(ε) Ο 371 εύχετο, χεῖρ ̓ ὀρέγων εἰς οὐρανὸν ἀστερόεντα Ω 97 ἀκτὴν δ ̓ ἐξαναβᾶσαι ἐς οὐρανὸν ἀιχθήτην, Θ 364 ή τοι ὁ μὲν κλαίεσκε πρὸς οὐρανόν, αὐτὰρ ἐμὲ Ζεὺς κτλ. These are typical, not isolated cases. Taken together they form one of the indications, which it is the purpose of this part of the paper to point out, that in a considerable number of his verses the Homeric poet regarded the end of the fourth. foot as a proper stopping-place. A new sentence or clause might be begun here, or, if he chose to continue the same clause to the end of the verse, he had in stock a number of words and phrases by which no new point was added, but the thought was beautified or explained. II. The second characteristic of the bucolic diaeresis which marks it as similar in kind to the main caesura in its influence on the connection of thought is the fact that for successive verses it is possible to omit the last two feet without disturbing the narrative, e.g.: Μ 131 τὼ μὲν ἄρα προπάροιθε πυλάων (ὑψηλάων) ἔστασαν ὡς ὅτε τε δρύες ούρεσιν (υψικάρηνοι), (οἶος, σὺν γάμβροισι κασιγνήτοισί τε σοῖσιν·) τῶν νῦν οὐ τιν ̓ ἐγὼ ιδέειν δύναμ ̓ οὐδὲ νοῆσαι), ἀλλὰ καταπτώσσουσι, κύνες ὡς ἀμφὶ λέοντα). III. The tags which are suitable to follow the bucolic diaeresis are very numerous. Here again there is so much material that only a brief indication of its character can be given. (a) All the most prominent divinities and many heroes whose names consist of not more than three syllables have epithets of such length and quantities that the name and epithet together just fill the last two feet of the verse.1 1 μητίετα Ζεύς, εὐρύοπα Ζεύς, εὐρύοπα Ζῆν, πότνια "Ηρη, Πάλλας ̓Αθήνη, Φοίβος ̓Απόλλων, "Αρτεμις ἁγνή, ὄβριμος "Αρης, χάλκεος "Αρης, ὀξὺν "Αρηα, θοῦρον (b) In more than five per cent of the verses of the Iliad and Odyssey the subject, predicate nominative, object, or substantive modifier in an oblique case, with or without a preposition, just fills the last two feet of the verse. This class contains many familiar tags; e.g. θυμὸς ἀγήνωρ, ὄβριμον ἔγχος, νηλέϊ χαλκῷ, ἐν μεγάροισιν. (c) Some tags are used in several cases: The results of this examination of the use by the Homeric poet of the bucolic diaeresis, if accepted, will tend to weaken the argument for the origin of the hexameter which is based on the likeness of the caesura of the third foot to the pause at the end of the verse. For the same argument may be urged for the derivation from tetrapody and dipody1 since the poet's treatment of the bucolic diaeresis differs in degree only from his treatment of the pause in the third foot. It is not the purpose of this paper, however, to discuss the origin of the hexameter, but, in concluding, to raise the query whether it may not be conducive to a better appreciation of the poems to reason in the reverse direction? Instead of arguing from the use of the pauses to the derivation of the hexameter, may it not be more profitable to try to understand better the bearing of the musical or metrical pauses on the meaning and artistic effect of the verse? The poet's chief pause in the sense, as well as in the rhythm, is at the end of the verse. Next comes the caesura of the third foot, and after that in order of importance, the bucolic diaeresis. The treatment of these pauses is the same in kind. The "Αρης, οὔλον "Αρηα, δι' ̓Αφροδίτη, ὠκέα Ιρις, φαίδιμος "Εκτωρ, ὄβριμος Εκτωρ, Εκτορι δίῳ, Εκτορα δίον, δῖος ̓Αχιλλεύς, ὠκὺς (without πόδας) ̓Αχιλλεύς, φαίδιμος Αἴας, δῖος Οδυσσεύς. 1 See the article by E. von Leutsch in Philologus, XII (1857), p. 25 ff. pauses are used to divide the thought into units sufficiently short to be easily apprehended by the minds of those for whom they were composed, by the sense of hearing alone, and without the necessity for repetition as in the case of poems which are meant to be read. The burden of the narrative comes first in a whole verse, or a half-verse, or four feet. Then may follow in a whole verse, or the second halfverse, or the last two feet, the unessential but picturesque or explanatory part, without which, as Professor Seymour has observed, we should have prose, not poetry.' By the use of decided pauses in the sense at these (and other) metrical