THE GIFT OF .'?:v-.ri..0.5..4:.4.% ..S...\:^.\.V\p 7583 NUMERICAL PHRASEOLOGY IN VERGIL Cornell University Library PA 6825.C59 Numerical phraseology in Vergil ... / 3 1924 026 572 580 a dissertation submitted to the faculty of Princeton University in candidacy for the degree of doctor of philosophy BY CLIFFORD PEASE CLARK PRINCETON I9I3 NUMERICAL PHRASEOLOGY IN VERGIL a dissertation submitted to the faculty of Princeton University in candidacy for the degree of doctor of philosophy BY CLIFFORD PEASE CLARK PRINCETON I9I3 Accepted by the Department of Classics June, 1910 C PMNCETON, N. J. THE FALCON PRESS I9I3 PREFACE. Grateful acknowledgment is hereby made to those professors of the University of Heidelberg, of the University of Chicago, and of Princeton University, who, by their continued encouragement, helpful advice, and careful teachiiag have made this thesis possible. I am especially indebted to Professor Duane Reed Stuart of Princeton University, who first suggested to me the field of this dissertation, and throughout the entire work has shown infinite patience and the utmost kindness. To Professor Edward Capps of Princeton University, who has shown the warmest personal interest in my investigations, by his timely suggestions, by reading the manuscript, as well as by his general oversight while this thesis has been in press, I wish to express my sincerest thanks. I am also deeply grateful for the inspiration received from the personal and class-room association with Professor Frank Frost Abbott of Princeton University, through whose continued interest in my behalf I was led to pursue this investigation. Acknowledgement is also made to The Falcon Press for the many courtesies extended to me during the printing of this thesis. C. P. Clark. Dartmouth College, Hanover, N. H., September i, 1913. ^ The original of this book is in the Cornell University Library. There are no known copyright restrictions in the United States on the use of the text. http://archive.org/details/cu31924026572580 TABLE OF CONTENTS. • Page. Bibliography 7 Introduction 10 Chapter I. Fixed Numbers 13 1. Dependence on Originals or Models 13 a) Dependence on Homer 13 b) Dependence on Sources other than Homer .. . 14 2. Fixed by Ritual 15 a) Ritualistic Two 15 b) Ritualistic Three 23 c) Ritualistic Four 26 d) Ritualistic Seven 28 e) Ritualistic Nine 31 f ) Ritualistic Twelve 34 3. Fixed by Convention 36 a) Tradition and Myth 36 b) Arms, Equipment, etc 39 c) Conventional Epithets 40 4. Fixed by Historical Fact 40 Chapter II. Favored Numbers 42 1. Myths 42 a) The Myth of Hercules and Cacus 42 b) The Theseus Myth 46 2. Magic Numbers 53 a) Magic Three 53 3. Variations from Sources 57 a) Homeric Suggestion 57 b) Non-Homeric Suggestion, Poetry 61 c) Non-Homeric Suggestion, Prose 67 5 4. Round and Indefinite Numbers 7° a) Round Numbers 7° b) Approximate Numbers 7^ c) Indefinite Numbers 73 5. Special Uses 75 a) Donations 75 b) Emphasis 7^ c) Transference 77 d) Experience 79 Conclusion 85 Index 87 BIBLIOGRAPHY. The literature which deals with the subject of numbers is much more complete on the Greek than on the Latin side. W. H. Roscher has dealt in a series of articles most fully and admirably with certain Greek numerals, especially seven and nine. Weinhold, Kaegi, and others have discussed, broadly, certain phases of the question relat- ing to the Germans and to the Eastern peoples. Hirzel and Wolfflin have given us some useful information on round numbers, while Seibel has discussed numbers from the standpoint of their usage in the poets as affected by the meter. Much may be learned concern- ing the magic power of numbers by a study of Heim's Incantamenta Magica, and similar collections. My indebtedness to these and other sources is greater than I can estimate and is gladly acknowledged. In the following bibliography my desire has been to include, not only the most important contributions to the study of numbers, but, in some cases, certain articles which have only a remote bearing upon the topics discussed in this thesis. Other articles are referred to in the notes. Andrian, F. — Die Siebenzahl im Geistesleben der Volker, Mitth. d. Anthropolog. Ges. in Wien 31 (1901), pp. 225 fif. Belling, H. — Review of A. Ludwich's Homerischer Hymnenbau nebst seinen Nachahmungen bei Kallimachos, Theokrit, Ver- gil, Nonnos, und anderen, Zeitschr. f. d. Gym. 64 (1910), Jahresb. pp. 145 fif. Breal, M. — Hercule et Cacus, Melanges de Mythologie, Paris, 1878, p. I ff. Cauer, F. — Die romische Aeneassage von Naevius bis Vergilius, Jahrb. f. class. Phil., Suppl. 