nature to substantiate the one possible statement in both-that Doson married Philip's mother, Demetrius' widow. This statement, however, is also given by Justin, 28, 3, 9, and bad as Justin is, Trogus did not necessarily get it from Plutarch's source; and I think it is easier to explain the origin of this story if we suppose that it was true in a certain sense: Chryseis did marry Demetrius after Phthia's death and adopted Philip on her marriage. She was his 'mother' when Demetrius died. Doson then had every reason for marrying her, both on account of Thessaly (if she were really a Thessalian) and as the simplest way of securing the succession of her 'son' by adopting him himself; doubtless he had promised this to Demetrius, and it is merely another case of the well-known family loyalty of the Antigonids.

The question here examined is not without a certain historical interest, for it helps to explain Philip V. It has always been obvious that in one way he did not resemble an Antigonid; for all his qualities, he showed outbursts of a temper which paid no heed to consequences provided its object were achieved, and which led to such horrors as that at Abydus-' madness' Polybius calls it. No Antigonid had had a temper of that sort; even cruelty had been foreign to the house since the first Antigonus, and his misdeeds had been, like Cassander's, of the calculated kind-steps to power. Only one house in Hellas, the Epeirot, traditional descendants of Achilles, possessed that temper. It had dominated Olympias; Alexander's will-power could not always restrain it. Even Pyrrhus had shown it once, and lost Sicily in consequence. If we did not now know that there was Epeirot blood in Philip, we might have guessed it from his character.

ETYMOLOGIES.

Απλούς AND διπλούς.

W. W. TARN.

THE generally accepted explanation of the -λoûs (-πλóos) in these words, that it comes from the root pel- 'to fold' (Boisacq, Dict. Etym. s.v. diλóos), fails to account for the presence of the double o in -λóos. May not this -λóos be identical with πλoûs [#λó(F)os] 'voyage'? In other languages a word for 'voyage' or 'journey' has faded into a word for 'time' ('mal,' 'fois ')—e.g. Serb. jedan put 'once' (lit. 'one journey'); Lett. vien-reiz 'once' (lit. 'one journey,' reiz being borrowed from Low German, in which reise had already suffered this change of meaning, cf. Endzelin, Bezzenberger's Beiträge 29. 179 note).

As the Lettish word for simple,' viz. vienkaršs, is based on an adverbial phrase meaning 'once,' which does not survive in Lettish, but appears in Lith. vienąkart, so perhaps an Aoûs 'simple' is based on an obsolete phrase eva λó(F)ov 'once.'

"Epxoμaι AGAIN (cf. C.Q. XV., p. 44).

Acceptance of my contention that epxopa is connected with opxéopa and with ἄρχω (ἄρχομαι) would make it easier to understand how the same word comes to mean in Lithuanian 'to jump' or 'dance' (šókti), and in Lettish 'to begin' (sākt). The identity of šókti and sākt, which correspond phonetically, was suspected by Leskien (Ablaut der Wurzelsilben 374), and has been asserted by Endzelin, Zeitschr. für vergl. Sprachf. 43. 25.

RODERICK MCKENZIE.

ENNIANA.

THE notorious line, 'O Tite tute Tati tibi tanta, turanne, tulisti,' is assigned to Ennius by Priscian, Isidore, and the Explanator in Donatum, as well as by Pompeius in his second quotation of it. In the first, however (287 K), he, like Auctor ad Herennium, Charisius, Donatus, Martianus Capella, and Plotius Sacerdos, gives no author's name. There is thus some reasonable doubt whether the attribution is correct; and this doubt is greatly increased by the remarks of our earliest authority, the Auctor ad Herennium (IV. 12. 18): Vitabimus eiusdem litterae nimiam adsiduitatem, cui uitio uersus hic erit exemplo, nam hic nihil prohibet in uitiis alienis exemplis uti O Tite e.q.s.; et hic eiusdem poetae Quicquam quisquam e.q.s.' This clearly indicates:

1. That it was unusual for this author to borrow his illustrative examples (cf. IV. 2. 4 sqq., where he states that he prefers to make up his own examples instead of quoting from the best prose authors and poets as most rhetoricians do).

