§ 8. Umbraculum. Macrobius illustrates from Varro and Cicero (de Legibus and Brutus). Servius E. IX 41 (Dan.) has a different quotation from Cicero, umbraculisque silvestribus.'

§§ 9, 10, 11. Transmitto, defluo, discludo. Macrobius only.

§ 12. Deductus. Macrobius says deductum pro tenui et subtili eleganter positum est, illustrating from Afranius, Cornificius, and Pomponius. Schol. Veron. E. vI 5, deductum carmen, tenue, gracile, subtile. Serv. ib. deductum tenue: translatio a lana, quae deducitur in tenuitatem. Nonius, p. 289 (s.v. deducere), deductum dicitur molle et suave: Vergilius Bucolicis . . . 'deductum dicere carmen.' The expression deductum carmen is praised by Quintilian VIII 2, as 'proprie dictum, id est, quo nihil inveniri possit significantius.'

§ 14. Proiectum. Macrobius illustrates from Sisenna, and quotes the subst. proiectus-ūs from Lucretius: si secundum veteres, porro iacta. Nonius, p. 373, s.v. proicere: proiectum, longe iactum, extensum : M. Tullius de Signis 'sed quod erat eiusmodi loco, atque ita proiecta in altum.' Vergilius Aeneidos lib. III 'proiectaque saxa Pachyni Radimus;' and other instances are given from Cicero. Servius, A. III 699 (Dan.), proiecta, porrecta, extensa, ut (A. x 587) 'proiecto dum pede laevo,' a passage quoted by Macrobius.

§ 16. Tempestiva pinus. Macrobius only.

§ 17-22. Greek words.

§ 17. Lychnus. Macrobius quotes Ennius, Lucretius, and Lucilius. Serv. A. 1 776, lychni; Graeco sermone usus est, ne vile aliquid introferret.

19. Aethra. Illustrated only by Macrobius.

§ 20. Daedala Circe. This note I have shown (pp. liv, lv) comes from Verrius Flaccus.

§ 21. Reboant. Macrobius illustrates from Lucretius: Nonius, p. 79, s.v. bount, quotes the passage under discussion, G. III 223, illustrating also from Pacuvius and Varro, and remarking bount a boum mugitibus. Servius, G. III 223, says of reboo, est autem Graecum verbum. Nam apud Latinos nullum verbum est quod ante o finalem o habeat, excepto inchoo; quod tamen maiores aliter scribebant, aspirationem interponentes duabus vocalibus, et dicebant incoho. Festus, p. 30, boare, id est clamare, a Graeco descendit, p. 107, inchoare videtur ex Graeco originem trahere, quod Hesiodus omnium rerum initium esse dixerit chaos: see on cohum, p. 39. Diomedes, p. 365 K., inchoo inchoavi: sic dicendum putat Iulius Modestus, quia sit compositum a chao, initio rerum. Sed Verrius Flaccus in postrema syllaba aspirandum probavit : cohum enim apud veteres mundum significat, unde subtractum incohare.

It would seem from these notices that two etymologies were suggested for boo and incoho, a Latin (boves, cohum) and a Greek one (Boāv, chaos); and I should be inclined to infer that both words were discussed fully by Verrius Flaccus, from whom Macrobius may directly or indirectly have derived his note.

§ 22. Pausa. Nonius, p. 158, illustrates this word from Accius and Lucilius.

The remaining Greek words are discussed in Macrobius only.

23. Foreign words. Urus. Macrobius only.

Camurus. This note, as I have shown (see p. lv), comes, or may come, from Verrius Flaccus.

Macrob. vi v 3. Epithets. Petulcus: this note (see p. lv) may come from Verrius Flaccus.

§ 4. Liquidus, as an epithet of fire. Macrobius illustrates from Lucretius, adding liquidi simul ignis pro puro vel lucido, seu pro effuso et abundanti. Servius, E. vI 33, liquidi simul ignis, puri, id est aetherei ; (Dan. adds) quem Cicero ignitum liquorem dixit. Lucretius, 'devolet in terram liquidi color aureus ignis.' A. v1 202, liquidum (aëra) pro puro dixit. Nonius, p. 334, has a long note on liquidus, which he explains as suavis or dulcis, purus, mollis or fluxus. The three notes all seem to come from the same source, which is probably not later than the age of Trajan (see p. lxvi foll.).

$ 5. Tristis = amarus. Macrobius illustrates from Ennius: so Servius (Dan.) on G. 1 75: comp. Servius and Philargyrius on G. 11 126. Nonius, p. 409, triste, amarum: Vergilius Georgicorum lib. I 'tristisque lupini Sustuleris fragiles calamos.'