15 (1887), pp. 97 flf. Diels, H. — Sibyllinische Blatter, Berlin, 1890. Ein orphischer Demeterhymnus, Festschrift f. Gomperz, pp. I flf. Drexler, W.— Review of W. Kroll's Antike Aberglaube (Hamburg, 1897), Woch. f. klass. Phil. (1898), col. 1066. Fowler, W. — The Number Twenty-seven in Roman Ritual, Class. Rev. 16 (1902), pp. 211 ff. Harrison, J. — Delphika, J. H. S. 19 (1899), pp. 205 ff. Heim, R. — Incantamenta Magica Graeca Latina, Jahrb. f. class. Phil., Suppl. 19 (1891), pp. 463 ff. Hempl, G. — The Sexagesimal System and the Cradle of the Aryans, Class. Rev. 16 (1902), pp. 413 ff. Heinze, R. — Virgils Epische Technik, 2nd ed., Leipzig, 1908. Hirzel, R. — Ueber Rundzahlen, Ber. d. sachs. Ges. d. Wiss., phil.- hist. CI., 37 (1885), pp. Iff. Kaegi, A. — Die Neunzahl bei den Ostariern, Phil. Abh. f. Schweizer- Sidler (Zurich, 1891), pp. 50 ff. Krall, J. — Die persische Flotte, Wiener Studien 3 (1881), pp. 147 ff. Lersch, L. — Antiquitates Vergilianae, Bonn, 1843. Lessman, H.-'— Aufgaben und Ziele, Myth. Bibl. I. 4. Lucas, H. — Die Neunzahl bei Horaz und Verwandtes, Phil. 59 (1900), pp. 466 ff. Miinzer, Fr. — Cacus der Rinderdieb, Programm z. Universitat Basel, Basel, 191 1. Norden, E.— P. Vergilius Maro. Aeneis Buch VI, Leipzig, 1903. Peter, H.— Die Epochen in Varro's Werk, De Gente Populi Romani, Rhein. Mus. 57 (1902), pp. 466 ff. Roscher, W. — Die enneadischen und hebdomadischen Fristen und Wochen der altesten Griechen, Abh. d. sachs. Ges. d. Wiss., phil.-hist. CI., 21, no. IV (1903), ("Abh. I" in notes). Die Sieben-und Neunzahl im Kultus und Mythus der Griechen, Abh. d. sachs. Ges. d. Wiss. phil.-hist. CI., 24, no I (1904), ("Abh. 11" in notes). Die Hebdomadenlehren der griechischen Philosopher! und Aerzte, Abh. d. sachs. Ges. d. Wiss. phil.-hist. CI., 24, no. VI (1906), ("Abh. Ill" in notes). Enneadische Studien, Abh. d. sachs. Ges. d. Wiss. phil.- hist. CI., 26, no. I (1909), ("Abh. IV" in notes). Die Zahl 40 im Glauben, Brauch und Schrifttum der Semiten, Abh. d. sachs. Ges. d. Wiss. phil.-hist. CI., 27 (1909), pp. 93 ff. ("Abh. V" in notes). Schmidt, J. — Die Urheimat der Indogermanen und das europaische Zahlsystem, Abh. d. Berl. Akad., phil.-hist. CI., (1890), Abh. II. Scott, J. — Homer's Estimate of the Size of the Greek Army, Class. Jour. 4 (1908), pp. 165 ff. Seibel, F. — Quibus artificiis poetae Latini numerorum vocabula difficilia evitaverint, Munich, 1909. Seiffert, O. — Die Totenschlange auf lakonischen Reliefs, Festschrift z. Jahrhundertfeier d. Universitat Breslau, pp. iii ff. Sutphen, M. — Magic in Theokritos and Vergil, Studies in honor of Basil L. Gildersleeve (Baltimore, 1902), pp. 315 ff. Usener, H. — Dreiheit, Rhein. Mus. 58 (1903), pp. i, 161, 321 ff. Weinhold, K. — Die mystische Neunzahl bei den Deutschen, Abh. d. Berl. Akad., phil.-hist. CI., (1897), Abh. II. Wolffin, E. — Sescenti, mille, centum, trecenti als unbestimmte und runde Zahlen, Archiv f. lat. Lex. 9 (1894-5), pp. 177 ff. Zur Zahlensymbolik, mit Probeartikel Septem und Novem, Archiv 9 (1894-5), pp. 333 ff. Das Duodecimalsystem, mit Probeartikel duodecim und sexaginta, Archiv 9 (1894-5), pp. 527 ff. INTRODUCTION. The Purpose and Plan of this Study. This study was begun with the thought that it might be possible to determine, by a careful examination of the numerical phraseology in the works of Vergil, how far the poet followed in this respect his models or originals and how far he worked independently of them. Peter in his Geschichtliche Literatur uber die romische Kaiser- zeit (II, p. 281) notes how, among the Roman historians of the empire, the striving for rhetorical effect has influenced the transmis,- sion of definite and exact numbers. Historians avoided numbers altogether or sought refuge in round or indefinite numbers. Krall ' has called our attention to illustrations of this tendency in Herodotus, e. g., in 4.87, where the historian says that Darius equipped 600 ships to fight the Scythians. With the same number the Persians sailed against the lonians (6.9), while Datis and Artaphernes col- lected 600 ships for their expedition against Greece (6.95). Such a combination of numbers at once excites suspicion and we doubt their historical accuracy. On the other hand, Scott' thinks that Homer is substantially correct in his poetical statement of the size of the Greek army. It would be rash, however, to argue from this or any similar theory that the historians are more inaccurate than the poets; yet it may safely be said that we often pass too lightly over a poet's statement as unworthy of serious consideration. An exhaustive study of the use of numbers by the poets and