V. 54-56. No note is offered on the personages here mentioned: nor again is any explanation given of the allusions in vv. 109-110 (sustulit iras-aera tabificum). 457. Movitque Ceraunia nautis, brought

Ceraunia nearer to the sailors.' Hardly Latin: the Berne scholia say quoniam montes navigantibus moveri videntur. 795-6. Neuterque recedens Sustinuit dixisse vale. Rightly condemned by Kindler as absent from Montepess. and Voss. sec. (first hands), and as probably a gloss on the preceding words. VI. 382. Bebrycio, all the scholia. 475. Direxit (made straight) should certainly be written derexit: indeed there is some carelessness about orthography in this text, which allows nequidquam, querela, diriguit, negligo, tentare, and poenitet to survive here and there. VII. 2. The Berne scholia mention a variant lucifugus for luctificus. 141. "Erigitur is straightened."" Surely erigitur cannot bear this meaning; I had conjectured derigitur, when I found that it was the reading of the Berne scholia. 199. Kindler reads numen for lumen, perhaps rightly. 257-8. Haec eadem est hodie quae pignora quaeque penates Reddat, et emerito faciat vos Marte colonos. These lines are omitted in Voss. sec. and added outside the text in Montepess. Kindler points out that emerito Marte is not Latin, nor the variant emeritos Marte, and rightly condemns them. 325. Mr. Haskins translates iugulum 'slaughter': with what authority It is difficult not to suppose the word to be corrupt. Ignoti hostis seems most naturally to mean an 'unknown,' i.e. plebeian, enemy: compare 578. In plebem vetat ire manus monstratque senatum; then, reading imputet with the Berne scholia, we get the general sense that Caesar will count the slaughter of a plebeian as a crime. 398. Crimen civile: not 'the guilty result of civil war,' but a reproach to the citizens.' 451.

Thyestes the Berne scholia, rightly. 607. The variant mentioned by the scholia, succensor for successor, should have been mentioned. VIII. 332. Secundum Emathiam lis tanta datur = in favour of Emathia: not in accordance with the verdict of Thessaly.' 474. Monstra, cf. infr. 548'; 548. For monstra cf. 474. supr.' 837. Si saecula prima Victoris timuere minas, i.e. cum recens vicerat Caesar. Schol.' Surely the meaning must be if the former generations of Rome were afraid that the conqueror would violate, as he had threatened to do, the ashes of the conquered '; in allusion to such a fear, for instance, as moved Sulla to give orders for the burning of his body after death. IX. 435. Nulla sub illa Cura Iovis terra est: 'under that ground there is no thought of the sky': as the Berne scholia say, non curat si non pluat. In 437 immotis seems to mean not stirred by cultivation.' X. 110. Romana in saecula: fashion, manners.' Can saecula mean this, which is usually expressed by saeculum? or does not Romana saecula rather mean 'the generations of Rome?' True, Juvenal (IV. 68) says tua servatum consume in saecula rhombum: 'your ages;' this, however, is the grandiose language of flattery. Saecula is also used as = saeculum, fashion, in the Epicedion Drusi: but this is just one of the points which make against the antiquity of that poem. 219. Veterum, e.g. Euripidis.' Diels points out that the theory came from Anaxagoras, and was also known to Aeschylus.

These remarks are made in the hope that the book, which will certainly hold the field in England, and justly, for a long time to come, may be improved in a second edition. No superhuman labour would be required to this end. It is much to be wished, on all accounts, that a second edition may soon be in demand. H. NETTLESHIP.

Die Griechischen Bühnenalterthümer. Von ALBERT MÜLLER. Freiburg i. B. 1886. Pp. viii. 432. 10 Mk.

A. MULLER originally intended, as we learn from his preface, to treat his subject as an appendix to the third volume of Blümner and Dittenberger's new edition of K. Fr. Hermann's standard work on Greek Antiquities; but as the subject required a fuller treatment than it would have been possible

to give it in this shape, he decided to publish it as part 2 of this volume. The book, which contains twenty-two illustrations and plans, is divided into three chapters; the first (p. 1 to 106) deals with the structure of the theatre; first beginnings of dramatic performances; sites and ruins of theatres ; the ground-plan of a Greek theatre according to Vitruvius; stage, place of spectators, orchestra; acoustic properties; technical names for the different parts of a theatre; the

