In the third chapter some passages are examined which had been first translated from the Greek by a Roman poet, and afterwards handled afresh by Virgil. The only one of these which Servius notices is Aen. II 492 foll., and this is only to mention the parallel passage in Homer.1

VI.

In the eighteenth and following chapters of the fifth book of the Saturnalia Macrobius has elaborate comments on passages in which Virgil is said to have drawn upon recondite Greek sources. Let us briefly compare these with the corresponding notes, where there are any, in Servius.

Georg. I 17, pocula Acheloia: Servius (Dan.) has a very brief extract of these remarks.

Aen. vii 689, vestigia nuda sinistri Instituere pedis: Servius says merely 'traxit hoc a Graeciae more.'

Aen. 4 fin., nondum illi flavum, etc. Servius (Dan.), again merely abridging, says, 'Euripides Alcestim Diti sacratum habuisse crinem dicit, quod poeta transtulit ad Didonem.'

Aen. IV 513, falcibus ahenis: Servius is silent.

Aen. Ix 584, ara Palici: the same story is mentioned by Servius. Georg. I 100, umida solstitia atque hiemes orate serenas, etc. Here the whole note of Servius (Dan.) is virtually identical with that of Macrobius, though not so clear or accurate. The paraphrase in Macrob. v XX 14, 'cum ea sit anni temperies, ut hiemps serena sit, solstitium vero imbricum, fructus optime proveniunt,' is identical in both commentators : and both also quote the rustic verse, 'hiberno pulvere, verno luto, grandia farra, Camille, metes.'

Georg. IV 380, Aen. 111 66: carchesia, cymbia. These comments are not in Servius.

Aen. XI 532, Opis. Servius (Dan.) has the same words about Alexander Aetolus.

Aen. 1 42, ipsa Iovis rapidum iaculata e nubibus ignem. This note is

not in Servius.

Georg. III 391, munere sic niveo lanae. Not Servius but Philargyrius has this comment in a shorter form: huius opinionis auctor est Nicander: nec poterat esse nisi Graecus.'

The conclusion which I draw from this comparison is again that Servius, Philargyrius and Macrobius are drawing upon the same source. And that this source was one work, not several, is, I think, rendered probable by the uniformity of style which characterizes the whole of

1 Ennius is often quoted by Servius and the later commentators in illustration of points of grammar or language, and so it is with other older Latin poets. Although the literary debt of Virgil to Lucretius was fully recognized by the ancient critics (Gellius 1 xxi 7), Servius generally quotes Lucretius only for the purpose of illustrating points of grammar or philosophy.

2 Or rather his ancient interpolator. I do not however think it necessary to suppose that this writer is borrowing directly from Macrobius, as in other places he either ignores him or is quite independent of him. (See note on p. liii below.)

these notes as given in their fuller form by Macrobius. Add Macrob. I iii 10 on torquet medios nox umida cursus (Aen. v 738).

VII.

I now come to a number of remarks in the third book of the Saturnalia, in which Virgil's knowledge of religious antiquities is discussed. As before, I shall compare Macrobius and Servius on each note.

Macrobius III 1: this note, on purification by a running stream, is abridged in Servius on Aen. iv 635.

III Aen. v 237: porriciam and proiciam: again abridged in Servius.

III ii 6
III i 7

voti reus: 'vox propria sacrorum est,' etc. : So Servius. Aen. IV 219: aram manibus apprehendere, ara and asa. note is in Servius (Dan.) on Aen. IV 219 and VI 124. vitulari; laetum paeana: Aen. vi 657: Servius is silent. III i 15: faciam vitula: Ecl. 111 77: Servius in a note independent of Macrobius says, 'ut faciam ture, faciam agna.'

Much of this
III i 10

III ii 17: Aen. I 373, et vacet annales nostrorum audire laborum : Aeneas pontifex: Servius (Dan.) has the same note, but in a fuller form. III iii 2: sacrum, sanctum, profanum: Servius (Dan.) has the gist of this note on Aen. XII 779.

III i 8 religiosus, religio: Servius (Dan.) has the same note, but without mentioning Festus, on Georg. 1 269.

III iv 1 foll.: delubrum: Servius (Dan.) on Aen. Iv 56 has the same quotation from Varro, and on II 225 he quotes another note from Masurius Sabinus (Dan.).

III iv 6: Penates: so Servius (Dan.) on Aen. I 378, III 119, II 296, 325, III 12, 134.

III VI: hostiae: so Servius on Aen. 11 119 (Dan.), III 231, 456, v 483.

III V 4: litare: so Servius on Aen. 11 119 (Dan.).

III V 7: ambarvalis hostia: so Servius on Ecl. III 77, V 75, Georg. 1 345.

