refer the Early Latin lore of some items in the Liber Glossarum. It is the lore of Donatus or some other Virgil commentator, not the lore of Festus (e.g. 176, 20 and 207, 38, with Plautus citations, come from scholia on Aen. 2, 53 cauae cauernae, Geo. 4, 235 hibernas).

To attempt a detailed account of Ansileubus' sources for the DE GLS. items is inconsistent with the aims of this Study, and indeed impossible since Goetz has published only excerpts. But there is one suggestion worth making. If 'Ansileubus' was a monastery-teacher in Charlemagne's time (I do not dare to say 'of Corbie in Adelard's time '), it is quite possible that older collections, such as the composite Abstr.-Abol. and the larger Abstr. (we might call it 'Abstrusa major ') were available to him only in the form of extracts.

At any rate the value of Lib. Gloss. to an editor of Festus seems limited to:

I. A few, trifling variants in its record of Festus glosses of Abol., e.g.: (164, 17) Aestimium aestimationem puto dici (a conjectural emendation of Aestimiam aestimationes, the form in which the compiler of the AA Glossary found this item in his MS. of Abstr.-Abol., whereas the Sangall. form preserved at least the -ae of Festus' Aestimiae: aestimationes); (167, 21) Aplustria : ornamenta nauis (armamenta Vat., ortamenta Cass. Whether Sangall. and Lib. Gl. found the right form or conjectured it is not clear). These petty details may be reserved for my Study No. VI. And since Lib. Gl. is in touch with the English group' and they draw largely on a fuller Philox., there are also a few trifling variants in its record of Festus glosses of Philox. These too may be reserved for another place.

II. A few supplements to the Festus glosses of Abol. and Philox. (like those mentioned in Studies Nos. II., III., items which are left unnoticed here), just as 202, 21 (Fors fuat: fortuitu; futurum significat) may be a Terence gloss (on Hec. 610) of Abol. omitted in Vat., Cass. (though it may also come from a note on Virgil, Aen. 10, 108 fuat; as 227, 22 from a note on Aen. 1, 41). For example, our remains of the bilingual collection have merely Indigito: ovoμálw. But the Abavus Glossary has Indigitat: inuocat; and Lib. Gl. adds Indigitare: inuocare, Indigitem: inuocem; good material, one would think, for the reconstruction of the full Philox. gloss. And yet it is quite possible that some or all of the three supplements come from Virgil scholia (cf. Don.' ad Aen. 12, 794: alii ab inuocatione indigetes dictos uolunt, quod 'indigeto' est precor et inuoco). Here is a list of (more or less) possible claimants to a connexion with Festus, whether through Abol. or Philox. I. (174, 38) Caluentes: frustra calumniantes. 2. (175, 28) Carpophorus: deus paganorum Graece, quem Latini Frugifer dixerunt. 3. (180, 33) Collifana boues (?) opus facientes. Cf. Philox. 103, 25 Collifana: ρóßаτα iepá. 4. (182, 21-22) Concapito: concipito, Concapsit: conprehenderit. 5. (188, 4) Decudes: decuriones. Cf. Paul. 66 Decures, decuriones. 6. (188,5) Deculcarunt (-tarunt) ualde occultarunt (leg. -carunt ?). Cf. Paul. 66

1 Like 240, 39 Replum: species <uestis> muliebris. Ne peplum facere debuisset.

K

NOTES ON STRABO.

I. THE MEANING OF μονοτροφέω (STRABO 3. 3. 6).

THE word μονοτροφέω is apparently an ἅπαξ λεγόμενον, being cited by the dictionaries only for Strabo 3. 3. 6 : ἐνίους δὲ τῶν προσοικούντων τῷ Δουρίῳ ποταμῷ Λακωνικῶς διάγειν φασίν, ἀλειπτηρίοις χρωμένους δὶς καὶ πυρίαις ἐκ λίθων διαπύρων, ψυχρολουτροῦντας καὶ μονοτροφοῦντας καθαρείως καὶ λιτῶς. Stephanus' Thesaurus (Liddell and Scott) and Sophocles give only one meaning: to eat but one kind of food. Madvig, whom A. Vogel follows, emended to κομοτροφοῦντας, obviously being led to do so, in part, because he took μονοτροφοῦντας in the meaning cited above.