by the historians would, I am sure, yield much valuable material toward helping us to place a proper estimate upon many numerical state- ments, either now mistrusted or too rashly credited. It is, partly, with this aim in view that the present study has been undertaken. The primary object has not been simply to collect facts and statistics, interesting or curious in themselves, but rather to utilize whatever data may be forthcoming as a contribution to our appreciation of 'Wiener Studien, 3 (1881), pp. 147 ff. ^ Class. Jour., 4 (1908), p. 165. ID Vergil's genius and to our understanding of his poetic technique. Throughout the work my hope has been to obtain a further glimpse into the workings of the mind of a poet in whom inspiration was invariably affected by literary purpose and consciously controlled by definite methods and ideals of composition, who was, in other words, "doctus poeta." The main discussion falls naturally into two divisions, repre- sented by the chapters "Fixed Numbers" and "Favored Numbers." My first problem was to assemble under these two categories the pertinent passages, according as these seemed to contain numerals that were fixed by some outside influence that removed them from the operation of the poet's preference, or contained numerals that were due to the poet's unhampered invention. It remained then to determine, under the first main category, what the outside influence was, and under the second category, to discover if possible — a much more difficult and elusive task — the inventive motives at work in the poet's mind which led him to choose one number rather than another. The whole process has involved at every stage the important and often difficult matter of interpretation. The categories which seem to include most of the numerals that are fixed by influences other than the poet's preference are these: (i) The poet has felt himself bound by ritualistic usage to employ numerals which have become fixed under the influence of the customs and forms of Roman religion, beliefs, and superstitions. Again, (2) certain phrases have become conventionalized, and it is beyond even a poet's imagination or license to change them. History (3) and the facts of nature (4) also determine numerals beyond the possibility of arbitrary variation. In dealing with the second portion of this study. Favored Numbers, it has been by no means easy to bring a large number of cases under any one rubric, or to give for variations an explanation which would hold for a large number of passages. Many cases demand individual discussion, as, for example, (i) the Hercules- Cacus (A. 8.185 flf) and the Theseus (A. 6.14 ff) myths. It has been found necessary, in order to show the method of variation employed by the poet, to treat somewhat fully the origin of these myths, and to take cognizance of the possibility that various accounts may have been current at Rome when the epic was written. Some numbers, again, are used because of their (2) magic force or potency, and still qthers because of (3) the poet's fondness for Greek literature II and the numerals found there. Vergil's choice of (4) indefinite and round numbers is mostly confined to one hundred and to one thous- and. I have gathered under (5), Special Uses, characteristic pas- sages wherein Vergil seems to show a predilection for certain numer- als in the case of certain objects, as, for example, a definite number of presents of a certain kind, doubled numbers, and pairs. Some facts have been determined by the poet's experience, and in such cases he chooses the numeral which his knowledge of conditions in Italy has suggested to him. A method which Vergil occasionally employs may be termed transference: here is involved the shifting of a number used in connection with one act or object to a different act or object. 12 CHAPTER I. FIXED NUMBERS. I. Dependence on Originals or Models. a) Dependence on Homer: — It has many times been said — and truly — ^that Vergil is not a mechanical imitator. Such is the view of that most careful critic, Heinze,' when he says : "Denn das freilich lasst sich wissenschaftlich beweisen, dass Virgil nicht mechanish nachgeahmt hat." It is a fact, however, that Vergil some- times follows his originals so very closely that one hesitates whether to call such a passage mechanical imitation or simply a translation. In the following illustrations the force of this last assertion may be clearly seen : — A. 1. 162: hinc atque hinc vastae rupes geminique minantur in caelum scopuli. Od. 13.97: Svo Se irpo^XriTe'i iv avrm aKTol airopp&ye'i, Xi/ievo^ iroTnreirT'ijvlai. A. 2.503: quinquaginta illi thalami, spes ampla nepotum. //. 