2. In the particular case of the uitia, however, he did permit himself to borrow. (Hic is the emphatic word; in uitiis may be a gloss. The Teubner editor brackets hic.) I further suggest :

3. That the poeta from whom the lines are taken is Lucilius. It is well known (see Pompeius, 289. 10 K) that Lucilius did not merely draw up a list of one hundred solecisms, but described or illustrated them. And nothing is more natural than to suppose that those illustrations took the form of parodies of Ennius. It is hardly an exaggeration to say that Lucilius played the part of Aristophanes to the Euripides of Ennius; and such a subject as an essay on 'how not to do it' would give him an opportunity too good to miss (on the whole question, cf. Hardie, Res Metrica, p. 4). These parodies may have been taken seriously by some later grammarians, so that even an attribution to Ennius in our authorities need not make us accept certain monstrosities as genuine. It is to be noted that both Auct. ad Her. and Plotius Sacerdos, neither of whom mentions Ennius, give the line to illustrate a bad form, not simply as a figure as the others do. The metre of the second example in Auct. ad Her. (trochaic tetrameter catalectic) seems to have been a favourite with Lucilius, so there is no need to follow the editors in calling the line a dramatic fragment. A last suggestion which may be made is that with the realization that many anonymous lines which appear in the grammarians and writers on metre, and which have been attributed to Ennius, are in all probability the composition of the critics themselves (see above), we can begin the work of clearing Ennius' text of much that is unworthy of him.

Nonius 370 M: Passum, dispersum solutum: Vergilius Aen. lib. I. crinibus e.q.s. : Terentius in Phormione Capillus e.q.s.: Caecilius Synaristosis Heri uero e.q.s. Passum, extensum, patens ... Ennius Annalium lib. X. Aegro corde comis passis late palmis pater . . . passis ait palmis patentibus et extensis.' Of the numerous attempts to make a hexameter from the Ennius quotation, some are too far from the MSS. and others give an absurd sense. A glance at the context, however, shows that Nonius has already dealt with the meanings of passum, in which the word could be applied to comae; and his note following our line shows clearly that it is intended to illustrate the second meaning only, which applies well to palmae but not at all to comae. I therefore suggest that comis should be omitted as a mistaken reminiscence of the previous section (or possibly as the remains of a note intended to remind the reader that comis passis was a phrase that might be used as well as crinibus or capillis passis), and pater placed where comis now stands. Having been ousted from its proper position by the wrongly inserted comis, it was placed later in the quotation when its loss was discovered. I further suggest that pater, which seems to be invariably regarded as vocative, is nominative, and that the whole fragment might be restored:

aegro

corde pater passis late palmis <lacrumatus >.1

The subject of Book X. was undoubtedly the Macedonian Wars; and a very natural reference for our line would be to the protests or farewells of a father whose son was included among the hostages demanded by Flamininus after the battle of Cynoscephalae. Valmaggi (Q. Enni. I Frammenti degli Annali, p. 100) also suggests that the error centres in comis. His conjecture comes is not easy to understand, but his further remarks are perfectly sound. With the removal of comis all necessity to refer the fragment to a woman disappears, and carries with it whatever justification may have been claimed for L. Müller's theory that the line describes Sophonisba, and that, therefore, the narrative of the Second Punic War was continued into Book X.

Annales VIII.: Tibia Musarum pangit melos (Schol. Bern. ad Georg. IV. 72). This fragment is usually interpreted as a reference to either:

1. The Ouatio granted to M. Marcellus on his return from Sicily in 211 B.C. (Vahlen), or

2. A council of the gods or banquet in Olympus.

The first of these is greatly preferable to the second; but it does not take account of the word Musarum, which would be out of place in a description of the flutes of the Ouatio (see especially Plutarch, Marcellus XXII., where there is an elaborate contrast between the Ouatio and the Triumphus). And there is no adequate reason for assuming that the Ennius passage contained a specific contrast between the music of war (which appears in Vergil) and the music of peace (which the Scholiast illustrates from Ennius) of the sort indicated in Plutarch (l.c.). The other interpretation is supported whole-heartedly by Pascoli (Epos, p. 40), and not rejected by Valmaggi (op. cit., p. 86). But Pascoli, who quotes Il. I. 602 sqq. in support of it, misrepresents a perfectly simple passage, seemingly from his preconception that Musarum depends on tibia. I suggest, however, that this is unnatural; the only Muse who seems to play the tibia is Euterpe, the others using the barbiton or some other form of the lyre. The phrase Movo(á)wv péλos is actually found in Greek, and it is natural to take Musarum melos together in Ennius in the sense of ' melodious or poetic song.' This will enable us to find in the fragment a uniquely interesting reference to the hymn Ad Iunonem Reginam, composed in 207 B.C. by the veteran pioneer Livius Andronicus. It was most probably sung cum tibicine, like the early hymns in honour of the heroic dead, when it was performed by the choir of girls; hence the tibia of the text. The word pangit in this metaphorical connexion is most commonly used of the author; but the not unnatural transference to the instrument is helped by the fact that melos can mean either the words or the tune. The words may form part of a pium uotum addressed to Juno herself, and perhaps introduced by the line Optuma caelicolum, Saturnia, magna dearum, which is usually placed near them. The goddess' favour would thus be solicited by drawing her attention to the service being performed in her honour. Or the fragment may belong to the body of the

1 The suggestion of Vahlen (Enn. Poes. Rel. II., p. xxxix) that Nonius is in the almost invariable habit of quoting whole lines of Ennius, whether they make complete sense or not, seems to go beyond the facts. There are actually a number of cases where portions of lines are quoted by him from the Annales, both with and without other complete lines. Especially interesting for our present purpose are 197 M. 10, Saturno Quem Caelus genuit, and 66 M. 22, Rastros† dentefabres † capsit causa poliendi Agri, in both of which a single word from an incomplete line is preserved.

There seems, therefore, little difficulty in the arrangement Aegro Corde pater e.q.s. suggested

above.

2 If the conclusions of Cichorius (Römische Studien I.) that Livius was not the author of the hymn of 207 are accepted, the suggestions of personal interest made above will, of course, be invalidated. But even so, the general reference will remain the same, and the fragment can be understood of the special religious and literary event of that year.

narrative. But in any case some reference to Livius and his hymn was practically inevitable. In a very real sense the occasion was the most important in the early history of Latin poetry. The popular estimate of the hymn must have been very greatly heightened by the dramatic suddenness with which success followed it. For the first time literature was officially recognized by the community, for the first time its efforts were publicly rewarded (by the foundation of the famous but little understood Collegium Poetarum, with the Temple of Minerva on the Aventine as its meetingplace). Ennius must, from his residence on the same hill, have been thoroughly familiar with the Collegium. He was intensely aware of the dignity of the poet's calling; above all, he can hardly have failed to feel the warmest personal interest in the career of Livius, a native of the very town in which he received his education, the pioneer who may have died in the very year in which the younger poet came to Rome. It may well be that it was the example of Livius which first inspired Ennius to become a poet. His first known works are dated to 240 B.C., the year before Ennius' birth, his last to 207 B.C., three years before Cato brought Ennius to the capital. He was thus engaged in active composition throughout the entire period of the formation of Ennius' character and talents; and his fame, some reports of which no doubt reached the neighbourhood of his old home, must have powerfully impressed the young' Rudinus.' On all these grounds may we not see in our present fragment a reference to the poet's great predecessor?

=

=

Annales III.: Ostia munitast. Idem loca nauibus celsis Munda facit nautisque mari quaesentibus uitam. Munda instructa, ornata, seems, as a rule, to be the note of the editors, who further explain nauibus and nautis as ablatives constructed with munda. An interpretation so extraordinary could hardly maintain itself did the scholars who repeat it translate their own words into the vernacular. The meaning cannot be anything but 'He supplied the place also (or decorated it) with ships and seafaring men.' No parallel for munda instructa is quoted; the interpretation seems to be based on the secondary meanings of the nouns mundus and munditiae. I suggest that munda is used in a sense which, if not quite literal, is readily derived from the literal. The meaning is that Ancus cleaned out the river mouth and made it ready for ships and sailors; that is, we are dealing with a description of dredging operations to provide a good harbour for the new colony. It is notorious that the channel was continually liable to become silted up by the quantity of solid material which the Tiber carried down. By the time of Sulla continued neglect had practically ruined the approaches; and before Augustus established his empire merchantmen had almost ceased to attempt the dangerous task of putting in. Even after the establishment of the new Portus Urbis, it was found safer to convey goods by road from it to the Emporium at the foot of the Aventine rather than trust anything bigger than an ox-drawn barge to the choked and dangerous river-channel. Munda then means 'cleaned' or 'prepared by cleaning,' and hence ready for service' (a force which may suggest an origin for the phrase in mundo esse or habere); with this explanation nauibus and nautis are of course datives. Festus quotes the lines to illustrate the very point of the meaning of mundus. Unfortunately his own text is badly damaged, but the epitomizer supports my explanation, Mundus quoque appellatur lautus et purus (143 M), words both of which mean properly 'freed from dirt or foreign admixture.' A reasonable suggestion is that the note stated that mundus could be used (of a person) to mean "clean or "well-washed "; and it may well have added that the word could also be used by extension of a place that had been cleaned or put in order.' Cicero uses terra pura (Sen. XVII. 59) of ground that has been cleared of stones and weeds. In the same way munda facit means 'he clears out the silt.' ETHEL MARY STEUART.

