ROMANCE LANGUAGES is the name generally adopted for the modern languages descended from the old Roman or Latin tongue, acted upon by inner decay or growth, by dialectic variety, and by outward influence, more or less marked, of all the foreign nations with which it came into contact. During the Middle Ages the old Roman empire or the Latin-speaking world was called Romania, its inhabitants Romani (adj. Romanicus), and its speech Romancium, Vulgar Romancio, Italian Romanzo, from Romanice loqui = to speak Romance; in Old French nominative romanz, objective roman(i), Modern French roman, " a novel," originally a composition in the vulgar tongue. In English some moderns use Romanic (like Ger-manic, Teutonic) instead of Romance; some say Neo-Latin, which is frequently used by Romance-speaking scholars. By successive changes Latin, a synthetical language, rich in inflexions, was transformed into several cognate analy-tical tongues of few inflexions, most of the old forms being replaced by separate form-words. As the literary language of the ancient Roman civilization died out, seemingly ex-tinguished by the barbarism of the Middle Ages, all the forms of the old classical language being confounded in the most hopeless chaos, suddenly new, vigorous, and beautiful tongues sprang forth, ruled by the most regular laws, related to, yet different from, Latin. How was this wonderful change brought about ? How can chaos produce regularity? The explanation of this mystery has been given by Diez, the great founder of Romance philology. The Romance languages did not spring from literary classical Latin, but from popular Latin, which, like every living speech, had its own laws, not subject to the changing literary fashions, but only to the slow process of phonetic change and dialectic variety. It is wonderful how like the very oldest archaic Latin is to the youngest, Modern Romance. A great number of old sounds, forms, and expressions, which were discarded or disused by classical Latin, reappear in late vulgar Latin, and live on in the modern languages. Here especially the words of Horace come true :

" Malta renascentur, quae jam cecidere, cadentque Quae nunc sunt in honore voeabula, si volet usus, Quern penes arbitrium est et jus et norma loquendi."

The present article, embracing all the Romance languages, dwells chiefly on their common origin and formation. Much of their general history has been treated under LATIN ; only some points, especially phonetic, which need a fuller discussion, are taken up again here.

We will now briefly review the fate of popular Latin through its successive stages, showing everywhere the earliest appearance of the germs of Modern Romance.

I. First (Pre-Classical) Period: to c. 80 B.C. (Cicero).

Latin, like all other literary languages, began as a living Latin popular speech. There was during this first period practi- dialec cally little difference between the vulgar and the literary language. In the oldest historical time Latin was spoken only in the small territory called Latium. The greater-part of Italy proper was occupied by the Umbro-Oscan tribes, whose languages were Italic, related to Latin, yet so different as to be unintelligible to the Romans. The two most distinct types were Umbrian in the north and Oscan in the south. The chief difference between them is that Umbrian was in a much more advanced state of phonetic decay, and was in many respects a precursor of Italian and Romance, while Oscan was still more antique than Latin. When the territories where these dialects were spoken became subject to the Romans, about the beginning of the 3d century B.C., the language of the conquerors was intro-duced, but of course modified by the speech of the con-quered. Thus two groups of provincial dialects were formed. (1) The North or Umbrian and Sabellian Latin, with which Etruscan Latin was closely connected, was peculiarly important, since it spread southwards and ex-tended to the neighbourhood of Rome; thus Falerii, Prajneste, and Tusculum spoke it. Later it spread to Northern Italy. Being really a fuller development of the tendencies of the old popular Latin and easier to pro-nounce than literary Latin, at last from the surrounding peasantry it reached the people of Rome and became the source of the modern tongues. (2) The South or Oscan Latin was Latin with some slight phonetic modifications, which, in Modern South Italian, have lived through the levelling influence of the north dialect. Between the TJmbrians and the Oscans lay the Sabellians, occupying linguistically as geographically a middle position, yet somewhat nearer to the north dialect than to the south. To the west of Umbria lay the mysterious Etruria, whose language, preserved in numerous inscriptions, has long been an unsolved riddle and is still a matter of dispute, some considering it as utterly unconnected with Italic or even Aryan, some, as Deecke and Sophus Bugge, thinking it Aryan, intermediate between Greek and Italic, but partly decayed. In the last respect it has much in common with Umbrian, but its tendency to a rapid and slovenly utter-ance is still more distinctly traceable than in Umbrian.

Of the other languages spoken in old Italy, such as Messapian, Celtic, Venetian, and Ligurian, too little is known to enable us to form an estimate of their phonetic character; but in general we see the peculiarities of North Latin penetrating more or less everywhere, in the north of Italy and in Spain (subjected at the end of the 3d century B.C.), as well as in Gaul, of which the southern part, Pro-vincia (later Provence), was subjected first, and the rest, by Caesar, in the 1st century B.C. All these countries were rapidly Latinized ; but the provincial dialects did not always follow the phonetic development of the mother speech, just as American does not always follow the changes of English.

Old Popular Latin.

A. Phonetics.

Consonants. 1. Final consonants, especially s^m, partly t, d, r, were feebly pronounced and often dropped in writing : s in nomina-tives, as Cornelio (T. Scip. ), —even Ennius wrote certissimu' nuntiu' mortis ; gen. aedem Scrapi (in a law); ace. viro' nostros (Plaut.). In verses was not counted in the metre till the manhood of Cicero, who, though he had himself followed the earlier use in his youth, calls it " subrusticum," which shows that it continued in the country. This agrees with Umbrian, where s was often dropped, as Ikuvinu Iguvini (nom. plur.), agre agri (gen.), &c. ; with North Latin, as matrona Pisaurese (nom. plur.; at Pisaurum); maxomo maximus, zenatuo (gen. at Falerii) = Old Lat. senatuos ; militare (nom. ; at Tusculum) looks quite Italian. M was often dropped, as in the well-known hone oino duonoro optumo fuise viro (T. Scip.). It was, by the express statement of Quintilian, obscurely pronounced (see LATIN), and must have been very loosely articu-lated, the lips probably only approaching, not closing, much like Polish Dqbrowslci (Dombrowski), which gives a sound between a feeble m and a nasal vowel. In Umbrian this dropping of the m is frequent, as poplo for populum, and in Faliscan sacra, datu. T, on the contrary, was generally retained in Old Latin, but frequently dropped in Umbrian, as dede dedit, hdbe habet, habia habeat; likewise in North Latin, as dede dedit, dedro and dedrot for dederunt (T. Pis.); dede (Tibur) ; cupa and cupat for cubat (Fal.). Some-times d was used, as fecid (Prameste), in Osc. fefacid, deded. In the perfects, however, d is perhaps the older sound. I) is the final consonant that was first dropped, as in in altod marid (Col. Eostr.), a Gnaivod patre, but also longd vita (T. Scip)., later fre-quently. Pate(r), soro(r), colo(r), occur in Plautus with r dropped or slightly pronounced (not counted in the metre) ; in pate(r) venit (Terence ; see LATIN) ; in Fal. uxo, mate ; in jEquian uxo; and in Ital. /rate, suora, moglie. N was dropped in tame(n) sus-picor(Ter.); Umb. nome and numem, pointing perhaps to a weak nasal vowel; Ital. nome.

1 There is nothing to prove that this s was voiced, like Eng. z; probably it was voiceless and weakly articulated, much the same as in Andalusian loQi) -padre, for Span: lospadres, "the parents."

2. D after vowels sometimes became r, as ar ad, arvorsum adversum, arfuise, arger agger,—rustic ar me advenias (Plaut., True, ii. 2) ; it was retained in classical Latin in arbiter, arcesso, meridies. In Old Umbrian d constantly changes to a peculiar sound, here transcribed d, in Later Umbrian spelt rs, as ad, ars ad, petur-pursus quadrupedibus, dupursus bipedibus. This sound appears to be only a less trilled variety of the Eoman sound. In some few instances this old r reappears later, as in Mediaeval Lat. armes-sarius (Lex Salica), in Eoum. armesariu, Ital. drgine, Ven. drzare, Span. <tree»=Lat. arger (agger). In Modern South Italian r for d is quite common, as Neap, rureee, Sicil. durici = Itai. dodici, Lat. duodecim.

3. Zfor s is found in some unintelligible fragments of the Carmen Saliare, mentioned by Varrò (L.L., vii. 26), but in the text he says that the ancients used s for r. Coz(ano), but also Cosano, is found on two old coins from Cosa, though here z may be due to Oscan influence. On the other hand, veter, quirquir, were found in the ancient augural books, perhaps Etruscan Latin (?). But the ordinary form of Old Latin was s, as in asa, honos, honosis, flosem, Lases, Fusius, Papisms, Valesius. Not z, but s or ss was used to render Greek f, —Saguntum, comissor, massa ; in this last word (Fr. masse) the present pronunciation still testifies to the old voiceless sound. Thus the whole theory of the ancient Latin z becomes somewhat doubtful ; yet it is probable that it existed in the oldest prehistoric Latin, and that forms like *easum resum were pronounced with z, like the Osc. eizazunc egmazum (same meaning). Later on s (z) passed into r, following in this the Umb. urnasiaru, pracatarum, &c. In most words Latin now got r, even in some where Umbrian retained s ; comp. Lat. ara with Umb. asa, Lat. plenarius, ordinarius with Umb. plenasier, urnasier. In Lat. nasus—nares the relation must be somewdiat like that in A.S. ceósan, ceds—curon, gecoren (choose, chose, chosen), probably owing to an original difference of accent. In many cases where Latin retained s it was originally double, as in mussa, cassus ( = cad-tus), thus spelt by Cicero and Virgil (Quint., i. 7, 20) ; formonsus and formossus ; this is the reason why Italian has always * hard (voiceless) in cosa (Ven. cossa), doloroso, &c, as Ascoli has shown. Thus, every z or voiced s having passed into r, the sound z must have disappeared from the language, and it was only much later that it was reintroduced. The sound of z seems to be especially at home on Celtic ground ; the voiceless sound still exists in the south of Italy (Rome, Naples, Sicily) and of Spain. The English pronunciation of Latin words like morose, jocose, and of Greek words like dosis, crisis, is probably due to ancient tradition.

