a Spanish seventh (?) century MS, of the De Verborum Significatu. Quite a respectable extract is— (OS 3) Ostenta proprie uocantur quae se sine corporibus solidis hominibus ostendunt et oculis atque auribus obiciuntur, qualia quondam apparuerunt uenatio in caelo, arma, Furiae, nocte lux, interdiu tenebrae, In Verrius Flaccus’ source this paragraph (or something like it) stood beside: (Fest. 284, 30) Portenta rerum fieri dicuntur cum solida corpora quae raro se ostendunt apparent, ut cometae, turbines, barathra, tonitrua sereno caelo facta. (Therefore soltda should not be ‘emended’ into insolita.) And probably beside : (Fest, 146, 32) Monstra dicuntur naturae modum egredientia, ut serpens cum pedibus, auis cum quattuor alis, homo cum duobus capitibus, iecur cum distabuit in coquendo. Someone will ask: Why do you not attempt a reconstruction of the books used by Verrius Flaccus? Because Festus’ epitome retained only a fraction of the original work, and of Festus’ epitome less than a half survives. For A-L we have merely Paulus Diaconus’ miserable epitome of Festus. No; the task would be hopeless. Still, in my second volume of Festus, now approaching completion, I inform the reader of what appear to be kindred items. But let me suggest to any enterprising reader of this article a far more hopeful task, The great Liber Glossarum or Glossarium Ansileubt is at last edited (Glossana Laima, Vol. I., Paris, 1926). Much of its material was quarried out of a collection of Virgil scholia, based—I think—mainly on Donatus’ lost commentary. There is a task for not one enthusiast, but for a band of enthusiasts—the reconstruction of 118 FESTUS, DE VERB. SIGNIF. 284, 30 Donatus’ scholia. Ansileubus ordered his monks to chop up the scholia into as many glosses as possible. The strict alphabetical arrangement of the glossary has scattered the kindred glosses far and wide. For example, the scholium of Donatus on Aen. IV. 335 was something like this (cf. Serv. auct. ad loc.) : Promeritus apud antiquos dicebatur qui quid bene fecisset. Ergo promeritam ‘ profuisse,’ ‘ praestitisse.’ Contra commeritus qui deliquit, ut (Ter. Andr. 139) ‘quid commerui aut peccaui, pater ?’ and soon. In the glossary we find— (CO 397) Commeritus : qui deliquit, ut ‘ quid (!) (CO 1167) Conmerui : aut peccaui pater’ (!) (PRO 2354) Promeritus etc., Promeritam etc., and so on. By intelligent piecing together of the scattered and unintelligently chopped fragments we can restore the scholium with a fair amount of certainty. For a long time there has been a call for a new commentary on Virgil. To my mind the ideal commentary would be a judicious selection from Servius and Donatus the latter to be recovered from (1) the Serv. auct. scholia published by Thilo, (2) these Ansileubus glosses. W. M. Linpsay. IN (p. 47) in the f fabulas takes 1 acttonen means have W Terenc (like L In Ci D Laeliu only S funera a noua 5s should court, 1 Jus suggest veliqutt, of Cice lia into as lossary has f Donatus romeritam ut (Ter. itelligently inty. . To my | Donatus, ), (2 ) these sINDSAY. TERENCE AND SCIPIO. In the new edition of Fvagmenia Poetarum Latinorum (Teubner, 1927) Morel (p. 47) persists in attributing to Volcacius Sedigitus’ Liber de Poetis the lines preserved in the final portion of the preface to Donatus’ Commentary on Terence: ‘Scipionis fabulas edidisse Terentium Vallegius (sic) in actione ait hae quae uocantur,’ etc. He takes no notice! of Schoell’s admirable suggestion (Rhein. Mus. 57, 163) that tm actionem were the opening words of the first line. Actto, I think (unlike Schoell), here means the performance of a play. Valleius (or whatever he called himself) seems to have written a prologue for the reproduction of plays of Terence at the funeral of Terence’s patron, Scipio Aemilianus, and to have stressed the rumour that Scipio (like Laelius) was a part-author : In actione hae quae uocantur fabulae, Cuiae sunt? Non has, iura qui (qui iura codd.) populis recens (recentibus codd.) Dabat summo honore affectus, fecit fabulas ? Laelius is said to have written Haut. IV iv; but this prologue probably mentioned only Scipio, since it sought to justify the performance of comedies of Terence at the funeral of Scipio; who himself, thirty-one years before, had the Adelphi performed, as a noua fabula, at the ludt funebres of his father, Aemilius Paulus. Valleius (?), I take it, was a literary freedman of Scipio; at any rate, Vallegius should not be ‘emended’ to Volcactus. Schoell’s explanation of im actionem as ‘into court,’ ‘into judgment,’ is more plausible than his ‘ emendation’ of populis to amplis. W. M. Linpsay. 1 Just as he takes no notice (p. 66) of the the symbol relig, (or reli or vel or rl), a symbol suggestion (Class. Phil. 13, 7) that the puzzling like our ‘etc.,’ used by medieval scribes to indi- veliquit, at the end of a quotation ofa hexameter cate that a quatation was left incomplete, It of Cicero, is nothing but a wrong expansion of really denotes reliqua, not reliquit. SUMMARIES OF PERIODICALS. LITERATURE AND GENERAL. Athenaeum (Pavia). V. Vol. IV. A. Levi writes on the doctrines of Parmenides, examining especially the theory of dofa. Starting with an exposition of ‘Truth’ (Part I. of the poem), he gives an account of Parmenides’ conception of ‘Being’ as unified, constant, and unchanging, and distinguished from the plurality of sensible objects by being the only object of real Knowledge. If with Burnet one calls P. the father of materialism, one must concede him the title of metaphysician on the ground of his ontological distinction between the reality of true Being and the world of experience. L. examines the question as to how far the cosmologies of Part II. of the poem belong to P. or to others; what value he ascribes to them ; and why he expounds them if they are not those he holds. If Burnet is right in thinking that the exposition of the Pythagorean view was made for the benefit of his own pupils, it is difficult to see why there is no refutation, althougk he declares it false (fragm. 