7^; 
 
 rn 
 
 UNIVERSITY OF PENNSYLVANIA 
 
 A 
 
 The Sacra Idulia in Ovid's Fasti 
 
 A Study of Ovid's Credibility in 
 
 Regard to the Place and the 
 
 Victim of this Sacrifice 
 
 BY 
 
 HORACE WETHERILL WRIGHT 
 
 A THESIS 
 
 PRESENTED TO THE FACULTY OF THE GRADUATE SCHOOL IN 
 
 PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR 
 
 THE DEGREE OF DOCTOR OF PHILOSOPHY 
 
 NEWARK. NEW JERSEY 
 1917 
 
GIFT OF 
 
 
 
UNIVERSITY OF PENNSYLVANIA 
 
 The Sacra Idulia in Ovid's Fasti 
 
 A Study of Ovid's Credibility in 
 
 Regard to tiie Place and the 
 
 Victim of this Sacrifice 
 
 BY 
 
 HORACE WETHERILL WRIGHT 
 
 A THESIS 
 
 PRESENTED TO THE FACULTY OF THE GRADUATE SCHOOL IN 
 
 PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR 
 
 THE DEGREE OF DOCTOR OF PHILOSOPHY 
 
 NEWARK, NEW JERSEY 
 1917 
 
I*. 
 
 ([Oft* 
 
I am happy in having this opportunity gratefully to acknowl- 
 edge the valuable suggestions and kindly criticism and proof- 
 reading of Professors J. C. Rolfe, W. B. McDaniel, R. G. Kent, 
 Assistant Professors G. D. Hadzsits and H. B. Van Deventer, 
 and Doctors T. A. Buenger and E. H. Heffner of the University 
 of Pennsylvania ; also the helpful suggestions and kindly reading 
 and criticism of the first chapter given by Professor Kirby F. 
 Smith of Johns Hopkins University, who was Acting Director 
 of the School of Classical Studies of the American Academy in 
 Rome during my Fellowship in that Institution. 
 
 H. W. W. 
 
 381669 
 
PREFACE 
 
 The passages in Ovid's Fasti which relate to the sacra Idulia 
 give rise to three problems, namely, the place of sacrifice, the sex 
 of the victim, and the age of the victim. The scope of this thesis 
 is confined to the first two problems. The third is so intimately 
 related to a study of the iuvenci offered Jupiter on January 1, 
 that it has seemed best to reserve it for a future article, in which 
 the investigation will be directed as much to cattle as to sheep 
 and will, therefore, involve a close study of other sacrifices besides 
 the sacra Idulia. 
 
SELECTED BIBLIOGRAPHY 
 
 Corpus Inscriptionum Latinarum; cited CIL. 
 
 Daremberg et Saglio, Dictionnaire des antiquites grecques et 
 
 romaines. 
 
 Fowler, W. Warde, Roman Festivals, London, 1908 ; cited R. F. 
 Fowler, W. Warde, The Religious Experience of the Roman 
 
 People, London, 1911; cited R. H. R. P. 
 Franke, C., De Ovidii Fastorum Fontibus Capita Tria, diss., 
 
 Halle, 1909. 
 Henzen, Acta Fratrum Arvalium, Berlin, 1874; cited Act. Fr. 
 
 Arv. 
 
 Huelsen-Carter, The Roman Forum, 2nd ed., Rome, 1909. 
 Jordan, Topographie der Stadt Rom, Berlin, 1871-1907; Vol. I, 3, 
 
 by Ch. Huelsen, 1907 ; cited Jordan. 
 Keil, Grammatici Latini, Leipzig, 1857-1880; cited Keil. 
 Krause, De Romanorum Hostiis Quaestiones Selectae, diss. 
 
 Marburg, 1894; cited Krause. 
 
 Mau-Kelsey, Pompeii, Its Life and Art, 2nd ed., New York, 1904. 
 Merkel, R., Prolegomena to his edition of Ovid's Fasti, Berlin, 
 
 1841. 
 Neue-Wagener, Formenlehre der lateinischen Sprache, 3rd ed., 
 
 Leipzig, 1902 ; cited Neue-Wagener. 
 Overbeck-Mau, Pompeii, 4th ed., Leipzig, 1884. 
 Pauly-Wissowa, Real-Hncyclopddie der classischen Altertumswis- 
 
 senschaft; cited Paul.-Wiss. 
 Peter, Hermann, Ovid Fasten I, 4th ed., text and commentary, 
 
 Leipzig, 1907. 
 
.,.3 
 
 Peter, Hermann, Ovid Fasten II, critical notes, 3rd ed., Leipzig, 
 
 1889. 
 Platner, Topography and Monuments of Ancient Rome, 2nd ed., 
 
 Boston, 1911. 
 Schanz, Geschichte der romischen Litteratur, 3rd ed., 1907-1913, 
 
 in Ivan von Miiller's Handbuch der klassischen Altertums- 
 
 wissenschaft ; cited Schanz. 
 Smith, A Dictionary of Greek and Roman Antiquities, 3rd ed., 
 
 London, 1891. 
 Teuffel and Schwabe, History of Roman Literature, translation 
 
 from the fifth German edition by George C. W. Warr, Lon- 
 don, 1891. 
 
 Thesaurus Linguae Latinae, Leipzig ; cited Thesaurus. 
 Van Deman, Esther Boise, The Atrium Vestae, Washington, D. 
 
 C., Carnegie Institution, 1909. 
 Wissowa, Religion und Kultus der Romer, in Ivan von Miiller's 
 
 Handbuch der klassischen Altertumswissenschaft, Munich, 
 
 1912; cited Wiss. 
 
 The editions cited of the following authors, unless otherwise 
 stated in the text, are : 
 
 Festus and Paulus Diaconus, O. Miiller, Leipzig, 1889. 
 Gellius, Teubner text, Leipzig, 1903. 
 Macrobius, Teubner text, Leipzig, 1893. 
 Varro, de Lingua Latina, Goetz and Schoell, Leipzig, 1910. 
 
THE SCENE OF THE SACRA IDULIA. 
 
 One of the most interesting sacrifices mentioned in Ovid's 
 Fasti, and one which presents some of the knottiest problems, is 
 that of the ovis Idulis, 1 or sheep sacrificed on the Ides of every 
 month to Jupiter by the Flamen Dialis. 2 Where was this cere- 
 mony performed? Ovid (Fasti I 587-88) says, 
 
 "Idibus in magnfcastus lovis aede sacerdos 
 Semimaris flammis viscera libat ovis." 
 
 According to him, then, the burnt offering was made in an aedes 
 of Jupiter. Before turning to the question of what aedes is 
 meant, it will be of interest to discuss on what part of the sacred 
 premises the Romans were accustomed to slay a victim and burn 
 its exta, the portion offered to the god ; in other words, whether 
 Ovid's phrase in aede be ritualistically correct. 
 
 The Latin in with the ablative covers the meaning of both "in" 
 and "on" in English. Now while mention is made again and again 
 in the Ada Fratrum Arvalium of the immolatio or slaying of 
 victims in Capitolio,* the phrase means "on" the Capitoline, 
 rather than in the temple of the Triad ; for we have a passage 4 "in 
 Capitolio . . in tem[plo lovis optimi maximi? fratres Arvales] 
 Iov[i] o(ptimo) m(aximo) bovem marem immolarunt," where the 
 distinction is clearly drawn between the hill and the temple. More- 
 over, the words in templo are extraordinary, for in turning over 
 page after page of the Acta we continually read of sacrifice to 
 Jupiter in Capitolio, but here only is the temple distinctly 
 specified. Even more illuminating is the immolation "in templo 
 novo divo Aug(usto)" of a bos mas,* for on another page we 
 
 1 Fasti I, 56; Macrob. I, 15, 16; Paul. 104, 17. 
 
 * Macrob. I, 15, 16; F. I, 587-88. 
 
 * Henzen, Act. Fr. Aru., p. XXXVI, p. XLIII et al. The parenthesis 
 indicates the filling out of an abbreviation, and the bracket a break in the 
 original inscription. 
 
 4 Henzen, Act. Fr. Aru., p. CXXIL 
 ' Henzen, Act. Fr. Aru., p. LXIX, 
 
10 
 
 meet with the sacrifice to Divtis Augustus ad templum novum* 
 and again the words, "an[te templum novum divo Augusto 
 bovem] marem et div[ae Augustae vaccam i]mmolavit." T 
 
 At once the question arises, what is meant by templum. The 
 author in Smith's Dictionary of Antiquities 3 in discussing Augus- 
 tus* use of the two words templum and aedes in the Monumentum 
 Ancyranum 19, 20, and 21, makes the following observation: 
 
 "It seems to us a truer view that the use of templum for aedes 
 was coming in before the end of the Republic, and that Augustus 
 in speaking by name of pre-existing temples, uses the term which 
 originally described them, but in those which he has just built 
 uses the term now in vogue." 
 
 Vitruvius throws some light on the word, in III, 4, 4: 
 
 "Gradus in f ronte constituendi ita sunt uti sint semper inpares. 
 Namque cum dextro pede primus gradus ascendatur, item in 
 summo templo primus erit ponendus." 
 
 This is translated by Morris Hicky Morgan as follows : 
 
 "The steps in front must be arranged so that there shall always 
 be an odd number of them ; for thus the right foot, with which 
 one mounts the first step, will also be the first to reach the level 
 of the temple itself." 
 
 The writer of the article on templum in Daremberg et Saglio's 
 Dictionaire des antiquites grecques et romaines 9 thus comments 
 on the passage: 
 
 "Un passage de cet auteur semblerait meme montrer que ce 
 terme, en architecture religieuse, indiquait seulement 1'aire sur 
 laquelle s'elevaient les murs et les colonnes de Tedifice ; la surface 
 superieure du podium est pour lui le summum templum" 
 
 This view seems correct, namely, that Vitruvius means by 
 summum templum the upper surface of the podium, the area 
 on which the walls and colonnade rested, and upon which one 
 stepped on reaching the head of the stairway. Templum would 
 then be identical with podium. This is a strict interpretation of 
 the passage, but, by a natural extension, templum would include 
 not only the summum templum or upper surface, but the walls 
 
 9 Act. Fr. Arv., p. XLVIII, a. 39. 
 ' Act. Fr. Arv., p. LXIII, a. 57. 
 ' Vol. II, p. 773. 
 p. 107. 
 
11 
 
 and colonnade which surrounded it, and any sacrifice performed 
 in templo would probably have been offered in the cella or in the 
 portico. In buildings of the Maison Carree type, however, a strict 
 interpretation of the word templum would include also the steps 
 cut in the podium and leading to the upper surface. 10 
 
 Varro 11 and Granius Licinianus, quoted by Macrobius, 12 tell of 
 the immolation of victims "in" the Regia, and in the foundations of 
 a temple excavated at Alatri are traces of an altar in the pronaos. 13 
 The mere fact, however, that the preposition in is employed only 
 in rare instances would indicate that the place of sacrifice was 
 unusual, and especially is this seen in the contrast above noted in 
 the A eta Fratrum Arvalium between in templo and in Capitolio. 
 In the three sacrifices to Augustus we have in, ad, and ante 
 templum novum. Ad and ante appear to mean about the 
 same thing, and must be references to immolation outside 
 the templum proper and in front of the steps leading up the 
 podium to the portico. One more passage from the A eta Fratrum 
 Arvalium must be quoted: "in Capitolio ante cellam lunonis 
 reg(inae) . .fratres Arvales convenerunt et immolaverunt . . 
 I(ovi) o(ptimo) m(aximo) b(ovem) m(arem) a(uratum), 
 lunoni reg(inae) b(ovem) f(eminam) a (uratam), Minervae 
 b(ovem) f(eminam) a (uratam). 14 Here, whether ante cellam 
 lunonis reginae implies in the portico, on the steps of the podium, 
 or in front of the steps, it is alike definitely stated that the sacri- 
 fice was not inside the cella. 
 
 "See Mau-Kelsey, Pompeii, pp. 49, 64, 131; Overbeck-Mau Pompeii, 
 pp. 69, 70, 115, 116; Huelsen-Carter Roman Forum, pp. 90, 149; H. B. 
 Walters The Art of the Romans, plate IX. For steps outside the podium 
 and merely attached to it compare Mau-Kelsey, pp. 86, 173; Overbeck-Mau, 
 96; etc. Can it be possible that the references recording a sacrifice in 
 templo refer to such offering at an altar on the steps and naturally included, 
 according to the stricter interpretation of Vitruvius, in the templum ? 
 The architecture of the Regia (for the immolation of victims in regia 
 see notes 11 and 12) and the fact that at Alatri the traces discovered of 
 an altar were in the pronaos (cf. n. 13), would be against such a view, 
 as is also the construction of those temples whose steps are outside the 
 podium and whose altars stand in front of the steps. 
 
 11 L. L. VI, 12. 
 
 a Macrob. I, 16, 30. 
 
 u Rom. Mitt. IV, 144; VI, 350. 
 
 "Act. Fr. Arv., p. CXCVII, a. 213. 
 
12 
 
 The very construction, however, of a Roman temple, standing 
 as it did on a lofty podium and reached by a flight of steps, con- 
 stitutes a natural obstacle to the immolation of larger animals 
 either in the cella or in the portico. Roman reliefs 15 and the 
 miniatures of the Codex Vaticanus 3225 of Vergil, a manu- 
 script in rustic capitals, 18 represent the victim as being slain, or 
 about to be slain, at an altar in front of the temple, and at 
 Pompeii there still exists a number of such altars, placed either 
 in front of the steps or on one of the lower rounds, 17 while altars 
 of this nature have not been discovered either in the portico or in 
 the cella. 
 
 But Ovid's passage does not discuss the slaying of the 
 victim. It merely says that the exta (viscera) were burned in 
 the aedes of great Jupiter. 18 Did the slaying, then, take place 
 outside and the burning of the exta within? In such of the 
 reliefs as portray the actual killing of the animal the act is not 
 being performed on the altar but beside it. True, the victim 
 here shown is a bull, but there is no reason to suppose that the 
 procedure was otherwise in the case of sheep or swine. 19 Fur- 
 thermore, the miniature in a Vatican MS. of Vergil which shows 
 an altar on one of the lowest steps of the podium, displays a 
 fire burning on the altar. 20 The altars outside cella and por- 
 
 " Relief in the Conservator! of Marcus Aurelius sacrificing before the 
 temple of Jupiter Optimus Maximus (see Brunn-Bruckmann, plate 269) ; 
 relief on an altar in front of the temple of Vespasian at Pompeii (see Mau- 
 Kelsey, p. 107); Clarac, Musee de Sculpture, plates 218, 310; reliefs on 
 the columns of Trajan and Marcus Aurelius (see Cichorius, plates 10, 
 38, 62, 63, 66, 72, 76, and Petersen-v. Domaszewski-Calderini, plates 13A, 
 38B). The relief on the silver cup from Boscoreale in the cabinet of 
 Baron Edmund Rothschild (see Monum. Piot. V, plates 34-36). 
 