6.243 • avrap iv avrai •jrevT'^KOVT evearav daXafioi ^ecrrolo \i6oio. A. 6.582 : hie et Aloidas geminos immania vidi corpora. Od. 11.307: Kai p' ereKev Svo iralhe — /iivwdaBia) Se '•/evea-Qijv — ''Utov t avTiQeov TrjXexXeiTop t '"Ei(j)id\rrjv. A. 6.893 : sunt geminae Somni portae. Od. 19.562 : hoial yap re frvXai aiJt£vr]vS)v elalv oveipmv. A. i.g4: O terque quaterque beati. Od. 5.306: rpU fidxapei Aavaol Kal TerpaKK. " See, e. g., Heinze, Epische Technik, 2te Auf., p. 254. 13 A. 2.792: ter conatus ibi collo dare bracchia circum; ter frustra comprensa manus effugit imago, par levibus ventis volucrique simillima somno. Od. 11.206: rpU fjuh iep(Ov v 'OXufiirCwv. The Erinyes of Vergil and likewise of Aeschylus (Eum. 208) are the avengers of blood guilt, the punishers of matri- cides and of murderers. Vergil (A. 7.324, 454) also speaks of AUecto as the goddess who delights in grim war, in death, in anger, and in plots, while Tisiphone ( G. 3.552, A. 10.761 ) , sent forth from the Stygian blackness, leads in her train pestilence and dread, Morbos Metumque. So, too, Aeschylus in the Eumenides (478, cf. 938) represents the Erinyes as responsible for the visitations of disease and plagues. Numerous other parallels might be cited, but these are suiiEcient to show the very close dependence of Vergil upon the portrayals of the Furies found in Greek tragedy and especially in the early tragedy exemplified by Aeschylus. There is one point, however, in this agreement with Aeschylus that especially concerns us here, viz., the number of the Erinyes. Neither Homer nor Hesiod mentions any definite number and Aeschylus allows us to infer that there were several, making his chorus of Furies either twelve or fifteen in number. In exactly the same way Vergil (A. 6.572) speaks of the agmina sororum,"' and Pentheus (A. 4.469) sees the "' Forbiger ad A. 4.469 tries to show that the words agmen and agmina are consistent with the conception of the three Furies. Such an interpreta- tion is at least strained and it seems far better to suppose that Vergil in respect to the number is following Aeschylus, exactly as he has done in so many other particulars. 62 agmina Eumendum. Seneca also reverts to this older conception of the number, calling them Furiae, dira Furiarum agmina, turba Furiarum, etc. (Med. 958, 966, Thy. 78) . Vergil does not at all times, however, maintain consistency with this older idea, but very plainly follows the later modification, first represented by Euripides (Ores. 408), who makes the number three. A. 12.845 • dicuntur geminae pestes cognomine Dirae, quas et Tartaream Nox intempesta Megaeram uno eodemque tulit partu. The three as named by Vergil are Allecto (A. 7.324), Tisiphone (A. 6.571), and Megaera. The Furies are represented by Vergil under various forms, as we have already seen. In harmony with Aeschylus they are now harpies, now maidens with snaky locks; they are the Erinyes or Furies, pursuing their victims with implac- able hate ; they are the Dirae or the 'Apai of Aeschylus (Rum. 417) : they are the Poenae or the penalties of Justice, especially in their relentless pursuit of Sisyphus, Ixion, and Pirithoiis. Unlike Aes- chylus, Vergil never allows the Erinyes to become the kindly Eumen- ides; they nowhere show any such mitigation of their original im- placable character as Aeschylus portrays. In a word, it is to be noted that Vergil follows the early notions as to form and attributes of the Erinyes; that, as to number, while he sometimes adapts the earlier conception of a troop or host, yet at other times — ^but still following the Greek tragic poets in this respect — he adopts the definite number three. There is, then, a mingling of the numerical traditions, a wavering between the older Aeschylean version and the later Euripidean. Vergil has four passages in which mention is made of a hydra or the Hydra, i. e., the Lernaean Hydra. In two of these (A. 6.287, 8.300) no definite number is used in connection with the hydra's heads. But in the sixth book the shield of Aventinus is decorated with a Hydra girt about with a hundred serpents, evidently equiva- lent to a hundred heads : A. 7.655 : post hos insignem palma per gramina currum victoresque ostentat equos satus Hercule pulchro pulcher Aventinus, clipeoque insigne paternum centum angues cinctamque gerit serpentibus hydram. In the sixth book, however, the hydra that sits before the gate of Orcus is represented as a monster with fifty black throats: A. 6.576: quinquaginta atris immanis hiatibus hydra saevior intus habet sedem. This last is not actually the Lernaean Hydra, but only a similar monster, as was suggested by Heyne (ad loc). Euripides, in speaking of the Hydra, first {H. F. 419) uses the epithet fivpiOKpavot ; then {H. F. 1190) eKaTo Ovid (Met. 4.451, 10.22; Trist. 