In 3. 3. 7 Strabo specifically asserts that the mountaineers-and this includes Lusitanianseat goat-meat, bread made of acorn-flour, and butter ; and that while water is their beverage, they also drink wine and beer.

The word appears to have a meaning similar to that of μονοσιτέω, which means (I) ‘to eat but once in the day' and (2) F to eat alone. Now μονοσιτέω certainly means the former in Xenophon's Cyropaedia 8. 8. 9 : καὶ μὴν πρόσθεν μὲν ἦν αὐτοῖς μονοσιτεῖν νόμιμον, ὅπως ὅλῃ τῇ ἡμέρᾳ χρῶντο εἰς τὰς πράξεις καὶ εἰς τὸ διαπονεῖσθαι. νῦν γε μὴν τὸ μὲν μονοσιτεῖν ἔτι διαμένει, ἀρχόμενοι δὲ τοῦ σίτου ἡνίκαπερ οἱ πρῳαίτατα ἀριστῶντες μέχρι τούτου ἐσθίοντες καὶ πίνοντες διάγουσιν ἔστεπερ οἱ ὀψιαίτατα κοιμώμενοι. Also in the following from Athenaeus 2. 21: μονοσιτεῖν τε αὐτοὺς τοὺς Ἴβηρας πάντας) ἀεὶ (Φυλαρχὸς) λέγει. Note that this statement applies to the Iberians.

6

Again, neither μονοσιτέω nor any of its cognates ever refer, so far as I can discover, to one kind of food'; and the same is true of μονοφαγέω. And there is surely no such suggestion in the lone instance of μονοτροφία (Plato, Polit. 261D), which means merely 'a rearing singly' as against 'a rearing of a group (συντροφία); and both μονοσιτία and μονοφαγία are used in the same sense as μονοτροφία, but nowhere in the sense of one kind of food.

I take it, therefore, that in the passage quoted from Strabo μονοτροφέω means, not 'to eat but one kind of food,' but 'to eat but once in the day.'

II. STRABO 3. 3. 7.

κηρίνοις δὲ ἀγγείοις χρῶνται, καθάπερ καὶ οἱ Κελτοί.

The commentators and translators all assume as a matter of course that κηρίνοις is impossible. Bréquigny conjectures κεραμέοις, which is adopted by Corais. The latter, however, suspects that κερατίνοις is the right reading.

Friedemann proposes ξυλίνοις, which is adopted by Groskurd, Kramer, and Meineke. Meineke1 originally proposed xnλivois (the reading later adopted by the Didot edition of Müller-Dübner), but in his text he reads ξυλίνοις . Κελτοί, transposing the whole sentence from a position after στιβαδοκοιτοῦσι to a position before καὶ παρὰ πότον κ.τ.λ. C. Müller conjectures ἐρίνοις δὲ ἢ αἰγείοις (σάγοις) for κηρίνοις δὲ ἀγγείοις.

The arguments for κεραμέοις and ξυλίνοις rest upon Athenaeus 4. 36, which is a quotation from Posidonius: τὸ δὲ πότον οἱ διακονοῦντες ἐν ἀγγείοις περιφέρουσιν ἐοικόσι μὲν ἀμβίκοις, ἢ κεραμέοις ἢ ἀργυρίοις. καὶ γὰρ τοὺς πίνακας ἐφ ̓ ὧν τὰς τροφὰς προτίθενται τοιούτους ἔχουσιν. οἱ δὲ χαλκοῦς, οἱ δὲ κάνεα ξύλινα καὶ πλεκτά.

