''^ii ON THEENCLITIC 7V£ IN tMEY LATIN A DISSERTATION FOR THE ACQUISITION OF THE DEGREE OF DOCTOR OF PHILOSOPHY AT THE KAISER-WILHELMS- UNIVERSITAT STRASSBURG BY MINTON WARREN OF PROVIDENCE U. S. A. BALTIMORE & STRASSBURG 1881. • * ■ -»■■• /v,^^' }/^i [From the American Journal of Philology, Vol. 2, No. 5.] ON THE ENCLITIC NE IN EARLY LATIN. By Minton Warren. Within the last twenty years much attention has been paid to the peculiarities of early Latin. Striking deviations from the usage of the classical period have been pointed out in the regimen of verbs, in the signification and use of certain particles and con- junctions, and in the principles governing the subordination of clauses. A multitude of monographs have contributed valuable material for the construction of an historical syntax. Not to speak of others, Liibbert has traced the development of quom, Kienitz of quin, and Becker, in Studemund's Studien Bd. I, has made a most exhaustive study of the Syntax of Indirect Questions in Early Latin. The enclitic ne, commonly known as the Inter- rogative Particle, has not to my knowledge received a special treatment in recent times. It occurs about i loo times in Plautus and over 400 times in Terence. Hence a discussion of its use which should aim in any sense to be exhaustive would far exceed the limits of a journal article. I shall therefore in the following pages content myself with the attempt to show that in the earliest Latin the use of this particle was not confined to questions. I shall then seek to establish a probability in favor of the existence of two particles nc distinct in origin and signification. I proceed at once to treat the passages where ne is found in the MSS. with no interrogative force, for which reason the text has in most cases suffered violence at the hands of the editors. Mil. 309 (II 3, 38). The slave Sceledrus, on discovering that Philocomasium has eluded his vigilance, gives utterance to his agitation and alarm in the following words : Edepol facinus fecit audax. Hocine si miles sciat Credo hercle hasce aedi's sustollat tolas atque me in crucem. Hocine BD, hoc me C In line 310 I follow Fleckeisen and Brix ; see Tyrrell's ed. for MSS. readings. Parens keeps hocine and ' Lorenz says, Philol. 32, p. 302, Auch C hat hoc ine welches Klotz sogar vertheidigt. Bergk vermuthet hdce si mMs sciat. puts an exclamation point after sciat. Ritschl's change to hoc nunc is adopted by Brix, Lorenz, Fleckeisen and Tyrrell. Fleck- eisen however in Jahrb. 105, p. 71, proposes to read hoc enim. Langen treating oi enim, in Beitrage zur Kritik und Erklarung des Plautus, p. 270, gives hoc nunc the preference. Lest the position of hercle, which usually follows si immediately (cf Epid. 326), should excite suspicion, I call attention to J'ersa, 627 f. Tu si banc emeris Niimquatn hercle hunc mensem vortentem, credo, servibit tibi. Epid. 73(1 I, 73): Haecine ubi scibit senex Piippis pereundast probe. Hecine B, haeccine J, hec F, hecine E (cf. Goetz' ed. of Curculio p. ix.) Muller, Plant. Pros. p. 732, proposes to emend by reading haec hercle or haec hodie. Goetz is wisely conservative. Although he makes no explanation, he evidently does not regard haecine as introducing a question. Parens puts an interrogation point after senex. Mil. 565 (II 6, 82) : Orationemque SC. Egone si post hiinc diem Muttivero etiam quod egomet certo sciam Dato excruciandum me : egomet me dedam tibi. Although the Ambrosianus with BCD has egone, the editors with one accord substitute ego nunc. So too Tyrrell, without even mentioning in his critical apparatus the reading of the MSS. And. 478 (III I, 20): Hicine me si inparatum in veris nuptiis Adortus esset, quos mihi ludos redderet ? Hiccine BCD'G'P (^Hicine D^E, Hie G ex ras.) Some of the codices of Donatus give Hie si me im.paratum, but the best codex, the Parisinus, gives Hie m,e m,isi in parta, evidently a corruption of hicine me si inparatum. Consequently any change here is direcdy in the face of the best MS. authority. Bothe however 3oes not scruple to transpose me, and reads Hie iyiparatum me si. Fleckeisen, Klotz, Meissner, and Umpfenbach follow him. Wagner reads hie nunc me. Conradt, Metrische Composition der Como- dien des Terenz, p. 117, reads Hie ne, "da die Fragepartikel durchaus fortgeschafft werden muss." Spengel nevertheless in his edition makes bold to retain hicine, citing for its support Mil. 309 V' and 565. His theory of explanation, however, involving as it does an anacoluthon, I cannot in view of all the facts accept. It does not fit all the other cases. Mil. 936 (III 3, 62) : Bene ambula, bene rem gere. At egone hoc si ecficiam plane Ut concubinam militis meus hospes habeat hodie Atque hinc Athenas avehat si hodie hiinc dolum dolamus Quid tibi ego mittam muneris? Egone B, egonec CD. Here metrical considerations forbid the mere substitution of nunc for ne and the editors have been pushed to other devices. Muller, Nachtrage zur Plaut. Pros. p. 82, proposes ego hoc nunc si. Ritschl too by transposition inserts his favorite nunc so as to read At ego nunc si ecficiam hoc plane. So too Fleckeisen and Lorenz. Seyffert (Philol. xxix, p. 399) finds nunc here " huchst matt und iiberfliissig," and looks upon ne or nee as a mere dittograph of hoc. Brix approves of this view and would have us read (with hiatus) at ego hoc si ecficiam plane , which is simply a return to Bothe's reading (in my copy of 1821). I wish to make it apparent how much trouble the editors have taken to get rid of a simple ne, which, if a rational explanation can be found for it, must after all be