To Ulpian therefore two conceptions of the meaning of convicium were present, the one defined by concitatione (vocis), vociferatione, the other by conventu, collatione vocum, id quod in coetu dictum est. It is to the latter of these that he attaches the etymological suggestion of convocium: cum enim in unum complures voces conferuntur, convicium appellatur quasi convocium. But this is an etymological explanation only for the second of his alternative interpretations, not for the first. Apparently he was divided in mind between a true feeling for the meaning of the word as a loud angry expression of abuse, a concitatione vocis, and the meaning which was yielded by a current etymology of the grammarians. It is, I venture to believe, out of this etymological doctrine that he has built up his twofold classification of convicium, a distinction which is in no way implied either in the praetorian edict, nor in the usage of the word. In the nature of things convicia spoken by a group small enough and well defined enough to be held to individual responsibility would be subject to the same legislation as the convicium of a single person. But the instances in which the miscellaneous outcries of a mob, though directed against an individual, could constitute an indictment for verbal injury would surely be few and exceptional. The offense was that of an individual against an individual, and this is contained in the edict of the praetor. But even if the idea of plurality of utterance was entertained as one characteristic aspect of convicium, it was not excluded by the explanation a concitatione (vociferatione), as appears from the phrase (cited above) attached to this definition, sive unus sive plures dixerint.1

The fact is of course that con- in convicium had entirely lost, if indeed in this compound it ever had, the notion of "with" or "together," and as in so many other words served merely to indicate the thoroughness or completeness of the action contained in the significant element of the word, as in conficere, confringere, concludere, concedere, conticere, conticinium ("the time of night of complete silence")." Whether indeed convicium is at all connected with the stem voc- vōcis certainly open to question. Not to dwell upon the alternative etymology of the ancient grammarians from vicus3 (which Usener favored, 1 But cf. note 2 on p. 117 above.

2 See the valuable discussion of concludere and concedere by Ulrich Leo in Glotta, X (1920), 173 ff., and Brugmann, Kurze vgl. Gram., p. 564. The latter (loc. cit., Anmk. 1) notes that the frequence of such composita, especially with con-, is due to replacement of aoristic formations which had been lost in Italic. Thus tacere conticere corresponds to σιγᾶν σιγῆσαι.

Which Horace may have entertained, Epp. i. 17, 62: vicinia rauca reclamet.

and connects with the theme of his Italische Volksjustiz), there is not a little reason for looking for some other explanation of the word. The long i in -vic- has occasioned much perplexity and variety of opinion to comparative philologists, and no satisfactory explanation of it, starting from the stem voc-, has yet been found.1 Most recently Professor F. Wood (in Class. Phil., VII, 304) has connected the word with vinco, pervicax, and though I have no right to an opinion in a question purely linguistic, his explanation seems to me more satisfactory not only phonetically but also semasiologically. The Indo-Germanic base *ueiq- (in vinco) Walde renders by "energische, besonders feindselige Kraftäusserung," and it would not be difficult to reconcile all the usages of convicium with this fundamental idea. The notion of insistent, obstinate, irrepressible (abuse, or attack) so characteristic of the word, is conveyed chiefly by the prepositional element con-, as it is conveyed analogously in pervicax by per-.

The belief that convicium implies the utterance, or requires the presence, of a crowd or mob, is clearly erroneous. It is spun out of an assumed etymology, which Ulpian does not in fact entirely indorse, but merely advances in explanation of one aspect of his twofold conception of convicium. But while not accepting it unreservedly, he yet rests one leg of his structure upon it. This is the starting-point of the modern doctrine, which, failing to note the alternatives, has accepted the idea of a plurality of voices or persons as the unqualified teaching of the jurists. Convicium has necessarily no more to do with a plurality of utterance than has clamor, or the ancient pipulum and vagulatio, both of which are defined by convicium. To be sure a mob might shout insults at an individual, and these were convicia, not however, because they were shouted by a crowd or in chorus-quasi convocium, but because they were vehement expressions of hostile feeling-a concitatione, and meant to overwhelm (convincere).

NEW HAVEN

1 See Brugmann, Grundriss, I, 1, p. 134 and esp. p. 505; Buck in AJP, XVII, 270; Solmsen in KZ, XXXIV, 15.

* Convincere shows little variation from the meaning "refute," "convict." But from such an example as infamatis magis quam convictis (Tacit. Ann. xv. 71) one may see how a related word like convicium should have become fixed in a meaning short of proof or conviction. The speaker of convicia utters epithets-fur latro parracida: if pressed in trial, and proven, their object would be convictus. The Greek equivalent is éλéyxew, which connotes not merely proof or conviction, but also at times abuse and insult: conviciari (Stephanus, citing Suidas), éλéyxoμev åvtì toû kakoλoyoûμev, and Plutarch, Mor., p. 18, τοῖς ἐλέγχειν καὶ λοιδορεῖσθαι βουλομένοις. A curiosity, scarcely significant, is the gloss convicisse concitavisse, which suggest Ulpian's convicium a concitatione.

THE "LOST" MS OF CICERO'S DE AMICITIA

BY C. H. BEESON

A MS in the hands of a private owner is the bane of the paleographer and philologist and the despair of the bibliographer. It is not always easy of access, the facilities for studying it are frequently insufficient, the restrictions placed upon its use are often annoying, and the privilege of using it sometimes expensive; it is always likely through a change of owners to disappear for an indefinite period. All who are interested in it breathe a sigh of relief when it has found a permanent home on the shelves of a great public collection, whose keepers realize that the possession of such treasures carries with it a responsibility and are eager to facilitate the research of scholars in every way.

