nate accusative and the absolute constructions are also frequent. If the dative had been employed in the Aeschylean passage, the meaning would have been unmistakable, but that would have produced hiatus. Cf. Eur. Alc. 199, 652. In Hel. 186 we find a dative of manner (aláyμaoi σrévovσa). The ode under discussion begins with στένω σε οὐλομένας τύχας (which is a constantly recurring thought in the song) and ends with στένουσιν ἄλγος οἰκτρόν. Cf. Eur. Alc. 1038 f. Not only Prometheus, but the other characters constantly advert to the Titan's dynμara: 267, 268, 298, 302, 306, 326, 375, 397, 413, 471, 512, 525, 541. Cf. esp. 615 and 695 ff., and 934.

32. Notes on Plautus and Terence, by Professor Charles Knapp, of Columbia University (read by title).

Among the fifteen passages discussed were the following from the Adelphoe: (1) 20, 21. Proof was offered that sine superbia must be interpreted of the predicate of its clause, not, as recent editors have held, of the subject.

(2) 137. It was argued that the phrase si obsto is not, as editors seem to think, transparent, but that we must supply with it Aeschini factis or the like. The spirit of the remark Demea was about to make can best be got by comparing 989 ff., especially 992 ff. The latter passage, too, was discussed; the author held that sense demands that secundare, “give a favorable turn to," "bring to a happy issue," not obsecundare, "support," "further" (so Mss), shall be read.

(3) 160, 161. It was argued that at ita ut usquam fuit fide quisquam optuma can be explained only as due to a fusion of (1) at ita (leno sum = talis leno sum) ut ille fuit qui optuma fide fuit and (2) at optuma fide (leno sum) si usquam quisquam ita fuit. Mode I is essentially affirmative, mode 2 is essentially negative in spirit. Other passages from Greek and Latin writers showing similar fusion of different syntactical elements were cited and discussed.

...

(4) 163-166. The author sought to determine the bearing of the quom-clause in 166. It cannot give the reason for dabitur . . . hac. Nor can it be easily or naturally associated with those words in adversative force. He proposed, therefore, to remove the period commonly set after feceris, 164, to set a dash there, and another after hac, 166; then the quom-clause can be joined directly with ego meum. . feceris, 164, as causal in force, giving the reason for the threat contained in those words. In this view nollem factum . . . iniuria hac is an excited parenthetical commentary on vostra haec. Here, as so often elsewhere, emotional exaltation is attended by syntactical dislocation. All this throws important light on the text in 165, 166. The Mss text gives here a trochaic octonarius followed by an iambic octonarius. Bentley condemned the change of rhythm, and editors in general have followed him, emending in various ways in 165, 166 to make 166 also trochaic. The author held that all such alterations are futile. The Ms text, reënforced by the proposed punctuation, is extremely effective; it throws out into such sharp relief the vital part of the quoted words, indignum iniuria hac. The author thus arrived, quite independently, at the conclusions previously reached by Kauer; that scholar, in his revision of Dziatzko's annotated edition, had argued strongly for the retention of the Ms reading, though he gave no heed to the difficulty of interpreting the quom-clause if the ordinary punctuation is retained.

Ad. 202, 574-575, 770, were considered, besides passages in the Eunuchus and the Andria, and in various plays of Plautus.

33. Travel in Ancient Times as seen in Plautus and Terence, by Professor Charles Knapp (read by title).

The purpose of this paper is to gather together all the information supplied by the plays of Plautus and Terence concerning travel. Such an investigation has much interest; that it has value is a fact emphasized afresh to the author's mind by the following passage in Kroll's Die Altertumswissenschaft im letzten Vierteljahrhundert (1905): “Was uns gleichfalls noch immer fehlt, ist eine Geschichte des Reisens im Altertum (für die Kaiserzeit liegt da freilich die treffliche Behandlung in Friedländers Darstellungen aus der Sittengeschichte vor) und in Zusammenhang damit eine neue Arbeit über die Fuhrwerke der Alten..." (so Blümner, p. 370).

The plays give abundant evidence of the freest movement from place to place in the Greek world (most of the places mentioned in Plautus are parts of the Greek rather than of the Roman world; all those mentioned in Terence are Greek). We have here an interesting and instructive illustration of the wellknown dependence of the Roman comic writers on Greek models.