Odeion; uses to which theatres were put when not required for performances; the theatres of Athens and Attica. In the second chapter (p. 107 to 308) the parts of a performance are discussed-places of actors and chorus, decorations, doors of the stage, etc.; changes of scene, etc.; actors and acting; chorus; dresses in tragedy, satyric drama, and comedy; masks; spectators. The third chapter (p. 308 to 414) treats of the times when performances took place; the expenses of such performances; poets and instructors of the chorus; state copy of plays, etc.; proagon, judges, didascalia; dramatic performances outside Attica; σύνοδοι τῶν περὶ τὸν Διόνυσον τεχνικών. Then follow four pages of addenda and three indices. In the former Müller touches on various points raised in recent publications which he was not able to make use of for the body of the work, and which tend to upset generally received opinions. Lipsius e.g. has shown that the comedies were performed separately on one day before the tragedies, that the choregus in a dramatic contest competed in his own name, not in that of his puλý, and that in such a contest aot a tripod but a tablet was awarded as prize for the best performance. The views expressed by von Wilamowitz years ago (cf. Hermes 21, 4, p. 597 foll.) have been borne out by the excavations in the Dionysiac theatre at Athens, the results of which Dörpfeld has communicated to Müller; he says, before the fourth century there was only a large circular orchestra of polygonal stones, close to the old Dionysiac temple; there were no stone seats, but the spectators sat on the slopes of the Acropolis or at most on wooden benches; there was no solidly built σkný. Besides this there was a second orchestra in the market-place (to the west of the Areiopagus) which was probably built in the same way. Here it was always necessary to erect bigh scaffolding for seats, since there was no slope here.-Is this latter not almost certainly the opxýorpa mentioned in Plat. Apol. 26 D In Boeckh's opinion this passage refers to the sale of books in the orchestra of the Dionysiac theatre when no performances were going on; Fränkel in his edition of the Staatshaushalt (ii. p. 13* f.) objects to this view, and sees in the passage rather an allusion to performances in the theatre (at a drachma admission for the three days) of plays of Euripides which were saturated with the doctrines of Anaxagoras. L. C. Purser (Class. Rev. no. 5 and 6, p. 150), as I think, effectually meets Fränkel's objections, and A. Müller (p. 82 n.) is no doubt

right in following Schöne, who holds that the passage refers to the trade in books which was carried on in that part of the marketplace called opуýσrpa; Photius has s.v. ὀρχήστρα πρῶτον ἐκλήθη ἐν τῇ ἀγορᾷ, and Nicophon (Mein. fr. com. ii. 2, p. 852) points to the same locality, when he mentions the Bißλionwλai amongst the motley crowd of sellers of figs, leather, spoons, sieves, etc., whom we can only expect to find in the market-place. Whilst Müller's view on this point seems to me preferable to that of Fränkel, the latter (ii. p. 111* f.) is probably right in what he says about the ȧywvobérns. If that official, who remained in office a whole year, had had to defray all the expenses of all the choruses out of his own pocket (p. 340), it would have been singularly ungracious on the part of the duos to deprive him of the title of χορηγός, for ὁ δῆμος ἐχορήγει, ἀγωνοθέ Tηs o deiva, etc. In some instances (e.g. C.1.G. 1444) it would indeed seem that the ȧywvoOerns defrayed all the expenses, but on the other hand it is clear from an inscription from Iasus (Le Bas-Waddington, iii. no. 255) that the ȧywvolérns there did not do so, for the names of several citizens besides the αγωνοθέτης are given as having furnished funds. Moreover we find in inscriptions the phrase dywvoerovvros dià Bíov (Tralles, Colossae, Bull. de Corr. Hell. xi. p. 298, p. 354). Fränkel (ii. p. 125*), as it seems to me, is again right in following Thumser, who infers from Dem. 20, 18 foll. that metoecs could perform at Athens choregia on other festivals than that of the Lenaea; for metoec choregi in Iasus and Rhodes, see Le Bas-Waddington iii. no. 256, and Ross, inscr. ined. no. 278. In connexion with this question a few other points might have been touched on, such as the voluntary choregia e.g. of Hypereides (Ps. Plut. vit. x. orat. p. 848 E), for being trierarch at the time he was exempt; the official title Tù ȧywvápɣu in a Boeotian inscription (Collitz, Samml. d. gr. Dialekt.-Inschr. i. no. 812a). In other parts of the book too I have noticed some slight omissions; on p. 295 the νομοφύλακες are not mentioned as possessing pocopía (Lex. Rhet. Cantabr. p. 673, 27 foll. καὶ ἐν ταῖς θέαις ἐναντίον τῶν ἀρχόντων ἐκαθέζοντο cf. Photius s.v. oi voμopúλakes Tives); on p. 387 an inscription from Samos giving a list of victors at the Heraea (P. Gardner, Journ of Hell. Stud. vii. p. 147 foll.) is not referred to (l. 3 ὑποκριτὴς παλαιᾶς τραγῳδίας Δημήτριος, 1. 8 τοὺς ποιητὰς τῶν καινῶν Taywor τοὺς ποιητὰς τῶν καινῶν κωμῳSur); the first Amphictyonic decree (p. 394, n. 3) dates from shortly after 279 B.C., for