The complications that may arise in the case of such a MS are well illustrated by the Codex Parisinus of Cicero's De amicitia. This MS was discovered by Mommsen in the library of Firmin Didot in Paris in 1863; Mommsen published the results of his examination in Rheinisches Museum1 of that year, and his collation was later used by Müller in his preparation of the Teubner edition. After sixty years that represents the sum total of our knowledge of the best representative of the text of De amicitia! Reid examined the MS in 1883 in London where it was in the possession of Quaritch, but too late to use it for his edition of Laelius which appeared in that year. He planned to publish the results of his examination but apparently never did so.

2

Other editors, except Price, appear to have had no knowledge of the whereabouts of the Parisinus, and the historians of Latin literature are equally uninformed. Schanz has no information on the subject, and Teuffel's only comment is "Wo jetzt?" The references to it have been of the most casual sort, and everyone seems to have accepted the situation with resignation except Bassi, the editor of Laelius in the new Italian series, who indulges in some speculations on the subject:

1 XVIII, 593–601.

2 Addenda to his edition (Cambridge University Press), p. 165.

Sed posteaquam Mommsen has notitias de illo dedit, nescio quomodo, codex nusquam locorum amplius repertus est. Quo delatus est? Quis eum nostris possidet temporibus? Estne adhuc in Europa, an in Americam est transvectus, ubi homines docti ac divites omnia, quae praecipua atque optima videntur, empta in Museis conlocant, quibus iure meritoque gloriantur atque delectantur? Sed hoc, quod mirum, quasi non verisimile videtur, heu malum! verum apparet.1

The state of affairs is not so desperate as Bassi's fancy paints it. The MS is not in America; it has been for nearly a quarter of a century in the capital of the stronghold of classical studies, in Germany, in the Royal Library at Berlin.

In a review of Teuffel's Literaturgeschichte, published in this Journal four years ago, I made this statement on the authority of Professor Price, who had his information from Quaritch that the MS had been sold to Berlin, and of Holder, the kindly old librarian at Karlsruhe, who told me in 1910 that it was in the Royal Library at Berlin. It was with some distress at the possibility of having committed an error that I later read Simbeck's statement in the Preface of his new Teubner edition (1917), a copy of which had not been available at the time the review was published: "Nunc eum [i.e., Parisinum] possidet B. Quaritch bibliopola Britannus. Qui cum nullis precibus ut librum photographice depingendum curaret, moveri posset, Muellerum secutus lectiones varias, quae sunt magni momenti, notavi." This distress was changed to bewilderment when in the spring of 1923 an inquiry at Quaritch's met with the response that they did not possess the MS and never had possessed it! The comedy of errors was closed when a search in their records revealed the truth that they had in fact bought the MS at the sale of the Didot library and had sold it to Berlin in 1903. It now bears the signature Ms. Lat. qu. 404.

After all these years the MS deserves a re-examination, both because of its paleographical interest and its importance for the text. The following report is based on a study of photographs kindly furnished by the authorities of the Royal Library.

The MS has forty-three folios containing, folios 1-32, De amicitia, 32-43, a collection of sententiae attributed to Seneca, followed by the first two verses of the second carmen of Eugenius of Toledo. The last 1 Corpus scriptorum Latinorum Paravianum, No. 27 (1917), p. v.

2 XV, 98.

page, according to Mommsen, is illegible. The worn condition of the first and last pages shows that the forty-three folios have existed as a single MS for a long time, certainly since the thirteenth century, as a notation in Gothic script at the top of folio 1 proves, "Tulli' de amicitia pūbia qdam." The quaternion signatures also show that De amicitia formed the first part of the codex from the beginning. According to Mommsen, there is a library mark in a fifteenth-century hand which indicates that the codex once belonged to the cathedral library at Constance, "liber est ecclie Constan'." Mommsen naturally suggested that it therefore once belonged to one of the great collections formed by the Irish monks at St. Gall, the Reichenau, and elsewhere.

The pages are ruled with a dry point in long lines, twenty-one lines to the page, with two vertical lines to mark the outer margin of the script. The margins were originally unusually wide, but were much cut down in binding. This is indicated by the fact that the names of the interlocutors which stand far out in the margins have been partially trimmed off. Mommsen states that there is only one quire signature (f. 29), but the remains of a signature are plainly visible on 7, and still more so on 23; at the bottom of 15 the letter R is written, with faint marks which may possibly be the remains of a signature. According to Mommsen, the first folio of the first quaternion and the second and seventh of the last quaternion have been lost. The latter point is important because it shows that the omissions in Parisinus (Simbeck, 74, 1. 27-76, 1. 3, and 80, 1. 26–82, 1. 6) were due to the loss of a double sheet and that the codex originally had the complete text. The second and third quaternions are regular, containing eight leaves each.

The script is a beautiful ninth-century minuscule, having all the earmarks of the Tours scriptorium. The first line (QUINTUS MUCIUS) is written in large rustic capitals and uncials. There are twenty-one lines to a page, except folio 1 which has twenty and folio 32 which has only six lines of text. The longer sentences begin with capital letters and are punctuated with a point followed by a mark resembling the figure "7"; the shorter members are punctuated by a point. The scribe does not always follow the rulings, and the lines therefore often have a wavy effect; the vertical rulings, too, are often disregarded, producing a ragged effect on the left-hand of the script