are necessary to complete the figure. For the combination with this verb cf. his κοσκινηδὸν διατετρυπῆσθαι, Sat. Epist. 24. For the variant ptc. from TETρаív and тружάw, cf. the parallel pass. Dial. Inf. 11, § 4. Compare this, in fact, throughout with Plato. Finally, as an indication that Plato's words were in Lucian's mind, cf. σтeyavóv of the Gorgias with un σTÉYOVTOs in the Timon and στέγειν οὐ δυναμένου in the citation from Dial. Inf.

The interchange of ck and p is obviously easy.

2. Gallus, § 22: ὑπερβὰς τὸ θριγκίον ἢ διορύξας τὸν τοῖχον. The τοιχωρύχος is a familiar acquaintance, but what does vreрßás etc., mean? Were the houses of the rich constructed with sloping roofs and openings under the eaves, or above the wall, corresponding to the enclosed metopae of a temple?

The two well-known Euripidean passages may be compared.

In I. 7. 113, Pylades points out that there is room for a man to let down his body between the triglyphs, and in Orest. 1371 a slave escapes by one of these apertures.

Unless reference is here made to some such opening, — usual, perhaps, for the sake of the light, the alternative meaning would seem to be an entrance effected

through the tiles or opening of the flat roof itself. Cf. N. T. Luke 19.

The translation, lorica domus superata' (Reitz-Hemst.) assumes that Tò Oрyklov was a parapet built around a flat roof. But would not Lucian have used some less vague expression, such as διὰ τῶν κεράμων, οι διὰ τῆς στέγης, if this had been his meaning?

Hermann (Lehrbuch d. griech. Antiq. IV. p. 154, note 2) says: "der Giebelbau des Tempels ist ohne vorgängige private Bauweise nicht denkbar." This passage also may indicate that the slanting roof was not monopolized by temple architecture.

3. Icaromenippus, § 13: Éπì TŶS KAт vodókηs, ‘over (or 'at') my smoke-vent.' Reitz's translation, 'in fumario,' is, I think, clearly wrong. Pauly does better: 'unter meinem Rauchfange ein Trankopfer darbringen.' But why not take not only kажvodóкη but also the preposition in the most literal sense and give a more burlesque, and therefore more probable, coloring? Icaromenippus promises that, alighting on his roof, he will pour a libation over (ẻπí, not 'unter') the smoke-hole, that it may be wafted up to Empedocles.

To illustrate cf. the two passages in Herodotus where кaπvodóкŋ is mentioned. In VIII. 137, the sunlight streams in кaτà тǹy ка vоdóкŋν (evidently here a mere hole in the roof). In IV. 113, π with the gen., Lucian's exact expression, is used. Herod. here represents the Taurians as transfixing their enemies' heads upon a pole and setting them up on their house-tops and by preference over the smoke-vent - μάλιστ ̓ ἐπὶ τῆς καπνοδόκης. Lucian could not have missed this passage in Herodotus.

In Icaromenip. § 25, Zeus seats himself at one of the scuttles (TÌ TŷS πρÚTNS) in heaven's floor and bends over it to catch the incense. To illustrate this and the whole meaning of каπ vоdóкŋ, one could wish for the context in Pherec. Tyrannis, 2, where Zeus out of thoughtfulness for the altar loungers''made for them a very large smoke-vent.'

However the much-debated chimney-question, Ar. Vespae, 139 sqq., may be decided, it would not interfere with this view. Even if a real chimney be understood here, the translation over the smoke-vent' would, I believe, be the best; the burlesque element is in either case the same.

·

4. Use of ὅτι μή. In οὐδὲν γὰρ ὅτι μή, Somnium, § 9, ὅτι μή means ' nothing but,' and Icaromenip. § 9, 'nothing else.'

Williams annotates these two passages as instances of Lucian's careless use of μn for où, and refers to Professor Gildersleeve's article, A. J. P. Vol. I. 1. Heitland2 also, although stating that or is here the neuter of oσTɩs, says: “It will be noticed that the μý is, as often in Lucian, unbearable." This use of ÖTL μn (= 'except') is expressly mentioned by Gildersleeve (l.c.) as 'well known and legitimate,' and is used as partially explaining the extension of the combination elsewhere. In addition to the examples of this use of öτɩ μý, cited from Homer, Herod., Thuc., Plato, and Arrian, in L. and S. (vide sub öтi (neut. öσтis) II.), may be added (from Abicht's Herod.) Herod. II. 13 and 50, and (see La Roche, I. XVI. 227) Herod. I. 183, III. 155 and 160; Thuc. IV. 94. 2, VII. 42. 6.