4. N was often dropped before s, as cosol consol, cesor censor, Nasals. cosentiont (T. Scip.), mo(n)strare, Mostellaria (Plant.); even festram for fenestram (fen'stram ; Ennius) ; Alliesis dies dicebatur (Fest.); mcscs (inscr. Neap.), &c. ; nearly always formossus, for-mosus for formonsus (see 3). This agrees with Umb. Eikvasese and N. Lat. Pisaurese, Pomp, castresis, pesa, and reappears in Ital. Milanese, Genovese, mese, pesare, Fr. mois, peser. In Latin n generally did not quite disappear, but was feebly pronounced, very like French n in penser, or probably still more like the Polish nasal vowel in mieso, ' ' flesh, " geè, ' ' goose, " half vowel, half consonant. By this partial absorption of n the preceding vowel was lengthened, c6"sol. Cicero expressly mentions insanus, Gellius pênsito, the grammarians mans, mens, gens. The same was the case before/, as Infelix mentioned by Cicero ; n sometimes dropped before v, coventionid (Bacch., 187 B.c.) ; comp. Umb. kuveitu convehito, kuvertu convertito, Fr. couvent, Eng. covent, covenant. N is rarely dropped before t and d except in N. Lat. dedroit) (T. Pis. ) ; but this does not count, as the t was mute. Cicero testifies to indoctus, wdiich proves the full sound of n. The spellings tamtae, scntemtiam, damda, tuemdam, &c, of the Lex Julia (45 B.c.) are probably only analogical from cumdem, camdem, kc, pronounced n. M disappears in Poponi, Seproni, but perhaps only apparently, as Priscian tells us that medial m, as in umbra, had a middling force, not the obscure sound of final m. In most old spellings, as in the Scandinavian Runes, we frequently find nasals dropped before corresponding "mutes" or stops, where there is no suspicion of an obscured sound, e.g., in Run. LAT for land the nasal is simply understood. Likewise in Umbrian the older ustetu is shown by the Latin spelling to mean ostendu, &c. In Modern Romance only Celtic districts have produced nasal vowels.

5. if is sometimes dropped from the 2d century B.c. downwards— ' ' Parcissime ea [H litteraj veteres usi etiam in vocalibus, quum ocdos [read aedos] ircosque dicebant" (Quint., i. 5, 20); "in Latio rure edus, in urbe aedus" (Varrò), for haedus=Goïh. gaits, "goat"; likewise olerà, asta (Varrò) ; in Marsian Lat. Irtius for Hirtius, Ostilius ; comp. Umb. e,retu = heritu "velit." H must have had a rather feeble sound, something like French h aspirée.
6. Es is found assimilated in ncssum (rûsum), snssum (sûsum), &c., in MSS. of Plautus and Cato, and in some inscriptions, always mpirôsa (but prorsus) and céna, Umb. çersnatur " cenati," dos-suarius (Varrò) ; "sic et dossum per duo s . . . quidam ut levius enuntiaverunt " (Vel. Long.). Comp. Umb. Tursce, Tuscom "Tus-eum"; Ligurian Lat. suso (bronze tablet near Genoa, 117 B.c.). But this assimilation was never carried out consistently ; we have L. Lat. susum et jusum (sursum et deorsum), Ital. su e giù, older suso e giuso ; dosso, but corso, verso, orso (ursus), Fr. ours, Span. oso.

7. /was doubled between vowels, as in "antiqui maiius," which Priscian distinctly explains to have been pronounced with the sound of i consonant (Eng. y in you), and " Pompeiii ut si dicas Pom-pelli." Cicero spelt aiio, Maiiam; "quod si est, etiam jungetur [i] ut consonans" (Quint., i. 4, 11). Comp. also Osc. Pompaiians. Out of the older and popular majjus, pejjtts grew Ital. maggio, peggio, whereas classical Latin had mdjus, pejus. For filius= filjus, see 23 below, p. 664.

8. V occurs in Old Lat. aierplovere perpluere (comp. pluvia), confiovont (comp. floviom; fluvidus, Lucretius), sovo, suvo, suo, ingenuvae, clovacas. Comp. Umb. to ver and tuer, tui; Osc. suvad, &c. It reappears in Petronius, urceatim plovebat, and in Ital. pióvere, rovína, védova, Genova, Mántova, chiávica, Span. llover, Fr. pleuvoir.
9. In qu the u after q was only a lip glide, defined by the grammarians as "neither vowel nor consonant," as it did not count in the metre. It was therefore easily dropped altogether, as in Cinti for Quincti in an old inscription. This was especially peculiar to Etruscan Latin, as in the very old Aeeetiai Aequitiae, Tarcna, Etrus. Tar%nas, Tarquinius. Instances occur later every-where, as cocus for coquus, cottidie (Lex Julia), &c. ; L. Lat. cinque, cinquaginta (Edict. Diocl., 301); Ital. chi qui, che quid, cubccre coquere, cucína=vulg. Lat. coquina ( = culina), cinque, &c. But generally Italian retains qu with the old sound, quinto, quanta, Span, cuanto, but que, quinto ( = k), Fr. qu = k. Greek 10. The Old Romans, having no aspirates in their own speech, .aspirates, could not pronounce the Greek <j>, x, 8, but generally turned them into p, c, t, as Pilippus <&i\nnros, Prune Qpúvi;, colapus, Bacanal. In some words the old popular pronunciation continued later, as class, ampulla, purpura ; Silver Age, percolopabant (Petron.) ; Mediaeval, colopus, colpus, Ital. colpo, Fr. coup ; spaera, Old Ital. spera (Dante), Early Eng. spere (Chaucer), vulg. Eng. spear, "sphere"; Josepas, Ital. Giuseppe; Slepanus, Span. Esteban, Slav. Stepan.

11. Greek initial TT was often rendered b, as Burrus TIvppós, also adj., "burrum antiqui quod nunc dicimus rufum" (Fest.); hence, through bur(r)eus, Ital. bujo, "dark"; likewise buxus TTU£ÓS, Ital. busso, bosso, Fr. buis, "box tree"; from TTV%ÍS, class, pyxis, popu-lar and L. Lat. buxis, buxida, Fr. boite.

12. F after vowels is rare in Latin, but frequent in Umbro-Oscan, as Umb. prufe probe, Osc. amprufid improbe, Safinim Samnium (*Sabinium), corresponding in cognate words to Greek aspirates where Latin commonly had b : Lat. vafer, "sly,"orig. "varius"; comp. vabrum "varium" (Gloss. Isid.), Umb. vufru. The re-lation between Lat. rufus, Umb. rofu and Lat. ruber is very like that between nasus and nares, or that between veho and lingo, _ligurio as representing Gr. %_ We may suppose that / is the older and stronger common Italic form, sometimes retained in old popular Latin speech, partly by provincial influence. We find sifilare for sibilare mentioned as archaic by Nonius, as vulgar in the appendix to Probus, Ital. zufolare, Fr. sifter. I' is preserved in many pro-vincial names, as Safinius (Petron.) and Tifernum, Ital. Tiferno. We find this /again in the Italian tafáno tabanus, bufólo bubalus, Eng. buffalo. The Neapolitan attrufe, " October," has a very Osean appearance.

Vowels. 13. Old Latin often prefers short e, especially original and unaccented, to i, as in dedet,fuet (T. Scip.), cepet, refecet (Col. Rostr.), velet (Bacch.), Gondetios (inscr.), acetare = agetare (Fest.), dubenus dominus (Fest.), componeto (Cato), genetrix, mercto, calecare (inscr.), famelia, magester (Quint.). It stands for radical i, mostly orig. e, as in en, endo — Qv. iv, scmul (Plaut.)—comp. Ital. insieme, Fr. ensemble; Menerva (inscr., and mentioned by Quintilian). Sometimes it stands even for original i, as tempestatebus, aídiles (T. Scip.), navebos, navaled, but marid (Col. Rostr.), sei ques, si quis (Bacch.). This is quite Umbrian ; comp. kanetu canito, urfetam orbitam, fratreks "fratricus," facefele = \j. Lat. facibilis, Old Fr. fesible feasible, famedias familiae (nom. plur.), kvestretie quaestura, formed like Lat. segnities. In North Latin we find Etrus. Lat. Aeeetiai Aequitiae; Umb. Lat. Publece, menesterium, (T. Pis.) Apolenei, dede, Nome(lia); Mars, dedet, mercto ; and rustic "rustiei etiam nunc quoque viarn veham appellant" (Varro). E, being more .subject to be obscured than i, is not unfrequently dropped, as in cante for * cañete, class, canite (Carm. Sali.), Numtoriai (inscr. c. 290 B.C.), oinvorsei (Bacch.), unversum (Lucretius), caldus (Cato), ardus, frigdaria (Lucilius); always fert, fertc = (pépere. Popular Latin here approached the North dialect, where such syncopes were constant: comp. Umb. nomne nomine, termnu termino, postro postero ; they were still more violent in Etruscan, as Elxsentre Alexander, Sehtmnal Septimiená nata, Tar%nas Tarquinius ; Pis. Lat. dedro(t) dederunt, Lebro Liberum ; Umb. Lat. cedre cederé, &c. ;South Latin, on the contrary, favoured i (see below).