the Phocians did not belong to the Amphictyonic confederacy from 345 to 279 B.C., and lastly, the question as to the number of choreutae who voted in the trial of Orestes might have been more fully treated (p. 202, n. 3); Müller puts little faith in the statements of scholiasts (p. 108), yet here he says, Für die Eum. sind 15 Choreuten bezeugt, (schol. Aesch. Eum. 585)! did twelve vote, as O. Müller and Wecklein (Philol. 1884, p. 712 f.) think, or an uneven number, eleven according to Hermann, or nine, as Zawadzki (Die Anzahl der Ariop. in Aesch. Eum.) suggests for, as Kirchhoff has shown, Athena as nyeur of the court of Areiopagus possessed the right of voting, and the trial was looked upon as a precedent to establish the principle that even votes amounted to an acquittal. Yet these are small matters when compared with the full and lucid account Müller gives of a confessedly difficult subject.

HERMAN HAGER.

[APPENDIX TO DR. HAGER'S ARTICLE.] It is worthy of notice that the results of the recent excavations in the Dionysiac theatre can be corroborated, in one important particular, by literary evidence. In the Oedipus Coloneus, when Creon lays his hand upon Antigone, and asserts his right to remove her from Colonus, the Chorus approach him with menacing gestures, saying, τάχ ̓ εἰς βάσανον εἶ χερῶν. Creon answers, epyov, stand back!' The Chorus rejoin, σοῦ μὲν οὔ, τάδε γε μωμένου (835 f.). It is evident that the poet conceives them as standing close to him. Now, according to old notions, Creon would at this time have been on the raised λoyeîov, while the Chorus were below him in the opxýσrpa. The scene loses much of its natural force, if we have to imagine the Chorus as merely advancing, in an excited manner, to the foot of a platform on which Creon stands. But it becomes at once fully intelligible, if we suppose that Creon and the Chorus are standing on the same level; i.e., that the place from which

the actors spoke was in the time of Sophocles -and the Coloneus probably belongs to his latest years-merely a segment of the circular ὀρχήστρα. It is now believed that this opxnoтpa was enclosed (as noticed above), by polygonal stones, but had, as yet, no permanent σký behind it. A permanent raised stage of masonry, and a permanent σkηvý, appear to have been erected for the first time under the auspices of Lycurgus, circ. 330 B.C. I am informed by Professor D'Ooge, late Director of the American School at Athens, that the results obtained in the Dionysiac theatre have been confirmed in some points by a recent examination of the theatres at Oropus and Sicyon. It is also said to be probable that, in the circular opýστpa of the pre-Lycurgean period, the actors used only one entrance. The traditional three doors were a later development.

Here, however, we must not forget the ἐκκύκλημα. As the recent explorers of the Dionysiac theatre doubtless remember, the use of that contrivance is not merely an inference from the tragic texts, but is proved by the two parodies in comedy,-Aristophanes using the verbs ẻkkukλeîv and ẻσkukdeîv (Ach. 408, Th. 265). It seems to have been a low stage, pushed forward through the central door. (See A. Müller, in the work noticed above, pp. 142-148.) While granting, then, that there was no permanent σký in the time of Aristophanes, we must also suppose that there was some temporary screen, by which the imaginary interior of a house could be veiled from the view of the spectators. There is nothing in the recent results which excludes the use of such a temporary screen, any more than that of a Ocoλoyelov for the deus ex machina. But it seems as if we must give up the TeρíaкTOL, which can hardly have come in before the permanent σκηνή. Their case is different from those just noticed, since nothing in the texts requires them. Niejahr, indeed, had already denied the use of the Teρíaктoι in the best age (Quaest. Arist. Scaen., pp. 8-12.) R. C. JEBB.

The Gentleman's Magazine Library: Romano-British Remains. Part 1, edited by G. L. GOMME, F.S.A. (Elliott Stock), pp. xxiv. 297. 8vo. 78. 6d.