4.7.16), Pseudo-TibuUus (3.4.88), and Propertius (3.5.43), as well as the mythographer Hyginus (Fab. 151), abide by the traditional number three. Thus Vergil conforms to a well-established tradition and follows here, as often, the Greek tragedians. In the following passage Vergil, though differing intentionally, or possibly unintentionally, from Hesiod in details, has followed him in making the fifth of the month unlucky : G. 1.277: quintam fuge: pallidus Horcus Eumenidesque satae ; turn partu terra nefando Coeumque lapetumque creat saevomque Typhoea et coniuratos caelum rescindere fratres. Hes. Op. 802 : irefiTTTa'; 8' i^aXeaa-ffai, iirel y^aKe-n-at re koI alvaC. iv Tre/jLTTTr) yap ^acnv 'E/siz/ua? afi^ciroXeveiv "OpKOv yeivofievov,*" rov "Epi? re/ce tt'^/a' iiriopKovi. Vergil attaches unlucky attributes to the fifth day because, as he thinks, pallidus Horcus (i. e., Orcus, god of the dead), the Furies, and the Giants as well, were all born on that particular day. It is plain, also, that the "O/j/eo? of Hesiod, god of the oath, bane of per- jurers, has been confounded with Orcus, god of the lower world. Vergil makes another slight deviation from Hesiod in saying that the Erinyes were born on the fifth, not that they attended "0/s«o?, who was born on that day, if we credit Hesiod. Jahn" has given a very plausible explanation of Vergil's varia- tion, suggesting, on the authority of Pliny (A^. H. 18.32.75), that possibly Democritus is responsible for not a few of Vergil's vagaries. I am inclined to think that, whereas Vergil may have been influenced by both Democritus and Hesiod, yet in the end we find Vergil ex- pressing what seems to him to be the prevailing tradition, in con- formity with early Greek writers. In the third Eclogue, especially, Vergil has enlarged upon his model in a very striking manner. For the sake of enhancing the effect, he allows himself the privilege of making a statement which almost surpasses credulity, as for example : E. 3.5 : hie alienus ovis custos bis mulget in bora. Theoc. 4.3 : fi ird yjre Kpv^Bav to, ■jrodeaTrepa iraaa<; afi,i\yeL<; ; Here, evidently for the sake of heightening the taunt, the secret " Many favor the reading Tivv/Ums. "Rhein. Mus. 58 (1903), p. 419- 65 mflkings of the flock are increased by Vergil from once at evening to twice an hour, or, in other words, to a number hardly to be credited. Similarly, too, in still another passage : E. 3.34 : bisque die numerant ambo pecus, alter et haedos. Theoc. 8.16: ra he fidXa iroOeaTrepa •kolv't apiOfievvri. In Vergil the harsh parents number the herd twice daily, while one or the other numbers the kids separately. In Theocritus there is but a daily numbering of the flock, at eventide. Vergil has done as well as Theocritus, if not better, in setting forth the desirability of certain gifts : E. 3.29: ego banc vitulam (ne forte recuses, bis venit ad mulctram, binos alit ubere fetus) depone. Theoc. 1.25: alyd re rot Beoaa> BiSu/iaroKOV e? r/al? a/xeX^ai, a Sv e'xpia ip(a)i irorafieXyeTai, e? Svo ireXkaf. We see that Vergil offers as a prize a heifer that, though with twins by her side, yet gives in addition two pails of milk each day. Theocritus does not give the goat and kids outright, but merely offers all the milk the goat will give in three milkings, two pails full to a milking, though she has two kids by her side. Theocritus, to be sure, adds to the present of milk a large drinking-cup or bowl. Certainly Vergil has outdone Theocritus in the value of the prizes, if not in extolling the wonderful capabilities of his heifer as com- pared with the goat." Again, if we wish to compare Theocritus 8.86, we find that the present is a hornless goat, that fills the pail with milk beyond the brim, but with no kids by her side. Vergil, as it seems to me, has striven in all these passages to enlarge upon the suggestion offered by his model. In the following passage, also, where the suggestion is again taken from Theocritus, the enlargement common to the examples just quoted is visible: E. 2.40 : praeterea duo nee tuta mihi valle reperti capreoli. Theoc. 3.34: ^ fidv roi Xevxav BiSv/iaroKov alja (ftvXdaaa). There is yet another imitation of Theocritus in which Vergil has added a numeral, apparently for the sake of vividness and definite- ness of detail: " See Fritzsche, Theoc. ad loc., for a similar interpretation, and also Kynaston's note. 66 E. 8.37: saepibus in nostris parvam te roscida mala (dux ego vester eram) vidi cum matre legentem. alter ab undecimo tum me iam acceperat annus; iam fragilis poteram ab terra contingere ramos. Theoc. 