14. Of long i Lucilius distinguishes two kinds—close (i tenue), as in pila, " mortar," and open (i pingue), approaching to e, written ei, as in peilum, " spear," meilia, "thousands"; he might have added feilius, "son," comp. felare, 6*7X17. There was an original diph-thong ei, as in deicere, Osc. deicum, Gr. deiKvvvai, wdiich was early contracted into this middle sound, exactly like Eng. ey in money from Old Fr. moneie. This too is quite Umbrian, spelt ei and ee, c, as enetu inito, Arch. Lat. eneito, feliuf filios, screihtor, screhto scripti, preve privus, Ioveine — Ijovine, Iguvini; comp. the datives Juve patre, mehe, tefe (tibi), Osc. Diuvel, paterei; also Umb. Lat. /eZ=filius, and Pis. Lebro. In North Latin datives in e are gener-ally constant (Apolenei, Pis., uncontracted), as Junone matre, Jove, patre, Marte, kc., extending sometimes even farther south ; and Gallo-Lat. nise (Lex Rubr. de Gall. Cisalp.). Livy used sibe and quase (originally -ei). This exactly agrees with the rustic pro-nunciation vèlia, speca recorded by Varro, leber by Festus—"ab antiquis et ameci et amecae per e litteram efferebantur" (Fest.). And this again surprisingly agrees with the Modern Emilian (Romagnuòlo) pronunciation, as Bolognese méga mica, dég dico, mèli mille, plur. méla = Ita\. mila, véta vita, mare marito, prém primo. Thus this old rustie sound seems to have extended north-wards, but later to have been driven from its old home by the classic close i, which penetrated everywhere else, and is generally represented in all the Romance languages.

15. Old Latin often prefers short 0, especially original and un-accented, to u, as in consol (T. Scip.), tabola (Bacch.), pocolom (many old inscr.), popolum (Tab. Santina), epistola = ìiri.o-To\-i\, Patricoles, Hercoles, colpa (Prise. ), Volcanom (inscr. 3d century B. c. ), volgo and volt, &c. (Plaut. ; 0 after v continued in the classical age), sont sunt, cosentiont (T. Scip. ), denontiari (Tab. Bant. ) ; in termina-tions— Venos, epos, robose, filios, and Luciom, &c. (T. Scip.); in Umb acc. poplo, salvo, tertio ; even for orig. u—aseriato eest, observa-tum ibit ; radical—molta multa, ' ' fine," onse (h)umero, Ital. òmero, Span, ombro, somo sommo (orig. sup-mo), Ital. sómmo ; in North Latin moltaticod (T. Picen. ) ; in Fai. rnaxomo, zenatuo ; and in Ligurian Latin floviom, infumo, suso (bronze tablet near Genoa, 117 B.c.). 0 reappears in Late Latin and Italian mólto, dólce, kc Medial 0 is often dropped before I, as in omni poplo (Plaut. Pseud. ), teglarius (inscr.), Fostlus (Faustulus, 114 B.C.), Hercle (Plaut.), singlariter (Lucr. )—comp. Fr. sanglier, Ital. cinghiale—copiata (Lucr. ) ; in some cases, as poclum, periclum, the contraction does not appear till the empire ; at Pompeii anglata, subla, kc As Umbrian always has contraction, pihaclom, anglom, Treblanir, vitlu, katlu, stiplo, whilst Oscan distinguishes diminutive forms like zicolom, zicelei, ziculud from such forms as sakaraklom, pestlom, the strongest impulse must have come from the north, although there must have been a pretty general tendency to syncope every-where. In Italian contraction is the rule in popular words, as vecchio for vet'lus, Late veclus, occhio oe'lus, &c. In pòpolo, tàvola, isola, kc, and diminutives like rivolo and góndola, the original vowel has remained throughout. South Latin favoured u (see below).





16. The use of 0 for e after v is shown by Old Lat. oinvorsei, vortere ; in vostrum (Plaut., Ambrosian palimpsest), the original sound (from vos) occurs, which must have been universal in popular speech, since it reappears in Late Latin and all the Romance lan-guages—Ital. vostro, Fr. votre, kc. There was a tradition that Scipio Africanus had been the first to spell these words with e (Quint., i. 7, 25). Here Umbrian distinguished between e and 0, as in vestra, covertu (pres.), but corortus (perf.); Lat. versare is perhaps Span. vosar, bosar, "to vomit"; in other words e is now universal.

17. U occurred for i before labials ; but the old optumo has scarcely left any trace, yet Ital. has menomo besides minimo. Later u was pronounced with the sound of u, Gr. v, Fr. u, which at last was turned into i, as in optimus, Ital. ottimo. But in some few words the old popular form has survived—stupula, Ital. stóppia, Fr. étouble, ' ' stubble, " dissupare, Ital. sciupare.

18. V stands foro—" Frundes, funtes vetustissimi ; quae tamen a primoribus repudiata sunt, quasi rustico more dicta" (Prise.)— contrary to the general rule (see 15). Spellings like muns, muntem, funtcm abound in Late Latin, as puntifex (already in 98 A.D. ), 2eirTo/u.ou»<Tiif>(Plut., c. 100) ; whence Old Fr. and Old Eng. munt, "mount," Ital. mónte, pónte, fónte with close 0. In Sienese we have even pognere, ponto, Ital. and Florentine pugnere, punto (pungere, punctum) ; the connexion, however, of ó with Old Latin u is doubtful.

19. Greek v was pronounced u, as in Bruges Qpiyes, Burrus, purpura, gubernum, gubernare, cupressus,—the last four also class-ical ; otherwise the classical age adopted the Greek sound y, as in Cyprus. In popular Latin the old sound remained, and u was sometimes (before r, as in fore) even changed into 0, as in storax (Ter. ), ancora (Naevius). In Low Latin u is frequent : cuprum, " copper," for wdiich Pliny used Cyprium (aes) ; comp. bursa, Ital. bórsa, Fr. bourse, crupta, Ital. gròtta, tumba, Ital. tómba, &c.
In popular Latin there was a general tendency to contract the Diph-old diphthongs, in accordance with Umbrian and North Latin. thongs.

20. Au in North Latin began very early to be contracted into long open 0, as Pola (inscr. Picen., 218 B.c.). In Umbrian this is a constant law—ote aut, Ital. ò, Umb. toru tauros, Ital. tòro, kc. The only form of the oldest Roman Latin was au, as Taurasia, (P)aulla (T. Scip.), Claudi (Bacch.). When the Umbrian poet wdiose provincial name was Plot(u)s settled in Rome in the latter half of the 3d century B.C. his name was Latinized into Plautus. Meanwhile the North Latin ò began to penetrate into Latium as rustic, vulgar, and familiar ; Cato and Varro often use it in their books De Re Rustica, as coles (Cato), colis, orata (Varro) : " Orata genus piscis appellatili' a colore auri, quod rustiei orum dicebant, ut auriculas, oricalas" (Fest.). But in the classical age there was a good deal of reaction against this vulgarism, and there is a well-known anecdote of Mestrius Floras warning the emperor Vespasian against saying plostra, and being next day facetiously greeted by the emperor as Flaurus (Suet., Vesp., 22). At Pompeii copo took the place of caupo, o(p)scultat of auscultât. It is scarcely possible to fix the precise period at which b for au became general with the upper classes ; but, as it was never changed into ub like the original open o, the general contraction of au is probably later than that of ac to è, which was accomplished about the time of the introduction of Christianity. That the contracted sound is open o, not closed, may be inferred from the Italian ò in oro, toro, cosa, povero, in Old French chose, pbvre, as is shown by the rhymes, only in Mod. Fr. chóse, póvre (etymologically spelt pauvre). The Eng. cause, sauce, poverty, have kept the Old French sound ; the Spanish forms are cosa, oro, pobrc. The old diphthongal pronuncia-tion of au was never quite lost, but preserved in literary tradition all through the Middle Ages, as in Italian and Spanish literary words like causa. It even appears to have been popular in some Roman provinces, judging from the Prov. paubre, causa, in Mod. Prov. cbuzo (spelt by the French "coouso"), and the Poring, ouro, cousa (also oiro, coisa). Even Fr. chose, chou, presuppose causa, coulis, not cosa, colis. In Italian, words like cavolo, "cabbage," Paolo, "Paul," are rare exceptions. It is to be especially noted that cauda seems to be false, coda being the correct Latin form, which is corroborated by Ital. coda with close o. Some philologists, however, think ó is the old contraction, unconnected with ò.