§ 6. Auritus (auritos lepores, G. 1 308). Macrobius illustrates the word from Afranius, aurito parente.' Servius, G. I 308, auritos, maiores habentes aures. Horatius aliter aitDoctum et auritas fidibus canoris Ducere quercus,' sensum audiendi habentes: comp. Nonius, p. 129, inauritum, quod non audiat. Festus, p. 8, auritus a magnis auribus, ut sunt asinorum aut leporum. It may be that Verrius Flaccus in his original note had quoted both Afranius and Virgil.

§ 7. Turicremus (Aen. Iv 453, 'turicremis aris'). Macrobius illustrates from Lucretius: Servius (Dan.) says 'nomen mire compositum.'

§ 10. Velivolus. Macrobius quotes Livius Andronicus and Ennius: Serv. A. I 224, has a note nearly identical with his, but quoting Ennius only.

§ 11. Vitisator. Macrobius illustrates from Accius: Serv. A. vII 179, says vitisator: non inventor vitis, sed qui vitis genus demonstravit Italis populis.

§ 12. Noctivagus. Macrobius illustrates from Egnatius. Servius, A. x 206 (Dan.), says 'nomen mire compositum.'

§ 13. Nubigena. This word (like arcitenens and silvicola) is discussed by Macrobius only.

Before leaving these lists I would call attention to the fact that they show signs of having been extracted from alphabetical series: additus, agmen, crepito, horrere, tremulus, umbraculum (transmitto): defluo, discludo, deductus, proiectus, tempestivus: (lychnus) aethra, daedalus, reboo: camurus, Mulciber, petulcus: (liquidus, tristis, auritus :) turicremus, velivolus, vitisator: arcitenens, silvicola: noctivagus, nubigena. There is also a slight tendency to put words from the same authors together: thus agmen, crepito, horreo, tremulus, are all illustrated from Ennius; so lychnus and aethra; daedalus and reboo, petulcus and liquidus, from Lucretius: arcitenens and silvicola from Naevius.

These facts alone might fairly lead us to suspect that Macrobius is drawing upon glosses or philological works of respectable antiquity.

But the suspicion becomes something stronger when we find that some of the notes are traceable to Verrius Flaccus (daedalus, camurus, petulcus, auritus, and perhaps reboo), that others are common to Macrobius and Nonius, and others again to those two writers, with Servius and other later commentators. For I have endeavoured to show further on (p. lxiv foll.) that the Virgilian notes which are common to Nonius and the later commentators cannot be assigned to a later date than the age of Trajan. And the conclusion to which we are led in the case of the scholia, whose origin we can directly or indirectly trace, it is natural to extend to those of whose sources we are ignorant.

NOTE.

It was not until after these sheets had been sent to press that I was able to procure two pamphlets, by Drs. Linke and Wissowa, De Macrobii Saturnaliorum fontibus, Breslau, 1880. Dr. Linke, who goes much more fully than Dr. Wissowa into the question of the sources of the Virgilian criticisms in Macrobius, has come to the conclusion (1) that the additional notes in Daniel's Servius are ancient interpolations : (2) that the Servius of our commentary stands in no relation of dependence to the Servius of the Saturnalia; (3) that the ancient interpolators of Servius borrowed, in a great many instances, directly from Macrobius; (4) that there are some cases, nevertheless, where this cannot have been the case: (5) that Macrobius III 1-12 is taken from two different manuals, of uncertain date, each of which probably contained information borrowed ultimately from Verrius Flaccus.

With regard to (1) and (3) I would observe that the additional notes in Daniel's Servius may be interpolations, but that whether they are so or not, they are, in my opinion, taken not from Macrobius, but from a continuous commentary. For (1) they often extend without a break over continuous lines; (2) they sometimes give information which is not found in Macrobius; (3) they sometimes, in a very striking way, ignore what is to be found in him, as notably in the case of his sixth book (see p. xlv-1). I entirely agree with Dr. Linke as to the relation between our Servius and the Servius of the Saturnalia; with regard to Macrobius III 1-12 I am not convinced that he is right, as chapters 10, 11 and 12 may come from the Aeneidomastix.

THE ANCIENT COMMENTATORS ON

VIRGIL.

1. CAECILIUS EPIROTA.

IT was not long before the poems of Virgil began to afford matter for discussion to lexicographers, grammarians, and writers on antiquity. The first scholar who actually lectured upon Virgil was Quintus Caecilius Epirota, for information about whom we are entirely dependent upon Suetonius (De Grammaticis, 16). He was, it appears, a freedman of Pomponius Atticus, the friend of Cicero, and was born at Tusculum. His cognomen suggests that he may have been the child of Epirot parents, brought over, perhaps, from the estates of Atticus in Epirus. The daughter of Atticus was married to Marcus Vipsanius Agrippa, and Caecilius was tutor to this lady. On account of a suspicion which arose against him with regard to his conduct in this relation, he left the family of Agrippa and lived henceforth on terms of intimate friendship with the poet Cornelius Gallus. His character was so unfavourably regarded by Augustus that this intimacy was the occasion of one of the gravest charges brought against Gallus by the emperor. After the condemnation and death of Gallus, Caecilius opened a school for a few young men, to whom he lectured on Virgil and other contemporary poets. Whether this was before Virgil's death or not there is no evidence to decide. A verse written upon him by Domitius Marsus—

'Epirota, tenellorum nutricula vatum,'

seems to be pointed at the real or supposed effeminacy of his character.

