the adoption by one of some readings of the other, than of two copies of the same edition which were beginning to exhibit points of dissimilarity. And yet some advocates of unity of origin for the two codices go so far as to ascribe certain apparent instances of 'consensus' in error to the existence of holes in the pages of this supposed original, and to estimate the number of lines which each imaginary page must have contained. This is surely to forget that A and P4 are two out of a vast number of ancient copies of Plautus, belonging to different parts of the Roman world, with as much likelihood of being related to each other as two copies of Shakspeare, published, let us say, at the interval of a century or half a century, the one at Glasgow and the other at Melbourne. The great argument used by the supporters of such theories is the 'consensus' in error of the two ancient codices. They confront us with an imposing list of lines in which the reading of A and of P4 is the same, and apparently erroneous. Year by year these lists grow smaller; for, as our knowledge of Plautine diction and prosody grows, we recognize the correctness of this or that reading supported by the 'consensus' of A and PA. Before 1892, when Prof. Skutsch published the first volume of his Forschungen, with its interesting discovery of the suppression of final è in ille, nempe, inde, proinde, etc., in Plautus' verse, just as in all literature in atque (ac), neque (nec), neue (neu), lines like Stich. 175: quia inde iam á pausillo púero ridiculus fui used to form a considerable part of these lists. Rud. 538 will, I presume, be omitted from them, now that Prof. Skutsch has shewn us that auderem has its old pronunciation aviderem: Qui? Quía auderem técum in nauem ascéndere. The whole history of Plautine textual criticism in recent years has taught us that truth lies, if anywhere, in the 'consensus' of A and P4, and that the danger in tampering with a reading supported by A alone or P (or P1) alone is not nearly so great as the danger of discarding the combined testimony of the 'two witnesses.' No judge will arrive at a correct verdict who does not weigh the evidence. The evidence of AP4 must outweigh the single evidence of P. The practice of emending lines of Plautus without stating whether the reading which is impugned rests on the authority of P only, or of P4 only, or of A only, or of A and PA combined, obscures the conditions of the problem to the reader and encourages the writer to reject genuine readings too hastily. The whole weight of tradition supports the reading penitus (in its original sense of 'from inside') in Pseud. 132: atque ipse egreditur pénitus (intus edd.), periurí caput. Are we as much justified in substituting intus in this line as we might be in a line for which we had no better evidence than the Carolingian MS P? In Stich. 704, does not the 'consensus' of AP in the reading in lecticis rather point to some Plautine coinage like inlectice (adv.) of the type of accubuo (Truc. 422)? STICH. Nímium lepide in méntem uenit: pótius quam in subséllio Cýnice hic accipimúr quam inlectice (in lectis edd.). SAG. Immo enim nimio hic dúlcius. Must we not retain their reading stultitiis in Trin. 509, and give de the sense of 'after' or 'in consequence of' (as in Cas. 415, etc.)? nám is (sc. ager) de stultitiís (diuitiis edd.) meis solús superfit praéter uitam rélicuos. Should we disregard their testimony to the old trisyllabic form of ergo adv. (as iurigo of iurgo, purigo of purgo) in Poen. 1051? patrítus er<i>go hóspes Antidamás fuit? Should we ignore their indication of an O.Lat. fortasse est like necesse est in Poen. 1004-5? MIL. Fortásse medicos nós esse arbitrárier. AGOR. SY ést (Si ita est edd.), nega esse: nólo ego errare hóspitem. And is the phrase in ius uos uolo so impossible that we must suppose both A and P1 to be in error in Poen. 1225? quíd istic? quod faciúndumst cur non ágimus? in ius uós uolo (uoco edd.). Certainly, if we consider the number of lines supported by the 'consensus' of A and P4, whose reading has been justified through advance in our knowledge of Plautus, we shall be inclined to predict that nearly every line so supported will prove to be free from error, unless there be an error into which A and P4 have fallen independently, like the 'modernizing' of an archaic form, e. g. ridiculisissimos for ridiculissimos (Stich. 389), haplography, or some other equally obvious miswriting, such as illorum for Iliorum (Bacch. 951), atque euoca for atque uoca (Poen. 1116), Euolaticorum for E (the 'nota personae') uolaticorum (Poen. 474), optumi maxumi for opt. maxume (Men. 574), festiua mulier for festiuam mulier (Mil. 591). But to argue on the other side is much more easy, for one has ready to hand all the apparent instances of 'consensus' in error which have not yet received their explanation; and, although the number available is diminishing steadily, there still remains a sufficient quantity to provide a respectable case. A large list of instances is furnished by lines which shew hiatus. These, however, lose their force, if we are to believe (and I do not see how we can venture to disbelieve) Cicero's express statement that the early poets made extensive use of this license. To discuss the limits within which we may suppose Plautus to have used it would, however, take too much space here.1 Even if real cases of 'consensus' in error, these lines with hiatus would hardly justify the theory of so close a relation between A and PA as is assumed. Prof. Leo has shewn the likelihood that in the early Empire unrestricted hiatus was believed to be a feature of Plautine verse, and that a 'versus hians' would be accepted without question by all editors of Plautus. The strongest argument that has been produced in favour of the close relation of A and P1 is the appearance at Merc. 598 of two lines which belong to another part of the play (vv. 842-3). In P4 the passage stood so: (CHAR.) sed isne est, quem currentem uideo? ipsus est, ibo obuiam. 