 18 Published in Rome. Officina Danesi, Via di Bagni 1899, illustrations 
 13, 18, 22 (all three giving the altar at the foot of the steps), and 31 
 (giving the altar on the bottom rounds). 
 
 " See Mau-Kelsey, pp. 64, 83, 86, 108, 125, 166, 167, 171 ; Overbeck-Mau, 
 pp. 71, 95, 111, 115, 117; cf. also the altar on a lower round of the steps 
 of the temple in a miniature of a Vatican MS. of Vergil (see Daremberg 
 et Saglio, figure 409, p. 348 under arc). 
 
 "I, 587-8. 
 
 19 Cf. Arist. Peace, 1017 sqq., and the picture in Daremberg et Saglio 
 under ara, fig. 417, p. 349. 
 
 20 Daremberg et Saglio under ara, fig. 409, p. 348. Unfortunately they 
 fail to give the number of the MS. 
 
13 
 
 tico, therefore, cannot be explained on any theory that they were 
 used for the slaying of the victims. The exta must have been 
 burned upon them. Reisch, the author of the article on altars in 
 Pauly-Wissowa's Realencyclopadie, when noting that in the Greek 
 ritual there are but isolated examples of animal sacrifices in the 
 temples, makes the practical observation that it would have been 
 damaging to the cult images and the costly votive offerings, had 
 fat and flesh been regularly burned in the poorly ventilated cella. 21 
 The same criticism would apply to Roman worship. 
 
 The conclusion is, therefore, that the usual custom in animal 
 sacrifices was to slay the victim and burn its exta outside the cella 
 and portico *of the temple, and, if Ovid's words in lovis aede 
 refer to cella or portico, the ceremony of the Ides was either 
 different in this regard from the usual practise 28 or the poet has 
 not represented it correctly. Metrical reasons demand in aede 
 rather than ad aedem. Let us see now whether the poet has 
 informed us correctly. 28 
 
 I. 
 
 To what temple of Jupiter does Ovid refer when he says in 
 magni lovis aede? The thought at once arises of the great tem- 
 ple of Jupiter Optimus Maximus on the Capitoline, seat of the 
 most important cult of Jupiter in the Roman world, yet the epi- 
 thet magni, though applied to this divinity in Ex Ponto, IV, 9, 
 29-32, proves nothing, for the poet elsewhere confers it on Jupiter 
 as a mere epithet. 24 Now, as will presently be pointed out, the 
 
 * Under Altar, p. 1650. 
 
 33 But cf. in regia Varro L. L., VI, 12 and Macrob. I, 16, 30. 
 
 * In any case, whether Ovid's expression be true or untrue of this 
 particular ceremony, we have seen that instances do exist of the per- 
 formance of sacrifice in tentplo, on one occasion even in the temple of 
 Jupiter Optimus Maximus. Ovid's words, therefore, seem to me a note- 
 worthy bit of evidence in support of those references which reveal sacrifice 
 in aede on certain occasions. It is doubtful if even his carelessness and 
 poetic license would have misrepresented one of the most conspicuous and 
 widely recognized customs of Roman ritual, had he not been aware that 
 the rule allowed occasional exceptions. 
 
 24 Tristia III, I, 33-38, describing the house of Caesar on the Palatine: 
 "Singula dum miror, video fulgentibus armis 
 
 Conspicuos postes tectaque digna deo. 
 'Et lovis haec 'dixi 'domus est'? Quod ut esse putarem, 
 
 Augurium mend querna corona dabat. 
 Cuius ut accepi dominum, 'non fallimur', inquam, 
 'Et magni verum est hanc lovis esse domum'." 
 Fasti V, 40, of the Gigantomachia : "magnum bello sollicitare lovem." 
 
14 
 
 priest who performed the sacrifice on the Ides was the Flamen 
 Dialis, 25 and it seems to me that W. Warde Fowler 26 is justified 
 in his assumption that this priest had no special connection with 
 the temple of Optimus Maximus, which was under the charge 
 of an aedituus. Professor Fowler is arguing from the absence 
 of all evidence to connect the Flamen with this cult and from 
 his extreme antiquity in Latin religious history. 27 The two great 
 festivals of Jupiter at which he is known to have officiated, 
 namely, the auspicatio vindemiae and the sacra Idulia, were 
 celebrated to Jupiter the great Indo-European sky-god, in whose 
 control of the weather lay the possibility of a successful 
 vintage ; 28 to whom all Ides, the period of the full moon, were 
 sacred; 29 and who is hailed by the Salii in their hymn as 
 Lucetius, the giver or source of light. 30 The cult of Optimus 
 Maximus, on the other hand, did not arrive in Rome until 
 late in the history of the kingship, was of Etruscan origin, 
 and was distinctly political, a character which it always pre- 
 served. 31 Professor Fowler 32 seems inclined to reject the Flamen 
 Dialis as the officiating priest on the Ides, 33 and at any rate 
 continues to place the seat of this sacrifice on the Capitolium. 
 
 K Macrob I, 15, 16: "quam (ovem Idulem) hoc nomine vocant Tusci 
 et omnibus Idibus lovi immolatur a flamine"; Ov. Fasti I, 587-8, "castus 
 . . . sacerdos". 
 
 * R. E. R. P., p. 239 and n. 41. 
 
 * R. E. R. P., pp. 128 and 239. The origin of his office is ascribed to 
 Numa by Ennius ap. Varro, L. L., VII, 45 and Liv. I, 20, 1-2. 
 
 88 Fowler, R. F., pp. 85-8. 
 
 w Macrob. I, 15, 15 : "lure hie dies lovis fiducia vocatur, ctrius lux non 
 finitur cum solis occasu, sed splendorem diei et noctem continual inlu- 
 strante luna, quod semper in plenilunio, id est medio mense, fieri solet"; 
 18: "Ut autem Idus omnes lovi ita Kalendas lunoni tributas et Varronis 
 et pontificalis adfirmat auctoritas." 
 
 "Macrob. I, 15, 14: "Nam cum lovem accipiamus lucis auctorem, unde 
 et Lucetium Salii in carminibus canunt"; Wissowa, p. 114 and notes 2 
 and 3. Fowler, R. E. R. P., p. 129. 
 
 "Fowler, R. E. R. P., pp. 129, 237-9; Dionys. Ill, 69; IV, 59; Liv. I, 
 38, 7; Plut. Poplicola XIII; XIV. 
 
 "R. E. R. P., p. 239, n. 41 and n. 38 to p. 129. 
 
 * He cites the use of Sacerdotes by Festus, p. 290 ("eo itinere utantur 
 sacerdotes idulium sacrorum conficiendorum causa") and does not appear 
 to consider castus sacerdos (F. I. 587) a reference to the Flamen. Cf. 
 what he says (n. 38 to p. 129) on Hor. C. Ill, 30, 8: "dum Capitolium 
 scandet cum tacita virgine pontifex". There is no reason to suppose that 
 this last passage refers to the sacrifice on the Ides. 
 
15 
 
 As both Ovid 34 and Macrobius, 35 however, tell us that it was 
 offered to Jupiter, and Macrobius says a flamine, it appears 
 obviously impossible that any other flamen or priest of any sort 
 than that of Jupiter can be intended, especially as Ovid's epithet 
 castus sacerdos expressly fits the character of the Flamen 
 Dialis. I fully agree then with Professor Fowler in his hypo- 
 thesis that the Flamen Dialis, because of his extreme antiquity in 
 Latin religious history and his superintendence over the rites of 
 the primitive Indo-European sky-god, can have had no special 
 connection with Optintus Maximus, because this Jupiter was 
 political and a late importation from Etruria. It is thus impos- 
 sible for the sacra Idulia to have been offered both at this temple 
 and by this priest. I cannot, however, reject the priest, for the 
 passages above noted from Ovid and Macrobius are too con- 
 clusive in support of his connection with the sacra Idulia. I 
 must, therefore, reject the temple, for the words in magni lovis 
 aede are by no means a certain reference to the aedes of the 
 great political god, any more than to one of the other Jupiter 
 cults in Rome. 88 
 
 "Fasti I, 56; 587-S. 
 
 "I, IS, 16 (see n. 25 above). 
 
 M I cannot see how the Etruscan etymology which Macrobius gives for 
 the words "Ides" and ovis Idulis can be used as an argument against the 
 genuinely Latin origin of the Roman festival. Even if his etymology 
 be correct, there is no reason to doubt that the worship of the sky-god 
 on the day of the full moon can easily have antedated any name that 
 later crept into the calendar for that day. Macrob. I, 15, 14: Iduum 
 porro nomen a Tuscis, apud quos is dies Itis vocatur, sumptum est. 
 Item autem illi interpretantur lovis fiduciam, nam cum lovem accipiamus 
 lucis auctorem . . . iure hie dies lovis fiducia vocatur, cuius lux non 
 finitur cum solis occasu, sed splendorem diei et noctem continuat in- 
 lustrante luna ; . . . diem igitur qui vel nocturnis caret tenebris lovis 
 fiduciam Tusco nomine vocaverunt; unde et omnes Idus lovis ferias 
 observandas sanxit antiquitas"; and I, 15, 16: "quam (ovem Idulem) 
 hoc nomine vocant Tusci". A parallel rite may well have existed among 
 the Etruscans, but it seems very improbable that the word "Ides" was 
 derived from the Etruscan. Varro (L. L., VI, 28), says: "Idus ab eo 
 quod Tusci itus vel potius quod Sabini idus dicunt". Corssen (Uber die 
 Sprache der Etrusker, Vol. II, pp. 237-238), gives a number of corre- 
 sponding roots from the Indo-Germanic languages, all containing the idea 
 of brightness or heavenly light. Cf. Walde (Lateinisches etymologischts 
 Worterbuch, p. 375). 
 
II. 
 
 We must now consider two passages of vital importance in 
 determining the scene of the sacrifice. Festus (p. 290, Muller's 
 edition) and Varro (L. L., V, 47) both say that the procession 
 of priests passed along the Sacra Via in this festival. Festus' 
 words are : 
 
 "Sacram viam quidam appellatam esse existimant . . quod eo 
 itinere utantur sacerdotes idulium sacrorum conficiendorum 
 causa, itaque ne eatenus quidem, ut vulgus opinatur, sacra appel- 
 landa est a regia ad domum Regis sacrificuli, sed etiam a Regis 
 domo ad sacellum Streniae, et rusus a regia usque in arcem." 
 "Some persons believe that the Sacred Way is so called because 
 the priests use that street in performing the sacra Idulia. There- 
 fore, it must be named Sacra, not only, as is commonly supposed, 
 from the Regia to the house of the Rex Sacrificulus, but likewise 
 from the house of the Rex to the shrine of Strenia, and back 
 from the Regia all the way to the Arx." 
 
 The use of itaque, meaning "therefore," as well as the state- 
 ment which it introduces, show that Festus means that the pro- 
 cession passed along the entire length of the Sacra Via and that 
 thus the name should be applied to the whole street from the 
 sacellum of Strenia near the Colosseum and the Cannae, which 
 lay south of the modern Via Cavour along the Via dei Serpenti, 
 as far as the Arx. 37 Varro says: 
 
 "Carinae postea cerionia, quod hinc oritur caput sacrae viae 
 ab Streniae sacello quae pertinet in arce(m) qua sacra quotquot 
 mensibus feruntur in arcem et per quam augures ex arce profecti 
 solent inaugurare." 
 
 "Carinae afterwards cerionia (the spelling is corrupt), because 
 at this point the Sacred Way begins, from the shrine of Strenia, 
 
 " See Jordan I, 3, pp. 258-259, 262-263, and maps; Kiepert and Huelsen, 
 Forma Urbis Romae Antiquae, 2nd. ed. maps. 
 
 16 
 
17 
 
 and extends to the Arx. By the Sacred Way sacra are carried 
 every month to the Arx and along it the augurs, having started 
 from the Arx, are accustomed to take omens." 
 
 Against this last quotation it may be argued that the words 
 "sacra quotquot mensibus feruntur in arcem" are indefinite, and 
 might apply to some other sacra than the Idulia, but we know of 
 no other monthly festival which included a procession along the 
 Sacred Way to the Arx, and Festus' passage expressly states that 
 this procession took place on the Ides. The offering must then 
 have been made at some point along the Sacred Way, if not on the 
 Arx itself, on which the Sacrjed Way terminated. 
 
 III. 
 
 But Ovid's words are in lovis aede, and no temple of Jupiter 
 is known to have stood on the Arx. O. Gilbert (Geschichte und 
 Topographic der Stadt Rom, pp. 236-237) conjectured that the 
 aedes was that of Jupiter Stator on the summa sacra via* 8 and 
 near the arch of Titus. The same objections, however, apply to 
 this cult as to that of Optimus Maximus. Jupiter Stator was 
 not the sky-god but a military Jupiter, and his temple was vowed 
 in the third Samnite war and dedicated shortly afterward, long 
 after that of the great god of the Capitol. 89 The passages in the 
 first book of Livy (I, 12) and Ovid (Fasti VI, 793), which 
 represent this temple as vowed by Romulus, appear to be purely 
 literary tradition, for the promise is that of a templum, and 
 none could have been erected before 294 B. C. or the consul 
 would not have made the vow when his army was retreating 
 before the Samnites. Livy attempts to reconcile his later account 
 with the earlier by stating in Book X (37, 14-16), that Romulus 
 had vowed only a fanum* 1 but the contradiction none the less 
 
 88 Jordan I, 3, pp. 20-23. 
 
 "Liv. X, 36, 11; Wiss., p. 122, and notes 8 and 9. 
 
 40 Liv. I, 12: "Hie ego tibi templum Statori lovi, quod monumentum 
 sit posteris tua praesenti ope servatam urbem esse, voveo". 
 
 1 "In ea pugna lovis Statoris aedem votam, ut Romulus ante voverat ; 
 sed fanum tantum, id est locus templo effatus (sacratus), fuerat. Ceterum 
 hoc demum anno, ut aedem etiam fieri senatus iuberet, bis eiusdem voti 
 damnata re publica, in religionem venit". 
 
18 
 
 remains, and the application of the title Optimus Maximus to 
 Jupiter Stator in Livy's first passage 42 is still further proof of 
 the careless and purely literary character of Livy's account. 
 There is as good reason, therefore, to reject the Stator temple 
 as the scene of the sacra Idulia as that of Jupiter Optimus Maxi- 
 mus; for the Stator cult was still later, and for this reason, and 
 also because of its military character, was no more the cult of the 
 primitive sky-god than was that of Optimus Maximus, nor could 
 it have been in the hands of so ancient a religious personage as 
 the Flamen Dialis. 48 
 
 IV. 
 
