Horace's birth and burial, (3-4) the dates of his birth and death, (5) that he lived much in the country, (6) that he made Augustus his heir, and (7) that he did it orally. In the case of (7), in lieu of a citation of authority we have an explanation which lends plausibility to the statement. Even without attempting to explain why he fails to cite authorities for the other points,54 we may be assured, I think, that we have in the author of this Vita one who is concerned to put truthful statements before his readers. The second half of the long opening sentence of the Vita deals with the scriptus quaestorius, and refers to six matters: (1) Horace was concerned in the battle of Philippi, (2) his adherence was secured by Marcus Brutus, (3) he served as military tribune, (4) his side was beaten, (5) he secured pardon, (6) he became scriba quaestorius. The first five of these statements are sufficiently attested; 55 does it seem likely that the sixth is a fabrication?

Suetonius was a diligent searcher of records, and made at least some use of inscriptions.56 Christ thinks that the statements of the Vita regarding Horace's birth and death and his age at death were, in part at least, derived from the inscription on his tomb, noting that commentators and biographers were accustomed to consult the inscriptions on the tombs of famous men." 57 Very many sepulchral inscriptions indicating the holding of the scriptus quaestorius are extant.58 It is not absolutely impossible that the inscription on Horace's tomb showed that he had held this office, although, if the theory is correct that he purposely refrained from mentioning it in his writings, it

54 (1) His birthplace was of course known from Sat. 11, 1, 35. (2) His tomb was probably still to be seen in Suetonius's day iuxta Maecenatis tumulum. (5) was easily inferred from his writings.

55 (1) Carm. II, 7, 9 ff., III, 4, 26 ff., Ep. 11, 2, 49 ff. (2) Plutarch, Brut. 24: τοὺς σχολάζοντας ἀπὸ Ρώμης ἐν ἄστει νέους ἀνελάμβανε καὶ συνεῖχεν. Cf. Ep. II, 2, 43 ff. (3) Sat. 1, 6, 48. (4) Carm. II, 7, 9 ff., Ep. 11, 2, 47 f. evident fact, in view of his subsequent relations to Augustus.

56 Dennison, op. cit. p. 66.

(5) A self

57 Sitzungsber. d. bayr. Akad. d. Wissenschaften, 1893, Band 1, pp. 71 f. 58 See, e.g. the indices of Dessau, Inscr. Lat. Select.

is not likely that his friends would have had the fact recorded thus after his death.

It is almost certain, however, that lists of the scribes were preserved from year to year, and altogether probable that such lists were to be found among the public archives when the Suetonian Vita was composed. The portion preserved of the Lex Cornelia de XX Quaestoribus provides for the posting of lists of viatores and praecones,59 and there is little doubt that a similar provision was made in the lost portion of the law which dealt with the scribes.60 A senatus consultum of 23 B.C. which has been in part preserved 61 directed at least that the names of the sexprimi who headed the organization should be publicly posted each year. Various relics of these lists have been found from time to time.62 And more than this, a list of scribes published by Huelsen in 1902 63 appears not to be, like those previously known, a register of the sexprimi, but a complete list of the members of the college. In all probability, then, lists of the scribes on stone or bronze, or the originals on paper from which the former were prepared, or both, were accessible to the author of the Vita, and it was probably from this source-unless he got it from the inscription on Horace's tomb-that he derived his information. In neither of these cases, of course, would there have been any indication of the means by which the office was secured.

After carefully weighing the evidence, we may, then, believe that the statement that Horace held this office for a time is correct. This belief is to be based not upon Horace's writings, but upon the straightfoward character of the Suetonian Vita,64

59 C. I. L. 1, p. 109 (11, 40 f.): quorum viatorum praeconum nomina in eis decurieis ad aedem Saturni in pariete . . . <scripta erunt >. Cf. Mommsen, De Apparitoribus, p. 7.

60 Mommsen, De Apparitoribus, p. 10: His praemissis, quae etsi non traduntur de omnibus apparitorum generibus, tamen ad omnia referenda sunt, etc. 61 C. I. L. VI, 32272.

62 Cf. Dessau, op. cit. 1, p. 374, n. on inscription No. 1892.

63 Klio, II, (1902), p. 273, Inser. No. 60.

64 The exceptions taken by writers like Lessing, in his Rettungen des Horaz, to certain statements in the Vita do nothing, of course, to impair its trust

in which we possess information acquired by a competent investigator with reliable sources at his disposal, free from prejudice, and, so far as we can judge, concerned in this brief sketch to state the truth.

worthiness. The statements objected to are frankly cited by the author as matters of hearsay: intemperantior TRADITUR . . . scorta DICITUR habuisse, etc.