8. 50-51), nor yet why he confines himself to the theories of one school only. L. concludes that P.’s hypothesis in 4d£a is the statement of the world of appearances that has the greatest coherence, as it explicitly affirms the existence of the Non-ens, upon which the other cosmologies fall back in particular cases only, and without realizing its Nature. B. Lavagnini points out the self-imposed limitations of the poet in Aeschylus’ Persae. The practical difficulty of representing the contemporary victory to the victors (e.g. the naval battle) and the aidés, which would have made it impious, conditioned the dramatic form. The choice of a distant scene for the action, the absence of a real development or any characterization of important Greek characters, gave him, on the other hand, a freer hand in depicting the woes of the vanquished enemy. Still there is a gradation and a climax to the sense of impending disaster from Atossa’s dream to the messenger’s narrative. The preponderance of the lyric element is thus explained, The form he regards as nearer the more ancient tragedy. P. Fabri contributes an eclectic but not very conclusive discussion of the relative importance of accent and ictus in Latin metres. Classical Philology. XXI. 3. July, 1926. A. Shewan, Asteris and Dulichium. A reply, on geographical and linguistic grounds, to Brewster’s indentification of Asteris with Arkondi and the consequent propositions that Leucas is Same, not Dulichium, and that the regular route from the Peloponnese to the West and North went east of Ihiaki and not up the Ithaca channel. R. J. Bonner, The Administration of Justice undey Athenian Oligarchtes, ex- amines the scanty evidence for criminal and civil trials under the 400 and the 30, and suggests how ra rdtpia may have been interpreted in the judicial sphere. N. W. de Witt, Litigation in the Forum in Cicero’s Time, distinguishes the three areas of the Forum, f. infumum, medium f. and ad Castoris, each having its own character and asso- ciated with the three orders opiimates, equites and plebs respectively, and discusses the position of the ¢ribunalia and of the puteal Libonis (which he connects with the peregrins). B. E. Perry, On the Authenticity of Lucius sive Asinus, produces linguistic data, supplementing those collected by Neukamm, to illustrate the Lucianic character of the Asinus, and concludes that the lost Metamorphoses of which it was an abridgement (made by Lucian or another) was the work of L. himself. E. H. Sturtevant, Centaurs and Macedonian Kings. idurmros, the etymological equivalent of Thraco-Macedonian Kévravpos, is, like other Macedonian royal names, a Greek translation of a native name adopted Kaooadvé KevT-) = | Platonic Poen. 47 Phanag« I.G. i. 2. XX ft misrepre due to ‘ houses, tives. story, in Cicero a Ennius narrativ R. M. | theory Stoic in oneness Eugippu paronom Plethones tract all and Latt of Attic permane apparen books a1 Compoun assimila of 6 51 P. Shore as conti XX i astronon son, ‘ Sa Latin sz meaning their ‘ sa form ans of the px Athenian used to « of the P sympath Homeri: dotus, + translati he theory he gives ‘ant, and r the only terialism, itological nce. L. m belong ids them osition of ult to see , nor yet that P.’s greatest he other re. sschylus’ y to the impious, tion, the aracters, iquished disaster ‘he lyric tragedy. relative nguistic sequent rom the . Ithaca hives, @X- 30, and . W. de ; of the id asso- sses the vegvint). ic data, acter of gement Lentaurs sdonian e name ‘ SUMMARIES OF PERIODICALS 121 adopted by a Hellenizing dynasty. Trojan names show similar treatment, and Kaoodvipa is a half-translated word showing Thr.-Phryg. root xaov- (? cognate with xevt-)=1,-G. ga- (Lat. cavus, Skt. kamas). P. Shorey finds an argument against Platonic authorship in Ep. 1V. 320p. H. J. Rose compares Ar. Birds 1081 with Plaut. Poen. 472 ff. C. P. Bull urges (against Brandis) an early date for the incorporation of Phanagoreia in the native state of Bosporus. A. B. West discusses the place of IG. i. 256 in the lapss secundus of the tribute lists. XXI. 4. October, 1926. F. B. Marsh, Tacitus and Aristocratic Tradition. Apparent instances of prejudice or misrepresentation in T. (e.g. his accounts of Libo Drusus and C, Silvanus) may be due to ‘wrong evaluation of legitimate evidence,’ the traditions of the great Roman houses, probably gathered from personal contact with their contemporary representa- tives. Tenney Frank, Two Historical Themes in Roman Literature. (a) The Regulus story, in itself reasonable, may be accepted as historical on the evidence of Tuditanus, Cicero and Livy; (d) the influence of the contemporary and trustworthy account of Ennius on early Roman history explains the picturesqueness of some parts of the narrative, e.g. the