 Now there was another religious edifice on the Sacra Via, 
 against which the objections above mentioned do not hold good. 
 The Regia 44 goes back to the hoary beginnings of a community 
 about the Forum valley on the adjacent hills, and was, as 
 the name implies, first the house of the king. 45 Tradition 
 describes it as the abode of Numa, or at least the place where 
 he discharged his priestly functions. 48 Later on it became the 
 central point of the activities of the college of pontifices and the 
 official headquarters of the Pontifex Maximus, upon whom most 
 of the religious functions of the king devolved after the fall of 
 
 "I, 12: "Haec precatus, velut si sensisset auditas preces, 'hinc', inquit, 
 'Romani, luppiter optimus maximus resistere atque iterare pugnam iubet." 
 
 43 The article on Jupiter in Daremberg et Saglio (p. 711) states that 
 the quindecemviri sacris faciundis had charge of the cult of Jupiter 
 Stator: "Le culte de Jupiter Stator etait celebre sous la surveillance des 
 Quindecemvirs". The value of the statement is unhappily weakened by 
 the absence of any reference to substantiate it. If it be so, however, it 
 furnishes weighty additional testimony against the Stator temple as the 
 scene of the sacra Idulia; for these priests were of later origin than the 
 other Roman sacerdotal colleges (see Wiss., p. 534), had charge of the 
 Sibylline Books (Wiss., p. 536), and probably a superintendence over 
 foreign rites whose importation these oracles had directed, and did not 
 include any flamen among their number (Wiss., pp. 542-3). 
 
 44 Jordan I, 2, pp. 302-4; Wiss., p. 502. 
 
 "Festus, p. 279: "Regia domus ubi rex habitat." Cf. Wiss., p. 502, and 
 Jordan I, 2, pp. 423-9. 
 "Solin. I, 21; Plut. Numa XIV; Tac. Ann. XV, 41. 
 
19 
 
 the monarchy. 47 Here at the appropriate festivals offerings were 
 made to certain of the oldest Roman divinities, 48 and here on all 
 Nundinae or market days, says Granius Licinianus, 49 the Flam- 
 inica Dialis sacrificed a ram to Jupiter himself. This sacrifice 
 by the wife of the Flamen Dialis on the market-days is comparable 
 to that by her husband on the Ides ; and the Regia alone, therefore, 
 of any sacred edifice yet known along the Sacra Via or at its 
 terminus can be conceived of as the scene of the sacra Idulia. 
 
 But Varro says that the sacra of the Ides were carried to the 
 Arx. 50 His words are "sacra quotquot mensibus feruntur in 
 arcem." This statement fronvsuch an authority as Varro would 
 at once settle the question of the scene of the sacrifice, were it not 
 for Ovid's words in magni lovis aede, for, as mentioned above, 
 no aedes of Jupiter is known to have stood on the Arx. Ovid, 
 however, does not speak of the immolatio, but only mentions the 
 burning of the exta. 
 
 Can it be, then, that the victim was led in procession along the 
 Sacra Via to the Arx and there slain, but its exta afterwards car- 
 ried back to the Regia and offered in a chapel of Jupiter in that 
 building? We know of two parallels which might justify such a 
 theory. The first is that of the dog and sheep of the Robigalia, 
 which were slain in Rome in the morning and whose exta were 
 later offered to Robigus in the grove of that deity at the fifth 
 milestone on the Via Claudia, whither they had been borne in 
 solemn procession. 51 The second example is connected with the 
 Regia itself : the tail of the horse sacrificed to Mars in the Campus 
 Martius on the Ides of October was carried with all possible speed 
 to the Regia and the blood allowed to drip on its hearth. 52 Mars 
 indeed appears to have held a very prominent position in the cults 
 
 4f Wiss., p. 502; Huelsen-Carter, The Roman Forum, p. 193. 
 
 48 To Ops Consiva in her own chapel in the Regia (CIL. I 1 , p. 327; 
 Varro, L. L. VI, 21); to Juno on the Kalends by the Regina sacrorum 
 (Macrob. I,, 15, 19) ; to Janus on the Agonium by the Rex sacrorum 
 (Varro, L. L. VI, 12) ; etc. 
 
 *Ap. Macrob. I, 16, 30: "Flaminica omnibus nundinis in regia lovi 
 arietem soleat immolare". 
 
 M L. L. V.47 (see above). 
 
 61 Ov. F. IV, 907-36; Fasti Praenestini on Apr. 25 (CIL. P, p. 316): 
 Fowler, R. F., p. 89. 
 
 " Festus, p. 178; Fowler, R. F., pp. 241-2. 
 
20 
 
 of the Regia. It was here that his sacred spears were kept, 53 and 
 Jordan even goes so far as to assume that he was the protecting 
 deity of the place. 54 However this may be, Ovid's expression 
 in magni lovis aede is a curious term in view of the nature of 
 the building and the variety of divinities there worshipped, and 
 perhaps we have already sufficient reason for concluding that the 
 Regia had nothing whatever to do with the sacra Idulia. But 
 before convicting Ovid of making so gross an error as saying that 
 the offering was burned at an aedes, if the entire ceremony really 
 took place out-of-doors, I should like to present an elaborate mass 
 of testimony based on two lines of the Fasti themselves. In II, 
 69-70, Ovid is discussing sacrifices made on February 1 at three 
 places. The lines are : 
 
 "Ad penetrale Numae Capitolinumque Tonantem 
 
 Inque lovis summa caeditur arce bidens." 
 I shall first enter into a rather lengthy discussion to prove 
 penetrale Numae equal to "Regia." 
 
 V. 
 
 H. Peter in his edition of the Fasti* 5 understands penetrale 
 Numae as a reference to the Atrium Vestae in its restricted 
 sense, the building utilized as the dwelling of the Vestals ; for he 
 interprets Fasti VI, 263-4, as meaning that Numa lived in the 
 Dvmus Vestalium and not in the Regia. The words are: 
 "Hie locus exiguus qui sustinet atria Vestae 
 Tune erat intonsi regia magna Numae." 
 
 Regia he takes in the general sense of "palace." 
 
 There is reason to believe, however, that the Romans employed 
 the terms Atrium Vestae and Regia interchangeably, and even that 
 the two were at the outset parts of one and the same building. 
 Let us examine the evidence set forth by Dr. Esther Boise Van 
 Deman in her monograph, The Atrium Vestae. 
 
 The architectural history of the Atrium of the Republic 
 "extends from the early republican or even the regal period to that 
 
 "Cell. IV, 6; Wiss., p. 502 and n. 5. 
 
 -1,2, pp. 424-5. 
 
 " Notes on II, 69 and VI, 257 sqq. ; also critical note on VI, 263 sqq. 
 
21 
 
 of the early empire." This earlier Atrium was wholly destroyed 
 in the fire of Nero, to whom the first rebuilding dating entirely 
 from the imperial period is due. True, "many of the walls were 
 restored more than once before their final destruction and new 
 walls were added," but the level, orientation, and essential features 
 of the republican Atrium remained the same down to Nero's day. 8 * 
 The centre of this republican structure lying below the level oi 
 Nero was a square court corresponding to the atrium of a private 
 house. Portions of the walls of this court have been preserved. 
 From it the term atrium spread to the entire building, and within 
 it stood the aedes Vestae, just as the shrine of the goddess had its 
 place in every private atrium. 57 According to a recent article by 
 Dr. Van Deman, Methods of Determining the Date of Roman 
 Concrete Monuments (American Journal of Archaeology, Vol. 
 XVI, p. 393), the extant foundations and podium of the 
 aedes Vestae not only antedate Nero but fall in the Augustan 
 period between 14 and 12 B. C. Both these and the walls of 
 the central court thus appear to be the very masonry existing in 
 Ovid's day, and a glance at plan A of her Atrium Vestae shows us 
 the aedes half surrounded by the court. In the earliest period, 
 then, we may rest assured, the aedes was completely surrounded 
 by the court, and stood probably in its northwest corner. That 
 the Domus publica, or residence of the Pontifex Maximus, was 
 structurally a part of the Atrium, we are informed by Dio Cassjus, 
 (LJV, 27), who, in relating how Augustus when Pontifex re- 
 moved to the Palatine, says : 
 
 7Yjv jjLsvTOt TOU fettrtXeu; i(ov lep&v (sc. oouov) 
 Tat? as i Kaptisvotc ICDKSV eirei'Stj OJAOTO^OI; -cat? 
 oixiQfftffiv auicov r]v, 
 
 "but he gave the house of the Rex sacrorum to the Vestal Vir- 
 gins, since it was joined to their dwelling by a party-wall." 
 
 Dio has confused the names Rex sacrorum and Pontifex Maxi- 
 mus, but this does not impair the value of his testimony that the 
 house was ouoio'.yoc . 58 Dr. Van Deman believes that certain 
 remains immediately to the east of the republican court belong to 
 the Domus publica. At any rate the passage in Dio is conclusive, 
 and since we have seen that even in Ovid's own time the Atrium 
 
 " The Atrium Vestae, pp. 4-5, 9-10, 15. 
 " The Atrium Vestae, pp. 4-5, 9-10. 
 "Wiss., pp. 502-503, n. 7. 
 w The Atrium Vestae, p. 13. 
 
22 
 
 Vestae proper, or Domus Vestalium, and the Domus publica were 
 parts of one building, while the aedes Vestae virtually made up a 
 third part of the same, what is more natural than to infer that the 
 fourth factor in the group, the Regia, was also in the beginning 
 structurally united with the rest ? Such is the opinion of Dr. Van 
 Deman, 80 who states in a foot-note that the road now separating 
 the temple from the later Regia is not original. 81 She continues : 
 
 "With the gradual breaking up of the simple cult of which the 
 king's house had been the centre, and the growing independence 
 of the various priesthoods among which the several religious func- 
 tions of the king had been divided, the necessity arose for the 
 assignment to them of distinct official residences. At this time it 
 is probable that the parts of the Atrium became independent ; for 
 during the later Republic and the early Empire, in place of a single 
 complex structure bearing one name, there were recognized four 
 separate parts with as many distinct names, two of which were, 
 however, those applied earlier to the whole structure." 62 
 
 This indiscriminate application of the terms Regia and Atrium 
 Vestae seems to be attested by several passages. First of all Ljvy 
 (XXVI, 27, 3) says in describing a conflagration in the Forum: 
 
 "Conprehensae lautumiae forumque piscatorium et atrium 
 regium. Aedis Vestae vix defensa est," and again (XXVII, 11, 
 16): 
 
 "Locaverunt inde reficienda quae circa forum incendio con- 
 sumpta erant, septem tabernas, macellum, atrium regium!' 
 
 This blending of the two names is curious. Cicero, writing to 
 Atticus 63 of the latter's interview with Caesar, then Pontifex 
 Maximus, says : "Visum te aiunt in regia." The reference is evi- 
 dently to Atticus' performance of the morning salutatio, which 
 was made at the residence, not office, of the person receiving the 
 call. In Caesar's case this would be at the Domus publica, which 
 Cicero here designates under the general term Regia for the 
 whole building. A third passage is that of Ovid above quoted, 
 which Peter construes so differently: 
 
 ** The Atrium Vestae, p. 9. 
 " The Atrium Vestae, p. 9, n. 2. 
 * The Atrium Vestae, p. 11. 
 *Ad. Att. X, 3a. 
 
23 
 
 "Hie locus exiguus qui sustinet atria Vestae 
 
 Tune erat intonsi regia magna Numae." 64 
 
 Dr. Van Deman 65 cites this as a particularly illuminating exam- 
 ple of the identity of the two names. 
 
 If this is so, perhaps Tristia III, 1, 29-30, is to be understood in 
 the same way : 
 
 "Hie locus est Vestae, qui Pallada servat et ignem, 
 Haec fuit antiqui regia parva Numae." 
 
 "This place is Vesta's. . . This is the little Regia which once be- 
 longed to ancient Numa." 
 
 But, if regia and locus, that is Atrium, are not here used 
 interchangeably, then the only remaining conclusion is that regia 
 is employed in its restricted sense and refers to the building proper 
 of that name. In that case, even if Ovid did mean by Fasti VI, 
 262-4, that the Domus Vestalium was Numa's palace, as Peter 
 supposes, yet the passage from the Tristia would contradict the 
 poet, and commit him by his own statement to the current view 
 supported by Solinus and Plutarch, which gives that honor to 
 Regia." 
 
 We have seen that the Regia and the Atrium were probably in 
 the beginning structurally united ; or, at least, that the Romans of 
 the classical period appear to make use of either name indiffer- 
 ently; but even were this view incorrect, regia in Tristia III, 
 1, 29-30, cannot refer to any one part of the temenos of Vesta 
 except the edifice to which the term "Regia" was technically 
 applied. Therefore, since Ovid, like Solinus and Plutarch,* 7 
 couples the Regia with Numa's name, penetrate Numae in 
 Fasti II, 69, should be taken as a reference to that building and 
 not to the Domus Vestalium. 
 
 M F, VI, 263-4. 
 
 M The Atrium Vestae, p. 10, n. 4. 
 
 M Solin. I, 21 : "propter aedem Vestae in Regia, quae adhuc, ita 
 appellatur" ; Plut. Numa XI V : cefoifiaTo jtXtioiov TOU Tfjg 'Eotiag Uoov 
 TT)V xaXov|ivr)v TTJYUXV olovn, 6acaXEU>v olxima. 
 
 " Cf. Tac. Ann. XV, 41, "Numaequc Regia et delubrura Vestae," or 
 are these words a further illustration of the interchanging of the two 
 terms? 
 
VI. 
 
 But why this elaborate digression to prove penetrate Numae 
 identical with "Regia"? The lines are: 
 
 "Ad penetrale Numae Capitolinumque Tonantem 
 Inque lovis summa caeditur arce bidens," 68 
 
 that is, on the first of February, the date assigned in the poem, a 
 bidens was slain in sacrifice at each of three places, the Regia, 
 the temple of Jupiter Tonans on the Capitolium, and in arce. 
 Let us first determine the meaning of arx in this passage. 
 
 Correct technical usage would confine the application of the 
 word arx in Roman topography to the fortified northern spur of 
 the Capitoline Hill. Both poets and historians, however, when 
 writing of the Capitoline, frequently employ it more loosely. H. 
 Peter in his notes on Fasti II, 70, is of the opinion that lovis 
 summa arce" means the temple of Jupiter Optimus Maximus. 69 
 I can find but one parallel for this interpretation. It will be noted 
 in a moment. Arx is sometimes employed in reference to a build- 
 ing, but only where the idea of a stronghold or castle is more 
 plainly implied than in the present instance. 70 . It is really the 
 Latin equivalent of oxporcoXsq and in instances where it is used 
 of an actual building the derived meaning is almost invariably 
 dependent on the original thought of the lofty and easily defended 
 situation which the structure occupies. The parallel to Peter's 
 interpretation of Fasti, II, 70, is arce lovis in Fasti VI, 18, 
 which, because of the connection with 33-34, I take with Peter 
 
 " P. II, 69-70. 
 