FROM its foundation in 1731 till 1868 the Gentleman's Magazine was one of the chief archaeological journals in England. The Society of Antiquaries began its series of publications in 1770; the Archaeological Institute and Association about 1845; and since 1868 the Gentleman's Magazine itself has neglected antiquities. But before that date its editors included

archaeology among the many subjects which interested the cultivated gentleman, and the paper was a recognised organ, as indeed it was the only monthly organ, of those who wished to read or write about matters archaeological. Thus its volumes became a storehouse for antiquarian information of the most varied character, embracing notices of discoveries in England or abroad, descriptions of churches and old houses, and the like. The idea of collecting 'a volume of antiquities' from this valuable but unsorted treasury is no new one it was suggested by a correspondent

of Sylvanus Urban in 1753. But though Hübner and others examined the Magazine for special purposes, nothing more was done till Mr. Gomme undertook to edit a classified collection of its principal contents' down to 1868, whether archaeological or otherwise. Of this 'Gentleman's Magazine Library,' eight handsome volumes have already appeared, and nine or ten more are promised.

The last addition to the series is that bearing the somewhat misleading title Romano-British Remains, of which part 1 is before us. This is intendedthough, as I shall have to point out, the execution falls short of this-to contain all the notices in the Magazine which relate to the discovery of Roman remains in our islands. The extracts are mostly printed in full, and are arranged in alphabetical order of counties. If I have calculated right, the present volume includes some 330 of these extracts, taking us down to Somerset. The other English counties, Wales, Scotland, notes and index will follow in a second volume. It is interesting to observe that of the 330 extracts before us, nearly 200 are dated later than 1830. The first twenty years of the Magazine yield only seven extracts, while 1836-44 supply 82, and 1862-68, the last seven years, produce 56.

But before criticizing the book itself, a few words are due to the interesting introduction with which Mr. Gomme has prefaced it. Here he discusses the character of the Roman occupation, arguing that it was military and not social: Roman Britain, he says, meant only the luxurious occupiers of villas, the town merchants, and the garrison. This, of course, resembles the view taken by Hübner in his preface to the seventh volume of the Corpus; but Mr. Gomme is more concerned to combat the theories of Coote and Seebohm and deny the alleged continuity between Roman and English history. This he attempts to do by placing the break in the history at 410: the Romans, he thinks, left little trace of themselves in the land. He may be right-one cannot here discuss the question in full-but I fancy that there is a slight confusion in his argument. The continuity between Roman and Briton is one thing, that between Briton and Saxon another, and Mr. Gomme seems to use against the former evidence which really makes only against the latter. The result is that he disproves Mr. Seebohm's view, but does not prove his own. And I think he might have alluded to Bath. There Roman and Saxon dwelt on the same spot; but it is perfectly certain that Aquae Sulis lay waste and deserted for years before it became Akemanceaster. I trust he will pardon me if I point out two slips in his most admirable essay. It is not true that no indications of settled occupation have been found in Derbyshire' (see C.I.L. vii. 177, and Archaeol. x. 17), nor is it correct to say that 'Herefordshire contained one station'; probably there were three. I must demur also to the terms in which Mr. Gomme bids his readers consult the map in Mr. Scarth's Roman Britain. That map, which does duty for other books beside Mr. Scarth's, is the reverse of good, and I do not understand Mr. Gomme's mention of it.

For the book itself Mr. Gomme claims a twofold

importance. It is a representative record of the Roman remains actually found in England, aud it is a contribution towards a complete index of such remains. No one would deny the value of a work which could justly claim to be either of these things, but I regret to say that it is difficult to allow the claim in the case of the present volume. To some extent this is due to the circumstances of the case. The contributions to the Gentleman's Magazine were made rather at random. Many finds were not communicated to its pages; when they were, it was a