21. Ae, oldest form ai, occurs very early, as in aes, praedad (Col. Rostr. ), and was general at the time of the Gracchi. The difference between ai and ae is really very slight, the sound of i being open, which is very near "raised e" (both vowels occur in Eng. pity). The pronunciation must have been essentially the same as in the Ose. at, which in the Latin transcription of the Tab. Bantina wavers between ai and ae. The contraction of ae into è first occurs in the north, where it was constant, for example Umb. kvêstûr, prê, mësiro magistra, dat. asê arae ; North Latin— Fai. pret(or), Cesilia; T. Pis. Cesala ; Etrus. Lat. Quecili (= Cecili for Oaecilii), Onevia Gnaevia, Gnaeja, &c. From the north it spread to Latium, where it first occurs as a rustic pronunciation scoffed at by Lucilius, as quoted by Varrò: "Rustici Pappum Mesium, non Maesium, a quo Lucilius tcribit—Cecilius Prctor ne rusticus fiat." "In Latio rare edus, in urbe aedus" (Varrò). We find e for ae quite commonly used in the vulgar dialect of the 1st century A.D.,—at Pompeii Ictus, queres, etate, presto, tabule, que. In Rome e became the pronunciation of the upper classes only in the Chris-tian period, 3d and 4th centuries, as Emiliano, <f>e\ae=Sliae (inscr. 3d century), Cesar (4th century). Servius (4th century) says "ë is pronounced almost like ae," that is, with the same sound, only shorter (at least in theory), equus like acquits. Vowels 22. In Old Latin compound words often did not soften their of com- radical vowel ; the preposition being then an adverb, and each pounds, word having a distinct stress, the compound was separable, as in manum endojacito (XII. Tables) = injicito ; db vossacro, vos obsecro, and sub vos placo, vos supplico, in ancrent prayers; transque dato = "et tradito," in ancient laws. This was imitated by classic poets (tmesis)—inque ligatus (Virg. ). Only some of these separable com-pounds continued later on, such as the intensifying per, in per ecastor scitus (Ter.), nobis ista sunt pergrata perque jucunda (Cic, De Or.), per mihi mirum visum est (Cic, ib.), Plafoni pier fuit familiaris (Gellius). Instances of the unchanged vowel in verbs— écquid placeant, mé rogas ? immo vero hercle pérplacent (Plaut. ), and ea mihi perplacet (Cic), to which correspond exactly the Old Fr. par est bons, par me plaist. While the older Latin accentua-tion was affativi, ddmodum, these words seem at the time of Gellius to have been pronounced affdtim, admódum. We have here a first step towards the modern system of accenting the last element of compounds ; comp. the modern Comasco ammò, Roum. amù, "just now." Quomodo is continued in Ital. come, but is formed anew in the dialectic comodo, cmbd. In Late Latin and Romance the old system reappears unimpaired ; many of the old compounds, having been lost in popular speech, were formed anew with stress on the last element and unchanged vowel : for instance, class, displïcet, vulg. displacet, Ital. dispiace, Span, despldce, Fr. déplaît ; class. dccldit, vulg. aeeddet, Ital. o<-.Me (inf. accadere) ; relego becomes Ital. rilèggo, &c

Accent 23. Stress, which in Old Latin was often farther from the end and than in classical Latin, seems early to have become pretty strong, quantity, so as to induce the voice to hurry over unaccented syllables even though they were long by position (see LATIN), as magïstratus, minlsterium, vo/Mntate, in Plautus, where st,nt were pronounced as quickly as in Eng. voluntary, magistrate. The vowels, too, were hurried over ; that fenestra was pronounced as fen'stra is corro-borated by the Ennian festra = fenstra,—comp. Ital. mcsliero = niinïsterium. Scansions like sagiitis, simïllumae show that double consonants might count as long or short ; position was not regarded so nicely as in Greek till the classical age. The real reason may have been thé stress on the first syllable. In accented syllables double consonants cannot have been really short, considering the full length of Italian double consonants in bello, anno, saétta, &c. Likewise final vowels of iambic words were shortened, as novo, habe, puta, bene, homo (see LATIN). This is important, as marking the first step towards the Romance levelling of the old quantities.

This tendency, too, was carefully restrained during the classical age, but reappears early in the post-classical period. Another effect of the strong popular stress is the pronunciation of i before vowels as a consonant ; thus words like films in Plautus often count as two syllables, filjus (j = Eng. y). This is evidently the preparatory stage to the Romance figlio, &c, with palatal I,—I and j having been fused into one sound. This pronunciation, too, was carefully suppressed by the classics.

B. Vocabulary.—We append, in chronological order, a brief selection of archaic words, disused, vulgar, colloquial, or used with a disparaging sense, in the classical age, but reappearing later as quite usual and dignified expressions. The many modern deriva-tives should be noticed.

Livius ANDRONICUS (C. 240 B.c.): sortus surrectus, Ital. sorto, sortire, Fr. sortir de l'eau. PLAUTUS (254-184) : bucea, "mouth" (fam. Cic), bucea panis (Petron.), Ital. bócca, Fr. bouche; mina-ciae, "threats," Ital. minaccia, Fr. menace; calceolarius, "shoe-maker," Ital. calzolaio. Diminutives—vidulus, vicPlus,, "wallet," Ital. valigia (*vid'litia, Diez), Fr. valise ; auricula, Ital. orecchia, Fr. oreille ; apicula, Fr. abeille ; lusciniola, Ital. usignuòlo, rosi-gnuòlo, Fr. rossignol ; sororcula, Ital. sirocchia, commonly sorella ; vitellus, Ital. vitello, Fr. veau, Eng. veal ; agnellus, Ital. agnello, Fr. agneau ; putillus, "little boy," Ital. putello (putus fam. for puer, Virg.). Adjectives—bellus (later fam. Cic)—the Romance bello has quite superseded pulcher ; minutus, "small," populus minutus (Petron. ) = Fr. le menu peuple ; rivalis, originally a law term, figuratively of a rival in love (class, acmulus ; fam. Cic); ebriâcus, "drunk," Ital. ubbriaco; sucidus, "juicy," "dirty," Ital. sueido, sùdicio, Span, siicio. Verbs—ambulare, "walk," "go," familiar in all ages, hence according to Dr. Vilh. Thomsen the Romance andare, aller, Roum. umblà, Ladino amnar ; bajulare, "carry,"bajulus, "carrier," L. Lat. "educator," Ital. bàlio, "steward," bàlia, "nurse,"Fr. bailie?; "reach," "give"; mandûcus, "glutton," manducare, "chew," "eat," frequent in Old Latin—the emperor Augustus wrote familiarly manducavi duas bucceas=lt&\. mangiai due bocconi—in the Vulgate manducat et bibit, Ital. mangiare, Fr. manger; auscultare, "listen" (once Cic, mihi ausculta), Ital. ascoltare, Span, escuchar, Fr. écouter ; cantare, frequent in Old and classical Latin, the only word in Romance [cancre has been lost) ; adjutare, Ital. ajutare, Fr. aider (adjuvare lost) ; exradicare, eradicare, Ital. sradicare, Fr. arracher ; mendicare, Ital. mendi-care, Fr. mendier; batuere, baltuere, "strike," Ital. bàttere, Fr. battre (bàtto from battilo), L. Lat. battualia, Ital. battaglia, Fr. bataille; muttire, "mutter"—palam mutire, "speak" (Ennius)— L. Lat. subst. muttum nullum emiseris, "not a word," Ital. motto, Fr. mot; sapere, "understand"—sometimes very near to the modern sense, sapit scelosta rnultum (Plaut. ) = Ital. la scellerata sa molto, nullam rem sapis = lta\. non sai nulla—Ital. sapere, Span, sabér, Fr. savoir; comedere, "eat" (Cic. fam.), comedere numos = Fr. manger son argent, Span, comer; despoliare (Cic, Ep.; Liv. once), Ital. spogliare, Fr. dépouiller; comparare, "to procure," later "to buy" (Suet.), Ital. comprare. Greek words—colaphus (colapus), "buffet," " box on the ear " (fam. Quint., Plin.), percolo-pabant (Petron.), L. Lat. colpus, Ital. colpo, Fr. coup. ENNIUS (239-169): civitas, " city " (reappears in Petronius and later), Ital. città, Fr. cité, Eng. city; campsare, "double (a cape)," a âiraii \eybij.evov, recognized by Diez in the Ital. (s)cansare, "to avoid"; nitidare, Ital. nettare ; petra, "rock," frequent in Pliny and later —in the Vulgate tu es Petrus, et super hanc Petram aedificabo ecclesiam meain—Ital. piètra, Fr. pierre. CATO (234-149) : nascere nasci, Ital. nascere, Fr. naître; fracidus, "mellow," "damaged (olives)," Ital. fràcido, fradicio, "rotten." PACUVIUS (219-129): eausari, " to plead," not classical, reappears in the Silver Age in the sense of pleading as an excuse, still later in that of disputing, discussing, Fr. causer, ' ' to chat, talk " ; comp. Ladino plidar (plead), "to speak." LUCILIUS (c. 148-103): acceptor for accipiter, "a hawk," frequent in Low Latin, Ital. astóre, F'r. autour ; quirïtare, "cry "—ut quiritare urbanorum, sic jubilare rusticorum est (Varrò; fam. Cic.)—Ital. gridare, Fr. crier. Greek words—gubernum, "rudder," Ital. governo; schedium, "improvised or unpolished poem," Ital. schizzo, "sketch"; caballus, "jade," seems first to occur here—optat arare caballus (Hor.)—later "horse," "steed," Ital. cavallo, Span, caballo, Fr. cheval, L. Lat. caballarius, "chevalier," "knight"; cyma, "young sprout of cabbage " (later Plin., Colum.), Ital. cima, "top." Q. CLAUDIUS QUADRIGARIUS _—Gellius blames several expressions of his as vulgar or rare : diurnare diu vivere, comp. Ital. soggiornare, Fr. séjourner, Eng. sojourn, and aggiornare, ajourner, adjourn ; arboretum, "grove," Ital. arboreto, Span, arboleda. VAERÒ (116-28), especially in Be lie Rustica : belare baiare, Ital. belare, Fr. bêler ; olor odor, Span, olor ; capitium, "bodice," Ital. capezzale, "cape," "pillow"; nervium = vevpiov, Span, nérvio ; rubeus, "red" (later Colum., Pallad.), Fr. rouge, Span, nibio, "fair"; badius, "bay (horse)," a rare word, Ital. bajo, Fr. bai. LABERIUS (105-43) is blamed by Gellius for using obsolete, rare, and vulgar words : lavandaria lavanda, later "laundress," Ital. lavandaja ; gurdus, "stupid," a Spanish word, says Quintilian, and in fact we find Old Span. gordo, "stupid," Mod. Span, "fat,"—comp. pinguis Minerva; pittacium, "patch," "label," = iriTTâxiov, Span, pedazo, "piece"; nanus, "dwarf,"=vâvos, Ital. nano, Fr. nain; botulus, "sausage" (also Petron.), dim. botellus (Martial), later =" bowels," Ital. budello, plur. budella, Fr. boyaux. LUCRETIUS (94-55) : baubare, "bark," Ital. abbajare, Fr. aboyer, Eng. bay; russus, "red"—o a rare word, russa gingiva (Catull. ), mentioned as usual by Gellius —Ital. rósso, "red," Fr. roux, rousse, "red-haired." Old words mentioned by GRAMMARIANS: burrus (see 11, above); artitus = bonis instructus artibus, Prov. artista, "trade," Fr. artisan; gluto (Fest. ), glutto (Persius and later), Ital. ghiottone, Fr. glouton (comp. Lat. glut glut, Fr. glouglou) ; planca tabula plana, Fr. planche, Eng. plank ; sarpere putare, "to imp," Fr. serpe, "prun-ing knife."