 69 See especially critical note in his edition. 
 
 70 Liv. II, 7, 6, "alto . . loco arcem inexpugnabile fore" of the house of 
 Poplicola on the Velia; Tac. Ann. XIV, 31, "templum divo Claudio con- 
 stitutum quasi arx aeternae dominationis" ; Agric. 45: "intra Albanam 
 arcem sententia Messalini strepebat," of the imperial villa (cf. Octavia, 
 "receptus arce") ; Hor. S. II, 6, 16: "ubi me in montes et in arcem ex 
 urbe rcmovi"; etc. 
 
 24 
 
25 
 
 to signify the Optimus Maximus temple, 71 though the Thesaurus 
 understands it of Jupiter's heavenly abode as in such passages 
 as Fasti I, 85, "luppiter arce sua totum cum spectat in orbem." 
 The expression arx lovis or its equivalent, arx Tonantis, where 
 Tonantis is a mere epithet equal to lovis 72 occurs in other pas- 
 sages of Ovid. In Fasti VI, 349-350, we have : 
 
 "Nomine quam pretio celebratior arce Tonantis, 
 
 Dicam, Pistoris quid velit ara lovis." 
 
 For these lines there is absolutely no proof that arce Tonantis 
 means either the temple or 4he portion of the hill to which Capi- 
 tolium was applied rather than the Arx in its true sense of the 
 northern spur of the Capitoline Hill, since nothing is known of 
 this altar to Jupitor Pistor. 73 Arce lovis in Fasti IV, 635, the 
 scene of part of the sacrifice of the Fordicidia, is equally indefi- 
 nite, and can as easily refer to the true Arx. Tibullus (II, 5, 25) 
 furnishes a further example of the expression : 
 
 "Sed tune pascebant herbosa Palatia vaccae, 
 Et stabant humiles in lovis arce casae," 
 
 where there is no thought of the temple, or apparently of either 
 summit in particular, but the idea is merely that of the Capitoline 
 Hill in a general sense as compared with the Palatine. -But Peter 
 in his note on Fasti II, 69-70 74 cites Livy (XXVIII, 39, 15) 
 as evidence of the use of arx in the sense of Capitolium. The 
 words "lovi Optimo Maximo praesidi Capitolinae arcis" do not 
 compel this narrow interpretation, however. They are indefinite 
 and appear to me to be employed rather of the hill as a whole. 
 
 71 F. VI, 17-18: 
 
 "Ex illis fuit una, sui germana mariti, 
 
 Haec erat (agnovi), quae stat in arce lovis". 
 F. VI, 33-34: 
 "Si torus in pretio est, dicor matrona Tonantis, 
 
 lunctaque Tarpeio sunt mea templa lovi". 
 
 "Fasti VI, 33-4; 349; IV, 585; Heroid. IX, 7; Metam. II, 466; 
 Hor. Epod. II, 29; Statius Silvae IV, 4, 58; Achil. I, 1-2; Martial V, 
 55; V, 72; VII, 60, 1-2; etc. 
 
 73 Cf. Jordan I, 2, p. 50: "Unsicher bleibt der Standort des Altars des 
 Juppiter Pistor", and n. 51. "Das die von Ovid, Lactantius (I, 20, 33), 
 erwahnte ara lovis Pistoris auf dem Capitol stand . . ist unerweislich". 
 Preller (I, p. 194), while placing it on the Capitol says (n. 5) : "Das die 
 ara lovis Pistoris auf dem Capitol gestanden ist nicht bezeugt und folgt 
 nicht nothwendig aus Ovid". 
 7i See n. 69 above. 
 
26 
 
 If, however, they be taken in a restricted sense at all, Capitol- 
 inae arcis points naturally to the actual citadel, especially with 
 the word "defender," praesidi, on which the genitive depends. 
 Likewise the same expression in Livy VI, 20, 9, because of the 
 co-ordinate use of Capitolium atque arcem in the same passage, 
 can refer only to the hill in general: 
 
 "Identidem Capitolium spectans lovem deosque alios devocasse 
 ad auxilium fortunarum suarum precatusque esse, ut quam 
 mentem sibi Capitolinam arcem protegenti ad salutem populi 
 Romani dedissent, earn populo Romano in suo discrimine darent ; 
 et orasse singulos universosque, ut Capitolium atque arcem 
 intuentes, ut ad decs inmortales versi de se iudicarent." 
 
 The same is true of the passage in Tacitus Hist. Ill, 71, too 
 long to quote. Any siege or storming of the Capitoline Hill would 
 naturally involve both spurs, so Capitolina arx is likely to be noth- 
 ing more than a general term for the entire mount. In Livy's 
 account of Tarpeia's treason and the battle over the Sabine women, 
 arx certainly means the entire hill, unless it refer to the true 
 Arx alone, for we even find the words "quod inter Palatinum 
 Capitolinumque collem campi" (I, 11-12), the distinction being, as 
 in Tibullus II, 5, between the Palatine and Capitoline Hills. Ovid 
 himself affords one of the most striking examples of the use of 
 the word meaning the hill in general, where Mars in addressing 
 an assemblage of the gods speaks of their abodes in arce, that is, 
 their temples, which lay on both spurs (Fasti VI, 367). Compare 
 with this Livy (VI, 16, 2), "luppiter Optime Maxime lunoque 
 regina ac Minerva ceterique dii deaeque qui Capitolium arcemque 
 incolitis." One other passage in Ovid (Tristia IV, 2, 55-56), 
 besides that already mentioned (Fasti VI, 18), may be a reference 
 to the Capitolium alone, though it would appear to mean the hill 
 and not the building. The lines are descriptive of the offering 
 at a triumph, and are as follows : 
 
 "Inde petes arcem et delubra f aventia votis ; 
 Et dabitur merito laurea vota lovi." 
 
 It is Jupiter Optimus Maximus, of course, whose temple is re- 
 ferred to in delubra. Even in these lines, however, I do not see 
 why arcem cannot be applied vaguely to the Capitoline as a 
 whole, though the thought of the southern spur is certainly more 
 prominent. Now there are two other passages in the Fasti in 
 
27 
 
 which the same adjective modifies arx as in II, 70. The first 
 of these (I, 257-264) treats of Tarpeia's treason with the words 
 (261-262), "Sabinos ad summae tacitos duxerit arcis iter," which 
 seem to me certainly to refer, like Livy I, 11-12, if not to the true 
 Arx, at least to the Hill in general. That summon is here parti- 
 tive, meaning "the top of the Arx," is shown by comparing it with 
 Livy I, 22, which speaks of the battle between the Romans and 
 the Sabines as taking place in infima arce, "at the bottom of the 
 Arx," as described in I, 12. (Summae arcis cannot, then, here 
 refer to the temple. In "the second of these two Ovidian pas- 
 sages arx is used unmistakably of the northern spur of the Hill, 
 the words being (VI, 183), "arce in summa lunoni templa Mone- 
 tae facta," for the temple of Juno Moneta is well known to have 
 stood on the Arx. In view of these parallels I think there can be 
 no doubt whatever that in lovis summa arce of II, 70, cannot 
 refer to the temple, but rather means the top of the hill. 
 
 Since in Fasti II, 69-70, we have sacrifices at three places, the 
 Regia (penetrate Numae), the temple of Jupiter Tonans, which 
 stood on the summit of the Capitolium near the temple of Optimus 
 Maximus, and thirdly, on top of the arx of Jupiter, lovis summa 
 arce cannot here refer to the hill as a whole or to the Capitolium 
 proper, for the Capitolium has already been mentioned in the 
 second place. Therefore the last place cannot well be other than 
 the true Arx. The Arx was, like the Capitolium, sacred to Jupiter, 
 for it was here that the augurs performed their function of inter- 
 preting his will 75 and of sacrificing with secret rites; 76 and the 
 augurs were interpretes lovis Optimi Maximi. 77 The sacra Non- 
 
 78 Liv. I, 18, 6-1Q; Varro L. L. V, 47. 
 
 "Paul., p. 16 (under Arcani), "sive a genere sacrificii, quod in arce fit 
 ab auguribus, adeo remotum a notitia vulgari, ut ne litteris quidem man- 
 detur, sed per memoriam successorum celebretur". 
 
 " Cic. de Leg. II, 20. The title Optimus Maximus in this connection 
 may seem to associate that cult with the Arx. All that is probably meant, 
 however, is that with the advent of the cult of Optimus Maximus as the 
 supreme deity of the state, the augurs, whose science was consulted before 
 commencing any act of political importance, became associated with this 
 greatest of the Jupiters, but continued to function on the same hill-top 
 where they had always interpreted the signs of the weather-god. Varro 
 (L. L. V, 52), tells of another auguraculum on the Quirinal, doubtless 
 overshadowed after the establishing of the Etruscan dynasty, by that on 
 the Arx. 
 
28 
 
 alia were held on the Arx, 78 and we know furthermore that the 
 worship of Jupiter on hill-tops was a primitive custom in Rome, 
 as in the rest of Italy. 79 Moreover, this Jupiter, who was worship- 
 ped out of doors on hill-tops in the primitive religion and whose 
 will was interpreted by the augurs, was the great Aryan weather- 
 god, and the offering on the Arx was unquestionably to him. 
 Jupiter Tonans too was a weather-god. Therefore, since we have 
 sacrifices on February 1 on both Arx and Capitolium to Jupiter in 
 his aspect of a weather-deity, the sacrifice in the Regia on that day 
 must have been to the weather- Jupiter, just as was the ram 
 offered there by the Flaminica Dialis on the Nundinae. 
 
 We have seen, then, that on February 1 a bidens was immo- 
 lated to the Jupiter of the weather at the Regia, at the Tonans 
 Temple, and on the Arx. This implies that the entire ceremony 
 took place at each of the three places, and the victim could not 
 have been slain on the Arx and its exta brought down to the 
 Regia to be burned. There is absolutely no necessity, therefore, 
 for reading any such action into the ceremony of the sacra Idulia, 
 and the conclusion must be that the Regia played no part in the 
 festival of the Ides, but that when Varro says "sacra quotquot 
 mensibus feruntur in arcem" (Z,. Z,. V, 47, see above) he means 
 that the sacra Idulia were carried to the Arx, and that there the 
 victim was slain and its exta burned. 
 
 VII. 
 
 How, then, could Ovid have made such a mistake as to say that 
 the exta was given to the flames in magni loins aede, when no 
 aedes of Jupiter is known to have stood on the Arx? The answer 
 is not hard to find. It is extremely probable that Ovid, however 
 often he might have seen the procession pass along the Sacra Via, 
 never witnessed the sacrifice, for the celebration was a survival 
 from the oldest Roman religious history, and, we may infer, was 
 attended solely by the priests. The leading sources, 80 however, 
 
 "Varro L. L. VI, 28. 
 
 79 Wiss., p. 116 and n. 5; Fowler, R. P., pp. 228-9; R. E. R. P., p. 129 
 and n. 41. 
 
 80 1 shall deal with this question more fully in a subsequent article. 
 
29 
 
 which Ovid consulted on questions of Roman religion in his Fasti 
 were Varro's writings 81 and a lost book of Fasti by Verrius Flac- 
 cus, of which fragments have been preserved in its epitome, the 
 Fasti Praencstini, and in such passages of Festus' and Paulus' 
 abridgments of the de Verborum Significatu as show their deriva- 
 tion from a calendar of festivals. 82 The de Verborum Significatu 
 itself was not published in time to be a source for Ovid, whose 
 poem, except for the few passages revised at Tomi, must have 
 been written between the publication of Verrius' Fasti, 4-6 A. D., 
 
 81 See Peter's Edition, p. 16 ; Schanz, Geschichte der romischen Litteratur, 
 2 I, p. 314; Franke, De Ovidii Fastorum Fontibus Capita Tria (diss. 
 Halle, 1909), pp. 51-52. Merkel, critical notes to his edition. 
 
 " Schanz 2 I, p. 313 : "Fur Ovid war ein Handbuch notwendig, in dem 
 nach dem Kalender die Feste atiologisch behandelt waren; denn die 
 Annahme, dass sich Ovid erst den Stoff zusammensuchte, ist ganz unwahr- 
 scheinlich. Wie durch Combination geschlossen wird, verfasste der 
 beriihmte Grammatiker, Verrius Flaccus, ein solches Handbuch, und aus 
 diesem gelehrten Werk ist, wie Mommsen (CIL. P, p. 285), (p. 313, 
 p. 314), zu erweisen versucht hat, der praenestinischen Steinkalender, den 
 Sueton mit Verrius Flaccus in Beziehung bringt, nur ein Auszug. Dieses 
 Handbuch benutzte aber Verrius Flaccus auch in seiner Schrift, de 
 verborum significatu; cf. Winther, De Fastis Verrii Placet ab Ovidio 
 adhibitis, p. 42; Franke, p. 32. Aus der Uebereinstimmung zwischen 
 Ovid und Verrius Flaccus schloss Winther dass das kalendarische Hand- 
 buch des Verrius Flaccus, soweit die romischen Sagen in Betracht kamen, 
 die einzige Quelle sei. Gegen diese Hypothese nahmen Stellung H. Peter, 
 Ausg, I 4 p. 16 Anm. 2; H. Willers, p. 39; M. Rabenhorst, p. 70; H. 
 Willemsen, p. 32; P. Wessner, Berl philol. Wochenschrift 1910 Sp. 680. 
 Unrichtig ist in der Wintherschen Hypothese das Einquellenprinzip 
 outriert worden; treffend Wissowa, p. 271, lusto plura huic fonti tribuit 
 nimiusque fuit in aliis auctoribus excludendis; qui ut Varronem nisi 
 intercedente Verrio ab Ovidio adhibitum esse injuria negavit.' Aber 
 wenn wir einerseits das Einquellenprinzip verwerfen, so muss anderer- 
 seits doch daran festgehalten werden, dass die Grundlage der ovidischen 
 Dichtung ein Handbuch bildete und dass dieses hochst wahrscheinlich 
 von Verrius Flaccus herruhrte". See particularly the able treatment of 
 this subject by Karl Franke, De Ovidii Fastorum Fontibus Capita Tria, 
 Halle, 1909, too long to be quoted here, in which he points out that a 
 passage in Festus like that on the Quinquatrus, p. 254, where the words 
 "Minervae autem dicatum eum diem existimant, quod eo die aedis eius 
 in Aventino consecrata est", are appended to an explanation of the word 
 Quinquatrus, adds nothing to the explanation of the word and betrays 
 itself as having been transferred from a calendar, pp. 29-32. 
 
