Sin does not occur in Horace.

In several instances si is found with the force of sin where one condition is opposed to another (Serm. I, 3, 6: Epist. I, 5, 6; I, 10, 43; I, 12, 7; I, 17, 11; II, 1, 66). Twice when no expressed statement of condition precedes, but with the implication of a condition (Epist. I, 2, 37-39: Od. III, 29, 53-56). In Epod. I, 5-6, the second condition is introduced by si contra. In Serm. II, 3, 74-76, contra alone, with omission of conditional particle, follows a preceding condition.

Sic is sometimes used, especially in petitions and prayers, to anticipate or resume a condition, upon the fulfilment of which a promise or wish is made, or a blessing invoked (Serm. II, 3, 300-302: Od. I, 3, 1-8). So with the sic clause following the clause containing the condition in Od. I, 28, 23-27. Once also the condition is explicitly stated in a following clause with si (Epist. I, 7, 69–70). Ita is used in a like manner in Serm. II, 2, 124-125.

The paratactic conditional structure occurs in the writings of Horace in all moods and tenses except the Imperfect Indicative.

18. An Horatian Gloss, by Dr. William E. Waters, of New York University.

C.G.L. 5, 236, 9 contains the gloss proelium orant et proelio ditam caleno tum bibis uuam pro vino dixit, which Landgraf, Archiv 9, 407, has correctly emended to Prelum: Horatius: et prelo domitam Caleno | tum bibis uuam: pro vino dixit, This means that prelum is a word used by Horace, the passage in which it occurs being then cited, and that in that passage uuam is used for vinum. Neither Keller nor any other of the Horatian editors has taken account of this gloss in attempting to settle the vexed question whether tu or tum is the correct reading before bibis.

Agreeing with Lucian Müller (ed. Goetz, 1900) we may hold that Horace's Ode I, 10, is a blunt and frank reply to Maecenas, who had warned Horace that he expects to visit him at his Sabine farm; cf. Cruquius' commentary: Maecenas iturus in Apuliam significavit Horatio ei se ante profectionem convivam esse velle - de profectione in Apuliam mentio fit in Divaei codice. Down to the concluding strophe the spirit and thought of the ode are somewhat like this: You are intending to visit me? Well, you must take me as I am; you shall have cheap Sabine and plenty of it, its only virtue being that it was decanted by myself, the year all Rome rejoiced over your return to health. Here follows the concluding strophe, in which, adopting the reading tum bibis (emend. to bibes) of the gloss, the thought is Then when you have finished your visit you shall find a better wine on your journey; there are plenty of better quality.

All the Horatian Mss. and Porphyrion read tu; but Porph. on Sat. 2. 2. 48 quotes this passage from Od. 1. 20, reading tum in all probability, and not tu. In their first edition of the Odes, K. and H. have tum; in the Epilegomena, K. argues for tu, and this is his reading in the second edition of the Odes. In such uncertainty the gloss with its lectio difficilior has great value. It is far more likely that if tum was actually written by Horace the gloss should show tu than that if he used tu it should show tum. Tum, "thereupon," moreover contrasts the time in the two futures, and removes the necessity of Keller's emendation of the verb to bibas. For tum=" "thereupon" in Horace, cf. Zangemeister's index. Mea,

which follows in the same line, is no argument for the necessity of a preceding antithetical tu; cf. Odes 2. 17 ad fin.

Reddere uictimas

Aedemque uotiuam memento;

Nos humilem feriemus agnam.

19. The Temple of Zeus Bλos, Herodotus I. 181, by Professor H. C. Tolman, of Vanderbilt University.

According to Herodotus, Babylon was divided into two parts by the river Euphrates; in the midst (ev uéow, not a gloss) of one section lay the "royal palace," in the midst of the other the temple of Zeus Bîλos.

From the inscriptions we learn that the two principal sanctuaries of Babylon were Esagila (Borsippa Inscr. I. 15, et passim) and Ezida (Borsippa Inscr. I. 19, et passim): the former, sacred to Marduk (Borsippa Inscr. I. 16), lay within the walls of Babylon (Borsippa Inscr. I. 23); the latter, sacred to Nabû (Borsippa Inser. II. 16), was at Borsippa (Borsippa Inscr. I. 27), the sister town of the metropolis (Tin-Tir 11 Kan Ki, "second Babylon" K. 4399, obv. 24). In such veneration were the two temples held that "restorer of Esagila and Ezida " (Zani in E-sag-ila u E-zi-da) becam: part of the royal title (Borsippa Inser. I. 7). To which of these sanctuaries did Herodotus refer ?