11^5- VpO'O'&t)''' M^" eyarye reows, Kopa, dviKa irparov ^vdei ifia aiiv fiarpi, deXoia vaKivBiva airl xP^ iroieWai, Swep ia-ri Kal tiKLvav. "P. 32. 68 G. 1.208: libra die somnique pares ubi fecerit boras et medium luci atque umbris iam dividit orbem: exercete, viri, tauros, serite hordea campis usque sub extremum brumae intractabilis imbrem. Varro R. R. 1.34.1 : sexto intervallo ab aequinoctio autumnali incipere scribunt oportere serere usque ad diem nonagesi- mum unum. G. 1.215: vere fabis satio. Varro R. R. 1.34.2: fabam optime seri in vergiliarum occasu. G. 1. 160: dicendum at, quae sint duris agrestibus arma, quis sine nee potuere seri nee surgere messes : vomis et inflexi primum grave robur aratri tardaque Eleusinae matris volventia plaustra tribulaque traheaeque et iniquo pondere rastri; virgea praeterea Celei vilisque supellex, arbuteae crates et mystica vannus lacchi. Varro R. R. 1.22.3 • itaque, Stolo inquit, proposita magnitudine fundi de eo genere Cato scribit, oliveti iugera CCXL qui coleret, eum instruere ita oportere, ut faceret vasa olearia iuga quinque, quae membratim enumerat, ut ex acre ahenea, urceos, nassiternam, item alia: sic e ligno et ferro, ut plostra maiora tria, aratra cum vomeribus sex, crates stercorarias quattuor, item alia: sic de ferramentis quae sint et quot opus (multitudinem), ut ferreas octo, sarcula totidem, dimidio minus palas, item alia. So also in his instructions concerning the feeding of young calves Vergil's language shows less precision than Varro's: G. 3.174: interea pubi indomitae non gramina tantum nee vescas salicum frondes ulvamque palustrem, sed frumenta manu carpes sata. Varro R. R. 2.5.17: semestribus vitulis obiciunt furfures triticios et farinam hordeaceam et teneram herbam et ut bibant mane et vesperi curant. In these passages Vergil is nowhere incorrect in the details so far as they are introduced, but by omitting numerical definition and designation he often expresses himself with less precision than does his prose source. In one case only, i. e., when he says that the bees are generated within a period of nine days as over against Varro's 69 much longer period, is there any real difference in numerical detail. But, as we have already shown," Vergil wished to make his period correspond with that well-known nine-day interval so common among the Romans. Elsewhere, we may say, Vergil is true to tra- dition so far as that tradition serves him, but he feels free to omit details or to forego specific statement if such a course appeals to him and is in accord with his poetic purposes. This is precisely what we might expect to find when a poet is dealing with subjects that belong properly to the domain of prose. Round and Indefinite Numbers. a) Round numbers: — It seems to me that there are to be dis- tinguished two well-defined classes of round numbers in Vergil : the first, in which the numbers are evolved wholly from the poet's imagination; the second, approximations to the fact, tending to satisfy conditions already expressed or understood. The technique of the poet in the creation of numbers of the first class is best illustrated by his enumeration of the troops that go forth to battle. We find mille twice (A. 10.167, 178), tercentum twice (A. 10.173, 182), quingenti (A. 10.204) and sescenti (A. 10. 172) once each. It appears that the poet intends to give us reason- able, and at the same time different, numbers, and thereby to clothe his account with a semblance of historical accuracy. It is note- worthy that he favors the numbers which are multiples of one hundred, just as in Homer's Catalogue nearly all the numbers are multiples of ten. This favoring of the hundreds is further illustrated in A. 9.370, where three hundred horsemen are sent out from the Latin city under their leader, Volcens. In A. 8.518 two hundred horsemen are given to Pallas by Evander, two hundred by Pallas to Aeneas. One hundred men follow each of seven leaders in A. 9.162. It will be noted that in A. 1.703 one hundred female servants and one hundred male tend the tables, while but fifty pre- pare the food within. One hundred envoys go to the town of Latinus (A. 7.153), and one hundred come from Latinus to the Trojans {A. 11.331). Lausus (A. 7.652) leads from the city of Agyllina one thousand men, while Aeneas sends one thousand men to attend the funeral of Pallas (A. 11. 