C. Grammar (Plautine, when no author quoted).—Gender: collus collum, Old Fr. cols, now cou ; dorsus, he. Declensions : lade lac, Ital. latte (Petron. unum laetcm biberunt) ; vasum, Umb. vaso(r), Ital. vaso ; ossum, Ital. osso, Fr. os ; pauper,-a,-um (also Petron.), Ital. povero. Comparison; magis pulcer=Span. inas hermoso ; even magis majores, as vulg. Ital. più meglio, vulg. Fr. plus meilleur, vulg. Eng. more betterier). The ordinary Rom-ance formation with plus only appears in the 3d period. With ego ipsissumus, "my very self," comp. ipsimus, "himself," i.e., the master of the house (Petron.). Out of (ego)met-ipsimus grew Ital. medesimo, Span, mismo, Fr. même. Pronouns : dat. fern, iliac for fili, Ital. le ; cccum = eace eum, Ital. ecco—pater eceum advenit= ceco che viene il padre; ccciste, eccista=Old Fr. cist, ceste, Mod. Fr. ce, cet, cette ; cecilie, eccilla = Old Fr. cil, cele, Mod. Fr. celui, celle. In Italian these were replaced by cecum istum, Ital. questo, and cecum ilium, Ital. quello, analogous to the Plautine eceum ipsum, eeeam eampse, &c. ; quotumus, "which of the number"— quotumo die hue pervenisti ? Ital. lavorare a cbttimo, ' ' work by the job." Unus indefinite was sometimes very like the modern indefinite article—Huic iilia una est, "he has a daughter," Ital. quest'uomo ha una figlia. Ille was sometimes used very like the modern definite article—Imponit geminorum alteram in nave pater, ilium reliquit alteram apud matrem domi, Ital. lasciò l'altro colla madre a casa. Note also natus nemo in aedibus, "not a living soul," "not a mother's son," Span, nadie, "nobody," likewise from nata (res), Span, nada, "nothing," whereas the Catalan pre-fers res, Prov. ren, re, Fr. rien (rem). Homo is sometimes like an indefinite pronoun—Nequior nemo, quisquamst, quern homo aut amet aut adeat, Fr. on. Verbs : moriri mori, Ital. morire, Fr. mourir. Active for deponents—jocare, Ital. giocare, Fr. jouer, partire = Ital. ; fabulare, Span, hablar. Perfects in si—parsi, prae-morsi, Ital. morsi ; the formation in si greatly increased in Rom-ance. Habere with past part, often approaches to modern com-pound tenses, Res omnis relictas habeo, Ital. ho lasciato tutto. Adverbs : Aliorsum ire for alio, Prov. alhors, Fr. ailleurs. Pre-positions : de for genitive, dimidium de praecla mihi dare, Ital. la metà del bottino ; ad for dative, te ad patrem esse mortuom re-nuntiem, Ital. dire al padre ; cum instrumental, cum virgis caseum radere, Ital. con; (feinstr., quam ilia de meis opuleniiis fiat pro-pensior, Ital. di. Peculiar local uses are in Ephesum ire, in Epheso esse, ab domo, "from home, " ex Epheso, " from Ephesus "; ab, from (at) one's house, as Esno tu ab illomilite, servus ejus = from (at) the soldier's, foris conerepuit a vicino sene, Ital. dal soldato, dal vecchio vicino.

Syntax Syntax and Phraseology.—Pater tuns is erat patruelis meus ; and comp. Fr. le père est-il, Ital. il padre lo conosco. Pone aedem ibi phrase- sunt homines, Ital. vi sono degli uomini. Scio jam quid vis dicere, ology. Ital. so (quel) che vuoi dire. Ne time, ne fac, Fr. ne fais pas cela.

Totus gaudeo, tota sum misera, even Cic. falsimi est id totum, "it is all false, " Ital. è tutto falso, son tutta contenta, Fr. elle est toute joyeuse. Multum miseri, even Cic. fam. multum bonus, Ital. molto buono, Span, muy bueno, Port, muito bom. Bene morigerus (Plaut.), lene saepe (Enn.), bene mane (Cic), Ital. ben bene, ben tosto, Fr. bien bon, bientôt, bien souvent. Bene velie alieni, Ital. voler bene ad alcuno. Epityrum estur insane bene (Plaut. ), showing that the furieusement bien of the Précieuses was no novelty. Aequo animo stare, Ital. star di buon animo ; stdbam tanquam mortuus (Petron.), Ital. stava come morto, stare in letto, star seduto, &c. Dictum (ac) factum (Ter.), Ital. detto fatto, "said and done." Cum (bona) gratia, ' ' with a good grace " (Ter. ), Fr. de bonne grâce. Id restabat (deerat, ut), used ironically, (Ter.), It. questa mancava ancora, Fr. il ne manquait (plus) que cela.

II. Second (Classical) Period: 80 B.c. to 100 (150) A.D. Second At this time we begin to have an idea of what South Latin was, of period, which the chief characteristics were—(1) its conservative character) supported by the antique forms of the Oscan language ; (2) its pre-ferring i to e, as in sinatum, cinsuerint, cintum, rim, ris, diibus (Lex Jul. ; Heraclea) ; nigatis, tenimus, Lucritius, dicembres, ocilli (Pompeii), tenimus = Mod. Neap, lenimmo ; fruminto, siptim (Pute-oli) ; venirandae (Nap.); later sicundo, siptim (Borgia, 386); (3) its preferring u to o, as in furtunilla, postcru, aliu (Pompeii). These forms are constant in Modern South Italian. Now classical Latin was an approach to South Latin, and there arose a strong reaction against the vulgar North Latin forms. But even a purist like Cicero could not always abstain from using colloquial forms, and he sometimes gives us precious information about familiar pro-nunciations, as cxm nobis for cum nobis, showing the assimilation of mn to nn. He hesitates to use the new word medietas, which later became the ordinary expression, Ital. meta, Span, mitad, Fr. moitié. Catullus has involare, "steal," Ital. involare, Fr. voler, and basium, Ital. bacio (for basio, bagio, like cacio caseus), Span, beso, Fr. baiser, and the familiar caressing issa for ipsa, Ital. essa. He and Horace have piada, "street" (for platea = irXaTeîa), Ital. piazza, Fr. place. Vitruvius has octuaginta, Ital. ottanta, a pcregre = "from abroad," and the Greek words chalare, Ital. calare, zelus, Ital. geloso, Fr. jaloux, and schidia, "wooden chip," "splinter," Ital. scheggia. The transition from b to v begins, triumphant for -bit (Lex Jul.). Gradually the popular speech undermined the classical correctness of the brief Golden Age, and at the beginning of the empire again rose to the surface. The emperor Augustus was fond of talking slang and bad grammar, as simus for minus, and he wrote as he spoke in his familiar letters, although he was very hard upon a poor legatus consularis for barbarously writing ixi instead of ipsi (Suet., Aug., 88; comp. proximus forpropsimus).