story of Appius Claudius Caecus, which was accepted by Cicero. R. M. Jones, The Ideas as the Thoughts of God. The origins of the neo-Platonist theory of ideas as thoughts of God are to be looked for not in neo-Pythagorean or Stoic influences but in Aristotelian doctrines of thought as God’s activity and the oneness of vovs and vonrév. C. C. Mierow, Some Noticeable Characteristics of the Style of Eugtppius, illustrates E’s use of (i.) adverbs, (ii.) abstract nouns, (iii.) balance and paronomasia. J. W. Taylor, Move Light on Theodore Gaza. Apostolius’s Contra Plethonem (1461) is a reply not to Gaza’s Contva Plethonem (1459), but to an earlier tract alluded to in Bessarion’s De Essentia and datable to 1458. F.A. Wood, Greek and Latin Etymologies. W.G. Hardy, The Hellenica Oxyrhynchia and the Devastation of Attica. The Decelean ravages, being longer and more thorough and affecting permanent stock, made the Archidamian incursions seem insignificant; hence the apparent inconsistency between the accounts of Hell. Ox. and Thucydides’ later books and the statements of T.’s early books. Walter Petersen, Attic rots and tts Compounds. Simple rods (for rus) obtained its realism from analogy of its compounds, assimilated to old adjectives in -ovs (=-oos). H. J. Rose writes a note on the ‘joke’ of 6 511, W. A. Oldfather on the source of Lib. Or. 51. 2 in Dem. Mid. 2. 24, P. Shorey on Laws 822z. Blanche Brotherton defends the retention of Catull. 2. 11-13 as continuous with I-10. XXII. 1. January, 1927. F. E. Robbins describes Michigan Papyrus I., a new anonymous Egyptian astronomical treatise, printing the text and adding a commentary. G. L. Hendrick- son, ‘ Satuva Tota Nostra Est,’ reconciles Quintilian’s dictum with the obvious of all Latin satire, in form and matter, to Greek antecedents by emphasizing the limited meaning of satuvya in Q’s time: he does not deny to Aristophanes or the mimiambists their ‘ satiric’ quality but refers to a definite literary genre with a particular metrical form and usage. Gertrude Smith, The Jurisdiction of the Areopagus, traces the history of the powers of the A. from Draco to the fourth century. A. P. Dorjahn, Poetry tm Athenian Courts. Poetry was normally quoted by the litigant himself, and generally used to elaborate the sentiment, sometimes to decide an issue. A.S. Pease, The Loves of the Plants, examines the origin of the conceit which represents certain plants as sympathetic or antipathetic to one another. W. A. Oldfather writes a note on a Homeric quotation (a 38) in Epictetus (3, 1, 38) (who supports the variant of Zeno- dotus, wéuavre), J. W. Spaeth on the date of Martial 8.67, L. Van Hook on the translation of J/. 22. 418-20, and P. Shorey on Diog. Laert. 7. 81. 122 SUMMARIES OF PERIODICALS XXII. 2. April, 1927. Eliza G. Wilkins, ETTYA ITLAPA A’ ATH én Litevature. The evidence of the passages in which the Delphic maxim is quoted is indecisive between the general interpretation ‘ Do not commit yourself’ and ‘ Do not give surety. B. L. Ullman, The Added Letters of the Greek Alphabet, concludes from forms and names that these were first added to the Western alphabet in the order ¢, x, €: the Eastern inde- pendently developed £, borrowed from the Western ¢ and &é (for its x), and used Western x for the new y. J. W. Hewitt, The Terminology of Gratitude in Greek, attempts, from a study of the synonyms of ydpis, to distinguish and date the various aspects (pay, memory, etc.), under which the ideas of gratitude and its opposite were expressed. Lily Ross Taylor, The Cult of Alexander at Alexandria. The evidence of the older versions of Pseudo-Callisthenes shows that the Alexander Festival was on A.’s birthday and not on the anniversary of His death, and that the centre of the cult was his Temple and not his tomb. A. Shewan, Fishing with a Rod in Homer, repeats his argument that xépas Bods dypavAoo (Q 81, w 253) refers to the hair line, and Suggests in support that in ¢ 395, the only passage adduced for xépas=‘ bow’ in Homer, it refers to the bowstring. E. S. McCartney, Modifiers which Reflect the Etymology of the Words Modified, illustrates the various types of the Figura etymologica, with special reference to Lucretius, who reacted against the more obvious forms of the device and made it subtler. Rachel L. Sargent, The Use of Slaves by the Athenians in Warfare, I., discusses the duty of slaves and their proportion among the noncom- batants in land operations. P. Shorey, Ideas and Numbers A gain, repeats his contention against Adam that Plato’s mathematical ideas are not intermediate between ideas and sense, but ideas of numbers precisely corresponding to other ideas and, like them, admitting xowwvia dkAjAwv. A. M. Harmon writes a note on a conceit in A.P. 5. 74, W. C. Green on é dvdyxns opurdpeOa in Rep. 490 A, W.H. Kirk on horae momentum in Hor. S. 1. 1. 7, and R. S. Rogers on the name Vitia in Tac. Amn. 6. Io. Hermes. LXII. Heft 1. 