30 
 and his own banishment in the year 9." As the passage from 
 
 "If Festus, p. 347a2S, "ubi mine est aedis Concordiae inter Capitolium 
 ct Forum", is copied from Verrius without interpolation and refers to 
 the temple after its restoration in A. D. 10, or one year after Ovid's 
 banishment, we have here sufficient evidence that the de Verborum Signifi- 
 catu was published too late for Ovid's use. R. Merkel, in the Pro- 
 legomena to his edition of the Fasti, p. XCV, not only uses this refer- 
 ence as an argument, but makes capital of the fact that Festus seven 
 times draws on Ateius Capito in matters of pontifical law, who was 
 probably born about the year 34 B. C. and died in 22 A D.; (see Tac. 
 Ann., Ill, 75, Frontinus Aq. 102. Teuffel and Schwabe Hist, of Roman Lit- 
 erature, Warr's translation, I, 265, and Schanz 2 I, p. 532), for, if Verrius 
 drew on him, how could the de Verborum Significatu have been written 
 until Ateius had become old and mature enough to publish a learned 
 work? Ateius wrote two hundred and sixty-one books on human and 
 divine law, and if he followed Varro's order of writing those on human 
 law first, it would place the date of his work on pontifical law so much 
 later. Suetonius (de Grammaticis et Rhetoribus 17) says of Verrius Flac- 
 cus : "Decessit aetatis exactae sub Tiberio". Schanz 2 I, p. 506, says : "Wahr- 
 scheinlich fallt das Werk (de Verborum Significatu) in die Regierungszeit 
 des Tiberius". "When Ovid began the Fasti cannot be definitely deter- 
 mined," says Schanz, 2 I, p. 308, who adds: "at any rate after the love 
 poems were finished. The 4th book falls in the period following the 
 conflagration on the Palatine because of which a restoration of the tem- 
 ple of Magna Mater by Augustus became necessary", in the year 3 A. D. 
 (See Fasti IV, 348). "The nature of the work prevented its being com- 
 posed in a continuous chain". Peter, in the introduction to his edition, 
 pp. 10-11, says that at any rate the composition fell some time after 
 Augustus' restoration of the Julian Kalender in 8 B. C., and Book IV 
 after 3 A D. In a footnote, however (n. 4), he asserts that the question 
 has in the main been solved by Merkel's industry and acumen. Merkel 
 believes that Ovid began his work in the year 755 A U. C. (p. X C I V 
 and CCLV). He was banished in 9 A. D., and seems to have done 
 little with the poem after that date, beyond revising Book I. It is, there- 
 fore, highly improbable that the de Verborum Significatu was written 
 before his banishment. On the other hand Franke, pp. 4-5, shows con- 
 vincing evidence that Verrius' Fasti were written between A. U. C. 757 
 and 759. As Ovid undoubtedly used these as a source (see n. 82, quota- 
 tion from Schanz) and we have seen above that for another reason 
 Ovid's 4th book must have been written after 3 A. D. or 756, surely it 
 seems safe to conjecture that his poem, except its revised portions, was 
 composed in the four or five years between the completion of Verrius' 
 Fasti and his own banishment. I am indebted to Professor Kirby Flower 
 Smith for the suggestion that the publication of Verrius' book would 
 naturally stimulate Ovid to a similar essay in the field of poetry. 
 
31 
 
 Festus above quoted, 84 relating to the Arx and the sacra Idulia, 
 is an explanation of the name Sacra Via and the mention .of the 
 day and festival is incomplete and subordinate, it cannot well have 
 been transferred from Verrius' work on the calendar; but 
 although notes in the Praenestini pertaining to the Ides have been 
 lost, the poet must either have read a note in Verrius' book of 
 Fasti to the effect that on the Ides a sheep was offered to Jupiter 
 on the Arx, or else he derived his information from Varro and 
 read perhaps something to this effect : Idibus ftamen Dialis in arce 
 lovi vervecem immolat. At any rate, being a poet and story- 
 teller rather than a scholar, he seems undoubtedly to have 
 bestowed but a superficial reading on his learned authority, and 
 seeing the word arx, to have understood it in the loose, general 
 application to the entire hill, so often given it, 85 the thought of the 
 Capitolium and temple of Jupiter Optimus Maximus, however, 
 being more prominent in his mind, as in his own lines from 
 Tristia (IV, 2, 55-56), already quoted. 86 He has thus made the 
 mistake of assigning to the great political Jupiter of the Capitoline 
 temple a sacrifice made out of doors on the Arx to the primitive 
 weather-god, and has even gone so far as to record the offering 
 as taking place in aede, when we know that sacrifices in the 
 cella or the portico were extremely rare and that the usual 
 custom in temple cults was to burn the exta on an altar on or 
 before the steps of the podium. 
 
 84 p. 290 (Muller's edition). 
 
 M See above: Tibullus II, 5, 25; Ov. F. VI, 367; Liv. I, 12, 22; XXVIII, 
 39, 15. 
 " See above, and cf. also n. 71. 
 
THE SEX OF THE OVIS IDUUS. 
 
 My first chapter dealt with the question of the place of the 
 sacrifice, the conclusion being that, through a misunderstanding 
 of his source, Ovid has fallen into a statement that is decid- 
 edly incorrect. The present^task is to examine his two remarks 
 concerning the sex of the victim and their credibility in the light 
 of what is further known about this sacrifice and about Roman 
 sacrificial ritual as a whole. Ovid's two passages are Fasti 1 1, 56 : 
 
 "Idibus alba lovi grandior agna cadit", 
 and I, 587-588: 
 
 "Idibus in magni castus lovis aede sacerdos 
 Semimaris flammis viscera libat -ovis," 
 
 which are apparently at variance in that agna would seem to 
 represent the female, while semimas ovis is unmistakably not the 
 female but the mrvex. Is this inconsistency real or only on the 
 surface ? 
 
 Before a closer scrutiny of the problems in this individual 
 ceremony, it will be well to review the general prescriptions of 
 Roman pontifical law in their bearing on the victim of the Ides, 
 confining attention to the ritual of the Roman State, to the 
 exclusion of sacra privata and Graecus ritus, whose practices 
 were materially different from the purely Roman customs sur- 
 viving from Rome's prehistoric beginnings. 2 
 
 Our trustworthy sources for this study are inscriptions and 
 the works of Roman writers on pontifical law, in so far as frag- 
 ments of those works have survived to us, and, to a greater or 
 less degree, other Greek and Roman scholars and antiquarians. 
 
 1 Hereafter to be cited as F. 
 
 2 Wiss., p. 420, and n. 3 to p. 420. C. Krause, De Romanorum Hostiis 
 Quaestiones Selectae. Krause's thesis, though containing serious errors, 
 is the most convenient handbook for a study of Roman sacrificial victims, 
 and is frequently referred to by Wissowa. 
 
 33 
 
34 
 
 The poets are unreliable on questions of ritual, 3 unless, as 
 appears to be true in the case of the sacra Idulia, they are corro- 
 borated by scholarly prose writers, 4 but they often afford val- 
 uable testimony where the proof turns partly on questions of 
 language or of style. This is especially true of Ovid himself. 5 
 
 As is well known, Roman pontifical law laid down the rules 
 of ritualistic observance with an almost painful exactness and 
 devotion to detail. 6 One of its fundamental principles was that 
 male animals should be offered to male deities, female to female. 7 " 
 
 There are three striking exceptions 8 to this rule in the prose 
 
 8 Krause, pp. 22-23 and n. 1. 
 
 4 The problem of the present chapter is that Ovid F. I, 588 calls the 
 victim of the Ides a vervex, yet in I, 56 says grandior agna, while the 
 feminine form agna is supported by Paulus, p. 104, and by Macrobius I, 15, 
 16; see p. 4 below. 
 
 ' See Chapter I. Here I showed that the sacra Idulia must have taken 
 place on the northern spur of the Capitoline Hill, technically known as 
 the Arx, but that Ovid and other writers loosely employ the term arx of 
 the hill as a whole, sometimes the thought of the Capitolium, or southern 
 spur, being more prominent in their minds ; that Ovid, then, when he read 
 in his source that the sacrifice took place in arce, understood the phrase as 
 a reference to the hill in general but more particularly to the Capitolium, 
 and so incorrectly recorded that the sacrifice took place in the great temple 
 of Jupiter Optimus Maximus, which stood on that spur. 
 
 6 Cic. de Leg. II, 29 : "Nam illud ex institutis pontificum et haruspicum 
 non mutandum est, quibus hostiis immolandum cui maioribus, cui lactenti- 
 bus, cui maribus, cui feminis." Ateius Capito ap. Macrob. Ill, 10, 4. Wiss. 
 p. 413. 
 
 7 Wiss., p. 413. Krause, pp. 19-20, and the list of sacrifices on p. 31 sqq. 
 1 An apparent exception that would apply equally to all sacrifices is the 
 
 passage of Servius, ad Aen. VIII, 641 : "In omnibus sacris feminini generis 
 plus valent victimae. Denique si per marem litare non possent succidanea 
 dabatur femina; si autem per feminam non litassent succidanea adhiberi 
 non poterat," which has been refuted by Krause, pp. 30-31. Krause quotes 
 Livy XLI, 15: "Alter consul curam adiecit, qui, quod caput iecinori de- 
 fuisset, tribus bubus perlitasse negavit; senatus maioribus hostiis usque 
 ad litationem sacrificari iussit". Wissowa throws the weight of his author- 
 ity against Servius, and cites Krause in a footnote, pp. 415-416 and notes 
 7 and 8. 
 
35 
 
 writers, however, which I expect to prove to be only apparent: 9 
 the sacrifice of a capra to Vediovis (Gell. V, 12, 12) ; of an agna 
 to Jupiter at the auspicatio vindemiae (Varro L. L. VI. 16) ; 10 
 and of the ovis Idulis to Jupiter. As the proof for my position 
 on the first two is bound up with that for my position on the 
 third, which is the real subject of the present dissertation, I shall 
 proceed at once to a discussion of the ovis Idulis. 
 
 Four places have survived in Roman literature that pertain to 
 the victim of the Ides, namely, those of Ovid, above quoted, and 
 two prose passages, one in Paulus Diaconus 11 and one in Macro- 
 bius. 12 Paulus says : 
 
 "Idulis ovis dicebatur quae omnibus Idibus lovi mactabatur". 
 
 The words in Macrobius are: 
 
 "Sunt qui aestiment Idus ab ove Iduli dictas, quam hoc nomine 
 vocant Tusci et omnibus Idibus lovi immolatur a flamine". 
 
 Let it be noted that the relative pronoun modifying ovis in 
 both instances is feminine. Therefore, these two passages, with 
 that of Varro just cited on the agna of the auspicatio vindemiae, 
 taken in connection with grandior agna in Fasti I, 56, would 
 seem to constitute rather a strong argument in favor of the 
 offering of female victims to Jupiter, especially of a female sheep 
 on the Ides. 
 
 9 Krause, p. 20 sqq., contends that these exceptions are merely ap- 
 parent, but his arguments are in the main superficial, and he does not 
 seem to me to offer convincing proof. He fails to produce any testimony 
 that the agna of the auspicatio vindemiae was male, except (p. 15) to draw 
 the inference from Paulus, p. 6, that agna was of common gender and 
 to conclude (p. 22) in contradiction of this hypothesis that Varro (L. L. 
 VI, 16) had made a mistake. As regards the ovis Idulis (pp. 11-12), he 
 makes the assertion that ancient and modern writers wrongly agree that it 
 was female, and thus leaves out of account Ovid's positive statement that 
 the victim was a vcrvex, which Krause quotes but otherwise ignores ; he is 
 chiefly concerned with proving that the animal was the aries rather than 
 the vervex or castrated male. Altogether there is need of a thorough in- 
 vestigation to establish the nature of this sacrifice. 
 
 10 "Nam aliquot locis vindemiae primum ab sacerdotibus publice fiebant, 
 ut Romae etiam nunc; nam flamen Dialis auspicatur vindemiam et ut 
 iussit vinum legere agna lovi facit, inter cuius exta caesa et proiecta flamen 
 t porus vinum legit." 
 
 11 p. 104. 
 
 * I, 15, 16. 
 
36 
 
 But it can hardly be poetic license or carelessness that caused 
 Ovid in I, 588, to say that the animal was semimas, or castrated. 
 Agna, as will be seen later, is an ordinary term which Ovid might 
 easily have written for stylistic reasons. Semimas ovis, on the 
 contrary, is an unusual expression, and the poet is going out of 
 his way to employ it. But verve x stands in the Ada, Fratrum 
 Arvalium as the only one of the sheep kind to be offered to 
 Jupiter, while Janus on the other hand receives each time an 
 aries^ and the goddesses oves. So detailed a statement of pon- 
 tificial ritual, therefore, as semimas ovis Ovid would scarcely have 
 written, had not the word vervex stood before him in the source 
 which he used. Thus this line of Ovid is corroborated by the 
 Acta Fratrum Arvalium, our most authoritative epigraphic source 
 on sacrifices in ancient Rome, and this combined testimony is 
 arrayed against the apparent evidence of Ovid's line on the agna, 
 Varro's vindemia passage and what Paulus and Macrobius write 
 about the ovis Idulis. 
 
 Modern scholarship has failed to reach a definite conclusion 
 regarding the sex of this victim. While Klausen (Aeneas und die 
 Penaten, p. 930) says "agna opima," and Aust (Die Religion der 
 Romer, p. 168) "ein mannliches Schaf," Samter says merely "em 
 Schaf," 15 C. Julian in Daremberg et Saglio 16 "un mouton," and 
 Wissowa 17 "ein weisses Schaf (ovis Idulis) " while elsewhere he 
 is quite explicit, that at the Agonium, for example, the sacrifice 
 was "ein Widder" and at the suovetaurilia a "Schaf bock." 18 
 
 There are several passages in the surviving literature which 
 point to a common gender for ovis like bos, that is, that the 
 modifiers were either masculine or feminine adjectives and pro- 
 nouns, according to the sex of the particular animal under con- 
 sideration. 19 They are as follows : Festus (p. 286) : 
 
 11 Henzen, pp. CLXXXVI and CCXIV. 
 
 14 Cf. F. I, 318 and 333-34; Varro L. L. VI. 12. 
 
 15 In Paul.-Wiss., VoL VI, p. 2491. 
 "Vol. II., p. 1162. 
 
 17 p. 114 of Religion und Kultus der Romer. 
 "pp. 103 and 142. 
 
 18 F. IV, 631 : "Forda ferens bos est, fecundaque dicta ferendo"; Act. Fr. 
 Arv., p. CLXXXVIII bovem feminam; et al. Cato Origines, 103 : "Trim 
 boves; Ov. F., Ill, 732, "Deque triumphato viscera tosta bove"; et al. 
 