X.-Some Neglected Evidence Bearing on the Ictus Metricus in Latin Verse

BY PROFESSOR J. F. MOUNTFORD

CORNELL UNIVERSITY

Despite the controversy between C. E. Bennett and G. L. Hendrickson,1 despite the numerous articles published during the last twenty-five years, many of us are still uncertain how Latin verse was delivered by the Romans themselves, and how modern pupils should be taught to read it. Editions of the Latin classics are still being published in which students are told how to scan the poet's verses, but are given little or no guidance if they desire to utter the lines with their own lips. The author of such an edition can well defend himself by pointing to the fact that, in this matter at least, specialists in metrics are divided. Some authorities believe that the succession of long and short syllables is, in itself, sufficient to establish the rhythm of a line, and was felt to be sufficient by the Romans; they would accent the words of a verse precisely as they would accent the same words in prose.2 Others cling to the old, established method of demarcating the feet of a

1 Bennett's first article appeared in A. J. P. xix, 361-383; the discussion was continued in A. J. P. xx, 198-210; 412-428; 429-434.

2 Bennett's articles and his editions of the Latin poets did much to popularise this view, which had previously been advocated by Madvig. Its great appeal, undoubtedly, lies in the fact that it does not force us to violate the normal pronunciation of Latin words. Furthermore, the external evidence for an expiratory ictus is admittedly not overwhelming (see E. H. Sturtevant, A. J. P. XLIV, 319-338). It is remarkable that when Cicero (e.g. De Or. 111, 182; Or. 66, 222) and Quintilian (e.g. IX, 4, 74) discuss passages of verse embedded in prose, they seem to think only of the succession of quantities involved, and make no mention of any possible difference between prose and verse in the matter of accentuation. If an ictus were normal, it is strange to find that the only example Quintilian gives of a change of accent due to verse (1, 5, 28) is the change from volucres to volúcres; for that change is, as Quintilian implies, fundamentally one of quantity, to which the change of accent is incidental. Most recently, E. A. Sonnenschein (What is Rhythm?, p. 83) has given reasons for believing that rhythm may be felt without the use of ictus.

verse by an expiratory ictus metricus; if the normal prose accents conflict with the ictus, they are disregarded.3 A third group of metricians tells us that the normal word accents should be heard, but that an expiratory stress should also be evident at those places where the ictus of the verse is thought to fall.4 In other words, the very existence of an expiratory ictus is sometimes denied, and when its existence is admitted, there is a very serious difference of opinion concerning its exact relation to the word accent.

During the last few years, investigators have tended to put on one side what we may call the external evidence of Cicero, Quintilian, and the later Grammatici Latini, and have turned their attention to the internal evidence of the Latin verses themselves. They have analysed the positions of the sense pauses, they have hunted the caesura almost out of existence, and, fortified by a valuable array of statistics, have determined the places in a line of dramatic or heroic verse where the Roman poets seem to have aimed to place word accents. These various avenues of approach have led to very divergent results;

3 Since, in English and German, the feet of a verse are normally demarcated by an accented syllable, it has often been thought (e.g. by W. Thomson in his Rhythm of Speech, 1923) that such demarcations by stress are necessary to rhythm, and that consequently Latin verse must have had an ictus where the word accents failed to demarcate the feet. Supporters of this theory lay stress on the use of such words as ictus (e.g. Horace, C. Iv, 6, 36; A. P. 253; Quint. Ix, 4, 51; Diomedes, 1, 475 K; Ter. Maur. vi, 366 K), percussio (e.g. Cic. De Or. III, 182; 186; Hor. Sat. 1, 10, 43; Quint. IX, 4, 75; Bassus apud Ruf. vi, 555 K), pulsus (e.g. Quint. ix, 4, 136), and crepitus (e.g. Quint. IX, 4, 55); they point also to evidence for a difference between the rendering of prose and verse (e.g. Quint. 1, 10, 29; Pliny, Ep. 1, 16, 6). A passage of Aulus Gellius (VII, 7, 1-9), where the accentuation exádversum is apparently deduced from a line of Terence, is sometimes held to support this ictus theory. Of recent investigators, E. H. Sturtevant is the staunchest supporter of this view (see T. A. P. A. LIV, 61 and n.; T. A. P. A. Lv, 79).

4 W. G. Hale propounded this ingenious compromise (see P. A. P. A. xxVI, xxvi-xxix). This theory finds its chief difficulty in such lines as Aen. 1, 2 (Italiam fato profugus Laviniaque venit), where the ictual stresses and the word accents, taken together, are so numerous that it is hard, if not impossible, to avoid differentiating in favour of one set at the expense of the other. Such lines are by no means uncommon.