retained. I know of no better parallel for this particular passage than is furnished by Poen. I 3, 18 (420 Geppert). To appreciate tlie scene, one must picture to himself the youth Agorastocles quite overjoyed at the prospect of outwitting a pimp and getting possession of the lovely Adelphasium. Milphio, his slave and willing agent in the matter, has been too often wheedled by fair promises to put faith in them any longer, and hence, when his master begins a long-winded sentence promis- ing him his freedom, he interrupts him at every turn, and bids him go bring the witnesses. AG. Egone edepol, si istuc lepide ecfexis. MI. I modo AG. Ut non ego te hodie. MI. Abi modo. AG. Emittam manu MI. I modo. AG. Non hercle merear pro hoc. MI. Abi, abi modo AG. Quantum Acherunti est mortuorum. MI. Etiamne abis ? AG. Neque quantum aquai in mari est. MI. Abiturunes? The enigmatical character of the whole speech, which is too long to quote entire, is shown by the words of Milphio as he goes off in a rage, muttering : Nam isti quidem hercle orationi est Oedipo Opus coniectore. It has puzzled the editors as well. I have given the text of Geppert except that I have omitted the interrogation point after manu. The only emendation which specially concerns us here is that of egone egoyie found in ABCD to egone edepol. This may have been suggested to Geppert by Men. 1023 (1025 Bx), Ergo edepol, si recte facias, ere, me[d]emittas manu. Bothe omits one egone and makes no attempt to heal the halting verse. Egone is most likely a dittograph, and edepol is on the whole a probable substitution. If accepted, however, it precludes the idea-of a question, since it is not employed either by Plautus or Terence in interrogative sentences. Bothe in fact does not seem to regard egone as introducing a question, but indicates by leaders placed after manu that Agorastocles' speech is broken off. Parens reads egone ? egone ! si, etc., but his commentary shows that he caught the true sense of the passage. He interprets thus, " ego profecto te manumittam si istuc feceris," and again, " non hercle velim pro mercede mortuos omnes consequi, etc., ut non te mittam manu si . . . effeceris." An exactly parallel use of ui non occurs in Bacch. 11 84", Quern quidem ego ut non [hodie] excruciem, altenim tantum non meream. Hodie was added by Hermann. The MSS. have alterum ianium auri, which Fleckeisen and Ussing keep, the latter without inserting hodie'' (cf. also Stich. 24, Men. 218.) Both in Mil. 936 and Poen. 420, egone must be retained, but not as part of an interrogation. Agorastocles' words taken con- nectedly give then the following sense or nonsense : " I in good sooth, so you play cleverly your part, would not forego the giving you this day your freedom, to earn as many as are the dead in Acheron, nor yet as many as are the waters in the sea, nor as all clouds that are, nor as the stars in heaven," etc. The next example which I shall cite is very similar. Asin. 884 (V2,34): PA. Aiidin quid ait? ART. Audio. DE. Egone ut non domo uxori meae Siibrupiam in deHciis pallam quam habet atque ad te deferam Non edepol condiici possum vita uxoris annua. Fleckeisen and most editors put an interrogation point after ' In Bursian's lahresbericht for 1881, Heft 3, p. 39, I find that Brachmann follows Fleckeisen, but reads altrum. Perhaps we should read, Quern quidem egone ut non e'xcruciem, altrum tdntutn auri non meream, cf. Haut. 950, Syrum quidem egone, €te.7-see below deferam. The usual conditions under which the egone ut ques- tions are employed by Plautus are not here fulfilled, as is clearly admitted by Kraz (Stuttgart Program, 1862, p. 33), Ussing cites the parallel passage Cas. 400*^*^- (11 8, 68) : Tribus non conduci possum libertatibus Quin ego illis hodie comparem magnum malum, and reads Ego ne, but the ne is certainly enclitic. It seems to me that there is no more interrogative force felt in ut iion than in ^uin, although I admit that in an earlier period the construction itself may have grown out of a paratactic question, as has been proved for quin clauses. In Haut. 950 (V i, 77) the Codex Bembinus reads : Sed Syrum quidem egone si vivp adeo exornatiim dabo Adeo depexum, ut dum vivat meminerit semper mei. The MSS. of the Calliopian recension BCDEHGP have, accord- ing to Umpfenbach, Sed Syrum MEN Q^iid eum ? CHR Egone ? si vivo, except that G has ergo me. From personal examination, however, I can state that Cod. Par. 7899, Umpfenbach's P, has no interrogation point after egone, although it is commonly found in P after questions. Cod. Par. 7900 A saec. X on the other hand reads egone ? Nonius Marcellus cites the latter part of this verse twice to illustrate Depexum and Exornare (cf. Quicherat's ed. pp. 7 and 336). The MSS. have egone. Quicherat inconsistently reads ego nae in the first place and egone f in the second. Even as early as the fourth century, however, the corruption of the text (which was no doubt due to a misunderstanding of the use of egone) had taken place. Donatus in his commentary on Adelph. Ill 3, 46 quotes thus, Sed Syrum ME. Quid eum. ? Bentley, without know- ing the reading of the Bembinus, saw at once the impropriety of keeping egone ? as a question. I give his comment somewhat abridged. ^'^ Egone semper respondet verbo secundae personae ut Phorm. I 2, 7. Sed quid tu es tristis ? Egone ? Huic igitur loco non convenit. Repone Se'd Syrum.. Quid eum f Ego si vivo, eum adeo exornatiim dabo.'' Fleckeisen discusses the passage (Philol. II, p. 76, 1847) and proposes Sed Syrum M. Quid eum ? CH. neego sivivo, etc. In his ed. of 1865, however (not, as it seems, having ascertained the Bembine's reading), he gives, with hiatus, Sed Syrum quidem ego si vivo, etc. Wagner, both in his German and English edition of the play, adopts egomet, Klette's emenda- tion (Rhein. Mus. XIV, p. 467). Shuckburgh, the latest editor (1878), who however has never heard of Umpfenbach's critical edi- tion, follows Fleckeisen. For quidetn used before a conditional clause, to make emphatic a preceding accusative, I need only to refer to the passage cited above, Bacch. 