The East India House Inscription (E. I. H. VII. 36) shows that Esagila was closely connected with the royal residence. Again, from the term Býλov BaσiMeta of Clitarch we infer that a palace structure, probably an addition of later kings, was joined immediately with the sanctuary (Baumstark, Pauly-Wissowa, Real Encyclopädie der classischen Altertumswissenschaft, sub voce Babylon). Nabopolassar had erected a palace near Esagila from Imgur Bêl to the East Canal and from the Euphrates to the festal street, A-a-ibur-sâbû (E. I. H. VII. 36–63). Nebuchadrezar built a new structure (E. I. H. VIII. 54) between the bastions Imgur Bêl and Nimitti Bêl (E. I. H. VIII. 53), which was joined with the older edifice (E. I. H. VIII. 50), and which towered mountain-like over the city (E. I. H. VIII. 63, Hu-ur-ša-ni-iš). Late reports from Dr. Koldewey (Mittheilungen der deutschen Orientgesellschaft) connect the ruins El Kasr with this palace of Nebuchadrezar, and the adjacent mound Tel Amran with the temple Esagila. Adorned glazed tiles, reliefs from the palace wall, and bricks stamped with the name of Nebuchadrezar, confirm such an identification (cf. Berliner Philologische Wochenschrift, 19 Mai, 1900.) There is no doubt that it was to Esagila that Clitarch and the authority quoted by Strabo (XVI. 738) referred, for by the phrase rápos Toû Bλov of the Strabo citation we must un lerstand the temple's pyramidal terrace, E-temen-an-ki (E-temen-an-ki zi-ku-ra-at Babili, Borsippa Inscr. I. 23).

We cannot accept the common hypothesis (Hommel, Babylonisch-Assyrische Geschichte; Baumstark, Pauly-Wissowa, Real-Encyclopädie) that Herodotus under the term "temple of Zeus Bλos" designates Esagila. Our historian expressly stated that the sanctuary lay on the opposite river-bank to that of the palace. On this bank was Ezida, the massive temple of Borsippa adorned by Nebuchadrezar (as he himself relates, E. I. H. III. 36 fg.) with gold, silver, precious stones, and bronze. The inscription found in the corners of the terrace

tower E-ur-imim-an-ki (E-ur-imim-an-ki zi-ku-ra-at Bar-sip, Borsippa Inscr. I. 27-II. 30) has long ago identified the mound Birs Nimrud with Borsippa (I. R. 51; cf. Oppert, Exp. Mésop. I. 212). That this temple is the sanctuary of Zeus Bylos described by Herodotus appears from the following considerations:

(1) The identification of the El Kasr mound on the eastern bank places the temple on the western. (2) Herodotus speaks of the sanctuary as in the middle of its section of the city. After the destruction of the walls of Babylon and Borsippa by Darius it would seem to a foreigner that both localities formed one town, an error that could easily occasion Herodotus's exaggerated estimate of the size of the metropolis. (3) The layers of the débris at Borsippa show clearly a division into stories such as Herodotus (as well as Harpocration) describes. (4) Herodotus's measurements of the tower at its base (185 m.) correspond with those of the Borsippa terrace as given by Rich, Oppert, Ker Porter, Layard. The description of the great temple at Babylon by Harpocration (according to a newly published manuscript, de Mély, Académie des Sciences) gives a terrace tower of seven stories with its base 186 m. square (Académie Inser. 1900. Cf. Revue Archéologique, Nov.-Dec. 1900, Vol. 37). (5) The designation Zevs Bĥλos for Nabû of Ezida presents no difficulty, since the application of the title Bêl (Lord) to other gods than Marduk is abundantly proved by the inscriptions (cf. Winckler, Untersuchungen, Leipzig, 1889).

In fact, later results seem to justify the view which we have earlier accepted (Nikel, Herodot und die Keilschriftforschung; Tolman and Stevenson, Herodotus and Empires of the East), that Ezida of Borsippa and the temple of Zeus BXos are identical.

Adjourned at 12.45.

On Wednesday afternoon, July 10, the members of the Association enjoyed a delightful excursion in Boston Harbor as guests of Harvard University.

EVENING SESSION.

The Association assembled at 8.15 P.M.

20. A Study of Browning's Agamemnon, by Dr. Curtis C. Bushnell, of Syracuse University.

In the preface to his Agamemnon Mr. Browning says: —

That he has been literal at any cost save that of absolute violence to the English language; that this conservatism is all for which his perhaps fruitless adventure claims praise; that he expects the result of his work to be very hard reading if meant to resemble Aeschylus.

This plan is consistently executed.

The spelling of the original is followed, as Menelaos (42), Meneleos (674). Idioms are often literally rendered, as 1, 962.

In almost every line the word-order is imitated, - the leading peculiarity of the translation.

Sometimes the result is ambiguous, as 1102.

However, chiasmus is well rendered, as 113, 519.

The effect of repetition in the corresponding parts of lines is reproduced, as 34, 35; 863, 874.

So with synonymous expressions in the corresponding parts of successive lines, as 5-7.

So with two adjectives, one preceding, one following, the noun modified, as 734; 219, 220.

The word-order is reproduced in preference to the construction, as 2; 300, 301.

The iambic trimeter is rendered line for line except in two cases, the trochaic tetrameter invariably. Stichomythia is preserved.

The iambic trimeter is rendered by the English heroic line with double ending to reproduce the extended effect; trochaic tetrameters are rendered by trochaic tetrameters; the choruses by rhymed lines of irregular length.