THE UNIVERSITY,
EDINBURGH.

THE MANUSCRIPTS OF THE METAMORPHOSES OF APVLEIVS.

I.

THE later MSS. of the Metamorphoses1 of Apuleius' have received little attention. Hildebrand's edition, the last to give an extensive apparatus criticus, appeared in 1842, and seven years later Keil3 announced his belief that all the MSS. which he had seen in Italy were derived from Laur. 68. 2 (F), the famous eleventh century MS., written at Monte Cassino, and now at Florence. Since Keil, all texts have been based almost exclusively on F, with assistance from its twelfth or thirteenth century copy Laur. 29. 2 (4), in the enormous number of places where F is now illegible.

Keil's view was based chiefly on the destructive rent in fol. 160 of F (Met. VIII. cc. 7, 8, and 9), to which correspond a series of intentional gaps in 4, filled up by a later hand in the fourteenth century. The later MSS. are all either defective in these passages (there are two affected areas, on the recto and verso of the sheet), or have readings obviously connected with those added later to p. Keil held that this filling was conjectural and of no authority. His verdict has been generally extended to all later MSS., and it has been argued or assumed that all are derived, directly or indirectly, from F. Keil, however, unlike most recent editors, took the sensible view that the later MSS. might be of some value, considering the present illegibility of F. But he did not develop this theme.

In 1895 G. Rossbach and J. van der Vliet independently published articles, claiming that the later MSS. deserved more attention than they had received. Van der Vliet claimed only this, that the filling in of the lacunae in some later MSS. was not borrowed from p's fourteenth century supplements, but was a genuine tradition going back to F untorn. One of the objects of my articles is to prove that in this van der Vliet was right. Rossbach, partly on similar grounds,7 maintained that the Dorvillianus (8), now at Leyden, was not merely superior in some ways to o, but was actually independent of F, and was of non-Italian descent (a view which is wholly untenable). Next year van der Vliet attacked Rossbach's view of 8, but argued in detail that the later MSS. fell into two classes, the better of which, including

1 The Apologia and Florida are in the same position as the Met. They are included in F and 4, and rarely occur in MSS, apart from the Met.

2 I have followed recent editors in keeping the traditional spelling Apuleius (with one p). because it does not seem quite certain that this spelling is wrong; the evidence of inscriptions suggests that Apuleius may have been characteristic of Africa (see Thes. L. Lat. s.v. Appius, a reference which I owe to Dr. Postgate). F has Apuleius, and so have the best MSS. of the philosophical works.

3 Obs. crit. in Catonis et Varronis de r. r. libris, Halle, 1849, pp. 77 sqq.

Lowe, however (The Beneventan Script, 1914, p. 70), assigned it to the thirteenth. Rostagno and Schiaparelli adhere to the old dating, and Lowe (Class. Quart. XIV., 1920, p. 155) suggests the date 1200. For all this, and for a refutation of Rossbach's view that is a Renaissance MS., see Lowe's article. Lowe thinks that o, like F, was written at Monte Cassino.

5 Professor Rostagno most kindly examined this writing for me, and made exhaustive comparisons. His conclusion is: Credo di non essere lontano dal vero attribuendo al sec. xiv.forse 2a meta-quei supplementi.'

6 Mnem., 1895, pp. 175 sqq.

7 N. J. j. Ph. u. P., 1895, pp. 570 sqq.

8 Mnem., 1896, pp. 210 sqq.