1184, and And. 164, quern quidem ego si seiisero. If MS. authority is to have any weight in the passages above discussed, egone stands in need of no defence. Its meaning may be illustrated by the early use of the asseverative enim to emphasize a conditional statement, Persa 236, Enim non ibis nunc vicissim nisi scio. Compare the similar use of the stronger compound enim, vera, Phorm.937, Enimve'ro si porro esse odiosi pc'rgitis . . where the threat is broken off by Demipho's asking Quid fades ? In seven of the examples already given it must have been noticed that ne precedes a conditional clause. Hidden away amid the critical apparatus of Terence I have found a tun which I think may now claim admission to the text. Adelph. 770 (V I, 8): Tun si meus esses. . . SY. Di's quidem esses Demea Ac tuam rem constabilisses. DE. exemplo omnibus Curarem ut esses. G alone has tu. Umpfenbach attributes the n of tun in P to a second hand. I myself could see no trace of a correction, and there is no crowding of the letters. As there is evidently no inter- rogation here, the editors with one consent have eliminated the n. The interrupted threat of Demea, to which Syrus gives so witty a turn, is taken up again in the words exemplo omnibus. The form tun shows convincingly what we might have inferred equally well from the forms hicine, hocine, that the ne in these formulas is enclitic and short, and therefore this ne is not to be identified with the prepositive ne, as has sometimes happened in the passages containing egone si. The next example is of a somewhat different character. Pseud. 371(13,137): Ecquid te pudet? BAL. Ten amatorem esse inventum inanem quasi cassam nucem. Ten is found only in the Ambrosianus, the other MSS. have te. Ritschl in his edition writes ted to obviate hiatus, tacitly recognizing the fact that the infinitive depends directly on pudet, and that there is no room here either for a question or an indignant exclamation. In Neue Plaut. Excurse, Heft I, p. 44, he prefers to keep ten, adding, " obgleich die Erklarung des Fragesatzes im dortigen Zusammenhange nicht ganz einfach ist." Bergk, in his treatise on Auslautendes D im alten Latein, p. 53, remarks, " Ich wUnschte R. hatte diese Erklarung gegeben ; denn mir erscheint das fragende ne dort ganz unstatthaft." He calls attention too to the fragment of Ennius quoted by Cic. de Div. I 31, 66 : Hoc dolet Men obesse, illos prodesse, me obstare, illos obsequi. where AV have men, which Ribbeck retains, Trag. Frag. p. 21. Bergk's own suggestion that the n is either a phonetic affix to prevent hiatus or a relic of the old accusative ending m, hardly deserves serious consideration. Fleckeisen keeps tene as question, Brix in Jahn's Jahrb. 115, p. 331, discusses the passage at length, giving as the proper reading for verse 370 : Niimquid aliud etiam voltis dicere? CAL. Ecquid te pudet? He too reads ten t.X.c., as a sort of jeering retort made by the leno in the form of a question dependent in thought xx^on pudet. If we compare however passages like Adelph. 432 Numquid vis ? DE. Mentem vdbis meliorem dari, and Mil. 617 f. : PE. Quid id est quod cruciat ? cedo. PL. Me tibi istuc aetatis homini facinora puerilia Obicere, etc. I think it will be admitted that a direct statement dependent on pudet understood is much more natural. Ten has the force of te enim or te vero, only somewhat weaker. This is made clear by two analogous passages, Capt. 566"'' (III 4, 36) : Tu enim repertu's, Phi'locratem qui superes veriverbio. and Pseud. 631" (II 2, 36) : Vae tibi : tu inventus vero, meam qui furcilles fidem. In dependent statement tu enim repertus might have been expressed by Plautus thus, ten esse repertum. Lorenz, although he reads Pseud. 359 (=371'') ted esse inventum, translates "Ja daruber dass du," and this ja gives the sense of ne or enim. In Pseud. 631, Plautus may have written Vde tibi : tiin inventus vero, which would relieve us of the necessity of either admitting hiatus after 1u or scanning tibi. The n would fall out very easily before inventjis, as it has in Pseud. 371 in all the MSS. except A. With the verbs invenio and reperio the use of some affirmative particle may have been especially common. Another instance is furnished by Eun. 930 (V 4, 8) : turn hoc alterum, Id verost quod ego mihi puto palmarium Men repperisse, quo mode adulescentulus Meretncum ingenia et mores posset noscere, etc. A has MEN at first hand, but some late scribe has struck out the N. The Vaticanus also has men. The editors read me, but Ttien must be kept, I think, as another instance of survival. Lest I may seem over-hasty in claiming for ne the force of enim, I wish to call in here the evidence of an old glossary. On p. 52 of Codex Par. 7610 (saec. XIII) 2d part, at the close of quite a long article on ne, I found these words : '■'Ne adverbium corripitur scilicet pro enim vel pro nonne hoc est interrogativum vel affirmativum. Dehortativum vero producitur et conjunctio similiter." (Exactly the same words may be read in Cod. Par. 761 1 (saec. XIII) p. 105, and Cod. Par. 7612 (saec. XIV) p. 115.) The testimony is une- quivocal. I wish I might trace it back to its first source. I could show, if space permitted, that much of the article on ne coincides in phraseology with Priscian, and represents good grammatical tra- dition ; but I have searched through Keil's edition of the Grammatici Latini in vain for any coupling of enim and nonne to explain different sides of 7ic. In the light of this gloss it is interesting to recall Fleckeisen's emendation of hocine to hoc enim. in Mil. 309, and Langen's proposal to substitute w«2/i/«^ .