598 Evr. Diuom atque hominum quae spectatrix atque era eadem es hominibus, 842 spem speratam quom obtulisti hanc mihi, tibi grates ago. CHAR. Nunc, quod restat, ei disperii: uoltus neutiquam huius placet; 599 tristis incedit,-pectus ardet, haereo,-quassat caput. 843 600 Eutyche. EVT. Eu, Charine. CHAR. Priusquam recipias anhelitum, 601 In A only the beginnings of the lines are legible. First comes a line beginning sed isne (v. 598), then a line (too long to be written in a single verse) beginning di-, then a line beginning spes, then a line beginning nuncq-, then a line beginning se... q-, then a line beginning tr—, then either one or two lines (perhaps a sceneheading) of which not a single letter can be read, then a line beginning Eutyche and ending quam recipi[as] anhelitum. The lines (vv. 842-3) diuom atque hominum quae spectatrix atque era eadem es hominibus, spem speratam quom obtulisti hanc mihi, tibi grates ago 1I have attempted to do so elsewhere, in the English Journal of Philology for this year. are suitable in the place where they appear later in the play, after v. 841 (the leaves of A which contained this part of the play have been lost), There Eutychus reappears on the stage, charged with joyful tidings,-not, as here, with a message of sorrow;-and it is argued that by some extraordinary mistake a scribe entered them in the margin or inserted them in the text at this place, and did not take the trouble to erase them. From a text marred by this blunder, it is said, both PA and A have been transcribed. Another explanation is possible—namely, that Eutychus, at his two appearances on the stage with his two messages, had much the same form of words put into his mouth by the dramatist, and that in P1 his utterance at his first appearance was by a blunder assimilated to his second utterance. If we could recover the rest of the two lines in A, they would, on this theory, exhibit their correct form. I do not think this piece of evidence for a close connexion of A and P1 is strong enough to overcome the mass of facts that speak against this connexion. And it is, so far as I know, the strongest piece of evidence that has yet been alleged.' W. M. LINDSAY. 1 Poen. 1168 seems to be correctly preserved by A and (in the main) by P4. In has the sense of 'like, after the fashion of': AGOR. Sed eccás uideo ipsas. HAN. Haécine meae sunt fíliae ? On Mil. 1419, Stich. 620 see Seyffert in Berl. Phil. Woch. XVI (1896), p. 234. In Poen. 331, why may not insecundo (cf. Auct. ad Herenn. IV 56) be formed from insequor in the same way as secundo from sequor? In Cas. 571 prius is the original scansion of the word, and contor, the simple verb of which percontor is a compound, is by no means impossible; in Stich. 223 Hercules te amabit is a most natural parenthetical exclamation to an imaginary bidder; in v. 243 of the same play eu ecastor seems to be 'extra metrum,' like attat in Cas. 619 (cf. Mar. Vict. 85); in Pseud. 306 iustus need not be altered, nor in v. 442 idné tu; in Mil. 254 quae mentibitur has the same construction as Ennius' uitam uiuitur (where uitam can not be acc. of time). III. CHRONOLOGICAL STUDIES IN THE GREEK TRAGIC AND COMIC POETS. Surprisingly little progress has been made since the first half of the present century in clearing up the dark points in the chronology of the minor Greek tragic and comic poets. This fact is at once a tribute to the epoch-making work of Clinton, Meineke, and Welcker, who all possessed in a remarkable degree the combination of wide learning and critical acumen necessary for the several important tasks to which they addressed themselves, and a testimony to the inadequate and often corrupt character of the available chronological data with which later scholars have had to be content. It is safe to say that practically all has been done that can be done in the way of new combinations of the old material, and not always to the advantage of our science. If one will but take the pains to glance at the current and the older handbooks under the names of the poets whom we propose to discuss, one will find abundant illustrations of how opinion has periodically swayed first away from the conclusions reached by the scholars above mentioned, then back again to them, the same ancient notices doing duty in either case. In the scholarly and well-balanced articles from the pens of Kaibel and Dieterich which are now appearing in the Pauly-Wissowa Encyclopädie we generally find a complete survey of all the data, and conclusions which, on the whole, are not assailable in the present state of our knowledge. If I shall take issue with any of these conclusions it will be, with possibly one exception, because of evidence which has hitherto not been applied to these questions. On any other grounds it would neither be profitable nor justifiable to attempt to reopen the discussion for the sake merely of recording a personal opinion. The new evidence to which I allude is to be found, for the most part, in the catalogues of victorious tragic and comic poets which I discussed in a recent number of this Journal, now found together under No. 977 of the second volume of the Attic Corpus. The order in which the names occur in these catalogues was determined by the date of the first victory of each poet. If, then, we |