1927. J. Geffcken, Ein Prinzip antiker Evzaihlungs- und Darstellungs- Kunst. Principle consists in the gradual exposition of an incident, etc., in a series of separated passages usually increasingly definite and detailed. Lead to a more conservative treatment of Greek texts. J. Sykutris, Jsokvates’ Euagoras. Analysis and discussion, showing how the treatment follows the rhetorical rules, largely invented by I. for the encomium, hence its limited historical value. H. Willrich, Augustus bet Tacttus. T.’s treatment of A.’s government in general; family policy; death. Illustrates T.’s injustice to A. and misrepresentations of ‘his character and policy. C. Hoppe, Hevon von Alexandrien. Illustrates his ability and importance as mathematician and physicist, and fixes his date as c. 150-100 B.c. O. Hense, Zu Seneca’s Bniefen. Emendations, etc., to nineteen passages. O. Weinreich, Die Christiantsterung einer Tibullstelle. Defends the authenticity of distich from Capua (Diehl, Inscr. Lat. christ. 1.2211), and shows the sources of its various parts. Miszellen: W. Baehrens, Zur Textgeschichte des Xenophonteischen Kynegetikos. Discusses character and origin of shorter Prooemium preserved in Cod. Vindob. IV. 34 (W). L. Deubner, Zu Aristophanes’ Fréschen, 1.207. Cites parallel to Barpdéywv xixvwv from Pap. Osloenses I., p. 12, No. 1, ll. 233 sq. LXII. Heft 2. 1927. F. Klingner, Virgils’ Erste Ekloge. Discusses technique and subject-matter, and place in gradual development of V.’s conception of Octavian as divine peace-bringer from the fifth and ninth through to the fourth, and eventually to Georgics and Aeneid. R. Minscher, Zur Mesudischen Liedform. Shows that ‘ mesodic’ structure (Hermes 59, pp. 204 sq.), whereby efodos is placed between strophe and antistrophe instead of followi Lyric | from h on the of evic and in could | docum were { when baerbuc U. Ka (spring in Me emend later « Suet. J KUpLe a: Fragm elegy. Deubr L S. schliss and pi (2) to. gramn U. v. of Chi Plato, Harde Discu: Polyb W. Si raised conuiut cena T Hexa tion ( indepe U. v. Ethics Pp Groni: passag adtxias unusu cf, Be 3 ‘ dar) Polluy lence of the the general L. Ullman, } that these stern inde- ), and used e m Greek, the various posite were evidence of ival was on of the cult ney, repeats r line, and =‘ bow’ in Reflect the etymologica, is forms of le Athenians le noncom- ; contention n ideas and like them, A.P. 5. 743 zomentum 1D Principle f separated onservative discussion, d by I. for bet Tacttus. Illustrates C. Hoppe, atician and a’s Bniefen. evung einer Inscy. Lat. . Baehrens, and origin ubner, Zu Osloenses I., natter, and ace-bringer und Aeneid. Hermes 59; instead of SUMMARIES OF PERIODICALS 123 following them, occurs, though rarely, in Sophocles and Euripides, and even in the Lyric poets. M. Wellmann, Timotheos von Gaza. Discusses two series of excerpts from his zoological writings, and the relation of these to earlier and later treatises on the same subject. H. Dessau, Zur Inschrift von Skaptopava. Sceptical view of evidence drawn from formulae by Mommsen, Preisigke, Wilcken as to petitions and imperial replies. Doubts (1) that originals were proposits; (2) that petitioners could only obtain reply by recourse to ‘ exhibited’ documents; (3) that Egyptian documents were normally forwarded through praef. Aegypt.; (4) that these Abelis were propositt at Alexandria, Shows that last rests on evidence only from year 200, when (not 202) Severus was in Egypt. W. Kolbe, Die Seleukidendva des I. Makka- baerbuches. Maintains use of era beginning 311 B.c. (spring) against views of U. Kahrstedt [normal era 312 (autumn), 166 (spring)—146/5, national Jewish era 312 (spring)]. K. Strecker, Ovidianische Verskunst im Mittelaltery. Classical metric neglected in Merovingian Age, but observed in twelfth, thirteenth centuries. Shows that emendations on Ovidian principles in Pyramus and Thisbe poem are confirmed by later examination of a Berlin MS. Miszellen: F. Lammert, Laevius Meltssus ? Suet. De Gram. 3. Nosuch person. G. Breithaupt, Uber zwei Anredeformen bei Epiktet kvpie dyyeAe not Antisthenes ; xvpre 6 Geos not Christian—both magical. E. Fraenkel, Fragment einer sympotischen Elegie. Solon, Frag. 13 (Bergk), really from a banquet- elegy. W. Kranz, Bdrpayos Tvpivos. Plato, Theaet. 161c, text supported by Deubner’s note (see above). LXII. Heft 3. 1927. S. Luria, Zur Geschichte dev Praskvripte in den Attischen Voveukleidischen Volksbe- schliissen. Seeks objective criterion of date in form of Prescript. Finds it in position and presence of Archon’s name, etc., (1) before Cleisthenes’ name as postscript ; (2) to 487, mention of Boule and grammateus ; (3) to 421, always praescript—prytanis, grammateus, epistates; (4) to 403, prytanis, grammateus, epistates, archon. U. v. Wilamowitz-Moellendorff, Lesefnichte CCXVIII.-CCXXX. Heraclitus: Ion of Chios: Aesch. Agam. 350: Antiphon: Eurip. Herac. 945-953 : Peivithous fragments : Plato, Epp. 344a: Aristoph. Eg. 407: Menand. Ilepix. 144: Plutarch, Op. mi.: Harders, Okellos: Riese, Anth. Lat. 199. W. Aly, Die Entdeckung des Westens. Discusses voyages, etc., in the Atlantic before 500 B.c., and dates the sources of Polybius, Pliny, Avienus on this subject. Prints Greek text of Hanno’s inscription. W. Siiss, Zu Lucilius. (1) Considers some grammatical and orthographical questions raised by fragments of Bk. IX.; (2) discusses view that Bk. XX. contained a convivium Granit, and was model for the cena Nasidieni (Hor. Sat. I. 8) and Petronius’ cena Tvimalchionis. Ed. Fraenkel, Die Vorgeschichte des Versus Quadratus. Unlike Hexameter and Anapaest, the trochaic Septenarius not definitely a literary importa- tion (by Livius Andronicus), but already traditional in Latin; not, however, an independent invention, but borrowed from Greek popular verse. Miszellen: U. v. Wilamowitz Moellendorff, Neleus von Skepsis. Name restored in Aristot. Ethics 11. 