^ for mihi enim. f in Casina II 6, 14 (cf Ceitrage p. 267.) Seyffert too would substi- tute in Rud. 1003 Ita?ie vero ? for Ita enim vero ? which does not occur elsewhere. In my judgment we must also recognize the enclitic nc in the following passage. Cure. 138 ff. (I 2, 47) : Tu me curato ne sitiam: ego tibi quod amas iam hue adducam. Phaed. Tibine ego, si fidem servas mecum vineam pro aurea statua statuam, Quae tuo gutturi sit monimentum. Tibi ne ego, BEJF. To the separation of tibi and ne in the MSS. no great importance can be attached. It often occurs where the ne stands in a question, cf Cure. 419 tu ne in B, 82 ei ne BE. M filler, Plaut. Pros. p. 405, reads with transposition, ne ego tibi, etc. Fleckeisen (cf Philologus II, p. 107) omitting tibi reads ne ego, si Jide'm \ttc\ mecjim servas aured pro statua. Mahler, in his dissertation De Pronominum Personalium apud Plautum collo- catione, has shown that ego tibi is the normal position unless special emphasis is to be given the pronoun of the second person. Here tibi is emphatic, and the ne belongs more properly to it than to ego. If ego were emphatic and ne the asseverative particle of which Fleckeisen treats, we should expect this order, ne ego tibi. There is no certain proof, as we shall see later, that 7ie is found postpositive with pronouns. The reading tibine is still further supported by the fact that a conditional clause si Jidem servas follows ; but even without this it might be defended by the following remarkable passage, where ne, undoubtedly enclitic, occurs and is retained by the latest editor, Goetz, Epid. 541 (IV i, 14): Plane hicinest is, qui in Epidauro primus pudicitiarn mihi pepulit. The MSS. have hicine or hiccine ; is has been added by Goetz. Camerarius emends to hie ille est. Prof. Studemund has been kind enough to furnish me with his own reading of the Ambro- sianus, which, as it differs somewhat from that of Loewe, I will here give. The verse is divided up between two lines. At the end of the first LANE Kic . . . s . can be deciphered with space enough for HICINEST. In the second line, after a brief space, inepidauro VIRGINI PRIMUS PUDI . . . lAM PE . . ULIT. PepuHt waS probably miswritten/>^;^M/zy. '^\\k\ex primiis , which is found in the other MSS. also, or virgini, must be due to a gloss. I prefer to keep />rmwj' and to regard 540 and 541 as anapaestic septenarii. 540. Certo east, quam in Epidauro memini me pauperculam conprimere. Phil. Plane hicinest qui mihi in Epidauro primus pudicitiarn pepulit. The parallelism between these two emphatic statements will be perceived at once. There is not a shadow of a question present. The hicine is simply a stronger hie, if you choose hie enini or hie vero. Other corresponding cases of hicine I confess I have not found. The usage must have been obsolescent even in the time of Plautus. Later recensions may have removed the few cases of its occurrence, and substituted other expressions in their place. Geppert, so far , as I know, is the only modern editor who for the sake of the verse has inserted ne where there is no interrogation. I am unable to learn, from the books at my command, that he has anywhere explained the usage, or supported it by such examples based on MS. authority as I have given. For the sake of completeness I lO give the passages which in a note on Trin. 589 (see his ed. p. 163) he has thus emended: Pseud. 348 huncine, 410 hucine, 954 illici- nest, 1 175 hicine ;, Rud. 778 huncine, 1357 illicine ; Stich. 435 huncine ; Trin. 590 istucine. I cannot undertake to discuss these passages here. It is no part of my present purpose to bring forward the letter w as a rival of the ablatival d with which some editors have so liberally be- sprinkled the text of the early poets. I shall be satisfied if I succeed in shielding it in the passages where it does occur. To do this more effectually, I propose to show that an enclitic ne with affirmative force is recognized by the ancient grammarians, as may be proved from their works and from glossaries founded upon ancient sources. I have already given one such proof above. Priscian (Keil II, p. loi) says : " Dubitativae sunt, quae dubit- ationem significant, ut an, ne correpta, necyie ... (I omit the examples) frequentissime tamen eaedem interrogativae sunt, ut Virgilius in III' Aeneidos : Hec torts Andromache Pyrrhin' con- ubia servas ? Idem in X :^ tanton' me crimine dignum ? Haec eadem invenitur et pro confirmativa ut Horatius in IP sermonum : Clarus erit, fortis, iustus, sapiensne etiam et rex. Idem in I : ^ O seri studiorum, quine putetis Difficile et mirum Rhodio quod Pitholeonti Contigit. Terentius in Andria : * Nuncine demum istud verbum in te incidit. Hie enim ne conjunctio nee interrogativa, nee dubitativa, sed con- firmativa est. Virgilius in X : " Tantane me tenuit vivendi, nate, voluptas Ut pro me hostili paterer succumbere dextra? Est enim pro etiam.'" It will be noticed that in all these passages it is the enclitic 7ie of which Priscian is treating, and not the asseverative ne to which (Keil II, p. 479) he assigns the meaning -o -(hu and the circumflex accent. Priscian's use of the terms affirmativa and conjirmativa 1 Aen. Ill 319. i* Aen. X 668. 