The

chance whether the correspondents knew Latin. The unscholarliness of most English archaeologists is no better illustrated than in these pages, though there are of course honourable exceptions, like Mr. Roach Smith. However, these defects are unavoidable. The value of the book is far more seriously damaged by faults which could have been avoided and which are indeed not yet irremediable. worst of these is that the collection is by no means complete. I have gone through thirty out of the first ninety volumes of the Magazine, and have tested the book in other ways, and the result is not satisfactory. Thus the book gives seven extracts on Cheshire, and omits at least four others of importance (1764, p. 510; 1862, March, p. 319, July, p. 59; 1868, March, p. 765). It gives seven extracts on Bath and omits three notable ones (1739, p. 622; 1804, 2, p. 1006; 1829, 1, p. 31). It gives three extracts for Manchester and omits three (1807, 2, p. 1,009; 1820, 1, p. 351; 1832, 2, p. 359). Again, it includes nothing from the year 1764 there are certainly two passages worth quoting (pp. 248, 510). It includes nothing from 1800; I have noticed three passages (pp. 513, 1,095, 1,137). Almost the first discovery noticed in the Magazine (1735, p. 217), the remarkable silver dish found at Corstopitum, is absent. Now these figures are only representative, and I fear they entirely dispose of Mr. Gomme's claim that the volume is a contribution to an index. I do not wish to blame Mr. Gomme for the omissions. We are nowhere told who searched the magazine for extracts, and he is probably not responsible for this part of the work. Nor should I wish to judge such omissions hardly: I know by bitter experience how easy it is to make them. But that does not alter the fact that the book is too imperfect to be of use. I hope Mr. Gomme will have the Magazine searched afresh, and give us a supplement in Part 2. There must be some sixty extracts omitted at the lowest estimate.

Conversely, I am inclined to think that Mr. Gomme has sometimes included remains which are not Roman. I see no evidence to show that the chains, &c., mentioned as found at Horsefield (p. 94) are really Roman, aud I have still stronger doubts about the brass lamp from Hinckley (p. 167). The illustration given in the Magazine (1801, 2, 797), but not repeated by Mr. Gomme, is not at all like a Roman lamp.

Indeed, the absence of the original illustrations is a serious defect. Frequently one is perfectly helpless without them. One extract, for example (1774, p. 512), reads 'Figs. I. and II. were found at Cilchester (sic) in Hants, and are judged to be great curiosities.' The value of this is not very obvious when no figures are given. The case is even worse where, as is not seldom the case, the text is not merely meaningless but actually misleading. Thus the future antiquary who trusts in this book, and copies the inscriptions on pp. 22-25, will go grievously astray; but any one who will consult the original magazine (1752, p. 105) will find tolerably correct illustrations. No doubt a reprint of these illustrations would be costly, but without that, or an equivalent description, the usefulness of the work is much impaired.

These are

Another question concerns the notes. to follow in Part 2, but so far as we can judge from Mr. Gomme's references, they will be much too few in number. For example, antiquaries should be warned that the 'tessera' from Littleboro', on p. 260, is no tessera at all, but an oculist's stamp. This is another case where the original illustration would be helpful, to say the least.

I regard it as a great pity, too, that the coin-finds are omitted, and reserved for a volume of 'Numis

matics.' They are quite out of place in such a volume: no numismatist cares to hear that thirty or forty third brass of the Antonines were found at such and such a place. On the other hand the information is invaluable to the student of Roman Britain. would suggest that a list be added to Part 2 of all the coin-finds, indicating simply the place of discovery and the supposed dates.

I

Perhaps it is unnecessary to add that there are a few slips of less importance, such as occur in most books. Thus a Housesteads inscription is printed under Cumberland (p. 26). The references to the Corpus are very often omitted (e.g. pp. 22-24), and are given rather clumsily. One would prefer C.I.L. vii. 171' to 'Huebner p. 49.' There are also a few misprints.

The conclusion of the whole matter is that the book is very unsatisfactory, but that the defects can be greatly remedied in Part 2. The most disappointing feature is the incompleteness. Mr. Gomme in his preface lays great stress on the value of the work as a contribution to an index of all the Roman remains in England, and it is rather trying to find that so far this contribution is worthless. He complains that the Index Society have been deaf to appeals to undertake the work. I am afraid that, unless Part 2 of the Romano-British Remains is better than Part 1, the Society will continue to be deaf. And yet the work is sorely needed. I suspect, however, that more has been done towards it than Mr. Gomme thinks, and that what is really wanted is unity. At present we are all by ourselves doing the same things and wasting labour. Thus that indefatigable antiquarian, Mr. Thompson Watkin, had indexed the Gentleman's Magazine for his own use long before Mr. Gomme's book was thought of. With division of labour a new Britannia Romana ought to be possible, only the labour must be thorough and scholarly. Absolute completeness may be unattainable, though Mr. Watkin's works on Cheshire and Lancashire, and Hübner's Corpus, show that one can come very near it. But an imperfect index is as bad none, and the contribution now before us is decidedly imperfect. F. HAVERFIeld.

as

DIONYSIOS OR LONGINOS ON SUBLIMITY OF STYLE.1

IT is a matter of uncertainty who was really the author of the treatise 'On sublimity of style' which used generally to be attributed to Cassius Longinos, the teacher of Zenobia. That it was written by him there is positively no evidence; but that it was written by some rhetor whose name was either Longinos or Dionysios is probable from the fact that the title of the work in the best MS. is Διονυσίου ἢ Λογγίνου περὶ ὕψους : and thus those who attribute it to Dionysios of Halikarnassos or Dionysios of Pergamos are more likely to be right than those who attribute it to Plutarch.