7, 1205a 17. A. Turyn, érurxoreiv und érurxid fev. Mnemosyne. LV. 3. 1927. P. H. Damsté has notes on Ammianus Marcellinus (XIV.-XIX.). B. van Groningen, ’Aéix«ia, argues that in Pseudo-Aristotle, Oecon. II. (p. 1348b 9-16), in the passage, trois re woAiras xatidwv ovoas Sixas woAAds, cal peyddas éx woddAov Xpovov ddixias tovrous 81d rodeuov, Sixarripiov KaBicas mpoeirev, x.7.r., the word adi«ia has the unusual meaning of an undecided suit. For the adjective dducos= without a lawsuit, cf, Bekker, Anec., p. 343, ddixos 6 pr Exwv Sixny Sid 7d pr SexdferOai run, and “Apxerros dats oddos, Sixasos &Sixos ; also Gdixos ypépa in Eustathius, Athenaeus, Lucian, and Pollux. In the passage from Ps.-Aristot, he would read with Goettling Sixas 7. xat 124 SUMMARIES OF PERIODICALS fe éx mwoAAod xpovov dédixias. B. van Groningen, On an Egyptsan Dedscatory Inscription. The inscription comes from Coptus, and consists of four fragments, one of which is in the museum of London University and the remainder in the museum at Alexandria. Van G. supplies a restoration of the text with comments. He assigns the inscription to the reign of Alexander Severus. C. Brakman has a short appendix to his article on Commodianus (Munem. LIV., 1926, p. 121), in which he criticizes certain views of Weyman, Beitr. z. Gesch. d. Christlich-Lateinischen Poeste (Miinchen, 1926). H. J. Rose, De Jove Latiavi, discusses the origin of the false assertions of Christian writers that the Romans were accustomed to offer human sacrifice to this deity. R. postulates an elegiac poem on the Roman sacra, which was read and misunderstood by Tatian and Tertullian, who thought that a bestiarius was sacrificed at the Latin festival This poem was also misinterpreted by Porphyrion. C. Brakman has an instalment of critical notes on Livy XLI.-XLV. J. D. Meerwaldt discusses various fragments of the comic poets which refer directly or indirectly to Demosthenes. He finds further satirical references by unknown poets imbedded in stories told of Demosthenes by Plutarch and other writers. H. D. Verdam, De Platonis Dialogo Thrasymacho qui vocatuy, argues, in opposition to von Arnim, that the first book of the Republic is an integral part of that work implied in the later books, and consistent with them, not an early and originally independent dialogue. J. H. Thiel, De Amtiphontis ovatione prima—assuming that the speech deals with a real case—argues that the accuser has failed to prove that his stepmother has poisoned his father. The evidence favours the view that the potion was administered by the éraipa sua sponte in hope of recovering the affection of her master. P.H. D. contributes two humorous poems in Latin hexameters. Neue Jahrbicher fiir Wissenschaft und Jugendbildung. III. 5. 1927. H. Berve, Ionien und die griechische Geschichte. The Ionian revolt and the Delian League were unfortunate episodes inspired by the ideals of Greece proper: the King’s Peace of 386 was a blessing to the Ionians, who cared little for political freedom, and could not flourish without close union with the interior. F. Geyer, De Tvagth im Schicksal des makedonischen Volkes. Philip might have been content with a Macedonian-Greek Mediterranean: Alexander drained Macedonia of men, and left her too weak for the real incorporation of Greece. Philological Quarterly (Iowa). VII. 1. 1921. B. L. Ullman, Latin MSS. in American Libraries. A supplement to the lists given in P.Q., 1922, p. 100, and 1926, p. 152. The list contains a codex of Horace of eleventh to twelfth century (Pierpont Morgan). The other codices are all of fourteenth to fifteenth century. R.C. Whitford, Juvenal in England, 1750-1802. Philologus. LXXXII.4. 1927. A. Rupprecht, Die demosthenische Prooemiensammlung. The collection represents fragments from the literary remains of Demosthenes. Writer discusses historical and literary importance of the ‘proems.’ Appendix on the MSS. tradition of the collection. W. Ruppel, Politewma II. Continuing earlier article, reviews use of word in inscriptions from circ. 200 B.c. to circ. 200 A.D. Summary of results. F. Schemmel, Die Schulzeit Kaiser Julians. A connected narrative with the authorities printed in footnotes. Miscellaneous: N. Wecklein; Discusses the Ephymmta in Aeschylus. Question affects the dating of the Supplices since that play shows affinities with the Ovesteia in this matter. R. Bliimel; in Petronius, Satyr. 62 mata (= macta) vita are imperatives, tau an interjection, the whole phrase in origin a spec- tator’s cry to a gladiator. LXXXIII. 1. 1927. O. Hey, ‘Apapria I. Use of the word down to Plato. K. Praechter, Platon Praéformist ? Criticises the attribution of this theory to Plato by Michael of Ephesus (Comm Epicu sugges Rufus Castigi einey F ninth Aristo L S text a TQO4, | club wv O. He and 15 not me Sopho T. Bir Pliny Apule! Cyprie Ennoc quod V outside any re Gelehrs mediae Querie antva I tion fr Corn. Sqq. ; | aiperew proble attribu and m Si E A prciu: RI H E. Ka Morali R. He poratec Athen 1 events, (agains an exc DICT. J. Me: materi: Dedscatory 1ents, one museum nts. He s a short which he hen Poeste the false xr human va, which bestsarius reted by 1.