3 Hor. Sal. II 3, 97. * Hor. Sat. I 10, 21. *Cf. And. V 3, II and 14, and IV I, 59. * Aen. X 846. II may be better understood by reference to the following passages, Keil II, pp. 85, 103, 156, 243, 253, 287, 337, et al. He includes under them the Greek '^ and the Latin nam, enim, ergo, etiam, in particular phases of their use ; e. g. for enim in affirmative sense he very properly cites (Keil II, p. 104) Adelph. II i, 14 enim non sinam.. As confirm.ativa he also enumerates (II, p. 85) profecto, scilicet, quippe, videlicet and nempe. Now nempe is one of those particles in the pronunciation of which the nicest discrimination is required to settle upon the proper voice-inflection, and we get no comfort here from the editors. The very passage cited by Priscian from Persius to illustrate the affirmative sense of nempe, namely Sat. Ill i, nempe haec assidue, is given by Jahn and Duebner with a question mark, by Hart with an exclamation point, by Gildersleeve, Pretor and Heinrich with a colon. The scholiast too regards it as a question and attaches to nempe the sense of num,quid non. Acron in his commentary on Horace, Sat. I 10, i, says : " Nempe aut interrogantis aut confirmantis." In Trin. 427. (II 4.25): LE. Nempe quas spopondi. ST. immo ' quas despondi' inquito. Parens, Ritschl, and Wagner have a period after spopondi ; Fleckeisen, Geppert and Brix a question mark. Riley trans- lates "These, I suppose that I was security for?" Bonnel Thornton renders very briefly " I engaged for." Shall we then cry out that Parens, Ritschl, Wagner and Thornton are perfect dolts because they did not see, what ought to have been as plain as day, that nempe quas spopondi is a question ? or shall we call Brix an ignoramus because he did not divine that a Roman spend- thrift would certainly have dropped his voice before he reached that melancholy word spopondi? Yet Hand's treatment of Priscian is not more fair when he says, Tursellinus, vol. IV, p. 73, " Nemo vero incautius in hac re versatus est quam Priscianus. Nam postquam exposuerat ne et dubitationem significare et in interrogatione poni, loquitur de vi confirmativa, collatis exemplis prorsus alienis et male expositis ita, ut ad extremum addat : est enim pro etiam.." Let us look at these ' exempla prorsus aliena ' a little more closely. It is well to premise that when Priscian calls ne a dubitative particle he does not deny that lustitiaene prius mirer belline laborum ? is a question, and when he speaks of ne as confirmative, he refers to the intrinsic value of the particle, and to the peculiar coloring which it gives to any utterance, and he is 12 not at pains to tell us what inflection we are to give the voice. He leaves it free to the scholiast of Persius to decide that nempe is a question, but he has felt nevertheless the true force oi nempe. The Vergil passage beginning tantane, speaks for itself. It is a question and Priscian recognized it as such, and yet he interprets ne by etiam, which he elsewhere styles confirmative. So sapiensne, if I mistake not, would mean for Romans ' wise forsooth,' and might be so pronounced as to convey to one person the feeling of questioning doubt, to another of ironical assurance. Acron says of it, "Aut interrogatio audientis aut dicentis dubitatio." To my mind ne has quite the force of nempe in Hor. Ep. I i6, 31 : ' nempe Vir bonus et prudens dici delector ego ac tu.' An instructive Plautine parallel for quine is found in Epid. 449 (1114,13): Ego sum si quid vis. MIL. Nempe quem in adulescentia Memorant apud rages armis, arte duellica Divitias magnas indeptum ? Here of course the relative clause is not one of characteristic re- quiring the subjunctive, but this does not aflect the force of nempe. Acidalius changes for metrical reasons nempe quem to quemne, in which he is followed by Goetz. We shall see later on that Scaliger identifies this ne oi quemjie with the ne of /««^ (com- monly written tu ne), which every one admits to be affirmative. Porphyrion moreover says, " ne adjectum, ut egone, hine ; abundat ne syllaba." Acron quotes Priscian, but is evidently in the same confusion in which most of the modern editors are. For the very latest theory on this subject see Keller, Epilegomena zu Horaz, p. 507,' where quine is said to be the fuller form for quin and qui a modal-instrumental. Does Keller mean to assign to quine the force of ' how not'? Really the most sensible comment which I have found is that of Cruquius. " Videtur autem mihi," he says, " esse Atticism us pro "? yz '' qui utique, qui certe, nam ea ' " Dieses quine ist, was bis jetzt noch Niemand beachtet zu haben scheint, die vollere Form fiir quin, nichts anderes ; und qui ist somit nicht eigentlich der Nominativ, sondern der Instrumental-Modalis." - Cf. Lucian Charon 24 : '0 rfjc avolag, 61 ye ovk laaatv — and Ovid Fast. II 45 : A ! nimium faciles qui tristia crimina caedis Fluminea tolli posse putetis aqua! ^3 particula ye familiaris est Demostheni in ea significatione." Not that I admit any influence of Atticism. But it maybe claimed that at least in the Terence passage Priscian has most wofully blundered. His memory has indeed played him false, for he has mixed up three passages in one, but his interpretation is sound. No one will fail to be convinced of this who shall compare And. 683, Nihil ad te, DA. Quaere. PA. Em, ntincin demum ? DA. At iam hoc tibi inventum dabo, with Cas. 421 (III I, II), Meminero ST. Em ! nunc Mint te demum