stops in varying combinations monotony was avoided. VII. - Donatus's Version of the Terence Didascaliae. BY DR. JOHN C. WATSON, CORNELL UNIVERSITY. THOUGH forty years have now elapsed since Dziatzko first wrote on the Terence Didascaliae,1 his treatise is still the most accurate and complete on this subject. Later writers have done little more than point out his errors in minor points, some of which he admitted. But the great value of his discussion should not blind us to the fact that our knowledge of the Didascaliae is still unsatisfactory in certain respects. It is my conviction that Dziatzko, both in his original papers and in his later writings, committed certain errors which have never been clearly recognized. Moreover, he himself admitted the existence of certain difficulties for which he could find no satisfactory explanation. The following paper, therefore, has been written partly to controvert certain views still generally accepted on the authority of Dziatzko, but chiefly to explain, at least in part, the difficulties for which no one as yet has been able to account. The Didascaliae are found in three well-defined recensions: first, the famous Bembine manuscript of Terence, containing Didascaliae for all the plays except the Andria; secondly, the whole body of later, or yd Mss, which also have lost the Didascalia of the Andria; and, thirdly, in paraphrases by the early commentator Donatus,2 such paraphrases forming part of the praefationes to the commentaries which exist for all the plays except the Hautontimorumenos. In form, and usually in content, these three sources are in such close agreement as to demonstrate their ultimate common origin. Material differences, however, in the information given by the three recensions render it difficult to determine with certainty 1 Rhein. Mus., XX (1865), pp. 570–598; XXI (1866), pp. 64–92. 2 The writer accepts the paraphrases as the work of Donatus and believes them free from serious changes by later hands. See p. 155 f. the readings of the latest common original. The study devoted to this subject by a long line of scholars has produced a text which is doubtless correct in most respects, but the origin and significance of many of the points of difference have never been understood. The need of a satisfactory explanation is obvious, for almost without exception the basis of every appeal from the accepted text has been the discrepant information given by the different sources. Differences in the three recensions can be sharply classi fied according as they are due, or are not due, to the repetition of plays. From statements in the life drawn from the Didascalia of the Eunuchus, it is known that the original Didascaliae were more complete than they are in any of our sources. Other evidence shows that they contained information not only about the first performances of the several plays, but also, to some extent, at least, about subsequent repetitions. In the process of transmission, ancient scribes, like modern editors, tried to drop references to any reproductions of the plays, and to preserve those relating to their first appearance. This tendency was not always strong enough to reduce the Didascaliae to items about single performances. It failed almost completely in the Hecyra, doubtless because of the peculiar history of this play. Perhaps the best example of information about more than one performance is found in the names of four aediles given by Donatus for the Andria. The name of a third aedile in the praefatio to the Hecyra is sometimes similarly explained. This is also the easiest way in which to account for the name Mummius, a third consul, apparently, given for the Eunuchus by all the later Mss. Finally, it is only by the theory of the repetition of plays that we can understand why two domini gregum are named instead of one. In attempting to restrict the Didascaliae to information about the original performance, scribes failed to see that one dominus gregis was in charge of a later repetition. Believing that the second name represented an actor (possibly the prologus) in the original troupe, they retained both names. In several instances it is apparent that scribes blundered, |