Travel is undertaken regularly, it may be said, in connection with business; there are very few references to travel undertaken for the mere love of travelling, animi causa. Illustrations of travel for business, in the narrower sense of the term, are afforded by the long trading trips (lasting two or even three years) frequently mentioned. A good deal of travelling was done in connection with warfare; one realizes to what an extent the citizen soldiery of Athens, for example, became acquainted through wars with the outer world. Akin to such journeying is the travelling of persons who were legati publice missi. The amours of the miles gloriosus and others involve much travel, either on the part of these personages themselves or on that of their messengers and the meretrices. Another chapter can be written on the travels of persons stolen in childhood by runaway slaves or pirates; they often undergo remarkable experiences. Much travelling is done also by their kinsmen as they seek to find those lost years before.

The paper, in its final form, will contain remarks on the geography of the plays, on the costume worn by travellers, the baggage carried by them, etc.

34. When did Aristophanes die? by Dr. Roland G. Kent, of the University of Pennsylvania (read by title).

The year 385 seems to be agreed upon as the approximate date of A.'s death (so Croiset, Hist. Litt. Gr. III.2 531; Christ, Gr. Litt.-Gesch.2 292; Kaibel in Pauly-Wissowa, Real-En. II. 972). This is based on Hyp. iv, Ar. Plut.; Anon. Vita Ar. § 12; Schol. ad Plat. Apol. 19 C; Suid. s.v. 'Apapús; and the fact that A. appears as a character in Plato's Symposium. It is certain that after the Plutus (388) A. composed only the Cocalus and the Aeolosicon, and that he gave them to his son Araros for presentation to commend him to the favor of the public. Now the ordinary interpretation of Suidas (l.c.) in regard to Araros didáĝas

то πρŵтоν 'OXνμmiádi pa' is first presented a play of his own composition in Ol. 101 (so Christ, l.c., Kaibel, op. c. II. 381), but it cannot mean more than first presented a play in his own name in Ol. 101, and hence refers to one of the above-named plays of A. They cannot therefore have appeared before the Lenaea of 375. Did they appear in A.'s lifetime? Naturally he would have desired to aid his son in their production; but there is another reason for thinking so. While both plays are of unquestioned authenticity, the Aeolosicon appeared in two versions (Novati Life; Athen. 372 A; Schol. ad Hephaest. I. p. 56 Gaisf.), the second of which is certainly by A. There is no authority for supposing that Araros made the revision; hence A. survived its first performance at the Lenaea of 375 or later, long enough to revise it.

As for A.'s appearance in the Symposium, it is not necessary to suppose that this implies that he was dead when the dialogue was composed (384 or later); the fantastic views which are there put into his mouth may be a retort for his satire upon the Platonic state in the Eccl. and for the mention of Plato as Aristyllus (nickname of Aristocles, his real name) in the Teλeμŋooĥs.

A. died therefore not before 375. Presumably he lived not much longer. Even then he is only a trifle over seventy if his birth is placed in 445/4, or a little above eighty if it is placed in 455/4, as the writer believes that it should be. This paper has appeared in the Classical Review, XX (1906), pp. 153-155.

35. Note on the Standpoint for the Study of Religion in Homer, by Professor Arthur Fairbanks, of the University of Iowa (read by title).

The study of the different phases of social life in Homer is necessarily difficult for the student who recognizes that the Greek epic is the result of a long process of development, since not only metre and language but the picture of life as well must have been influenced by this process. At some points we can see that the account of religion would be subject to forces which would not affect so easily language or metre; e.g., the migration to Asia Minor must have interfered with religion more than with language, for religion is closely bound to locality. Yet it is untrue to the historic method for scholars to apply totally distinct methods to the two lines of study. It is commonly taught that the epic language was not spoken at any one place and time, although it includes no "manufactured" forms or grammatical usages; that it is so consistent that it is difficult to trace any evolution in assumed strata of the poems; that it came to be understood in many parts of Greece where it was difficult for those who spoke one dialect to understand those who spoke dialects not closely related. In other words, the language was distinctly "epic," created by the poets by assimilation from different sources. Are we not justified in assuming that the same principle holds true of the picture of social life? That the picture of religion, in like manner, does not reproduce the religion of any one place or one period, though it includes no absolutely new creation of the poet; that its consistency is due to the poet's unconscious art; that it came to be understood all over Greece, when the worship of one cult-centre would often be foreign to that of another cult-centre ?