-XLV. r directly unknown writers. pposition hat work riginally that the that his 1e potion n of her 5- +1927. e Delian per: the political eyer, Dee at with a and left the lists Horace re all of 02. presents istorical n of the use of results. thorities jmnia in + shows 62 mata @ spec- , Platon +phesus SUMMARIES OF PERIODICALS 125 (Comment. in Arsstot. Graeca, Vol. XIV., pars. III., pp. 25, 20 sqq.), and notices Epicurean view on subject. T. Birt, Marginalien zu lateinischen Prosatkern, 1. Critical suggestions on the text of Cicero, Ad Herennium, Caesar, Varro, Vitruvius, Curtius Rufus, Petronius, Seneca. T.O. Achelis, Die hundert dsopischen Fabeln des Rinucct da Castiglione. A detailed examination of this Latin version. C. Mengis, Fragmente einey Fretburger Prudentiushandschrifit. Description and text. The MS. dates from the ninth century. Miscellaneous: O. Schroeder discusses Astvopha in Euripides and Aristophanes, LXXXIII. 2. 1927. S. Luria, Ein mslesischer Mannerbund im Lichte ethnologischey Parallelen. Discusses text and interpretation of famous inscription (compare Wilamowitz, Berl, Sitz-Ber. 1904, etc.) relating to club of poAroi at Miletus. Argues that political importance of club went back to very early times. Comparative ethnology furnishes many parallels. O. Hey, “Apapria II. The Aristotelian usage with special reference to Poet. 1453a 10 and 15. Since Aristotle there, as normally, uses the word to imply intellectual error, not moral fault, there is no contradiction when he illustrates from the Oed. Rex of Sophocles. Oedipus’ error lay in forgetting original purpose of his visit to Delphi. . T. Birt, Marginalien zu lateinischen Prosaikern, 11. Critical suggestions on the text of Pliny the Elder, Quintilian, Pliny the Younger, Tacitus (Dialogus), Suetonius, Apuleius, Justin, Tertullian, Minucius Felix, Scriptores Htst. Aug., Ammianus, Cyprian, Julius Valerius, Martianus Capella, Arnobius, Palladius, Donatus, Ambrose, Ennodius, O.Immisch, Der Hain dey Anna Pevenna. In Martial 1V. 64, 16 read et, quod Virgine crudiove gaudet (compare VI. 42, 18 cruda Virgine). The nemus standing outside the walls, crudus means ‘ fresher than when tapped later.’ Context excludes any reference to magical value of menstrua. P. Lehmann, Reste und Spuvren antiker Gelehrsamkett in mittelalterlichen Texten. Indicates information to be obtained from mediaeval commentaries (generally of Irish origin), illustrating from one on Orosius. Queries source of line cited by Sedulius Scottus and Walter Map—pauper poeta nesctt antva Musarum, Miscellaneous: E. Ziebarth; Publishes and comments on inscrip- tion from Dystos relating to mortgage. O. Viedebantt: Metrological notes on (1) Corn. Nepos, Atticus 2, 6 (read septem modiu); (2) Josephus, Arch. III. 320 sqq.; (3) Metvol. Script. I., p. 208, 5-11. F. Zimmermann: In Charito I. 8, 1, for aiperews read adéoews = vemissio (a medical term). N. Wecklein: Deals with textual problems in Homer. A. von Blumenthal: Discusses temple of Apollo at Delphi attributed to Trophonius and Agamedes with special reference to H. Afoll. 294 sqq. and meaning of ovdés in 1. 296. L. 298 should follow 1. 299. Supplement, Band XIX., Heft 3. 1927. E. Brandt, Untersuchungen zum vomsschen Kochbuche: Versuch einer Losung dey A pictus-F rage, pp. 164. Rheinisches Museum fir Philologie. LXXVI. 1927. Zweites Heft. H. von Arnim, Die Echthest dey Grossen Ethtk des Anistoteles. In reply to E. Kapp’s criticisms A. argues again for his view that Aristotle wrote the Magna Movaha and then as a revised edition the Eudemian Ethics. (To be continued.) R. Helm, Hieronymus und Eutrop. An important study of the new material incor- porated by S. Jerome in his translation of Eusebius. (To becontinued.) W. Judeich, Athen und Theben vom Konigsfrieden bis zur Schlacht bet Leuktva. A critical account of events, defending the tradition about Athenian action in the liberation of Thebes (against Grote) and that about the short-lived peace of 374 (against von Stern), with an excursus on the battle of Leuctra. Th. Birt, Was heisst BAZIAEY2? Was heisst DICTATOR ? connects Bacircts with Bdfw, ‘ speaker,’ and dictator in Latin towns. J. Mesk, Libamsos TIEPI AOYAEIAY. This speech (XXV. in Forster) contains material from Bion, but the form seems to be due to Libanius himself, as a comparison SUMMARIES OF PERIODICALS with VI. shows. Miszellen. J. Mesk, Zu Martial VI. 14. Put a question mark in l. 4 after Laberi. C. Fries, Ad Cie. or. Phil. 1. 14, 35. Read omns potest actate from omni potestate. B. Warnecke, Ludi barbari. Native Roman players, in opposition to histriones, who produced paliatae after their introduction. Fr. Marx, Critica hermenceutica. I. x@por in Ditt. Syil.* 1078 means xopoi dvéipov. II. On ad Hevenn. IV. 50. 63. III. Suet. Calg. 50.2 is parallel to the probably Suetonian statement in S. Jerome about the madness of Lucretius. M. adds remarks about the arrangement of the liber usvorum sllustrium. 