The ingenious emendation of C. Müller rests upon Diodorus Siculus 5. 33: φοροῦσι δὲ οὗτοι οἱ Κελτίβηρες) σάγους μέλανας τραχεῖς καὶ παραπλήσιον ἔχοντας τὸ ἔριον ταῖς αἰγείαις θριξίν. Müller's ἔρινοις, however, is not to be found elsewhere, I believe. Müller's reading admittedly has the merit of fitting the context, thus obviating what Meineke regards as a necessary transposition, and the thought which he secures might be that of Strabo.

Yet it is at least very doubtful whether the text here should be changed either by transposition or conjecture. A wide reading of Strabo shows that his work is not always characterized by a logical succession of ideas, especially in his running descriptions, while at least two arguments may be advanced in favour of the ηpivois itself: (1) Strabo is given to the habit of recording the rather unusual characteristics and possessions of a people or country (more so, I think, than others who wrote upon the same subjects), e.g. ' the silver feeding-troughs' (3. 2. 14). (2) It is not at all improbable that the people in question, or the Celts, made use of waxen vessels for certain purposes. The Galatae ate honey, and they made a certain kind of beverage, says Diodorus Siculus (5. 26), with the wax, τὰ κηρία πλύνοντες τῷ τούτων ἀποπλύματι χρῶνται. And I am informed by a colleague who is versed in such matters that waxen vessels not only would be very substantial and would harden more and more with age, but would impart to wine a peculiar flavour. If this be the case, it is altogether likely that the crude mountaineers of whom Strabo speaks employed waxen vessels for the storage and serving of wines.

III. STRABO 2. Ι. 36.

The reading of all the MSS. is : μὴ διδομένου δὲ τούτου (the assumption of Hipparchus), κενόν ἐστι καὶ τὸ ἐφεξῆς δείκνυσθαι δοκοῦν, ὅτι συνισταμένου ὀρθογωνίου τριγώνου πρός τε Πηλουσίῳ καὶ Θαψάκῳ καὶ τῇ τομῇ τοῦ τε διὰ Θαψάκου παραλλήλου καὶ τοῦ διὰ Πηλουσίου μεσημβρινοῦ, μία τῶν περὶ τὴν ὀρθήν, ἡ ἐπὶ τοῦ μεσημβρινοῦ, μείζων ἔσται τῆς ὑπὸ τὴν ὀρθήν, τῆς ἀπὸ Θαψάκου εἰς Πηλούσιον. Gosselin says: ' Il est facile de voir que le texte est corrompu dans cet endroit, où il ne peut être question ni du parallèle de Thapsaque, ni de méridien de Péluse, mais seulement du méridien de Thapsaque et du point 2 Ind. Var. Lect. p. 956.

1 Vindiciae Strabonianae, p. 29.

où ce méridien coupe le parallèle de Péluse. Il manque aussi quelque chose au raissonnement de Strabo; car je ne vois pas comment l'un des côtés du triangle rectangle dont it parle pouvoit se trouver plus que son hypoténuse.'1 Penzel, Groskurd, Forbiger, C. Müller, and Meineke likewise hold that the text is corrupt, and proceed to exchange the positions of Oavákov and Πηλουσίου, suggesting that it is perhaps the positions of παραλλήλου and μeonμßpivov that should be exchanged. But, in my opinion, the Greek is clear and consistent as it stands, and the transpositions unwarranted.