At Pompeii and Herculaneum we find a town dialect fully de-veloped, half South Latin, half vulgar Roman. Final consonants were dropped, for instance m in the ace. puella, laudata ; t in the verse "quisquís ama valia, peria qui nosci (? = nescit) amare (Pomp. ) ; s rarely, valea = valeas (Pomp. ). H, the right pronuncia-tion of which had become a mark of education, just as in English (Rusticus fit sermo, si aspires perperam ; Nigid. Figulus), was con-stantly dropped, as in abeto, abuaerit (Pomp.), abiat (Here). Nn took the place of nd : verecunnus (Pomp. ), later innulgen. (Abella, 170 A.D.), agennae (Puteoli), were generally Italic ; comp. Umb. pihaner piandi (distennite, dispennite, Plant.), Osc. opisannam operandam ; now South Italian—Roman monno, Neap, munno, Sicil. mwrmu—Ital. mondo, Lat. mundus. Ss was used for cs (x) as in words in -triss for -trix (Pomp.). <J> popularly became^), as Aprodite, but with those who tried to speak fine /, just as they made the Greek v an i—Iacintus, Amarillis, Dafne, Filetus, Ital. físico, fisonomía, &c E for oe seems first to occur during the empire, for example, pomerii (8 B.c.), pomerium (49 A.D.), Phebus (Pomp.). There is no certain evidence of the sound of Germ, o, Fr. cu ; wo see that the first vowel is absorbed by the second. According to Diez, oe becomes close e, ae open e, which is true for popular words like Ital. pena, Fr. peine, but untrue for Latinisms like fête, fèdo, amèno, cèto, commèdia, tragedia, Fèbo, oe in Low Lat. being constantly confounded with ae. Original oe never becomes ie, as ac does (ciclo, fieno).

Vocabulary.—We remark at Pompeii exmuccavit emunxit, Ital. Vocabu-dial, smoccare, "snuff (the candle)"; mi similat, Ital. mi somiglia, lary. Fr. il me ressemble, elsewhere only in Late Latin. PETRONIUS : bisaccium, Ital. bisaccia, Fr. besace; nesapius, "unwise," Ital. sávio, sággio, Span, sabio, Fr. sage ; berbex, Fr. brebis ; peduclus, Ital. pidocchio, Fr. pou (panus facit diminutivum panucula, Fest., Ital. pannocchia—comp. L. Lat. genuculum, acucula, Ital. ginocchio, agocchia, Span, hinojo, aguja, Fr. genou, aiguille); striga, "witch," Ital. Strega; fatuus, Fr. fat; basiavit me spissius, Ital. spésso, "often." Instances of Petronius's grammar are—vinus, caelus, comp. Old Fr. vins, ciels ; jacebat tanquam bovis, Ital. bove, bue; hoc vetare nec Jovis potest (Old Lat. Jovis pater = Jupiter), Ital. Giove ; munus excellente, neuter, shows the L. Lat. declension, Ital. eccellente ; habet unde—'A a de quoi (vivre) ; unus de nobis = uno di noi (see also LATIN). PLINY mentions sanguisuga as vulgar for hirudo, Ital. sanguisuga, Fr. sangsue ; he has gyrare = Ital. girare, "turn," "roam about." PERSIUS (34-62) has stloppo túmidas rumpero buceas, "a slap," L. Lat. sclopus, sclupare, Ital. schioppo, "gun." QUINTILIAN (c. 40-118): ôvvarbv quod nostri possibile nominant, quae ut dura videatur appellatio, tamen sola est, Ital. possibile, &c. JUVENAL has bucea foculam excitât, Ital. fuoco, Fr. feu, and TACITUS spatha, Ital. spada, Fr. épée (see also LATIN).

III. Third (Post-Classical) Period: 100 (150) to 300 (350).

This period is represented by the latest pagan inscriptions, by Gellius, Apuleius, &c. On the decay of classical Latin, see LATIN. In pronunciation only few decided changes appear.

1. / is inserted before s impurum, as iscripla (Afr., 197), et<nreipeiTLû='ispirito (Rome, 269); later-this was universal, but in Italian only after consonants—non istà (colloq. non stà), in ispirito, in iscritto ; Span, always, as espíritu, está, escribir, escuela ; Fr. esprit, écrire, écolo.
2. Tt is put for pt, as Selcmb. (Helvet., 219), Setimus (Afr., 3d century), later frequent. This assimilation originally came from the north ; comp. Etrus. Lat. Nethunus (th=t), Etrus. Neduns, Neptunus, Etrus. Lat. Setumnal, Etrus. Sehtmnal = Septimienà nata (inscr. biling.), Settime, Setimi, Septimius, Umb. screihtor, screhto, Ital. Settimo, Nettuno, atto aptus, cattivo, scritto.

3. Tt takes the place of et, as in Beneditus (204), latluea (Edict. Diocl., 301). This too came from the north ; vitoria occurs on an ancient mirror, probably Etruscan. Comp. Umb. subahtu and subator, formally=subacti, rehte recte, where h, originally sounded, had become mute or nearly so, which is further proved by the spellings ambrehtu, ambreta, ambito, &c. ; comp. tetto?n(e), probably = Ital. tetto tectum, just as Ital. retto rectus, atto actus, otto octo, notte noctem. This change was restricted to Italy ; elsewhere ct became it, as Fr. fait factum, Port, feito, Span, hecho ( = *feitjo). Likewise cs (x) became in Italy ss, as in Lat. cossim, vissit, elsewhere is, as Span, and Port, sois, Fr. cuisse, laisser.

4. G before a single e, i becomes palatal (like the old-fashioned Eng. gyarden, &c.) or nearly (Eng. y in you), so that both are confounded—magestati (inscr. before 243) ; cogiugi, conjugi. The sign for Jin the Gothic alphabet, taken from Latin G, shows the pronunciation of the 4th century.

5. Ngn for gn, as congnato, singno, belongs perhaps to this period—connato, &c, is also found—E. Ital. sengno, now spelt segno, with palatal n, degno dignus, regno regnum, &c, Fr. digne, &c. ; sometimes »(»), as in conóscere, Fr. connaître. There are traces of some other transitions that appear to have been completed only in the next period. Vocabu- 6. Vocabulary.—APULEIUS has sapidus, "savoury" (comp. in-lary. sipidus), Old Fr. sade, whence Fr. maussade (mal-sade) ; morsicare = Ital. ; follicare, "breathe" (like a bellows), Port, folgar, Span. holgar, "rest," "lounge''; masticare, Ital. masticare, Fr. mâcher; minari (equumbaeulis), "threaten," hence "drive," Ital. menare, Fr. mener ; cambiare (better than cambire), Ital. cambiare, Fr. changer; victualis adj., in the 6th century subst. victualia (nom. plur. ), Old Fr. and E. Eng. vitaille, Eng. vittles, spelt etymologically victuals ; aucilla, ancella for avicella, the existence of which had been denied by Varrò, L. Lat. aucellus, Ital. uccello, poet, augello, Prov. auzels, Old Fr. oisel, Mod. Fr. oiseau. GELIUS AURELIANUS (2d or 3d century) has testa, "skull," frequent in the 4th century, Ital. testa, "head," Fr. tête. LAMPRIDIUS (C. 250): papilio = pavilion, pullicenus, Ital. pulcino, Fr. poussin; pipio, "pigeon"; platea, "place." TERTULLIAN, first Christian author : rememorare, "remember," aeternalis, "eternal," compassio, "compassion," and many other modernisms, such as plus miser=Ital. più misero, Fr. plus misérable. DIOCLETIAN : fata, parca, Ital. fata, Fr. fée. SOLINUS (c. 3d century): repatriare, "to repair (to)." APICIUS (e. 3d century) : excaldarc, Ital. scaldare, Fr. échauder ; spatula, Ital. spalla, Span, espalda, Fr. épaule.

IV. Fourth Period : 300 (350) to 500 (550). Fourth This period extends from the introduction of Christianity-period, to the Middle Ages and the great migrations. Christianity marks an important epoch in the history of the Romance languages. Pagan literature was abhorred, and classical traditions lost. Popular speech got the upper hand ; the gospel was preached to the people in the people's own language, and the New Testament translated into vulgar Latin. Several phonetic changes which formerly had been wavering and uncertain became rapidly fixed and decided. The whole language was revolutionized.

1. Palatal g or j was developed into dj (like gy in Hung. Magyar =palatal d + Eng. y) or dzh (Eng. j), sometimes dz, and was con-founded with original di—Zerax, Hierax (Cumae, 202) ; TiovKiae (pagan inscr.), comp. Diuliali (568), Madias (364), also Mazas = majas, Gianuaria (503), also Zanuari, &c. Italian still retains the sound of dzh in giungere j ungere, Giove Jovem, genere, giacere, &c. Original di in Aziabenico (Afr., 195), medius, pronounced medsius according to Servius (4th century) and Consentius (5th century), Ital. mezzo, pronounced med-zo (but giorno from diurnus).