One of the first things to be observed about this striking treatise-for it is striking in ability, in delicacy of judgment, in cultured urbanity-is the high spiritual tone which we can detect, and sometimes are not obliged to detect, behind the scientific analysis. Take the following passage, for example, which occurs in a comparison drawn between a Plato who often slips and a Lysias who is always

correct.

'Of what did those divine ones who aimed at the

1 Διονυσίου ἢ Λογγίνου περὶ ὕψους. Edidit O. Jahn, 1867. Iterum edidit J. Vahlen, 1887. (Bonn.) 2 Mk. 40.

highest effects of literary composition take regard ! As well as of many other things they had regard of this-that nature namely did not judge man to be a mean or ungenerous animal, but that introducing us into life, into the universe, as into a great assembly, to be spectators of her contests, and to be keen competitors in their rivalries, she implanted in our souls from the beginning an irrepressible love of whatever was great and nobler than ourselves—τοῦ μεγάλου καὶ ὡς πρὸς ἡμᾶς δαιμονιωτέρου. And therefore the whole world, even, is not enough to satisfy the speculation and apprehension of human understanding, but our reflexions often travel beyond the bounds of our environment; and if one consider life in all its aspects, and behold the abundance of greatness and beauty in its sundry spheres, he will soon be aware to what things we are born. And thus it is the guidance of nature that we follow when we admire not small streams, translucent and of service though they be, but admire the Nile, the Danube, the Rhine, and yet more the Ocean. . . .'

In this passage there is no attempt at effect, there is not the faintest touch of rhapsody; the writer does not exceed the strict limits of the matter in hand. But we can feel the high-spirited tone behind the Hellenic reserve and penetrating it.

The

Whether the author's name be Longinos or not, he certainly deserves the fame for critical acumen which attached in later times to a rhetor of that name, just as Dionysios was noted for his much learning. judgments and the critical observations contained in this treatise are so good, so penetrating, that when we have read a few pages we feel complete confidence in the writer's powers; they are also full of surprises -a subtle remark meets us when we fancied that the criticism was complete.

For example having explained in a phrase worth remembering that os, the grand style,' is an echoed reflex of grandeur of soul, μεγαλοφροσύνης anhxnua, he observes that in some lofty moods silence is more than all echoes-the silence for instance of Aias among the shades:

ὡς ἐφάμην δ δέ μ' οὐδὲν ἀμείβετο· βῆ δὲ μετ ̓ ἄλλας ψυχὰς εἰς ἔρεβος νεκύων κατατεθνηώτων.

Again in explaining the difference between real and false grandeur of style, he quotes some lines from the Arimaspeia of Aristeas (a history of the Hyperboreans), describing the perils of sailors, of which one

verse was

ὄμματ ̓ ἐν ἄστροισι ψυχὴν δ ̓ ἐνὶ πόντῳ ἔχουσιν.

'It is manifest,' he says, 'that this description is less fearful than flowery —πλέον ἄνθος ἔχει τὰ λeyoueva-déos. This criticism is happy, and is without a superfluous syllable: the writer always knows where to stop. The urbanity is also to be observed; another critic would not hesitate to call the lines frigid-4vxpá, whilst our critic with an urbanity that is really more crushing says, they are 'less fearful than flowery.' He then takes an example of true os from Homer to set beside the false of Aristeas:

os

ἐν δ ̓ ἔπεσ ̓ ὡς ὅτε κῦμα θοῇ ἐν νηῒ πέσῃσι λαβρὸν ὑπαὶ νεφέων ἀνεμοτρεφές, ἡ δέ τε πᾶσα ἄχνῃ ὑπεκρύφθη, ἀνέμοιο δὲ δεινὸς ἀήτης ἱστίῳ ἐμβρέμεται, τρομέουσι δέ τε φρένα ναῦται δειδιότες· τυτθὸν γὰρ ὑπὲκ θανάτοιο φέρονται. The idea of the last line Aratos tried to express more vividly by

ὀλίγον δὲ διὰ ξύλον ἄϊδ ̓ ἐρύκει.