126 LANGUAGE. Indogermanische Forschungen. XLV. (1927). Whole volume dedicated to Thurneysen on his seventieth birthday. J. van Ginneken writes a long and important article, with a wealth of botanical and zoological as well as linguistic illustration, to demonstrate the existence of biological factors in sound change—inherited tendencies as revealed in modifications, mutations, and crossings. According to A. Meillet, the pattern for dépocus which has replaced *époua (cf. Arc. éfeAavvora) was dépwyse rather than ri?yws. F. Sommer examines instances from the literature and inscriptions of independent subjunctive and optative used side by side in Greek (with or without av, xé), differentiating between the two moods, and concludes that Arc. ayevdjwv av must be not subjunctive but optative (in ‘future’ sense); -7-wy he explains as graphic for -ei-wv, -wy 1 sg. following the pattern of the subjunctive -w, with the addition of -v of the secondary tenses. E. Fraenkel, on Lithuanian syntax: the incorporation (in crystallized form, by the omission of verb, etc.) of clauses, originally parenthetic and independent, into the structure of the sentence; the genitive of the possessor and of definition with ellipse of word of reference ; compendious comparison ; adverbial case forms governed by prepositions; expressions of comparison. N. van Wijk discusses iterative verbs and other forms denoting repeated action inO.Ch.Sl. Ina paper on semantic change M. Leumann points out that such change is dependent upon grammatical and syntactical conditions, not on etymological—i.e. a word not standing alone, but from its use in a sentence, (1) will not always be given precisely that shade of meaning by the listener which the speaker intended, (2) subsequently will be used (by the listener and by others after him) with the shifted meaning. Similarly, the second element of a compound will lose its original force, not alone, but in the compound. L. maintains that there is no true change of meaning in figurative expressions so long as they are felt as figurative. E. Sievers seeks to establish varieties of intonation in respect of quantity and accentuation (1) for the eight different cases, even in un- declined words (as in Modern German), or where syncretism has taken place (as in Classical Greek), and similarly (2) for the different genders, and (3) for different stem classes of nouns. He gives numerous illustrations of his account of these varieties of intonation from Sanskrit, Greek, Latin, and Germanic. W. Porzig, after making a distinction between an older category of ‘ kinds of action,’ each ‘ Aktionsart’ having its own stem-formant, and a later category of ‘ Aspekt,’ several formants being grouped together under each aspect, examines the verbal sk-formation [the sk-nouns being regarded (a) as later than the verbs, (5) as not forming any clearly defined class, and (c) as distinct from g-extensions of s-stems, in which each element has its separate significance], and concludes that the verbal sk indicates an event as taking place in steps or stages. H. Jacobi: Prakrit #h from Sanskrit sz, though a vulgar s/, which J. shows to be attested in certain proper names. A. Debrunner discusses metrical lengthening and shortening in Homer: insists, against Brugmann and Gintert, on metrical lengthening in words of the type of o:fupwrepov—oreios (€ 194, cf. the irregular owéos) follows the rhythm of omeiovs (68, 226); peuadros, etc, (-w- after vowels), KNpUK= ( a (u —- at metrical variation etre, but dre, or wl tive *-oy nominati (-@ befor in Frenc to distin; paramou tions of Greek, a lative’) : ipsus t ¥ Greek a ‘ profanu fidem fa ‘tear alc *odiory ‘induitor Idem, To but only a borrov ‘auTds, Ss genitive class of simple v was set distribut Germani Sanskrit, Benfey’s kat Adxw latter ad connecte Gothic h simahs fc expected H. Amm the olde: fac, née fa but ‘mo etc., nolsi ‘weak’: ‘high,’ * lindan—i (Hesych ‘basket’ darbha- ‘ of the s mark in ate from sition to nencutica. . 50. 63. , Jerome at of the ledicated botanical tence of fications, hich has Sommer junctive entiating yjunctive wy I Sg. scondary ed form, ent, into ion with yoverned ve verbs c change ical and mut from aning by : listener element ind. L. so long ation in n in un- >e (as in ent stem varieties making ’ having ts being sk-nouns ed class, separate place in 7, which metrical ntert, on cf, the -w- after SUMMARIES OF PERIODICALS 127 vowels), started in forms like re@vndérwv; metrical shortening in £epév (unexplained), xnpuxe (after P 324), Tpoty (uv v— rather than --—-, to avoid ~—--), Marx’s law (u -- at end of verse for words of molossus shape) rejected; his examples all false— metrical shortening before ox- in words with iambic opening (not, pace Hirt, an I. Eu. variation between ox- and «- carried over into Greek). Idem, no asyndeton with ere, but write ed re ‘and rightly,’ except where (edre-clause following) either edre= dre, or where there is confusion with nite. Idem, eipiora properly accusative (nomina- tive *-of) with Znyv (as in half the recorded instances of the accusative), and hence nominative and vocative, F. Marx would read in Lucr. V pelageque at 35, guacsita (-@ before pr-) at 5, excising -gue. L. Spitzer writes on ‘split’ prepositional phrases in French (‘avec sur la figure grave un bon et doux sourire’), E. Hermann seeks to distinguish between various types of subjective aspects (speaker’s point of view paramount) and objective aspects (difterence of fact); he gives diagrammatic illustra- tions of the distinctions which he makes, with examples from Russian, German, Greek, arid Latin. W. Havers examines ‘enumerative’ (or perhaps better ‘ accumu- lative’) methods of expression, often with apparent asyndeton and pleonasm (cam ipsus t voga), especially in verbs of motion, taking, beginning, etc. E. Schwyzer, Greek and Latin etymologies: BéBnAos, *be(gh)- ‘outside,’ BnAds ‘limen,’ hence ‘profanus’; ériopxos is 6 éri dpxw Bds, 6 rdv dpxov rarjoas, i.e. ‘ perfidus, qui per fidem fallit’; pds, *Fpa(s)a (cf. amrnupa, dmovpas for *-Fpda, *-Fpds) from *ura /é- ‘tear along, move rapidly,’ pawrwvn either an abstraction from *jdiorwvos or for *sdurry avy; Stium [: d-mentum ‘enveloping skin,’ 6- from *oy(z)-, Umb. anoushimu ‘induitor ’) i.e. properly ‘a going clothed or shod,’ hence freedom from toil (nudus ara). Idem, Toch. éle-smilne ‘smiling,’ I.Eu. *smei-, no evidence of Toch.