2. Ti was assibilated before vowels ; there are no certain pre-Christian instances. In the 4th century the Gothic laiktjo= lectio shows that the change had only begun ; ci and ti were sometimes confounded, as in KÓÀTIOS calceus (Plut., c. 100 ; Ed. Diocl., 301), ocio (Rome, 389) ; we have ti for si in aecletiae (early Christ, inscr. ). In the 5th century the change appears as fully accomplished ; the grammarian Pompeius (probably a Maure-tanian) expressly teaches the pronunciation Titius as Titsius, and Consentius (a Gaul) etiam as eziam, i.e., etsiam. In the 6th century we have the Gothic Icavtsjon, cautionem. It is probable that this change too originally came from the north, especially from Etruria, where it was very old : we have Etrus. Ventia Venza, Arndia Arnza, Etrus. Lat. Venosi, Vensius, Arnzius (while Umbrian has tertio, Martie, &c), Mars. Lat. Martses, evidently the source of the Latin Marsi. At length this pronunciation reached the capital and became general, mostly expressed by c, as in eciam, 1er-dus, milicia, jnsticia, stacio, nacionem, &c. The pronunciation te is confirmed for Spain by Isidore in the 7th century. This sound is still preserved by Italian in words like piazza, grazia, nazione, giustizia, likewise by Old French, as naciun, place, whereas in Spanish ts has coalesced into a lisped s, as in gracia, nación, and in Portuguese, Provençal, and Modem French it has been assimi-lated into a hard (voiceless) s, as in plaça, plassa, place.

3. Ci was assibilated before vowels. In the 4th century we have the Gothic faskja, which shows no change. Schuchardt is probably right in supposing that ci was assibilated a good deal later than ti, —an inference from the comparative rarity of instances of change, and from the silence of the grammarians. From instances like ocio (see 2) it would appear that ci and ti had the same sound, but this cannot have been the case, as Italian still keeps up a differ-ence, as in piaccia placeat, piazza platea. Yet there may have been a beginning of a palatal affection, making the two sounds somewhat like each other. In the 6th century we have the Gothic unkjanc (unciarum) and owiaa in the Ravenna charters, but on the other hand 7rpeKeia=}}recio for pretio, Urbitcius (533, which, however, may be a miseorreetion, c being added instead of substituted for t), Mauriscius on a Gaulish coin, from the same period,—making it probable that the assibilation had been accomplished. In the 7th and 8th centuries spellings like onzia for uncia are quite common.

4. The most important change of all was the victory of stress Accent over quantity and tones. The popular songs of the classical period and show a decided prevalence of accent. In the soldiers' songs from quantity, the end of the 3d century stress has quite superseded quantity and is the ruling principle of the metre :—

" Umis liómo mille, mille, mille decollavimus. Tantum villi nemo habet quantum fudit sanguinis." This is just like the metre of early Christian hymns—"Dies irae, dies ilia," and "O Rex aeterne, Domine," ka. The old quantities had been levelled. St Augustine (354-430) says that in his time only a few grammarians had kept up the distinction between mSrae and more, and that the Africans confounded os and as. Servius (4th century) could only find out quantity from stress in words like impius, amicus. Consentius says that some, especially Africans, said piper for piper and orator for orator. This does not necessarily mean that all short accented vowels before single consonants had become long ; probably both short and long coalesced to a medial quantity, rather short than long, and there was no very marked difference of quantity between accented and unaccented vowels. This appears plainly from the nature of Romance poetry, where stress has only a moderate importance—namely, in the most prominent parts of the verse — while the chief principle is the number of syllables. In Italian, especially in the Tuscan dialect, we may hear bambino with short accented i, the unaccented o being rather longer than the accented i, just as Eng. y in pity, when emphatic, is longer than i. In Spanish words like mano, primero, señora the accented vowel is generally as short as the unaccented or even shorter, which does not prevent the Spaniards from count-ing it as " dos tiempos " (duas-moras) ; the unaccented final syllables are often rather longer, as in Eng. pity. In Italian the longest syllables are those containing two or more consonants after the accented vowel, as tempo, paziènza, bocea, gatto ; the chief difference between fatlo and fdto is the length of the t in the former. In Spanish and French even long consonants are shortened, as Span. boca, gato, Fr. chatte ; jette and achète are equally short ; vowels are generally short or medial, as une belle ville (not like veal, as the English make it), and are long only by compensation, as in âne, chante (where the nasal is absorbed by the vowel). The old distinction of tone necessarily disappeared with these, the acute or high even (Roma) and the circumflex or high falling tone (Rôma) being dependent on the quantities.

5. Vocabulary.—AUSONIUS (Burdigala, 309-392) : burrae, nugae, Ital. burla, "joke." AENOBIUS (Afr.,c. 330): coquina, Ital. cucina, Fr. cuisine. FIRMICUS MATERNUS (C. 340 ?) : computus, Ital. conto, Fr. compte. PALLADIUS (C. 350): cdtus, "cat"; species=~Fr. épices. AMMIANUS MARCELLINUS (end of 4th century) : molina, Ital. molino, Fr. moulin ; impedicare empêcher ; pilare = expîlare, Fr. piller, Ital. pigliare ; proba, preuve. VEGETIUS (end of 4th cen-tury) : burgus, castellum parvum, Ital. borgo, Fr. bourg. JEROME (331-420) : camisia (Celtic), Ital. camieia, Span, camisa, Fr. chemise; carricare, Ital. caricare, Span, cargar, Fr. charger; impostor=Eng.; rancor, "rancour." THE VULGATE : grossus, Ital. grosso, Fr. gros (see also LATIN). SERVIUS (C. 390) : falco, Ital. falcone, Fr. faucon. LACTANTIUS (d. c. 330) : meridionalis. AUGUSTINE (354-430) : com-binare ; vanitare vanter. CODEX THEODOS. (438) : exagium, Ital. saggio, Fr. essai ; paraverëdus, L. Lat. palafredus, Ital. palafreno, Fr. palefroi ; acucula, aiguille. CASSIODOEUS (beginning of 6th century): modernus, "modern." CODEX JUSTINIAN. (527-565): amicabilis, ' ' amiable. ' '

V. Fifth Period: 500 (550) to c. 900 (1000). This period, which extends from the migrations to the Fifth first appearance of the modern Romance languages, is the perio4 age of Low or barbarous Latin, as written especially in charters and laws. While Christianity had brought vulgar Latin into the foreground, barbarism quite made away with literary Latin. Latin was not only dead but for-gotten. The old sounds and forms had been lost, and the scribes, ignorant of classical grammar, tried neverthe-less to make a show of learning by putting Latin forms at random. This makes the language look much more barbarous than it really was. It is sometimes very hard to find out the living popular form under the Latin varnish, and harder still to find out what is proper to each country, as Low Latin is very much the same everywhere. The following were some of the chief peculiarities. Charac- 1. Old Latin forms reappeared, especially in Italy. Final con-teristics sonants, especially m, t, s, were dropped or wrongly put. In Italy of fifth the nom. annus (-os), the acc. annum (-__), the dat. ahi. anno, period, the acc. plur. annos, coalesced in anno ; as this was the real pro-nunciation, the scribes confounded the Latin forms. There being no longer any difference between in locum and in loco, both were confounded ; even in locus, -per locus, kc, were written. The pre-valence of the type anno on the one hand and the growing use of the preposition de for the genitive on the other made the genitive anni diseppear, and anni (with which annis coalesced) was used solely fr : the plural, likewise fiori, monti. Sometimes, as in Old French, the objective form anno, Old Fr. an, was used as a geni-tive, of which JIótel-Dieu is still a remnant. In the same way Ital. rosa became the common singular, rose the common plural form. Out of Italy, final s being retained, the plural form was annos, rosas, florcs, which is the Spanish (alios) and Portuguese form, and partly the French, as roses, with fienrs and ans as accusatives. In Provencal and Old French s was kept in the nom. sing, masc, ans, but in the nom. plur. the oldest form was i, retained in some Old French monosyllables—li illi, dui, trei, twit from tutti, used in the Gloss, of Cassel ; in polysyllables this i was lost, and an remained, likewise fior (*fiori; ____. sapienti, Gloss. Cass.), munt. The verbs canto, cantos, cantat in Spanish only lost t ; in Ital. cantas through cantis (caused by the affinity of s to i) became canti. In French both s and t were kept—citante, chantes ; cliantet, later chante ; est, vit, vient, and from the analogy of these Mod. Fr. aime-t-il, chante-t-il. Many old tenses were lost, as ama(ve)ra7ii, which even in Latin sometimes had the sense of the simple past (see LATIN), only in the oldest Fr. rovéret, roga(ve)rat = rogavit. Ama(v)i be-came Ital. amai, Fr. aimai, and was used as simple past (perfectum historicum) ; the praesens perfectum was expressed by a new tense, as habeo amatum (of which early germs are found), Ital. ho amato, Span, he amado, Fr. fai aimé. The old future was destroyed and a new future formed out of amare habeo (of which early traces are found), Old Ital. ameraggio, Mod. Ital. amerò, Span, amare, Fr. faimerai, &c. In consequence of the general loss of body, short words dwindled down so much that they became unfit for use, and were superseded by fuller words or reinforced by composition ; this is especially the case in pronouns and particles. As ad became a, Lat. a, ab could no longer keep its ground, and was replaced by de, likewise vis by fortia, &c. (see LATIN). Sic disappeared and was replaced by iste, which remained uncomposed in Span, este, and in the others was composed with ecce, cecum (see above). Abante, Mod. Ital. avanti, Fr. avant, occurs in a pagan inscription. Lat. unde became de unde, Ital. donde ; in Spanish this was further composed with ad, as adonde, and as this in time came to mean " where " a new de had to be added,—de adonde, kc.