37 
 
 "Etiam in commentariis sacrorum pontificalium frequenter est 
 hie ovis et haec agnus ac porcus, quae non ut vitia, sed ut antiquam 
 consuetudinem testantia debemus accipere," 
 and Gellius (XI, 1, 4) quoting Varro: 
 
 "Quando nunc quoque a magistratibus populi Romani more 
 maiorum multa dicitur vel minima vel suprema, observari solet, 
 ut oves genere virili appellentur; atque ita M. Varro verba haec 
 legitima, quibus minima multa diceretur, concepit: 'M. Terentio, 
 quando citatus neque respondit neque excusatus est, ego ei unum 
 ovem multam dico/ Ac nisi eo genere diceretur negaverunt 
 kistam videri multam." 
 
 Again Paulus (p. 195) says: 
 
 "Ovem masculine genere dixerunt, ut ovibus duobus non 
 duabus." 
 
 From these examples it will be seen that the use of ovis as a 
 common gender noun suggests that the excerpts from Paulus (p. 
 104), ovis quae, and from Macrobius (I, 15, 16), ovis quam, refer 
 to female victims. But I shall show that the ovis Idulis was a 
 male, and hence the common gender theory does not furnish the 
 correct explanation of these two prose passages. 
 
 Equally suggestive with this study of nouns of common gender 
 is a study of the gender of pronouns with nouns of unquestion- 
 able gender. I find no examples of feminine relatives modifying 
 masculine nouns and but one reliable instance of the opposite 
 process. 20 There is also, however, a passage in Varro's de Lingua 
 Latina* 1 in which the manuscripts give the reading quorum 
 after a feminine antecedent, though the vulgate and two earlier 
 editors 22 print quorum. This antecedent, it should be stated, is 
 haec, plural of the demonstrative ; not the neuter plural, however, 
 but an old form of the feminine. 23 In spite of the fact that it 
 contains certain serious corruptions, the passage may throw light 
 on the two prose references 24 to the ovis Idulis, and shall, there- 
 fore be quoted : 
 
 20 Pompeii Commentum, p. 249 (Keil V, p. 206) : "Quis tu es mulier qui 
 me hoc nuncupasti nomine". 
 
 21 V, 98. 
 
 22 Mulier and Lindemann. 
 
 23 See Miiller's edition on the passage and Sommer Laut und Formen- 
 lehre, p. 423 in 2d ed. 
 
 24 See preceding page, p. 35, and notes 11 and 12. 
 
38 
 
 "Aries qui team dicebant ares, veteres nostri ariuga, hinc 
 ariugas. 25 Haec sunt quorum in sacruficiis exta in olla non in 
 veru coquuntur, quas et Accius scribit et in pontificiis libris 
 videmus." 
 
 The first thing to be noted here after the hardly explainable cor- 
 ruption qui, is the gender of earn, which, if it is really an accusa- 
 tive form of the demonstrative is ea id, refers to aries. The vul- 
 gate, followed by Miiller, prints quod eum. In view of the cor- 
 ruption of the line, however, I shall not attempt an argument on 
 this pronoun, but wish merely to direct attention to it. The 
 point desired to emphasize particularly is that, although the 
 word ariuga or arviga was bestowed a great part of the time 
 on males, its gender is shown by the forms haec and quas to 
 be feminine. The word hostia, though covering both male and 
 female victims, is well known to have but one gender, feminine, 
 and after a careful perusal of the surviving passages relating to 
 arviga, 26 I can see no other meaning in the word than that it was 
 a special kind of hostia, referring very frequently to the male, 
 but properly modified at all times by the feminine of the pronoun. 
 This arviga passage constitutes, as it were, a bridge over which 
 we pass to the real solution of the problem confronting us in 
 this chapter. 
 
 There is a class of animal names known as epicenes from the 
 Greek e7rtx,otvo<; which have but one word form and one gram- 
 matical gender to denote both sexes, some words of this class 
 being masculine, some feminine, as hie passer, haec aquila 21 
 
 25 The text followed is that of Goetz and Schoell. The most serious and 
 unmistakable corruption is in qui and earn, for which Goetz and Schoell, 
 though printing the reading of the best MS., offer the emendation quidam. 
 The more recent MSS. and the Thesaurus give arviga rather than ariuga; 
 Paul. (p. 100) harviga. 
 
 26 Varro L. L., V, 98 : the part quoted above and "In hostiis earn dicunt 
 ariugem quae cornua habeat"; Paul., p. 100; Velius Longus (Keil, VII, 73, 
 9) "Ar(v)iga quae est hostia." 
 
 27 Lane, Latin Grammar, revised ed., paragraph 411, gives the following 
 definition : "Epicenes have one ending and one grammatical gender, though 
 applicable to animals of either sex. Thus aquila, eagle, is feminine, though 
 it may denote a he-eagle as well as a she-eagle; anates, ducks, feminine, 
 includes drakes." See also for definitions Diom, I, 276 (Keil, I, 301, 10) ; 
 Charisius, II, 89 (Keil, I, 153, 15) ; Donatus (Keil, IV, 375, 22) ; etc., and 
 cf. Neue-Wagener Formenlehre, I, p. 926 sqq. 
 
39 
 
 One of the many good definitions is that of Charisius, I, 8 (Keil 
 I, 17, 10) : 
 
 "Adicitur. . .genus quod Graece erctaotvov dicitur, Latine pro- 
 miscuum, ut haec mustela, aquila. Nam etsi mas sit mustela vel 
 aquila, tamen feminine genere tantum dicitur. Item hie passer 
 quamvis masculine genere proferatur, tamen etiam femininum 
 genus significat." 
 
 Neue-Wagener 28 say "the names of most animals belong to 
 this class, generally of the smaller animals, but especially those 
 in which there is no occasion to differentiate the male from the 
 female." A few animals of considerable size are epicene, how- 
 ever, for volpes, as will presently be seen, is decidedly so, and 
 panthera is shown to be so by Cicero, by Pliny, by Fhaedrus, and 
 finally by Varro. 29 Furthermore a marked differentiation of sex 
 is sometimes made, as, for example, in the expression et mas 
 et femina aquila which does not indicate a different gram- 
 matical gender, 31 and in Pliny's expression volpis masculae, 32 
 where the form of the adjective masculae shows that the gram- 
 matical gender of volpis is feminine, but the adjective itself 
 proves the animal to be unmistakably male. But the most com- 
 mon use of epicenes is in passages where the animal mentioned 
 
 'Formenlehre, I, p. 927 sqq. They furnish quite a long list (p. 927) of 
 each gender, including among the feminines panthera, volpes, bidens (the 
 sheep), avis, anas, aquila, merula, and apis. 
 
 29 Pliny, N. H., VIII, 62 and 63 ; Phaedrus, III, 2 : "Panthera et Pas- 
 tores", where nothing is implied as to the animal's sex. Cicero, Epist. ad 
 Pam. VIII, 9, 3 : "Fere litteris omnibus tibi de pantheris scripsi : turpe tibi 
 erit Patiscum Curioni decem pantheras misisse, te non multis partibus 
 plures ; quas ipsas Curio mihi et alias Af ricanas decem donavit, ne putes 
 ilium tantum praedia rustica dare scire". For the Varro passage see p. 
 40 above. 
 
 30 Varro, L. L., VIII, 2, 7. 
 
 81 Neue-Wagener, I, p. 919. Priscian, V, 8, 42 (Keil, II, p. 169) states 
 that in the earliest times aquila was of common gender; but in classical 
 times, at least, we find no more epicene word in the language. It is quoted 
 as a stock example of the epicene in all the definitions above quoted or 
 cited (see note 27 and note 28). In Serv. ad Aen. I, 394, Interpol. Serv. 
 ad Eel., VI, 42, Suet. Aug. 96, etc., it is an epicene with nothing implied 
 as to sex. 
 
 98 N. H., XXVIII, 166; for gender, also see Phaedrus. 
 
40 
 
 may be either male or female, or in the case of a group, the refer- 
 ence may be to beasts of both sexes; in other words, where the 
 reference to sex is wholly indefinite. This is best shown by a 
 long passage from Varro (L. L. IX, 55 sqq.), 33 which it will be 
 well to bear in mind throughout the remainder of this dissertation. 
 
 "Negant, cum omnis natura sit aut mas aut f emina aut neutrum, 
 (non) debuisse ex singulis vocibus ternas figuras vocabulorum 
 fieri, ut albus, alba, album; nunc fieri in multis rebus binas, ut 
 Metellus, Metella . . ., nonnulla singula, . . . ; dici cor- 
 vum, turdum, non dici corvam, turdam; contra dici pantheram, 
 merulam, non dici patherum, merulum, . . . Ad h(a)ec dici- 
 mus omnis orationis, quamvis res naturae subsit, tamen si ea in 
 usu(m) non pervenerit, eo non pervenire verba; ideo equus 
 dicitur et equa: in usu enim horum discrimina; corvus et corva 
 non, quod sine usu id quod dissimilis natura (e). Itaque quaedam 
 al(i)ter olim ac nunc; nam et turn rimnes mares et feminae 
 dicebantur columbae, quod non erant fci eo usu domestico quo 
 nunc, (nunc) contra propter domesticos usus quod internovimus, 
 appellatur mas columbus, femina columba." 
 
 These animal words are not discussed as nouns of common 
 gender and of one declension, as hie or haec corvus and turdns, hie 
 or haec panthera and merula. The evidence proves them unmis- 
 takably to be epicenes. First of all let us note the testimony of 
 other passages on the animal names chosen. Varro himself says 
 (R. R. Ill, 5, 6) : 
 
 "Turdi, qui cum sint nomine mares, re vera feminae quoque 
 sunt. Neque id non secutum ut esset in merulis, quae nomine 
 feminine mares quoque sint." 
 
 Here, though omitting the designation epicene, he shows em- 
 phatically that such is the usage of the words turdus and merula. 
 Again, corvus is defined by name as epicene (promiscuus) by 
 Consensus (Ars, Keil V, 30) : 
 
 "In promiscuis . . . sub uno articulo uterque sexus signifi- 
 catur. Nam cum dico masculino genere corvus, neque nomine 
 neque articulo confusionem generis separare possum: tarn enim 
 femina quam masculus corvus masculino genere enuntiatur. Item 
 
 33 Text of Goetz and Schoell ; cf. with Servius' Commentary on Donatus, 
 p. 1782 (Keil, IV, 408). 
 
41 
 
 cum dico cornix, sexum nulla ratione discernere possum, quoniam, 
 sive masculus sive femina sit, feminine genere cornicem appello. 
 Nam in corvo f emininum et in cornice masculinum genus intellegi 
 necesse est." 
 
 That panthera too is epicene is perceived both by its use in 
 the various passages above cited 34 and by its choice in the lines of 
 the de Lingua Latina (IX, 55 sqq.) to accompany coruus, turdus, 
 and merula. But Varro's own words prove that he is discussing 
 epicenes, not nouns of common gender. It will be best to restate 
 this portion of the passage *in translation : 
 
 "We say that, although all language is based upon distinctions 
 arising from conditions in the natural world, yet if these natural 
 conditions are turned to no practical employment by man, such 
 distinctions fail to be expressed in speech ; therefore we have the 
 words equus and equa, for there is a distinction in the way these 
 two words are employed ; on the contrary, we do not say corvus 
 and corva, for a discrimination of sex is here of no practical 
 benefit. For this reason certain words were formerly differently 
 used from the manner of their employment to-day; at one time, 
 for example, all pigeons, male and female, were called columbae, 
 because they were not put to that domestic use to which they 
 are now given over. Now, however, for domestic purposes, we 
 make a distinction, calling the male pigeon columbus, the female, 
 columba." 
 
 If Varro were here discussing common gender, hie and haec 
 corvus, for instance, the natural condition of the bird, male or 
 female, would be every whit as distinctly expressed by the pro- 
 noun as if the termination in -a 'had been employed to differ- 
 entiate the female and that in -us had been confined to the male. 
 
 Since beyond doubt this passage of Varro is a discussion 
 of epicenes, attention must once more be called to my state- 
 ment made immediately before the passage was quoted, namely, 
 that epicene names for animals commonly show that there 
 is no definite thought of sex in the writer's mind. This is 
 the whole point of Varro's definition; that man does not 
 make a linguistic division based on the sex of animals, unless 
 there is practical need of a distinction. The names of most 
 
 See n. 29. 
 
42 
 
 animals, therefore, are epicene, and the use of epicenes seldom 
 offers any direct information concerning their sex. Varro, how- 
 ever, has neglected to mention that his rule of vagueness must 
 admit of occasional exceptions, in which an epicene is proved to 
 be of a sex directly opposite to that indicated by its grammatical 
 gender. A step in this direction appears in the scholiast to 
 Germanicus' Aratea, "lovem in aquilam transfiguratum" 35 and 
 in Apuleius' Metamorphoses III, 25, "non avem me sed asinum 
 video." In both of these selections the bird is certainly a male, 
 because in either case a male has been metamorphosed. Yet in 
 all extant literature aquila and avis are of feminine gender. 88 
 The most perfect illustration of the employment of an epicene 
 in unquestionable reference to the sex opposed to its grammatical 
 gender is volpis masculae in Plir^y (N. H. XXVIII, 166), 
 which has already been noted. This development produces the 
 purest epicene, and is the logical outcome of the more common 
 usage in which the sex is not distinctly specified. It would seem 
 possible only in the case of true epicenes, if such a term may be 
 employed ; that is, of words which have only one grammatical 
 gender, like volpes, merula, corvus, panthera, aquila, and avis. 
 Now aquila has been shown to be a true and unquestionable 
 epicene. Therefore, when we read in Varro (L. L. VIII, 2, 7} 
 et mas et femina aquila, there is every reason to agree with 
 Neue-Wagener, 38 that the expression does not indicate a differ- 
 ence of grammatical gender. It is not to be regarded, then, as 
 identical with the similar epithets applied to bos, which has been 
 seen to be of common gender, but aquila mas is in a class with 
 volpis masculae of Pliny. 39 I likewise believe that this is the 
 very manner in which ovis mas is employed by Varro (L. L. V, 
 98), "si cui ovi mari testiculi dempti et ideo vi natura versa, 
 
 35 Breysig, p. 160. 
 
 36 See Thesaurus and cf. note 28 and note 31. 
 
 17 Later on it will be shown that certain words which are by no means 
 true epicenes, are sometimes employed after the manner of epicenes, when 
 nothing is implied as to their sex. 
 
 38 Seen. 31. 
 
 89 See n. 32. 
 
43 
 
 verbex declinatum," and by Ovid in the expression 40 semimaris 
 ovis. In other words, it will presently be shown that the word 
 ovis is epicene. 
 