nullum scitum scitiust. The ne in itself is as much affirmative in the one case as e7iim in the other. Here too we have an excellent illustration of the near- ness of ne to enim, and the real worth of the gloss already given. And this brings me to the consideration of other glosses. Although Ritschl, Loewe and others have abundantly shown how much light may be shed on the signification of a word by a neglected gloss, as yet only a beginning has been made in the utilization of the riches which the old glossaries contain. It is to be hoped that before many years Loewe, whose admirable design is set forth in the Prodromus Corporis Glossariorum Latinorum (Leipzig, 1876), will provide us with the material. I have been able to examine for my special purpose most of the important glossaries to be found in the Bibliotheque Nationale in Paris, and in the Libraries of Berne and St. Gall. For the sake of convenience I will here name the principal of these. Foremost of all is the famous Liber Glossarum Cod. Sangermanensis 12, 13, Parisinus 11529 and 11530 (cf. Loewe in Prod. p. 225, Wilmanns in Rhein. Mus. XXIV, p. 367, and Usener in same vol. p. 387). For some reason Prinz, who supplied Wilmanns with data in regard to this glossary, mentions only the early part of it as in existence. The glossary does not, however, as he states, close with E ; the second tome, containing the letters F-Z, being designated by the catalogue number 11530. It con- tains 246 leaves, on the first of which is written, Antiqui glossarii pars secunda S. Germ. lat. 12, 2. The whole work is thus desig- nated in the catalogue, " Glossarium antiquissimum Ansileubi putatur." Usener, whose patience in such investigation is well known, could get no definite information in regard to this worthy bishop. The codex belongs to the eighth century. I shall refer to it as A. Cod. Bernensis i6, saec. IX = 258 IX = b li 224 (< X = c a A ,92,1 (t IX = d Parisin us 7640 (( X=e (( 7641 (1 X=f i( 7642 (( Xl=.g ** 7643 // in tyw-.'rj , zwr^ ; -va in (hi-^a derselbige, ->- in Ti->-(i(;, zi-v-i, Ti-^d, lat. ne, nae, traun, versichernd, 7ia-m in quisnam, wer doch, n in nu-n-c u. s. w., Goth, -n in hun, ains-hun.') The Latin language seems to have gone its own way in the devel- opment of this stem, and to have been more prolific than any of the cognate tongues. I owe to M. Breal the following clear state- ment regarding it (see Memoires de la Societe de Linguistique de Paris, I, p. 195,) : ''na avait a I'origine un sens demonstratif. II a donne I'adverbe interrogatif num et avec le c enclitique nmic ace. m.; la conjonction nam ace. f., la particule interrogative nc dont la flexion casuelle a tout-a-fait disparu, ou qui peut-etre n'en a jamais eu. La forme secondaire ni a donne I'accusatif nem dans nempe. De plus notre pronom s'est conserve comme enclitique sous les formes nam et num. dans quisnam et etianmumy Etymologists generally have derived nempe from nani-pe without taking the trouble to explain on what phonetic principle nam -\- pe gives nempe while nam ■\- que gives namque. Breal, as we see, assigns to nem a distinct origin. What if nem itself be the fuller form of the enclitic ne "which Breal admits may have lost a case ending? The Latin language would then have started with three particles of similar formation, nam, netn, and num, all of which were used in interrogations, though not confined to them; while nem. from the greater frequency of its enclitic use had lost its final 7n before the literary period. The independent existence of 7iem is tacitly admitted by Corssen, Ausspr. II, p. 640, when he says : " Ehe die enklitische Partikel -pe an ne^n- antrat, war das m auslautend, also schwach nachklingend ; diesen schwachen verschwindend kurzen Laut behielt es auch vor der Anfiigung -pe, wie das m von enim in der Tonverbindung von enim vero. So ward nempe mit verschwindend mattem und kurzem labialen Nasallaut gesprochen fast wie *nepe und so bei den Biihnendichtern gemessen " Nepe is found, too, in the MSS. in Trin. 328 in BCD, where Ritschl remarks m.emorabili indicio pyrrhichiacae mensurae, and in Trin. 966 in D. The vanishing of the nasal sound is perhaps still more clearly 28 proved by a gloss which I have found in the Cod. Bern. 224, saec. X, p. 217, neppe : eerie. Neppe is found according to Ribbeck in Verg. Georg. Ill 259 in c = Cod, Bern. 184, saec. IX, to which our gloss may refer. The independent existence of nem receives further support from a gloss of Festus (M tiller, p. 162), iV^-mut, nisi etiam vel nempe usus est Cato de pot. /^'-ibunicif, cum ait : " nemut . . . aerumnas." The letters italicized are due to conjecture. Paulus (Muller, p. 163) gives simply 'nemut nisi etiam, vel nempe.' Meyer and Duebner, Oratorum Romanorum Fragmenta, p. 168, assign the mutilated fragment to Cato's oration ' de Tribunis Militum,' which was prob- ably delivered in the year 171 B. C. (see Jordan's ed. of the Fragments, Proleg. p. Ixxxiv). We cannot be absolutely certain however that Verrius Flaccus did find the words in Cato. Now, as the loose compounds sicut, velut, praeut and prout presuppose the existence of sic, vel, prae and pro, so a compound nemut pre- supposes the .existence of 7iem, and it is clearly quite improbable that nam -f- ut should ever give nemut. I need not rehearse here the familiar facts in regard to the dropping of m, final in the Roman folk-speech from the earliest times. If now we posit a form tunem, for an early period we might expect to see it reduced to tune and finally to tun, just as noenum becomes noenu and finally nan. And as non is the regular form even in Plautus, occurring hundreds of times, while noenum. has good MS. authority in only one passage Aul. 67 (I i, 28, see Ussing's note and Langen Beit. p. 263) although used as an archaism by Lucilius and Lucretius, so under the most favorable circumstances we could not expect more than two or three cases of the fuller form nem in the MSS. of Plautus. If I were asked to give examples where nem ut might have been used in a pre-Plautine period, I should give the following : Egone(m) ut te adversum mentiar, mater mea (Aul. 682). Egone(m) lit cavere nequeam, quoi praedicitur (Pseud. 