2. VERRIUS FLACCUS.

Verrius Flaccus, the compiler of the first Latin lexicon ever written, must have paid a great deal of attention to Virgil. His work De Verborum Significatu has, as is well known, survived only in the abridgments of Festus and Paulus. Even in these, a considerable number of quotations from Virgil is to be found; and I am inclined to think that several of the original glosses of Verrius may be partially reconstructed from later writers, notably from Nonius and Macrobius, who seem to have preserved them in a fuller form than Paulus or even Festus. Thus (1) Paulus has preserved the following gloss on daedalus (p. 68, Müller), Daedalam a varietate rerum artificiorumque dictum esse apud Lucretium terram, apud Ennium Minervam, apud Vergilium Circen, facile est intellegere. Macrobius vi iv 2 remarks that Virgil says

daedala Circe because Lucretius had said daedala tellus. It seems from this that Verrius must have had an article in which the daedala tellus of Lucretius and the daedala Circe of Virgil were quoted together. The case was probably similar (2) with Verrius' article on camurus. Fest. P. 43 says camara and camuri boves a curvatione ex Graeco кáμжŋ dicuntur. Nonius, p. 30, has the following note: camurum obtortum, unde et camerae tecta in curvitatem formata. Vergilius Georgicorum lib. III (v. 55), 'Et camuris hirtae sub cornibus aures.' Commenting on this line Macrobius vi iv 23 says camurus peregrinum verbum est, id est in se redeuntibus. Et forte nos quoque camaram hac ratione figuravimus. Servius, in his note on the passage of the third Georgic, says, camuris, id est curvis. Unde et camerae appellantur, and Philargyrius brings us very near to the gloss in Paulus, camuri boves sunt qui conversa introrsus cornua habent. I conjecture that these remarks all represent parts of a full note in Verrius Flaccus, in which camuri boves, camurae aures, and camera were discussed together. (3) On p. 206 Festus has a note on petulcus which he illustrates from Virgil's fourth Georgic (haedique petulci), from Lucretius, and from Afranius. It is instructive to find that Macrobius, in his comment on the line in the fourth Georgic, also quotes the same line of Lucretius in illustration of the word.

I have little doubt that had the work of Verrius De Verborum Significatu been preserved in its original extent, it would be possible to multiply these examples of comments drawn from articles in his lexicon in which Virgil was quoted. It is much easier to collect instances in which the De Verborum Significatu was used by late commentators for general purposes of illustration. (1) Take for instance the note in Festus p. 298 on the word summussi. Summussi dicebantur murmuratores. Naevius: 'Odi, inquit, summussos; proinde aperte dice, quid siet quod times.' Ennius in sexto Annalium: 'Intus in occulto mussabant,' et Ennius in Andromache . . . Mussare silere est: nam [Iuventius in Anagnorizomene], 'quod potes sile cela occulta tege tace mussa mane.' Philargyrius on Georg. IV 188, mussant: hic murmurant. Quae vox ponitur in tacendi significatione, ut apud Ennium in XVII, non possunt mussare boni qui facta labore Nixi militiae peperere.' Interdum autem pro dubito, ut (A. x11 657) 'mussat rex ipse Latinus, Quos generos vocet.' Mussant autem murmurant. Ennius in x sic ait, 'Expectans si mussaret quae denique pausa Pugnandi fieret.' Serv. A. XII 657 mussat, modo dubitat; Dan. adds, .. Veteres mussat pro timet. Ennius mussare pro tacere posuit. Clodius Tuscus: 'mussare est ex Graeco; comprimere oculos Graeci μvoa dicunt.' And Nonius, p. 427, distinguishes mussare and murmurare.

(2) Paulus, p. 368, on vescus. Vescus fastidiosus. Ve enim pro pusillo utebantur. Lucretius vescum dixit edacem, cum ait 'nec mare quae impendent vesco sale saxa peresa. Gellius XVI v 6 has words to the same effect, but Nonius, p. 186, seems to preserve a better form of this gloss, in which it is clear that Paulus or Festus have confused quotation and interpretation: Vescum minutum, obscurum. Lucilius lib. xxvi 'quam fastidiosum ac vescum cum Falcidio videre.' Vergilius Georgicorum lib. III (175) 'nec vescas salicum frondes.' Afranius in