Lucian may, therefore, be here relieved of the charge of being 'unbearable.' 5. The Arrangement of Guests in Lucian's Symposium (see the plan on p. xiv). Reitz3 and Wieland1 assume the triclinium arrangement. A careful reading of the text will contradict this assumption, and the accompanying plan is intended to meet the conditions in the text, viz.: (1) In § 8 Lucian says: "On the right as you entered, the women and they were there in full force took up all that bench, and among them the bride,” etc.

(2) §§ 8, 9. "On the side over against the door the rest of the company, each according to his rank. And first, beginning opposite the women, Eucritus," etc. [Twelve banqueters are here named.]

(3) These twelve with Dionicus and Alcidamas, who arrive later, are the only men expressly mentioned, and Reitz and Wieland in their arrangement assume that there were no more. Certain expressions, however, imply that there were other guests in the company. Cf. § 6. "Why should I tell you of the others? It is chiefly about the philosophers, I think, and the litterateurs that you want to hear." Again, in § 35, Lucian clearly implies that there were others besides the philosophers when he says: “The laymen (oi idɩŵraɩ) dined in a very orderly fashion." The single candelabrum (rò Xvxvlov) — which in § 46 is overturned, leaving the company in utter darkness. may, however, seem to forbid the assumption of any large number of guests, and the seeming contradiction involved in φῶτα εἰσεκεκόμιστο (§ 15) is to be explained as a reference to the separate lamps placed on the several trays of the large holder.

(4) The uninvited cynic Alcidamas is urged by the host to take a chair by Nos. 11 and 12.

(5) Dionicus the last comer, although invited, had no place reserved for him after the signal to lie down was given. He therefore squeezes in (§ 20) near, or next to, No. 12.

(6) From direct mention (§ 38) we learn that each pair of banqueters had a separate small table.

1 C. R. Williams, Selections from Lucian, John Allyn, 1882.

2 W. E. Heitland, Somnium, etc., Cambr. Un. Press, 1885.

3 Cf. Vol. IX. p. 359, where, too, he speaks of Histiaeus and Dionysodorus as lying in tertio lecto ultimi.' This would imply that there were no other banqueters beyond: but see below (3).

Wieland says: 'Ieder von diesen Triclinien hatte seinen eigenen Tisch,' but we know from the text that there were six small tables for the twelve male guests first mentioned.

(7) The couch or bench on which all the women sat together is called (§ 8) a Kλтýρ. In § 44 the same word is used of the couch apparently on the opposite side of the room, where Hermon sends Diphilus head-foremost ảnò Toû KλɩVTĥPOS. Kλivn, however, is used (§ 47) where the Cynic throws himself in a drunken sleep ἐπὶ τῆς κλίνης. Cf. ὁμόκλινοι, Hdt. IX. 16.

Only some such arrangement as that here proposed can meet the requirements of the text. The question of the length of the couches and whether they were placed closely together or continued on around the left-hand end of the room must, I think, be left open.

In regard to the place of honor1 a word may be said. The rich old father of the bridegroom-the guest of honor - lies first of all and next above the hostthus combining two of the points mentioned by Plutarch (Quest. Sympos. II. § 4 and III.). The bridegroom, too, might be regarded perhaps as in one of Plutarch's places of honor. He comes μeraíTaTos so far as the twelve guests ranged along Tò ȧvτíðupov are concerned. Becker, or the reviser (Charik., Göll's revision, 1877, Germ. ed. Vol. II. p. 305), makes a curiously vague or inaccurate statement. He says: "Auch bei Lucian (Conv. 9) liegt der Bräutigam neben dem Schwiegervater und Wirth." As a matter of fact he lies (No. 7) five places below them.