III V 8: invita hostia: so Servius on Georg. II 395, Aen. Ix 627.
III V 9: contemptor divum MezentiusThese notes are not in Servius.
III vi I : ̓Απόλλων γενέτωρ

III vi 9: Hercules victor: so Servius (Dan.) on Aen. vIII 363.

III vi 12: domus Pinaria: so Servius (Dan.) on Aen. vIII 270.

III vi 16: sedili: so Servius, but shortly, on Aen. viii 176.

III vi 17: aperto capite: so Servius on Aen. 111 407.

III vii : Pollio: so Servius (Dan.) on Ecl. IV 43, nearly word for word.

III vii 3 foll. telisque sacrarunt Evandri: so Servius (Dan.), partly word for word, on Aen. x 419.

III viii : ducente deo: Servius (Dan.) has a note of nearly equal fulness on Aen. 11 632, with a passage from Sallust which is not in Macrobius.

III viii 4 in astris: Servius is here silent.

III viii 6: Camille: so Servius (Dan.), word for word, on Aen. 11 543.

III viii 8 mos: on Aen. vII 601, Servius has a note quite independent of this, and indeed says that Virgil is not correct in his facts.

III ix 1: excessere omnes, etc.: this is abridged by Servius (Dan.) on Aen. II 351.

To these criticisms may be added the following remarks in Servius : Aen. IV 29 (Dan.): 'sane caerimoniis veterum Flaminicam nisi unum virum habere non licet, quod hic ex persona Didonis exequitur . . . nec Flamini aliam ducere licebat uxorem, nisi post mortem Flaminicae uxoris, quod expeditur quia post mortem Didonis Laviniam duxit.'

Aen. IV 103 (Dan.): 'sciendum tamen in hac conventione Aeneae atque Didonis ubique Vergilium in persona Aeneae flaminem, in Didonis flaminicam praesentare.'

Aen. IV 137 (Dan.): 'veteri caerimoniarum iure praeceptum est ut flaminica venenato operta sit.' A long note follows on the dress of the flaminica.

Aen. IV 166: 'prima et Tellus: satis perite loquitur. Nam secundum Etruscam disciplinam nihil tam incongruum nubentibus quam terrae motus vel caeli dicitur. Quidam sane Tellurem praeesse nuptiis tradunt, nam et in auspiciis nuptiarum vocatur,' etc. There is more of the same kind in the notes on Aen. IV 262-3 (Dan.), 339 (Dan.), 374, 518 (Dan.), 646 (Dan.); VI 210 (Dan.); VII 190; VIII 106 (Dan.), 363 (Dan.), 550, foll. (Dan.); x1 76 (Dan.).

But in the tenth, eleventh and twelfth chapters of Macrobius' third book we have some hostile criticisms in the style of which so many specimens have been already quoted under other heads. On Aen. III 21, it is remarked: 'Ecce pontifex tuus apud quas aras mactetur ignorat, cum vel aedituis haec nota sint et veterum non tacuerit industria.' The attack is replied to ; and both attack and reply are abridged by Servius on Aen. III 21 as follows: 'contra rationem Iovi taurum sacrificat . . . ubique enim Iovi iuvencum legimus immolatum . . . adeo ut hinc putetur subsecutum esse prodigium.'

The

Macrobius III xi 1: miti dilue Baccho; in mensam laeti libant. attack and reply are given in a shorter form in Servius on G. 1 344 and Aen. vIII 279 (Dan.), 'quaeritur sane cur in mensam et non in aram libaverint,' etc. But Servius has not the remarks on mitis and on mulsum in §§ 9 and 10.

III xii : Aen. VIII 285: on this alleged geminus error of Virgil Servius (Dan.) has a note in substance much the same as that of Macrobius.

III xii 10: Aen. IV 57: Virgil is said 'toto caelo errasse cum Dido sua rem divinam pro nuptiis faceret Legiferae Cereri, etc. Et quasi expergefactus adiecit Iunoni ante omnes,' etc.

Serv. A. IV 57 (Dan.): ‘Alii dicunt hos deos quos commemoravit nuptiis esse contrarios, Cereremque propter raptum filiae nuptias execratam, etc., etc. Male ergo invocat hos Dido, quae sibi nuptias optat Aeneae,' etc. The note is very long and full, but I suspect that

Macrobius, a fragment only of whose comment remains, has more of the original wording.

Compare also Macrobius I xv 10, with Servius on Aen. vIII 654 ; Macrob. I xvii 4, with Serv. on Aen. 1 8.

The result of the foregoing comparisons between Servius and Macrobius is this: that in the great majority of cases where Servius and Macrobius have identical notes, those of Macrobius are far the fuller, clearer, and more logical; that in the collections of parallel passages from Homer Macrobius has some which Servius has not, Servius many more which Macrobius has not, and there are many in common. Hence the natural inference is not (as Ribbeck thinks) that Macrobius was using a fuller form of the actual commentary of Servius than that which we now possess, but that both Macrobius and Servius were drawing upon older commentaries and criticisms.