 It will at once be objected that the passages of Gellius (XI, 1, 4) 
 and Festus (p. 286), which record the words of Varro and Ver- 
 rius respectively and have been quoted earlier in this work, 41 
 show the gender of ovis to be common. It must be observed, 
 however, that Varro so uses it only in an antiquated legal 
 formula, and that Gellius^and Festus mention the usage only to 
 explain it, because in their day it was strange and unusual. Not 
 only was this common gender for ovis a curiosity in classical 
 Latin, but, if we remember Varro's rule illustrated by columba* 2 
 I think I am justified in the conjecture that ovis began in the 
 earliest times as a feminine epicene for the sheep kind without 
 distinction of sex, just as it will be proven to be in historic times; 
 that this usage persisted, but in the course of development in the 
 still prehistoric Latin the usage in common gender, hie and haec 
 ovis, arose in those cases where a differentiation of male and 
 female was important, especially in religious ritual, where with 
 the advent of animal sacrifices and the offering of male victims 
 to male divinities and female to female, practical necessity com- 
 pelled a distinction in sex to be observed. While in historic 
 times, owing to the limitations of the Latin vocabulary, ovis was 
 retained as the particular term for the ewe, and dries and vervex 
 instead of hie ovis became the more common and conventional 
 terms for "ram" and "wether," ovis remained (as it had always 
 been) the general name for the sheep kind, like the English 
 sheep and the German Schaf. So much for my conjecture. It 
 will now be shown that in the classical period ovis was the generic 
 term for sheep and that its gender was feminine. 
 
 Cicero says (N. D. II, 63) : 
 
 "Quid enim oves aliud afferunt, nisi ut earum villis confectis 
 atque contextis homines vestiantur"? 
 
 This is supplemented by Isidorus, Origines XII, 1, 8-9: 
 
 "Discretio est autem inter armenta et greges: nam armentn 
 
 40 F. I, 588. 
 
 4J See pp. 36 and 37 above. 
 
 42 See p. 40 above. 
 
44 
 
 equorum et bourn sunt, greges vero caprarum et ovium . . . 
 Apud veteres initio non tauri sed oves in sacrificio mactarentur. 
 . . . Vervex ... a viribus dictus, quod ceteris ovibus sit 
 fortior." 
 
 Note especially the last words, "Vervex . . . ceteris ovibus 
 sit fortior," which, taken in connection with the Cicero passage 
 just quoted and with Columella VII, 4, 43 prove that the ancient 
 flocks, like those of to-day, 44 were made up of both ewes and 
 wethers, the latter being slaughtered while their flesh was still 
 tender, but after they had yielded a goodly supply of wool. In 
 these passages of Cicero and Isidorus, ovis can signify but one 
 thing, the sheep kind as a whole4 Isidorus moreover says in the 
 same selection: 
 
 "Ex his (sc. ovibus) quasdam bidentes vocant, eas quae inter 
 octo dentes duos altiores habent, quas rrtaxime gentiles in sacri- 
 ficium offerebant." 
 
 Here and also in the Cicero quotation (earum), the feminine 
 gender of ovis as well as its general meaning is well illustrated. 
 Further excellent examples are the Pseudo-Acron scholiast to 
 Horace (C. Ill, 23, 14) : 
 
 "Bidentes autem proprie dicuntur oves duos annos habentes, 
 sic vocatae ab eminentioribus dentibus, qui circa duos annos nas- 
 cuntur," and Servius (ad A en. VI, 39) : 
 
 "Bidentes autem, ut diximus supra (IV, 57) oves sunt circa 
 bimatum, habentes duos dentes eminentiores ; quae erant aptae 
 sacrificiis." 45 
 
 J "Plures autem in eiusmodi gregibus, quam in hirtis, masculos enutrire 
 oportet. Nam priusquam feminas inire possint mares castrati, cum bima- 
 tum expleverint, enecantur, et pelles eorum propter pulchritudinem lanae 
 maiore pretio, quam alia vellera, mercantibus traduntur". 
 
 44 Farmers' Cyclopedia, abridged agricultural records from publications 
 of U. S. Dept. of Agric. and Experiment Stations, Vol. I, p. 368: (On 
 eastern Nevada sheep raising) : "They (the sheep) usually reach the 
 shearing grounds . . . about the first of April. The sheep are then sepa,- 
 rated into ewe bands and wether bands, the ewes and their lambs of the 
 previous year having run together during the winter. . . Two crops of 
 wool are also obtained from each lamb, as they are not usually sold until 
 about two years old (Cf. Columella's bimatum) , which differs from the 
 practice in the Sierras". The Encyclopedia Britannica (1911 ed., article on 
 Agriculture) I, p. 408, speaks of wether sheep above twelve and under 
 twenty- four months old. 
 
 45 Also Paul, p. 4; Cell. XVI, 6, 12. Bidens is recognized by Neue- 
 Wagener as an epicene (I p. 927). 
 
45 
 
 These last passages have to do with bidentes, which are seen to 
 be sheep of an age sufficiently advanced to display a certain 
 phenomenon in their teeth. This phenomenon, however, is not 
 confined to the female, but appears equally in the male. 
 
 Ovis, therefore, is epicene, and, moreover, a true epicene, like 
 volpes, aquila , merula, avis, panthera, corvus, and others. 46 It is 
 epicene, because it has just been proved to be the generic term for 
 sheep without thought of sex, to have only one grammatical gen- 
 der, feminine, 47 and yet, as a true epicene, to be carried to its 
 logical development, so that on a few occasions, such as we have 
 encountered in Ovid, 48 and in Varro, 49 it is employed strictly of 
 the male. 
 
 If we return, then, for one moment from the study of epicenes 
 to our subject of the sacra Idulia, it will be recalled that Roman 
 ritual required the sacrifice of male animals to male divinities and 
 of female to female. The actual ritualistic custom is thus seen 
 to be quite in accord with the words of Paulus and Macrobius, 50 
 who, although they apply the correct gender of the relative pro- 
 noun in writing "Idulis ovis . . . quae" and "ab ove Iduli 
 . . . quam," do so with no more thought of sex than did 
 Varro (L. L. V, 98) in writing "Haec (sc. arvigae) sunt . . . 
 quas"; for arviga, though a hostia and of feminine gender, was 
 often used of the ram. 51 The sex of the ovis Idulis was male, 
 therefore, as is to be expected in a sacrifice to Jupiter. 
 
 It remains to discuss Ovid's use of the feminine form agna in 
 Fasti I, 56. . This in itself would not offer serious difficulty, for 
 had it suited his purpose, Ovid might not have hesitated to write 
 agiw,, even if bearing definitely in mind the female significance of 
 the termination in -a. Vergil in his famous line (Aen. VIII, 641), 
 
 w No matter that it often means the ewe, for we know that even so 
 widely recognized an epicene as avis frequently refers to the female bird 
 alone, e. g. Cic. N, D. II, 129 : "Gallinae avesque reliquae . . . quietum 
 requirunt ad pariendum locum"; Plin. N. H. X, 147: "Ceterae aves semel 
 anno (pariunt)"; X, 165: "nee alia (avis) plures (quam XX pullos parit)". 
 
 47 Except in a few legal survivals of prehistoric times. 
 
 48 F. I, 588. 
 "L. L. V,98. 
 
 50 Paul. p. 104; Macrob. I, 15, 16. See page 35 above. 
 
 51 See above p. 38. 
 
46 
 
 "caesa iungebant foedera porca," may employ porca in this man- 
 ner, inasmuch as the commentators mentioned as quidam by the 
 interpolator of Servius, explain porca on the ground of euphony. 52 
 Perhaps the quidam mentioned in the interpolation of Servius are 
 Quintilian and Porphyrio, the commentator on Horace. The lat- 
 ter says : 53 
 
 "Adtende feminino genere agnam maluisse dicere quam agnum 
 secundum illud Vergilianum : 'Et caesa iungebant foedera porca' ; 
 nescio quid enim quaedam eloquutiones per femininum genus 
 gratiores fiunt." 
 
 Quintilian's words are (fnst. VIII, 3, 19) : 
 
 "Quaedam non tam ratione quam sensu iudicantur; ut illud, 
 'caesa iungebant foedera porca/ fecit elegans fictio nominis ; quod 
 si fuisset porco, vile erat." 
 
 These three explanations of the feminine form in -a on the 
 ground of euphony are good as far as they go, 54 but they hardly 
 account for Varro's mention of the offering of an agna to Jupiter 
 at the auspicatio vindemiae or for what Gellius writes of the sacri- 
 fice of a capra to Vediovis. Once more it becomes necessary to 
 fall back on epicenes. 
 
 First of all it must be pointed out that there are certain words 
 which, though not strictly of the epicene class, in that they admit 
 both genders, are nevertheless often employed in the epicene 
 manner described by Varro, 55 namely, in a vague, general sense 
 referring to either or both sexes. Their gender is arbitrarily 
 chosen, being sometimes masculine, sometimes feminine, but there 
 is no thought of sex in the writer's mind. 
 
 52 Serv. Interpol. ad Aen VIII, 641 : "Quidam 'porcam' euphoniae gratia 
 dictam volunt". 
 
 58 C. I, 4, 12. 
 
 64 Servius himself (ad Aen VIII, 641), after recognizing that ritualistic 
 observance called for porco, gropes about in a maze of explanations for the 
 use of the feminine : "Falso autem ait 'porca' : nam ad hoc genus sacrificii 
 porcus adhibebatur. Ergo aut usurpavit genus pro genere ut 'timidi 
 venient ad pocula dammae', cum has dammas dicamus, item supra (631) 
 lupam cum artis sit 'hie' et 'haec' lupus. Aut certe illud ostendit, quia 
 . . . ' ' From this point on see note 8 above, which also notes the unre- 
 liability of this passage in questions of ritual. 
 
 55 L. L. IX, 55. See above p. 40. 
 
47 
 
 One word notably so employed is cants, which is ordinarily of 
 common gender because the natural difference of sex is so often 
 forced upon one's attention owing to the intimate association of 
 dog and man. When employed with absolutely no thought, how- 
 ever, of male and female, the gender of canis is determined accord- 
 ing to the writer's whim or preference. A few examples will 
 suffice : Varro (R. R. II, 9, 6) in discussing the purchase of dogs 
 uses the masculine gender: 
 
 "Magni interest ex semine esse canes eodem, quod cognati max- 
 ime inter se sunt praesidio," and again in treating of their food 
 says (II, 9, 8) : "Cibatus canis propior hominus . . . Dili- 
 genter ut habeat cibaria providendum. Fames enim hos ad 
 quaerendum cibum ducet." Once more in regard to buying dogs 
 he says (11,9,5): 
 
 "Videndum ne a venatoribus aut laniis canes emas, alteri quod 
 ad pecus sequendum inertes; alteri, si viderint leporem aut cer- 
 vum, quod eum potius quam oves sequentur. Quare a pastoribus 
 empta melior, quae oves sequi consuevit, aut sine ulla consuetu- 
 dine quae fuerit." 
 
 In this passage both genders are used, yet each time dogs are 
 mentioned the reference must be to both sexes. Toward the close 
 of the chapter Varro speaks of protecting the watchdogs with 
 collars armed with nails, 
 
 "quod si lupus aliusve quis his vulneratus est, reliquas quoque 
 canes facit, quae id non habent, ut sint in tuto." 
 
 In these words both sexes are referred to without doubt, and 
 furthermore, there immediately follows a discussion of the num- 
 ber of dogs necessary for the flock, and we meet with the words : 
 
 "Villatico vero gregi in fundum satis esse duo, et id marem et 
 feminam." If one may differentiate, then, between the word and 
 the usage, one would say that canis, while not a true epicene, 
 quite frequently receives that epicene usage, which Varro has 
 described, namely, that in which there is no thought of sex. 56 
 
 Neue-Wagener present a small list of animal names with the 
 double endings in -us and -a, which they say 57 are epicenes to the 
 extent that they may be used in either form for the species in 
 
 "L. L. IX, 55. See above p. 40. 
 IT I, pp. 927-930. 
 
48 
 
 general without reference to sex, although grammatically those in 
 -us are masculine and those in -a feminine. A good example of 
 this is seen in Pliny (N. H. VIII, 141) : 
 
 "Lacertae inimicissimum genus cocleis, negantur semestrem 
 vitam excedere. Lacerti Arabiae cubitales, in Indiae vero Nysa 
 monte XXIV in longitudinem pedum, colore fulvi aut punicei, aut 
 caerulei," where, whether the termination be in -ae or in -i, the 
 entire species is meant, including both sexes. Another word of 
 this type is simius, simia, the latter being the commoner, 58 but 
 simius appearing in Phaedrus' fable of the wolf, the fox, and 
 the ape. 59 
 
 Varro 60 tells us that the word columba was originally an epicene, 
 and that when the pigeon came into wide domestic use and man 
 found it convenient to differentiate between the sexes, the mascu- 
 line form, columbus, was introduced for the sake of distinction. 
 We find, however, that in those instances where this word is 
 employed indefinitely without sex distinction, the form in -a is 
 usually retained. 61 Sometimes, on the contrary, we meet with the 
 masculine, columbi. 62 Either form of this word can thus be 
 given the epicene usage, although the form in -a was originally 
 a true epicene. 
 
 If one will turn now to the word agna, and that with which it 
 is closely associated, porca, attention must once more be called 
 to my translation of Varro's definition 63 (p. 41 above), the 
 point of which is that man does not make a linguistic division 
 based on the animal's sex unless there is a practical need of such 
 distinction. Much of the time there is a manifest need of distin- 
 guishing the sexes of agnus and agna, porcus and pored, as in the 
 cases of equus and equa, columbus and columba, mentioned by 
 
 " Plin. VIII, 54, 215. 
 
 69 Phaedr. I, 10. 
 
 " See n. 56. 
 
 61 Only one or two examples will be quoted here: Plin. X, 35: "Columbae 
 et turtures octonis annis vivont". Varro R. R. Ill, 2, 13: "alterum (sc. 
 genus pastionum) villaticum in quo sunt gallinae ac columbae et apes et 
 cetera", and 14; III, 5, 7: "volucres . . . vernaculae ut gallinae ac colum- 
 bae". 
 
 Colum. VIII, 8, 1 : "palumbos columbosque cellares pinguissimos 
 facere"; and 10, 2: "(Turdi) locum aeque munitum et apricum, quam 
 columbi desiderant". 
 
 " R. R. IX, 55. 
 
49 
 
 Varro. 64 On the other hand, as in the case of columba, so the 
 reference in agnus or porcus is often only generic, with absolutely 
 no thought of sex. This was, we have seen, the original use of 
 columba, 65 and it was retained in classical literature after the form 
 columbus had come into use. 66 Although agnus and porcus are 
 used so very frequently of the male alone, there are a number of 
 passages in which they are employed in as epicene a manner as 
 columba is known to have been, for example, Cicero de Senec. 
 XVI, 56: 
 
 "Villa . . . abundaCporco, haedo, agno, gallina," 
 and Varro on several occasions: 
 R. R. II, 1, 20: 
 
 "Fere ad quattuor menses a mamma non disiunguntur agni, 
 haedi tres, porci duo," 
 R.R. II, 2, 15: 
 
 "Deinde matres cum grege pastum prodeunt, retinent agnos, 
 ad quos cum reductae ad vesperum, aluntur lacte et rursus dis- 
 cernuntur," 
 and II, 2, 17: 
 
 "Cum depulsi sunt agni a matribus diligentia adhibenda est ne 
 desiderio senescant." 
 