61). It seems as though in '" See Ritualistic Nine, p. 32. 70 these and in other similar cases centum, mille, and even quinquaginta serve the poet's purpose admirably, being definite in outward appear- ance and conveying to the reader the impression of being an accu- rate numerical statement. Hirzel " in his discussion of round numbers, recognizes the especial appropriateness of the numbers "ten", "hun- dred", "thousand" for poetic purposes. He says : "Diese AUgemei- heit ist der Grund gewesen, weshalb solche Zahlen bei den Dichteni besonders beliebt sind ; denn man kann sich ihrer bedienen ohne wie bei den auf die Einer bestimmten Zahlen Gefahr zu laufen, dass durch sie der rechnende Verstand mehr, als fur den Zweck der Dichtung wiinschenswerth ist, in Thatigkeit gesetzt wird." As in the case of the larger numbers, so with the smaller, cer- tain numbers prevail. When Vergil wishes to indicate the great wealth of Galaesus (A. 7.538) he ascribes to him five herds of cattle, five flocks of sheep, and one hundred plows. In E. 3.71 the present to the maiden is one of ten apples, and Sinon {A. 2.126) is silent for twice five days. Turnus appears at the city attended by twenty youth (A. 9.48), while the chosen chiefs sail forth to the aid of the Trojans in thrice ten ships (A. 10.213). There is a noticeable fondness for the multiples of ten, as Hirtzel has already indicated. b) Approximate numbers: — These satisfy known conditions, — at least they are not improbable numbers, in view of statements made earlier in the poem, — conforming to the numbers required by the structure of the poem. In a sense, then, these numbers, also, are products of the imagination, but the structural process of the poet's imagination having been carried through, the numerical details which follow are subject to the conditions already fixed by the poet and known to the reader. Such, for example, is A. 2.501 : vidi Hecubam centumque nurus Priamumque per aras sanguine foedantem ipse sacraverat ignis. The two words requiring comment are centum and nurus. How is it possible for Vergil to say there were one hundred daughters- in-law? There were approximately one hundred included in the band of women designated by nurus*'' The word includes daugh- ters as well as daughters-in-law, as is clear from the fact that natae "Uber Rundzahlen, Ber. d. Gesell. d. Wiss. zu Leip., phil.-hist. CI., 27 (i88s), pp. 3 ff- "' Cassandra, for example, was not there ; she is accounted for in 403. 71 (515) IS used to describe the same group. It is plain that neither centum nor nurus is strictly accurate, yet either would satisfactorily include all the women. It is possible, too, if the Andromache of Ennius were extant in its entirety, that a full solution of all the difficulties involved in Vergil's statement would be forthcoming. Cicero (Tusc. D. 3.1944) has quoted from that play a few verses from which we infer that the poet included in his theme some account of the affairs of Priam and his household. Vergil was undoubtedly following a form of the myth which had increased the number of Priam's children from Homer's more modest sum to a number which Vergil expresses by the approximate number centum. Homer (//. 6.244, cf. 24. 495) gives Priam fifty sons, and implies that there were twelve daughters, but nowhere does he refer to the number of fifty sons and fifty daughters. This latter number might have developed later in the following way : the fifty sons, their fifty wives, and the twelve daughters become fifty sons, fifty wives, and fifty daughters, i. e., one hundred nurus. A similar use of both mille and decern is to be found in several places in Vergil, wherever the number of ships or the length of the war is mentioned {A. 2.198, 9.148, et al.). That the precise number of Greek ships might have been inserted in the Latin hexa- meter is illustrated by the author of the Epitome of the Iliad (220) : bis septem venere minus quam mille ducentae. But precise numerical definitions like this would not have satis- fied Vergil's purpose at all. Again, in A. 8.716 we are told that Augustus dedicated three hundred shrines, a statement which the poet did not intend should record the exact truth but should express what we know to be the case, viz., that a large number of shrines and temples was restored and reconsecrated under Augustus. The poet, further, assumes a poet's right, referring to one time restorations which no doubt ex- tendel over many years. So also in A. 7.170: tectum augustum ingens, centum sublime columnis. When presents are to be made by Dido to the companions of Aeneas {A. 1.633), Vergil takes into account the approximate num- ber of ships and represents the Queen as giving twenty bulls, approx- imately one for each ship. So, too, when the deer are killed {A. 1. 