Is it possible to say with any degree of certainty to whom these works or any of them can be assigned?

Taking the hostile criticisms in Macrobius and Servius first, with the exception of those which can with certainty be assigned to Cornutus and Hyginus (see pp. lvi, lvii), I would observe that there are a number of precisely the same character and often worded in the same vigorous and acrimonious style; I mean those which deal chiefly with minute points of logic or narrative and less often with points of expression. Such are (1) the unfavourable remarks upon the order of the narrative in the Aeneid (p. xxxv foll.); (2) those in which Virgil is blamed for want of invention in his incidents, or for observing a wrong order and adopting an artificial style in his catalogues, or for forgetful repetitions of the same name, or inconsistency in his narrative, or divergence from Homer, or false taste, or bad mythology, or other minor faults akin to these (pp. xxxvii-xlii); (3) those in which Virgil is declared to have fallen below Homer in similar and other passages borrowed from him (pp. xlii, xliii); (4) those in which he is charged with ignorance of religious antiquities (p. xlvii foll.).

Now if I am right in saying that these criticisms are expressed in the same venomous but idiomatic style; if it be true, as it is so far as I have observed, that they are all directed against passages in the Aeneid (the only exception is an apparent one, Macrob. III xi 1, where Georg. 1 344 is quoted; but this is instantly followed by a line from the eighth Aeneid in octavo)—it is natural to infer that they come from the Aeneidomastix of Carvilius Pictor, which is quoted by Servius on Aen. V 521.

Besides this, two other works of hostile criticism are mentioned by Suetonius: the vitia of Herennius and the furta of Perellius Faustus.

It is possible, though I do not like to say more, that the criticisms quoted on p. xxx foll. were taken from the work of Herennius. As to the furta, it is very difficult to pronounce with any amount of assurance what was the scope and extent of the work. It may or may not have included collections of Virgil's plagiarisms from Homer and the Greeks, as well as of passages taken from Latin authors. But I am inclined

in any case to suspect that the passages from Latin authors collected in the sixth book of Macrobius came directly or indirectly from this work. It is remarkable that in this book there are apologetic remarks on the propensity of the ancient writers to steal from one another: 1 3, 'exprobrantibus tanto viro alieni usurpationem, nec considerantibus hunc esse fructum legendi, aemulari ea quae in aliis probes,' etc. Compare VI ii 33, nec Tullio compilando, dummodo undique ornamenta sibi conferret, abstinuit': a hostile remark admitted inadvertently, as so often, by Macrobius into a context where it is out of place. Now these general remarks about plagiarism would have been better in place at the head of the passages from Homer collected in the fifth book: and I am tempted therefore to suppose that they were suggested by observations on this question which Macrobius found in the works from which he got the instances quoted in Book vi. This work may or may not But it seems in any case to

have been the furta of Perellius Faustus. have been a work which Servius did not much use, for (except in the case of Ennius) he quotes from Latin authors mainly for the purpose of grammatical, or historical, or philosophical illustration.

The passages of neutral tone, in which Virgil's obligations to Homer are simply pointed out, it is natural to assign to the oμоóτnтes of Octavius Avitus; whether this is also the case with the passages in which Virgil is said to have drawn upon recondite Greek sources is, I should think, doubtful, nor am I at present able to offer any hypothesis on this point.

Turning to the passages where Virgil is defended against hostile criticism, it is natural to suppose that when his alleged plagiarisms from Homer, or alleged mistakes or want of management in his narrative are in question, the ultimate source of the notes both in Servius and Macrobius is the work of Asconius Contra obtrectatores Vergilii.

It is less easy to conjecture what were the sources of the minute verbal criticisms on which we dwelt at length in previous pages; but there is considerable presumption that some of them at least are as old as Verrius Flaccus. I have drawn out the following lists with a view of eliciting the points common to Macrobius with Nonius, Festus, Gellius, Servius, Philargyrius, and the Verona scholia.

Macrobius vi iv 2, addita: adfixa et per hoc infesta. Hoc iam dixerat Lucilius in libro XIV his versibus 'Si mihi non praetor siet additus atque agitet me.'

Servius A. VI 90, additus: est autem verbum Lucilii.

§ 3. Vomit undam: agmen of a river. These notes are only found in Macrobius.

§ 5. Crepitantibus flammis. Macrobius illustrates only from Lucretius. Nonius, p. 255, quotes the passage in Virgil in a note on crepare. § 6. Ferreus hastis Horret ager.

Macrobius illustrates from Ennius. Serv. A. XI 601: Horret, terribilis est: est autem versus Ennianus vituperatus a Lucilio dicente per irrisionem, eum debuisse dicere horret et alget.

Tremulum lumen. Macrobius only.