Curious results follow from the combined influence of the English rhythm and of the retention of the word-order, as 37, 1373.

Repetition is rendered by repetition, as 971-4, 299-300.

The variety of the Greek is reproduced, as in the interjections of Cassandra's speeches. Etymologies are rendered, as 121, 1232, 1346. Plays on words are brought out, as 686 ff., 699. Amplification is almost absent.

The translation is in point of clearness very uneven, as 737 ff., and considerably more obscure than the original, - - a fact chiefly due to the retention of the word-order.

The unfavorable criticism evoked (Living Age, 190:564: American Supplement to Encyclopaedia Britannica, "Robert Browning": Edinburgh Review, 147: 409 ff.) would have been greatly tempered if allowance had been made for Mr. Browning's motive. This was variously assigned by the critics, as:

1. "To compensate for the inexactness of other translators" (Edinburgh Review, 147: 409 ff.). This consideration had weight, as the preface shows, but the work might have been exact without its remarkable peculiarities.

2. "To show that as a dramatist Aeschylus is inferior to Euripides" (Mrs. S. Orr, Handbook to Works of R. Browning, p. 119). In that case Mr. Browning would have given the writings of the two a similar treatment.

3. "To rebuke claims made for the Greek authors as models of literary style" (Mrs. S. Orr, Life and Letters of Robert Browning, Vol. 2. See under year 1877). But Mr. Browning, in his preface to the Agamemnon, expresses the hope that this work may give evidence that "the Greeks are the highest models of expression."

4. "To disparage the Agamemnon and Aeschylus" (Edinburgh Review, 147: 409 ff.). But elsewhere Mr. Browning frequently mentions Aeschylus in terms of praise. He is "godlike," his works are "marvels" (Aristophanes' Apology); he is classed among "the tragic triad of immortal fames" (Balaustion's Adventure); the "song which saved at Salamis" is "veritable Aeschylus . . . old glory" (id.); "How please still-Pindar and Aeschylus!" he exclaims (Epilogue to Pacchiarotto); the "lore" which tells the story of

"that king

Treading the purple calmly to his death,"

is an "old delight" (Pauline).

1

The real motive is stated by Mr. Browning himself when he calls the work "a perhaps fruitless adventure"; in other words, an experiment. This experiment, so consistently carried through, was the natural sequence of his earlier translations, the Alkestis and the Herakles.

The Alkestis (published in 1871), the Herakles (1875), and the Agamemnon (1877) are three stages of an intensifying endeavor to reproduce the original in its entirety, particularly by adherence to the word-order, the Herakles being intermediate in fact as well as in time. The Alkestis, a far closer rendering than at first sight appears, paraphrases more freely than the Herakles, a quite exact translation, while the Agamemnon is almost absolutely literal. The Alkestis makes no distinction in rhythm between dialogue and choral passages, translating both by the heroic line; the Herakles and Agamemnon come closer to the Greek effect, translating choral passages by rhymed lines of irregular length. The Alkestis and Herakles are alike in employing the heroic line to translate the dialogue, while the Agamemnon adds a syllable for the reason stated above. In the Alkestis 51% of the stichomythia is translated line for line, in the Herakles 97%, in the Agamemnon 100% A study with reference to line for line translation and word-arrangement of 1200 lines selected from the dialogue of all parts of the plays, 400 from each play, shows the following results. Of those from the Alkestis 30% are rendered line for line; of those from the Herakles 40%; of those from the Agamemnon 85%. For the same 1200 lines the number of lines of translation with precisely the same word-order as the text, - disregarding postpositive particles and prepositions following their cases, is of the 400 of the Alkestis 7.5%, of the 400 of the Herakles 17.5%, of the 400 of the Agamemnon 35.5%.

I conceive that Mr. Browning-having with great success in the Alkestis and Herakles gone far in strict literalness, especially in fidelity to word-arrangement and line for line translation - determined as an experiment to proceed in the same manner and to the very limit with this new play.

The experiment proves that to reproduce the Greek word-arrangement involves less violence to the English language than is usually thought. But in the Herakles Mr. Browning had already gone as far as was possible in this direction. For the reason that in them, as in their originals, content and form are harmonized, the Herakles and Alkestis represent Euripides more truly than the more literal Agamemnon does Aeschylus.

21. Notes on the Nominative of the First Person in Euripides, by Professor Mortimer Lamson Earle, of Barnard College, Columbia University.

The subject treated were more accurately designated as the substantival nominative of the first person in Euripides. The current doctrine of the nominative persons in Greek is concisely put as follows in Hadley-Allen, § 603, a: "The only nominatives of the first person are è̟yw, vw, nμeîs; of the second person, oú, opw, vueîs; all other nominatives are of the third person." But what should be said of such nominatives as 0eà Kúπpes in Hipp. 2? Such self-introducing and selfidentifying nominatives of proper names as subjects of verbs in the first person are to be found also Androm. 5, 1232, Hec. 3, 503, Troad. 2, Bacch. 2. In view