516). Hicgne(m) lit a nobis iioc tantum argenti aiiferat (Piiorm. 955). Niinc agitas sat tiite tuarun\ rerum ; egone(m) ut opem te mih i Ferre putem posse inopem ? (Bacch. 637). Ritschl omits ne and does not consider it a question. Dombart in Bliitt. f. bayer. Gymnas. u. Realschulwesen, 1880, p. 40, claims that nam was originally asseverative by nature : " nam, enim und vero sind ursprtinglich versichernd und bedeuten ' in der That.' Daraus entwickelte sich fruher bei nam spater bei enim eine begriindende und erlauternde Bedeutung." As the comic poets use nam quis for the classical quisnam, so nem, as is proven 29 by nempe, was not always enclitic. In an early period we might conceive of it taking the place of nam in the fpUowing sentence, Cist. IV I, ID. I give the reading of Pareus : Nam hercle ego illam anum inridere me ut sinam? satiu'st mihi Quovis exitio interire. Terence could only use nem as an enclitic and for him the m was irrecoverably lost. Compare for instance with the above passage, Eun. 771 : Hancine ego ut contumeliam tam insignem in me accipiam Gnatho ? Mori me satius est. Primarily the difference in meaning between nam and nem in such questions could not have been very great. Nam and enim often border very closely on one another, and again diverge widely in their use. Quia enim, so frequently found in Plautus and Terence, is quite different from the archaic quianam, but quite like quiane (for quianem), as I have tried to show above. So we need not be surprised to find uti7i differing both from utinam and from ut enim, in Epid. 277. EP Ut enim praestines argento, pn'usquam veniatfiliuSy cf. Cas. 165, Poen. 845. Utine in fact is nothing more nor less than an inverted nemut, to which it stands in the same relation that curnam stands to the Plautine nam, cur. I cite here all the passages of its occurrence in Plautus and Terence. Epid. II 2, 41 (225): Utin inpluvium indiita fuerit ? AP. Quid istuc tam mirabilest ? Merc. Ill 3, 15 (576): Senex hirquosus, tu aiisculere mulierem ? Utine adveniens vomitum excutias miilieri ? Rud. IV 4, 19(1063): DAE. Gripe 'animum advorte ac tace. GR. Utin istic prius dicat? DAE. Audi loquere tu. GR. Alienon prius. Phorm. V 6, 34 (874): Somnium : utin haec ignoraret suom patrem ? GE. Aliquid credito. Hec. II I, 2 (199) : Pro deum atque hominum fidem, quod hoc genus est, quae haec est conjuratio! Utin omnes mulieres eadem aeque stiideant nolintque omnia. But I must defer till another time any special discussion of the interrogative use of the affirmative ne. The cases where ne 30 strengthens a relative, of which there are some twenty-five in Plautus and Terence, are perhaps the most interesting, inasmuch as here the close relation between ne and nempe is most clearly seen. Here too as in the case of nempe we shall often have diffi- culty in deciding whether there is any real question involved. I find for instance that Bothe anticipates my view in not regarding Cist. IV 2, 6 as a question. I cite from his edition, without making myself responsible for his metre : Quamne in manibus tenui atque accepi hie ante aedis cistellam, ubi ea est, Nescio ; nisi ut opinor, loca haec circiter excidit mihi. Tyrrell in his excellent edition of the Miles, at verse 62 points out that this usage is very similar to the Hibernicism sure, " sure they both asked me." He reads quae me ambae obsecraverint, without a question, where Bentley and Scioppius read quaene ... .^ At 973 he reads quae cupiat, where Ritschl and Brix quaen cupiat ? In both cases I should keep 7ie without a question. Ussing in Epid. 444 (449 Gz.) keeps nem,pe quein in adulesce'ntia. Goetz, as we have seen, unwilling to admit nempe^ under the ictus, reads with Acidalius quem,ne. If we accept this reading, the change to nempe quern must be referred to an early revision, whose author recognized the connection between nempe and ne{in). Doubtless a Roman soldier in the first Samnite war might have said nem quem, just as our Plautine hero in v. 462 says nam quid parcam ? I have before alluded to certain points of contact between the Plautine use of enim. and of 7ic. I look upon enim not as a com- pound of nam, but of nem, or rather nini, to which it stands in the same relation as equidem- to quidem. I say of nim, for I think it probable that, as we have of is both the accusatives em and im, so nim may have existed side by side with nem, although in less com- mon use. A closer scrutiny of the MSS. may reveal a very few cases of ne = nem, but I hardly expect it. Even if some slight vanish- ing nasal sound were heard in the time of Plautus, it would probably not have been represented by any written sign, and later scribes certainly would not Introduce a sound which had died away in the language. I attach no importance, therefore, to a corruption like ei nemiruTn for eine aurum in Trin, 960 BCD, as aurum often becomes mirum in MSS." Should nem be found in A, that would ' The metrical question is still an open one, cf. Pseud. 353, 1189 Rud. 565, Bacch. 