 Agnus and porcus in all these selections must refer to lambs 
 and pigs in general without thought of sex. 
 
 There survive also some interesting passages showing a com- 
 mon gender for agnus and porcus in the earlier times and in 
 ritualistic survivals. The first of these (Festus, p. 286) has been 
 already noticed in connection with the antiquated common gender 
 usage of ovis,* 7 but must be quoted here for present purposes : 
 
 "Etiam in commentariis sacrorum pontificalium frequenter est 
 hie ovis et haec agnus ac porcus, quae non ut vitia sed ut antiquam 
 consuetudinem testantia debemus accipere." 
 
 Perhaps second in importance is Paulus (p. 6) : 
 
 "Agnus ex Graeco dpLvo? deducitur, quod nomen apud maiores 
 communis erat generis, sictit lupus/' 
 
 "/?. R. IX, 55. 
 "/?./?. IX, 55. 
 
 " See IL 59. 
 
 " See p. 37 and p. 43 above. 
 
50 
 
 If the epithets mas and femina are to be taken as indicative of 
 common gender, which appears to be their most frequent usage, 
 porcus is of common gender in Cato's de Agricultures, 
 CXXXIV. The selection is too long for quotation. Although 
 pertaining to private ceremonial, it is highly ritualistic and most 
 definite in recording the details of sacrifice and the formulae of 
 prayer. In it porcus femina and pore a are employed several 
 times interchangeably. 68 Festus also says in reference to a law 
 of Numa (p. 222) : 
 
 "Si tanget (sc. pellex aram lunonis) . . . agnum femi- 
 nam caedito." 
 
 But most significant for the present argument are Festus' words 
 on page 189, in which he seems to be drawing his authority from 
 the pontifical books : 
 "lanui Quirino agnum marem caedito." 
 
 This expression has its counterpart in the inscription recording 
 the Ludi Saeculares of 17 B. C., now in the Terme Museum. 69 
 That section which relates to the sacrifice to the Moerae contains 
 the words: 
 
 "uti huiusjsacrifici acceptrices sitis VIIII agnarum feminarum 
 et VIIII caprarufm feminarum propri]arum inmolandarum ; 
 harum rerum ergo macte hac agna femina inmolanda estote fitote 
 v[olente]s propitiae," etc. 
 
 The words hac agna femina exclude any other interpretation 
 than that of common gender for agna in this passage, although it 
 is the only surviving passage in which agna is so employed. 70 
 
 If now it be permissable to turn for a moment to poetry, a line 
 of Ovid will be found of particular interest (Fasti, IV, 648) : 
 
 88 With this passage cf . Cicero de Leg. II, 22, 57. 
 
 69 C I L. VI, 4 2 , 32323. 
 
 70 Feminam agnam quoted by Augustine from Leviticus (Quaestionum 
 in Heptateuchum Libri VII in the Corpus Scriptorum Ecclesiasticorum 
 Latinorum, Vienna 1895, 3, 3, p. 236) is a translation of the original 
 Hebrew, not a Latin usage. As quoted by Augustine the passage reads : 
 'Offeret pro his quae deliquit domino, pro peccato quo peccavit feminam 
 ab ovibus agnam', and with this cf. the translation made at London in 
 1580 from the Hebrew: "Itaque adducito reatum suum Jehove propter 
 peccatum suum quo peccabit foeminam e grege pvem", and King James 
 translation into English: "And he shall bring his trespass offering unto 
 the Lord for his sin which he hath sinned, a female from the flock, a 
 lamb". Augustine's comment is: "usitata locutione 'agnam feminam' 
 dicit, quasi possit esse non femina". 
 
51 
 
 "Agnaque nascendo saepe necabat ovem." 
 While the form in -a may well be chosen by the poet for 
 stylistic purposes, yet it refers as clearly to the male as to the 
 ewe lamb and forms a striking counterpart to the epicene use of 
 agnus and porcus in Cicero 71 and Varro 72 quoted above. Prose 
 passages in which agna and porca are employed as epicenes are 
 few, but that of Paulus (p. 235) is highly important: 
 
 "Porci effigies inter militaria signa quintum locum obtinebat, 
 quia confecto bello inter quos pax fieret, caesa porca foedus fir- 
 mare solebant." 
 
 Porci effigies seems to be a clear reference to the boar, for 
 Pliny 73 says that the aper held fifth place among the animals on 
 military standards; caesa porca may be a reminiscence of 
 Vergil 74 ; but in whatever way this passage is regarded, the thought 
 
 71 de Senec. XVI, 56 ; see p. 49 above. 
 
 72 R. R. II, 1, 20; II, 2, 15 and 17; see p. 49 above. 
 "N.H. X, 16. 
 
 "The theory of the slaughter of female animals to Jupiter in treaty- 
 making (see Neue-Wagener I, p. 925-926) is not well supported by the 
 existing evidence. Porcus, to be sure, was of common gender in early 
 religious formulae, but that does not affect the sex of the pig slain in the 
 conclusion of a treaty, for Livy, I, 24, 8, says: "Turn illo die, luppiter, 
 populum Romanum sic ferito, ut ego hunc porcum hie hodie feriam (cf. 
 IX, 5, 3; XXI, 45, 8; Cic. de Invent. II, 30, 91) ; also in Varro, R. R. II, 
 4, 9, the mention of this sacrifice in the same passage with that of a pig at 
 the marriage ceremony in no wise proves the word feminine, for the 
 wedding victim was chosen not because of its sex but because of some 
 mysterious connection between the pig and the pudenda muliebria, as seen 
 in the Greek use of ^otpo? in the comic poets (cf. Aristoph. Ach. 774) 
 and the close relation between SeXcpac and SsX^u? (see L- Meyer, Griech. 
 Etym. Ill, p. 256.) Varro's passage is: 'Initiis Cereris porci inmolantur, 
 et . . . initiis pacis foedus cum feritur, porcus occiditur, et nuptiarum 
 initio antiqui reges ac sublimes viri in Etruria in coniunctione nuptiali nova 
 nupta et novus maritus primum porcum inmolant Prisci quoque Latini 
 etiam Graeci in Italia idem factitasse yidentur. Nam et nostrae mulieres, 
 maxime nutrices, naturam qua feminae sunt in virginibus appellant 
 porcum, et graecae choeron, significantes esse dignum insigne nuptiarum". 
 The porcus of the wedding ceremony should be the male, as the deity who 
 presided over the confarreatio was Jupiter (see Wiss. pp. 118-119). The 
 initia Cereris, which Varro mentions, were undoubtedly the Eleusinian 
 mysteries, and sacrifices at the Eleusinia were naturally celebrated accord- 
 ing to Greek ritual. The pig had an intimate connection with Demeter, 
 goddess of earth fertility (see Andrew Lang in The Nineteenth Cen- 
 tury, Apr. 1887, p. 562), just as it had with the fertility of females. The 
 whole thought, therefore, in these cases mentioned by Varro, is rather 
 of the pig species, and although in at least two of the three sacrifices which 
 he mentions the victim was actually male, he probably employed porcus 
 as an epicene without thought of sex. 
 
52 
 
 of sex seems decidedly weak, and I am disposed to see in it an 
 epicene usage closely analogous to that of lacertae, lacerti in 
 Pliny's account of the lizards. 
 
 Thus we see that porous and agnus are used as epicenes ; 75 that 
 in the antiquated speech which survived in ritualistic formulas, 
 porous and agnus are used as possessing common gender; 76 that, 
 therefore, porous and agnus are used of both sexes interchange- 
 ably. Porca is also used as an epicene interchangeably with 
 porous. 77 Agna is an epicene in Fasti, IV, 648, 78 and is defined 
 with remarkable precision in a ritualistic inscription, 79 as though 
 there could be doubt about the sex. In the light, therefore, of 
 Varro's plain statement that columba was used as an epicene, I 
 believe that he so employed agna in his account of the sacrifice 
 to Jupiter at the auspicatio vindemiae, and had in mind no thought 
 of the animal's sex. 80 Since, therefore, we have seen that Roman 
 pontifical law demanded^ the sacrifice of male victims to male 
 divinities, there is no reason to doubt that the agna of the 
 auspicatio vindemiae was a male. 
 
 The second apparent exception to the general rule of sacrifice 81 
 has thus been shown to be nothing more than apparent. The first, 
 an offering of a capra to Vediovis, 82 is easily disposed of in the 
 same manner. Cato, quoted by Varro, 83 employs the word as an 
 epicene in speaking of the habits of wild goats. Varro himself 
 goes further and uses caprae of wild and domestic goats alike, in 
 a close connection with the epicene word oves: 
 
 " See notes 71 and 72. 
 
 76 Festus, pp. 286, 222, and 189 ; Paul. p. 6 ; see pp. 49-50 above. 
 
 77 Paul. p. 235 and see p. 51 above. 
 
 78 Can praecidanea agna defined by Festus (p, 223) be epicene? True 
 it occurs in the same passage with the definition of praecidanea porca, 
 which Festus says was an offering to Ceres, but the words, "praecidanea 
 agna vocabatur quae ante alias caedebatur", do not seem to me to point 
 to any single divinity. If this be true, the offering might be to a god as 
 well as to a goddess, and alias could well mean alias hostias. 
 
 TO See n. 69 above. 
 
 80 See n. 10 above. 
 
 81 Cf. pp. 34-35 above and n. 10. 
 
 82 p, 35 above. Gellius sas : "Eum deum (i. e. Vediovem) plerumque 
 Apollinem esse dixerunt; immolaturque ritu humano capra eiusque ani- 
 malis figmentum iuxta simulacrum stat". For Krause on this sacrifice, see 
 his thesis, p. 21. 
 
 M R. R. II, 3, 3. 
 
53 
 
 "In Originum libro Cato scribit haec: 'in Sauracti et Fiscello 
 caprae ferae stint, quae saliunt e saxo pedes plus sexagenos.' 
 Oves enim quas pascimus ortae sunt ab ovibus feris: sic quas 
 alimus (caprae) a capris feris ortae. . . . Quidam etiam dant 
 operam ut ex insula Melia capras habeant, quod ibi maximi ac 
 pulcherrimi existimantur fieri haedi." 
 
 Isidorus, likewise, so employs the word in a passage quoted 
 before : 84 
 
 "Discretio est autem inter armenta et greges: nam armenta 
 equorum et bourn sunt, greges vero caprarum et ovium." 
 
 Finally, in the inscription recording the Ludi Saeculares of 17 
 B. C., caprarum is limited by feminarum, as though the noun were 
 insufficient in itself to establish the sex, a treatment which is 
 identical with that of agnarum feminarum in the same passage. 85 
 
 If we return from this long digression to prove that an epicene 
 usage existed for agna and capra in prose, and glance once more 
 at Ovid's line (P. I, 56) : 
 
 "Idibus alba lovi grandior agna cadit," 
 
 it will be recalled that agna might have been written for stylistic 
 purposes. 86 Now, however, the point has been reached where 
 it will readily be perceived that the poet's use of agna rests on a 
 more solid foundation than mere euphony. 
 
 The fact is that, while both Vergil 87 and Ovid doubtless selected 
 the forms porca and agna with a metrical and stylistic purpose in 
 view, yet their choice of the feminine form for animal names 
 was in perfect accord with the epicene usage in prose discussed 
 and practiced by Varro, 88 and so did no violence to established 
 principles of the language. Without doubt, the two poets gave 
 no thought whatever to the actual sex of the animal. There are 
 many examples 89 in the poets of an epicene use of the forms 
 terminating in -a, one of the most striking being Ovid's line above 
 quoted (P. IV, 648) : 
 
 "Agnaque nascendo saepe necabat ovem," 
 but before closing it will be sufficient to call attention to one or 
 
 "Orig. XII, 1, 8; see pp. 43-44. 
 
 M See p. 50 above, and n. 69. 
 
 *" See above pp. 45-46. 
 
 m Aen. VIII, 641 ; see pp. 45-46 above. 
 
 88 L. L. IX, 55 ; see p. 40 above. L. L. VI, 16. 
 
 w Cf. Virg. Ed. II, 21 ; Stat. Theb. VII, 397; et al. 
 
54 
 
 two more passages which illustrate Ovid's use of the epicene. In 
 the first of these, a true epicene, ovis, a common gender noun, bos, 
 and a first and second declension noun with masculine termina- 
 tion for male and feminine for female, caper and capella, are 
 employed as epicenes, for the poet is relating the origin of the 
 sacrifice of the several animal kinds, irrespective of sex (F. I, 
 357-384) : 
 
 " 'Rode, caper, vitem ; tamen hinc, cum stabis ad aram, 
 
 In tua quod spargi cornua possit, erit.' 
 Verba fides sequitur. Noxae tibi deditus hostis 
 
 Spargitur adfuso cornua, Bacche, mero. 
 Culpa sui nocuit, nocuit quoque culpa capellae: 
 
 Quid bos, quid placidae commeruistis ovesf 
 Quid tuti superest, animam cum ponit in aris 
 
 Lanigerumque pecus ruricolaeque boves?" 
 In these lines, although there is some thought of the masculine 
 in caper (1. 357), caper and capella (1. 361) are used interchange- 
 ably, and capella refers to the entire goat kind without sex dis- 
 tinction. This interchangeable use of caper and capella is also 
 found in Ovid's long, aetiological account of the L,upercalia (F, 
 II, 267-452), where capella in the earlier line (361: "Cornipedi 
 Fauno caesa de more capella") is used as an epicene, but line 441 
 of the later story, 
 
 " 'Italidas matres/ inquit, 'sacer hircus inito,' " 
 proves that Ovid knew the victim (caper, 90 capella) to be a male. 
 The Lupercalia passages constitute a perfect parallel to those on 
 the sacra Idulia. 91 
 
 Thus we have seen that Ovid's two statements regarding the 
 ovis Idulis are perfectly consistent, no matter whether he says 
 agna or semimas ovis. It was the male, not the perfect male, but 
 the vervex; for, as has been shown, 92 the Ada Fratrum Arvalium 
 prescribe the sacrifice of verveces to Jupiter, and Ateius Capito 
 in his first book on sacrificial law (quoted by Macrobius III, 10) 
 says : "lovi tauro, verre, ariete immolari non licet." Moreover, 
 Ovid would hardly have made so specialized a statement as 
 semimaris, had he not read the word or its equivalent, vervex, 
 in his source. 
 
 11 "Ille caprum mactat". 
 81 Capella = caper: agna = semimas ovis. 
 02 See n. 13. 
 
381669 
 
 UNIVERSITY OF CALIFORNIA LIBRARY