7. A Study in the History of German Metrics, by Professor Julius Goebel, of the Leland Stanford Jr. University.

I. Accent. Among the German grammarians of the sixteenth century who made the first attempt at a scientific inquiry into the nature of German versification, Johann Clajus (1535–92) takes the principal place. In his Grammatica Germanicae linguae (1578) he devotes two chapters to German metrics in which he exhibits a clear understanding of the vital differences in verse-structure that separate the Germanic languages from the ancient tongues. Besides, he defines in the same chapters with remarkable precision the law upon which the metrics of all the Germanic languages are based: the law of accent. While the critical efforts of Clajus produced little effect upon contemporary German versifiers, the little book by Martin Opitz, Buch von der deutschen Poeterey (1624), also marked an epoch in the history of German metrics. In the seventh chapter of this famous book he says:

'Nachmals ist auch ein jeder Vers entweder ein Iambicus oder Trochaicus; nicht zwar, dass wir auf Art der Griechen und Römer eine gewisse Grösse der Sylben können in acht nehmen, sondern, dass wir aus den Accenten und dem Thone erkennen, welche Sylbe hoch und welche niedrig gesetzt soll werden.'

It seems that Opitz in establishing this fundamental fact concerning German metrics did not know of Clajus. Nor is there any evidence of the influence of Opitz on Friedrich von Spee, who, in the preface to his Trutznachtigall (written previous to 1635), makes the same discovery of the law of accent in German versification. Spee goes even beyond Opitz and approaches the modern view by claiming that the accent of the verse must be that of the living speech. Die

1 Jowett's trans. of eσxarov, in Plato's Symposium, as at the end of the table' is infelicitous for other reasons and possibly misleading to the English reader acquainted only with the Roman arrangement of the tables.

Quantität aber, das ist die Länge und Kürtze der Syllaben ist gemeinlich vom Accent genommen, also dass diejenigen Syllaben, auf welche in gemeiner Aussprache der Accent fällt, für lang gerechnet seind und die andre für kurtz.'

For more than a century the influence of Opitz's book made itself felt in a great regularity of German versification, a regularity too painfully regular for the genius of the German language. The negative result of Opitz's rules — that verses constructed after the principle of quantity were impossible in the German language — found, however, less attention than one would expect from poets and the writers on metrics. This becomes evident especially during the eighteenth century when the revival of Humanism kindled the desire of imitating the muchadmired Greek metres. Happily the great poets like Klopstock, Goethe, and Schiller were possessed of too fine a feeling for the genius of their language to allow themselves to be ruled entirely by metrical law-makers like Joh. Heinrich Voss. Still we can notice in their metrical practice a constant struggle between their German Sprachgefühl and the rigid requirements of the classical metricians. We have an amusing document of this in the famous strophe by

Goethe:

Ein ewiges Kochen statt fröhlichem Schmauss!

Was soll denn das Zählen, das Wägen, das Grollen?

Bei allem dem kommt nichts heraus,

Als dass wir keine Hexameter machen sollen,

Und sollen uns patriotisch fügen,

An Knittelversen uns begnügen.

In 1832 Lachmann's famous essay, Ueber althochdeutsche Betonung und Verskunst was published, a treatise which, by a careful inquiry into the nature of accent, laid the foundation for all the future investigations of Germanic metrics. Starting from the fact that in all the Germanic dialects the principal accent is placed on the first syllable of every word, Lachmann establishes the following laws of accentuation:

(1) Wenn in drei- oder mehrsilbigen Wörtern des Alt- und Mittelhochdeutschen die erste, d. h. die betonteste Silbe lang ist, so hat die zweite den nächst hohen Accent; (2) ist dagegen die erste kurz, so hat die dritte den Nebenton.

Although later investigations modified and corrected these laws, by their discovery, Lachmann was enabled to unfold the secrets of Old-Germanic versification, especially in Otfrid and in the Hildebrandslied. Starting from the verse of Otfrid, Lachmann pointed out that the Old Germanic verse consisted of four arses or accents, showed why the thesis in certain cases could be omitted, and, by a careful examination of the various parts of speech, established the rules which governed the position of the accent. An excellent account of Lachmann's principles, supplemented by original research and subtle metrical observations, may be found in the Deutsche Verskunst nach ihrer geschichtlichen Entwickelung by Vilmar-Grein (1870).

While Lachmann had confined himself chiefly to the word-accent, Max Rieger in his essay Die Alt- und Angelsächsische Verskunst (Zeitschr. f. deutsche Phil. Vol. 7) made the sentence-accent and its relation to Germanic versification the object of his investigations, reaching the result that alliterative verse had but two and not four accents.