188, and Muller Plaut. Pros. p. 433 ff. ^ Cure. 10, Egone apieularum is cited by Priscian as ego nam, etc. be a different matter. I hope no one will call me rash or incon- sistent if following the MSS. quite closely I propose nim as a possible reading in two passages which have sorely vexed the editors. The first is Trin, 922, where B has ancharesancharmides mim charmides. C and D have Anchares ancharmides min charmides. See Ritschl's critical note and Brix. 3d ed. for the various emendations which have been proposed ; vtim has been changed to numne, numnam, anne, ain, num and enim. My own view receives its best illustration from Terence, Phormio 307. Demipho exclaims in anger : hominem conmonstrarier Mihi istum volo aut ubi habitet demonstrarier, to which Geta, as if pretending not to know certainly, replies, Nejnpe Phormionem ? Langen, in his very valuable article on nempe, Beitr. p. 125 ff, translates ' Du meinst wohl den Phormio ? ' and remarks : " Der Hinweis auf Phormio ist nicht so sicher das Plautus hier nempe gebraucht haben wiirde." Now it seems to me unquestionable that an earlier generation might have used nem or nim Phormionem where Terence uses netnpe, and with strong affirmative force. It is perhaps a solitary survival of this use which we have here. Ribbeck in Rhein. Mus. XXVII, p. 179, proposes : SYC ad hoc exemplumst : Char. CH Chares ? an Charmides ? SYC enim Charmides: Em istic erat. I should read nim for enim, attaching to it the same sense. Paleographically nim = min of CD. Moreover I agree with Brix 3d ed. that the future erit must be kept.' The other passage is Merc. 767, CO. Ni(m) metuis tu istanc. LYS. Sapio: nam mihi unicast. Ni metuis libri, Num metuis Camerarius Ne metuis h. e. metuisne Meursius. Ritschl reads nempe metuis. It may be a mere accident that nempe is not found in Plautus associated with the verb metuo. Enim is thus found, Pers. 319 : Enim m,etuo ut possim. reicere in bubile, ne vagMur, cf Cas. 281, Mil. 429. Some may therefore prefer to substitute enim as Ritschl has nempe for ni, but I think we shall do better to keep the simple ' Or following the MSS. still more closely we may read ; SYC. ad hoc exem- plumst an Chares an Charmides CH nim Charmides ? SYC Em istic erit. For an — an cf. Epid. 223, cf. Langen Beit. p. 266. 32 particle, from which, according to my view, the others are derived. It may have been even in Plautus' time an archaism in this usage, as noenu was for Lucretius. The bantering tone is very evident : ' Sooth you're afraid of her.' I am moreover emboldened to keep nim by the following gloss found in Cod. Bern. A 92, i, saec IX, p. 14, nim: ni, nisi, si non, of which no doubt the earliest form was nim: ni, nisi and si non having been added later to explain tii. It is easy to dismiss such a gloss with a shrug of the shoulders as the absurd attempt of some ignorant scholiast to explain away a corruption in the MSS. But it is unfair to pronounce sentence upon a gloss of which we do not know the context. No doubt many of the ' happy emendations ' of modern times, did we but know it, are only ingenious attempts to explain what for the ancients needed no explanation. So hie nunc has been substituted for hicine, and egomet for egone. Spengel in True. II 6, 52 f reads : Is te dono. PHR. Poenitetne te, quot ancillas alam Qui etiam alienas siiperadducas, quae mihi comedint cibum? Quin etid men super adducas BCD. Haupt (Hermes III, p. 229) proposes to read Quine examen super adducas. Dombart (Philol. XXVIII, p. 735) Quin etiam mi in- super adducas, which reading I accept as being nearest to the MSS. But I differ from Dombart inasmuch as I regard quin, with Haupt, as equal to quine, i. e. quine(m). It is very like Horace's ' quine putetis.' Now Kiessling, whose eminence as a Plautine critic no one will deny, comparing Eun. 1013 and Rud. 579, proposes (Jahrb. 97, p. 634) ni eiiatn examen superadducas, and we must admit with Fleckeisen that the emendation is ' very tempting.' But if my view of the passage is correct, it is quite the same as glossing ne{m) or ni(m) with nisi. So, too, in Bacch. 637 (already cited) we might substitute nisi etiam for egone{ni) ut, and still fairly represent the sense of the passage. I cannot believe that Verrius Flaccus was guilty of a worse blunder than this would be in his gloss, nem ut : nisi etiam velnempe. MiNTON Warren. .■«Mi?f!v|'.>'.-n The foregoing Dissertation was originally presented in Latin to the Philosophical Faculty of the Strassburg University and by them accepted in 1879. Owing to the writer's sudden recall to America, its publication was unavoidably deferred. In its present English form (due to practical reasons connected with its insertion in the American Journal of Philology) it has received some modifications. These reprints have the sanction of the Philosophical Faculty of the Strassburg University. CORRIGENDA ET ADDENDA. p. 3, 1. 10 ff. from bottom, put ... for period before MI. in each verse. 7, 11. 2, 4- 9 „ „ read inventus. 7, 1. 8 „ „ read repertus. 9. 1. 20 „ „ read KIC(INEE)S(T). 19, 1, 13 „ „ read Mihine? 20, 1. 17 ■ y, r, on special examination I find that J has tune} (sic.) 22, 1. 20 from top, omit sentence beginning : Perhaps, According to Studemund only V i.e. the first three strokes of the M oi Mihine} can be read in the Bembinus. 28, 11. 10, 11, 12 from bottom put interrogation-point at end of each verse. 29, 1. 8 from bottom dele period after prius. 29, 1. 3 from bottom add Hec. I, l, 9 (66). Utin eximium nemirum habeam ? SY. Neminem : ■,i<^'X . "•*' '■ ■"<•. ti f ^''f'*^. :;>ii£iS2^j^