1 92) the number equals that of the ships. Of the smaller animals, sheep and swine, a much larger number would be necessary ; 72 hence the poet uses centum, an approximation which very nearly, if not precisely, meets the needs of so large a number of sailors/" c) IndeUnite numbers: — These are round numbers more or less indefinite in meaning, and may be translated either by "several", a limited number, or by "many", an unlimited number. For ex- ample in G. 1. 410: tum liquidas corvi presso ter gutture voces aut quater ingeminant. The ravens croak three or four, i. e., several times. The wailing women beat their breasts three or four times in A. 4.589, cf. 12.155. So also in A. 1.94, when the poet says that thrice and four times happy are they who died beneath the walls of Troy. It may properly be said in all these cases that the poet wishes his readers to under- stand that the numerals are but approximate round numbers, equiva- lent to "several", yet with a limitation. The idea of "many" or "very many" is usually expressed by centum or mille when the numeral is used. A typical instance of this usage is A. 6.625 (cf. G. 2.43) : non, mihi si linguae centum sint oraque centum, ferrea vox. The immediate inspiration of Vergil's expression, as Servius (ad loc.) observes, is a fragment of Lucretius (inserted by Lachmann as 6.840), which our author has repeated almost word for word, using ferrea vox where Lucretius used aerea vox. The prototype of the topos, however, as Macrobius {Sat. 6.3.6) has pointed out, and Servius as well, is Hom. II. 2.489 ff. : ovS' ei fioi BeKa fiev 7roi. In A. 11.285, when the embassy returns to Latinus and tells of Diomedes and of the speech that was made to the ambassadors, Vergil puts into Diomedes' mouth the phrase : si duo praeterea talis Idaea tulisset terra viros, ultro Inachias venisset ad urbes Dardanus, et versis lugeret Graecia fatis. When Vergil wishes to express the wonderful fertility of the fields and the extraordinary productivity of the herds he says : G. 2.150: bis gravidae pecudes, bis pomis utilis arbos. In A. 4.228 Venus is represented as having twice snatched Aeneas from the hands of the Greeks, and in A. 3.476 Anchises was twice rescued from the ruins of Pergamum. We should compare the following with a thought similarly expressed in Homer : A. 9.161 : bis septem Rutuli, muros qui milite servent delecti; ast illos centeni quemque secuntur purpurei cristis iuvenes auroque corusci. //. 9.85 : en-T eaav '^yefi6vepeva)v irat? aX.{a P- 71 p. 62 p. 40 P- 38 p. 78 p. 62 p- 63 p. 48 P- 39 Book 10. Book 8. 42, p. 47, P- 130, p. 168, p. i8s, p. 20s, p. 207, p. 330, p. 288, p. 291, p. 300, p. 399, p. 447, P- 448, p. 461, p. 496, p. S18, p. 565, p. 630, p. 661, p. 697, p. 714, P- 716, p. Book 9. P- 75 p. 46 36 47 Z1 47, 42 43 45, 56 19 74, P- 75 63 37 59 47, P- 40 75 70 47 Z7 39 15 41 25, P- 72 78 30, p. 78 48, p. 71 132, p. 75 148, p. Zl, p. 72 155, P- 37 161, p. so, p. ^^ 162, p. 48, p. 70 263, p. 47, p- 76 272, p. 57 370, p. 70 549, p. 75 586, p. 56 599, P- n 635, p. 11 706, p. 14 799, P- 57 p. 48 167, p. 70 172, p. 70 173, p. 70 178, p. 70 182, p. 70 202, p. 37, 204, p. 70 213, P- 47, P- 71 30 28 59 566, p. 47 68s, p. 56 761, p. 62, p. 75 783, p. 59 844, p. 41 873, P- 56 88s, p. S6 328, 517, 565, Book ii. 61, p. 72, p. 133, p. 167, 188, 210, 28s, p. 290, p, 326, 331. 397, 402, 629, 785: Book 12. 34, P 125, P 155, P 162, p, i6s, 270, 580, 754, 84s, 925, 70 59 47, P- 59 75 23, P- 56 25 77 37 47 70 75 77 56 22 77 75 73 35 39, P- 47 23 39 77 75 63 59, P- 78 Georgics I. 15, 47, 160, 172, 208, 215, 231, 74 80 47 69 69 14 88 Georgics i (Continued). 232, p. 47 233, P- 14 24s, P- 15 258, p. IS 277, P- 6s 284, p. 83 286, p. 84 410, P- 73 489, p. 40 Georgics 2. ' 43, P- 73 150, p. 77 398, p. 80 535. P- 37 Georgics 3. 18, p. 74 32, p. 41 60, p. 81 87, p. 15 113, P- 38 174, P- 69 190, p. 81 381, p. 40 404, P- 15 552, p. 62 Georgics 4. 88, p. 82 95, P- 47 231, p. 82 292, p. 38 29s, p. 68 336, p. 31 453, P- 28 454, P- 27 473, P- 75 493, P- 55 507, p. 30 528, p. 28 538, p. 27 S44, P- 27, p. 32 547, P- 27 550, p. 46 552, p. 32 Eclogue i. 42, p. 35, P- 48 Eclogue 2. 21, p. IS 36, p. 61 40, p. 66 42, p. 48 Eclogue 3. Eclogue 5. Eclogue 8. S, p. 65 6s, p. 21, p. 60 37, p. 67 29, p. 66 67, p. 48, p. 49 73- P- 54, P- 55 30, p. 47 34, p. 66 36, p. 76 CiRIS. 44, p. 76 Eclogue 6. 71, p. 71 247, p. 75 104, p. 55 75, P- 61 371, P- 54 89 "Q ^m