If this assumption be granted, the study of epic religion should follow the same lines as the study of epic language. It is necessary first to study the picture of

religion in the poems with all due regard to what we may learn from other sources as to different "strata." The results of this study cannot be directly used for the religion of one epoch or one place, any more than the results of such a study of epic language or metre. Secondly, we may ask what modifying influences must be assumed as acting on the bards. Evidently the account of the gods and of worship is cut loose from local religious centres and given such a universal form as will suit poetry sung in many places. Again, the deeper phases of religion are not suited to the banquet occasion with which this poetry is associated. Perhaps the "rationalistic" atmosphere of the epic, its disregard for magic, some forms of divination, etc., is due partly to the attitude toward this phase of religion among the "princes" who were entertained by the bard. Thirdly, we may be able to connect some parts of this picture of religion with data from other sources, before and after the epic, and thus give it its true place in the history of Greek religion.

36. Can Ancient and Modern Views of the Minor Sapphic and Other Logaoedic Forms be reconciled? by Dr. Herbert W. Magoun of Cambridge, Mass. (read by title).

The object of this paper was to show that the difference between ancient and modern ideas of the Minor Sapphic and other logaoedic forms is chiefly one of viewpoint. The rhythm actually used in the days of Horace may have been, and probably was, essentially the same as that now employed. The reasons for this supposition are as follows: First, all logaoedics were composed in 4/4 time. The evidence on this point is conclusive. Second, all such measures contained rhythmical elements. This also can be abundantly proved. Third, the metricians confessedly omitted those elements. Fourth, pauses did occur within the lines, Schmidt et al. to the contrary notwithstanding. Native testimony on this point must outweigh modern conjecture. Fifth and last, the analyses that have come down to us are metrical, and therefore devoid of the rhythmical elements, which are necessary to complete the bars.

scheme :

The Minor Sapphic has the structure (Latin, standard form): UL-IL AUU_IUL. Stripped of its rhythmical elements, this gives the _|_uu_| ~ _ _. Adding the possible alternate short syllable in the fourth place (Greek form) and the syllaba anceps, gives the result: ~_~l_uu_ | ~ _ ~, which is exactly the analysis found in Hephaestion. The alternate short may occasion some trouble in the scheme; but it occasions none in practice, if the sense is properly observed. A balancing element usually a pause - always occurs in the bar. Observing the apparent trochees, Schmidt evidently surmised that the time was 3/8. He accordingly analyzed the line as (Greek): _~l_31-uu| _ ~ | _ ~, or (Latin): _~l_> |_twl_ul _, ignoring the fact that the final syllable, at least in Latin, is generally long. Others, however, modified the Latin scheme and treated the last two syllables as, _ | _ ^, by syncopation.

Schmidt's (Latin) analysis and the above 4/4 scheme have two things in common; namely, both recognize the fact that the third syllable is regularly longer than the fourth and that the fifth takes more time than the sixth and seventh. In the Greek the place of the caesura is not fixed, and the rhythmical elements

are used with much greater freedom, in the matter of position, than in Latin. The cyclic dactyl, so-called (Greek 3/8 scheme), has no justification.

The lack of agreement at the close, in the 3/8 and 4/4 analyses, seems to have been due to a desire on Schmidt's part for uniformity. A similar reason may be urged for the non-agreement, in some parts of the other forms, of the 3/8 and 4/4 analyses. The renderings actually used by Schmidt and other scholars were probably in 2/4 time, if not in 4/4. Correct 3/8 time is almost never used in practice. A 2/4 rendering results from the 3/8 schemes, because a slight deliberation is used in scanning, which amounts to the use of minute balancing pauses between the words and syllables. They are too brief to be noticed; for they are not over one-sixth of a second in length for ordinary speech. The 4/4 renderings and analyses will be found to satisfy all the essential requirements of both ancient and modern ideas on this subject.

The Latin forms are the more regular of the two, and they may be taken as the standard in consequence. Even these, however, show frequent irregularities. In the Greek, the rhythmical elements, including the caesuras, are constantly shifting their positions, and almost every line must be considered by itself. It was for this reason that the metricians confined themselves strictly to the conventional feet, which remained constant. In the Asclepiadean group, the forms with divided bars (see below) are Latin. The Greek may have preferred the other arrangement. The divided bars, indicated by the double lines (||), correspond to modern musical usage. The analyses (Latin standard lines) are as follows:

A few changes may be necessary in minor details. For example, there are reasons for thinking that the Asclepiadean group, in Latin, always ended either as ~~LI~_ ^ || or as