-Kelt. relationships, but only of the distribution of this root. P. Kretschmer sees in Hom. ¢idos ‘own’ a borrowed, or possibly a proto-I.Eu., word with Anatolian connexions (cf. Lyd. dis ‘avrds, self,’ bslis ‘suus,’ bil ‘suus,’ cf. idr-repos, -raros, and for --formant Hittite genitive of personal pronoun in -/). W. Meyer-Liibke points out that in the small class of nouns in -tum (proper to compound verbs) formed directly from the simple verb stem (e.g. gaudium) all have the -d- (-dh-) formant; the type, he suggests, was set by prandium, prandere. T. Frings and J. Niessen write on the geographical distribution and history of the words for Easter, Wednesday, and Saturday in Western Germanic. L. Siitterlin: two pages of etymologies (thirty-six words, Greek, Latin, Sanskrit, German). J. Wackernagel, while admitting certain difficulties, revives Benfey’s explanation of heus as 2 sg. impv. ‘hear !’—for Hesych. ads: avrés. Kpjres kat Adxwves he would write (two glosses): ads: avros. Kpijres, ads: ots. Adxwves, the latter ads being then cognate with auris.—éxarn-BdAos, -BeAérns (no digamma) not connected with éxnBdAos (F-), but with éxardv (for -n- cf. €AagdnBoAos)—Latin censére : Gothic hazjan ‘praise,’ Sanskrit sams- (cf. Vedic -éds-)—sentence haplology: Sanskrit simahi for sim imahi, but observe that the accentuation (not *simahi as would be expected from imahe) has followed that of other -imaht endings (e.g. vamsiméhs). H. Amman collects prohibitions in Greek (Homer) and Old Latin (Plautus) to deduce the oldest types (Homer almost entirely y» rote type, Old Latin no normal type, ne fac, ne faxis, cave faxis corrective, ne factas—except from slaves—not truly prohibitive, but ‘ motivating,’ ne feceris, cave feceris preventive, nolt facere corresponds to ne fac, etc., nolito facere to ne feceris, etc.). H. Giintert, Greek etymologies—Cretan adds ‘weak’: Vedic dgu- ‘small’ (*al-nu), ddéw, ddAivew (Soph. fr. 995)—Cretan Avrros ‘high,’ *leudhtos (cf. Gortyn. pérros *medhios), Sanskrit rédhatt ‘increases,’ Gothic lindan—ipov- opjvos. Kpires (Hesych.): Latin susuvrus—xdorov fvAov. ‘A@apaves (Hesych.): Latin castrare, Greek xed{w ‘split’ (cf. G. Holz: Latin clades)—ddpry ‘basket’: conflation of two roots *tarp- (cf. Hesych. réprn) and *darbh- (Sanskrit darbhd- ‘tuft of grass," W. Russ. dorob ‘basket’). F. Slotty continues his studies of the sociative etc. use of 1 pl. for 1 sg. in Greek, Sanskrit, and Modern I.Eu. 128 SUMMARIES OF PERIODICALS tongues (Cech, French, Italian). J. Vendryes supposes Gothic to have preserved two distinct I.Eu. terminations for genitive plurals -dm and -m, the former in o-, é-, and «-stems afterwards replaced (after loss of -o- and -m, cf. gutun ‘Gotorum’ in the’ Pietrossa inscriptions) in o- stems by ~ proper to #-stems (from *-2jdm). R. Meissner. writes at length on the Germanic forms and meaning of the name Hamlet. P. Kah q the tradition of pointing in Hebrew texts, 3 Revue de Philologie. LXXVI. 1. 1927. C. Dubois, L’Oliviery et ’'Huile d’Olive dans Vancienne Egypte. 11. The Re Epoch. ‘The cultivation of the olive for its oil seems to have been almost entir restricted in Roman times to the Arsinottes Nomos. Other oils (especially €A«< paddyivov) were extensively manufactured. With regard to olive oil, the evidencil points to a monopoly of sale, not of manufacture. C. Saumagne, Sur la Lot agraire’ de 643/111, Essat de Restitution des Lignes 19 et 20, p. 50. Proposes restorations (too; long to quote here) which agree with and explain Appian’s account (B.C. I. 27): of the Lex Thoria, and contradicts Mommsen’s assumption of the abolition of the uectigal. P. d’Hérouville, Une Formule cicéronienne qui a fait Fortune. Examples, ranging from Cicero to St. Gregory, of expressions like cogitanti mit... (Cica De Or. I. 1). J. Carcopino, Encove le Catalepton. A rejoinder to E. Galletier, (Rev. Phil., April, 1926). i LXXVI. 2. 1927. 4 L. Robert, Etudes d’Epigraphie grecque. Restorations of many inscriptions,” chiefly from Caria. F. Gaffiot, Quelques Passages des Lettres de Caelius a Cicévon, Defends the text of M. in numerous passages, and seeks to correct the edition of | Antoine on several points, P.Chantraine, Le Réle des Désinences moyennes en Gree ancien. In Homer no precise distinction in sense between Active and Middle forms” is found; Middle forms are usual in secondary tenses and in participles. The contrast between the two sets of forms became more and more accentuated. The distinction began tentatively with some formulae—e.g. roAczov roceiv and roetoBar— | and was carried out most thoroughly and most delicately in Attic; the various dialects developed it independently. S. Lambrino, Archaeology in Rumania. A review of the work of V. Parvan. e preserved © ner in o-, é-, | m’ in the © R. Meissner | P. Kahle :7 ee The Roman ost entirely” ially €Aacov! he evidence | Lot agvaire} rations (too B.C. I. a7)a ition of the? Examples, | {pa (Cic. 3 . Galletier = nscriptions, | a Ciéron, % edition of © nes en Gree | iddle forms © iples. The | ated. The ™ rovetoGar— | he various © A review |