Eratosthenes has roughly estimated the distance between Thapsacus and Babylon at 4800 stadia, reckoning the distance along the circuitous course of the Euphrates; and the distance between Pelusium and Thapsacus at 6000 stadia. Now the reductio ad absurdum is the form of argument which Hipparchus regularly uses against Eratosthenes, and Strabo quite as regularly defends the latter against the scientific use of estimates which were professedly rough. So here. Hipparchus, in the opinion of Strabo, subjects the abovementioned estimates to a geometrical test, applying them to lines that represent meridians and parallels of latitude, at the same time injecting a proposition of his own, namely, that the parallel of latitude of Pelusium is 2500 stadia farther south than that of Babylon. On this basis Hipparchus easily computes that the distance between Pelusium and Thapsacus must be more than 8000 stadia-he might have said, more accurately, about 8500 stadia. And further, by adding Eratosthenes' 4800 to his own 2500, as though both lay in a straight meridian line, Hipparchus reduces to absurdity Eratosthenes' assertion that the distance from Pelusium to Thapsacus is 6000 stadia; for thus the meridian-leg of the right-angled triangle (7300) becomes greater than the hypotenuse (6000).

To illustrate, draw a rectangle ABCD and its diagonal AC. Let A be Pelusium, B the intersection of the parallel of Thapsacus with the meridian of Pelusium, C Thapsacus, and D the intersection of the meridian of Thapsacus with the parallel of Pelusium. Then, according to Strabo, Hipparchus reasons that Eratosthenes' estimates for CD (= a corresponding meridian line through Babylon) and AC make AB (CD) greater than AC, since AB is really 7300 (4800 + 2500), whereas Eratosthenes makes AC only 6000 when it should be over 8000. But, according to Strabo, Hipparchus' 7300 is unfair for three reasons: (1) The 2500 is not granted by Eratosthenes; (2) the 4800 of Eratosthenes does not lie in a meridian line, but follows the circuitous course of the Euphrates; and (3) the limit of the 4800, Babylon, lies 2000 stadia to the east of the meridian of Thapsacus.

CORNELL UNIVERSITY,

November 25, 1916.

HORACE L. JONES.

1 Note in the Corais-Du Theil-Letronne edition, vol. i., p. 231.

NOTES ON LATIN POETS.

LVCRETIVS I. 469-470:

namque aliud terris, aliud regionibus ipsis

euentum dici poterit quodcumque erit actum.

In 464-8 Lucretius states the view of the Stoics that euenta had an independent existence. Their argument was: 'We say "Helena rapta est," but Helen died long ago (468), therefore the accident rapta must possess an independent existence '-the old fallacy of confusing esse as copula and esse as predicate. Lines 469-470 are Lucretius's first answer, which, as Giussani says, is little better than the fallacy he is refuting. He draws a distinction between the men to whom the events happen and the place where they occur. Events are accidents (euenta), not only of the men, but also of the place, and if the men pass away the place remains.

With this explanation of Giussani's the only difficulty is in 469, which scholars from Lambinus downwards have regarded as corrupt on the ground that terris and regionibus mean the same thing. Munro's Teucris will not do, for, as Giussani points out, quodcumque erit actum in 470 shows that Lucretius is stating a general proposition and not referring merely to the case of the Trojans. The other emendations are too remote from the MSS. to account for the corruption. Bailey keeps terris with this explanation: 'Lucretius . . . says first, in a rather frivolous spirit: "Well, we can say that it is an 'accident' of the place, or if you object that that too has changed, of that part of space (Translation, p. 283), and translates: For, firstly, we might well say that whatsoever has happened is an accident either of the countries or even of those regions of space.' But such a distinction can hardly be maintained. Moreover aliud . . . aliud . . . ought to mean 'some things can be called accidents of Teucri (or saecla or terrae), others of regiones,' whereas Munro, Giussani, and Bailey alike take it to mean that the same event can be called an accident first of the one and then of the other; which one would expect to be expressed by alias . . . alias. . . . Hence it seems more probable that terris means 'the whole world' as in Aen. I. 460 'quae regio in terris nostri non plena laboris?' We can then give aliud . . . aliud. . . its proper force by translating: 'For it will be possible to call whatever happens an accident in one case of the whole world, in another of the precise district in which it occurs.' That is, if we cannot predicate an event of the men who are dead, at any rate we can predicate it according to its character either