2. G became assibilated before single e, i. That during all the preceding periods _ had the sound of _ is proved by Old Latin spellings like dekem ; by Greek transcriptions, as "_____, ~Kaicrap ; by the German renderings Kaiser (Old Eng. cdsere), Keller, " cellar," Kirsche, " cherry," &c, and by the use of _ for k in Old English, as in cynne, "kin," cennan, "ken"; by late Latin spellings like ques-quenti, quiescenti, pache (Rome, 408), chingxit (Gaul, 676), vachis (Lucca, 722)—in Italian still ch — k, as chi quis, che quid ; and lastly dulkisma (Pisaur., 410), ofikina (Gallic vase, end of 6th century; see Romania, xiii. 485). Yet e before e, i must early have had a more palatal sound than before _, _, u, something like the old-fashioned English pronunciation kyard for card. But it was not until very late that the difference became so marked as to be ex-pressed by different letters : the palatal sound began to be assibilated into something like Italian ce, ci, Eng. eh (as in child from Old Eng. cild), about the 6th century, although perhaps not equally early everywhere. We find Kctrave and tzitane (Ravenna charter, 591), and paze (inscr. about the same time). Ital. c= Eng. eh in pace, dieci (in Central Italy mostly pronounced sh\ cielo ; Span, paz, dicz, cielo (lisping s from ts) ; Fr. eiel (Old ts, Mod. s). Accented S. _ for accented _ became frequent. As original I and _ were no and un- longer different in quantity, they had to be distinguished by quality ; accented acc0rdingly vinum became Ital. vino, minus (minus, open i), Ital. vowels. méno, with the same vowel as in véna from vena ; thus Latin i and I coalesced ; likewise we have scrivo scribo, bévo bibo, sevo sebum, fido fidus, fède fides, créde credit- scritto scriptum, detto dictum.

"We find traces of this in the Pomp, veces, Ital. vece. In Low Latin menus, fedes, kc, are constant. But many who had a little more learning retained the Latin spelling minus, pronouncing méno(s), and by this were induced to write constantly i for close e, especially accented, as habire, vedire, cridere — Ital. avère, vedére, crédere; vinderc, stilla=venderé, stêlla, Ital. venderé, stélla, Piedm. esteila, Old Fr. esteile, Fr. étoile. In some few instances this may represent a provincial i, as in Sicil. aviri, or special cases like Fr. tenir, plaisir, merci.

i. Unaccented e for î, which is frequent in Old Latin and not unfreqnent in the post-classical age, returns in Low Latin with re-doubled force. Thus we find fedelis (404), later Domenecus, septemus, decemus, anema, deposeta, genetor, cápete, soledos ; in the Ravenna charters vecedomeno, vendeeare, ordenata, &c. ; in Prankish charters decemo, Fr. dime, Domenecus, venerabelis, nobelis, lacrema, caretate, veretate, femena, placetus, kc. This penetrated even into Central Italy, as in the Tuscan Latin of the 8th century, dedet, placetum, homenis, inviolavelis. In Middle Italian uomeni, in Modern Italian Domeneddio, ospedale, are rare remnants of the earlier form. Gener-ally classical i has prevailed in Florentine and Italian, as uomini, femmina, anima, asino, ordine, spirito. In the Sienese dialect e has remained, as ordenare, cardenále, and in North Italian— Venet. àmeni, femena, menestra, àrdene, dseno ; Lomb. omen, asen, ordenári, kc. In Spanish it is not unfrequent—orden, úmedo, nove-dad, corredor ; comp. Prov. lagrema, semenar, conoissedor, and Old Fr. aneme, pronounced ánme (the spelling points to an older stage).

5. Unaccented medial e was often dropped, though less in Italian, except in the north, as in L. Lat. dulkisma, dulcissima (Pisaur., -410), answering to modern Romagnuôlo forms like 'sirfssm=illus-trissimo ; generally L. Lat. domnus, domna=Vcs.\. donna, "lady," "woman." In French all proparoxytones are contracted; for in-stance, Old Fr. (asinum, *as'no) asne âne, (anima, *an'ma) anmc âme, (hominem, *hom'ne) homme, (femina, *fem'na), femme.

6. Accented o for ü, not unfrequent in the preceding period, was constant in Low Latin. While Lat. lucem became Ital. luce, Lat. crucem, nucem (niicem, open u) became cróce, noce, like voce from vocem ; comp. Justus, Ital. giusto, Fr. juste, but augüstus, Ital. agosto, Fr. août. This caused many to write u for close o, as oxure uxore, genelure genitorem, although this may partly represent a provincial u, as in Sicil. amúri. In Old Fr. we have honour, favour, and still amour, and in L. Lat. cwr¿e=cohortem, *corte, Ital. corte, Fr. cour, Eng. court.

7. Unaccented o for ü as in Old Latin is frequent, as in L. Lat. tabola, popólo, secolom, regola, volontate, Ital. tdvola. pbpolo, kc. ; even for original u, as mano, spirito.

8. Unaccented medial o was often dropped before I, in which case the unfamiliar group t'l was changed into cl, as in L. Lat. oclus oculus, Ital. bechio, Span, ojo, Fr. ceil ; veclus vetulus, Ital. vècchio (Old Ital. veglio), Span, vièjo, Fr. vieil, vieux ; also L. Lat. tabla= Span, tabla, ^qpZom=Span, pueblo. In French syncope is a law— table, peuple, sangle cingulum, &c.

9. Final unaccented vowels in Italian on the whole obey the general laws. By the loss of the Latin final consonants all words end in a vowel, except such as per, con, non, in, un bel giorno, buon giorno. In South Italian i, u stand for e, o. In Spanish c is some-times dropped, as sed sitim, Ital. séte ; pared parietem, Ital. párete ; fácil, orden. In Portuguese o is pronounced u, but often only whispered ; e is nearly always whispered or mute. In North Italian, Ladino, Catalan, Provençal, and French all final vowels are generally dropped, except a, which remains in Provençal, as port, but porta (Mod. pbrto). In Catalan a is mostly pronounced as open e ; in French this becomes obscure e, which in modern French is mute except in such cases as table ronde.

10. Accented ie stands for ë, no for S. As ê and ë, S and o were levelled in quantity, they were already distinguished in quality, as in Ital. bene bene, véne venae ; Portuguese has got no further than this. But generally this was not distinct enough for the wants of the speakers, and unconsciously è (in which è from Lat. ae was included) became iè, and b (in which ô from Lat. cm was not included) became ub, both diphthongs being generally accented on the last vowel—Ital. and Span, viene, Fr. vient venit ; Ital. bubno. Span, bueno ; Ital. subno, Span, sueno (generally sonido) sonus, different from Ital. sano sunt, Span. son. This change must be very old, as it is found in nearly all the Romance languages. In Spanish, Provençal, and French ub was changed into ué, and this in Modern French to eu, ceu, contracted into the sound of o. In Spanish the use of diphthongs is extended to position—tiempo, siete septem, Ital. sètte, hierro, Ital. ferro, cuerpo, Ital. carpo, muerte, Ital. morte. Mark the displacement of stress in popular Latin— pariétem for pariëtem, Ital. párete, Span, pared, Fr. paroi ; filiblus for filiolus, Ital. figliuólo, Span, hijuelo, Fr. filleul.
11. Of mediae (voiced stops) for tenues (voiceless stops), especially between vowels, we find some few earlier instances, as grassus for ciassus, Ital. grasso, Fr. gras. This is generally rare in Italian Latin, as cubitus, cubitus, Ital. gómito, and is still compara-tively rare in Italian. Yet it occurs in some of the most familiar words, as ago aens, lago, luogo, segare, pagare (pacare, "to satisfy," "_pay"), pregare, budello, parentado, contado (comitatus), contrada, spada, strada ; p from b becomes v, as in riva, riviera, povero, savio ; before r—padre, madre, segreto, lebbra. This recalls similar changes in Umbrian, and was perhaps originally North Italian and from thence penetrated into Tuscany. In Spanish the voiced sounds are the general rule, as fuégo, Ital. fuòco, amigo, Ital. amico, amado, fluido, Ital. amato, finito, sabér, Ital. sapere, &c, yet there are some exceptions, such as poco (paucum). In Old French the soft sound is the constant law, but this mostly disappears later, as in aimé, aimée (see FRANCE, vol. ix. p. 632 sq. ). A general Romance case is placitum, Old Ital. piatto, Ital. piato, Span, pleito, Old Fr. plaid, through *plagito, *plajeto, which is singularly like the Umbrian feitu, feeiu for facitu. A limited case is that of such words as Fr. raison, Ital. ragione from rationem, probably first changed to ratsióne, *radsióne, thence to rasione, which sometimes occurs in Late Latin, and often in Early Italian, Sjian. razon ; comp. Ital. cagione occasionem.

12. I and e unaccented before vowels, especially post-tonic, became j, as filjus (Plautus), olium pron. oljo. J" coalesced with the preceding consonant, making it palatal, as in Ital. figlio, aglio, figlia, maraviglia (mirabilia), battaglia ; Span, maravilla, batalla ; Fr. fitte, mervcille, bataille (in North French the palatal sound is replaced by i or j), &c. N—Ital. vigna vinca, Spagna Hispania, ingegno, castagna, campagna, bagno balneum ; Span. Espana, cam-pana ; Fr. vigne, champagne. In some cases j was assibilated, as Span, granja, cstranjero, Fr. songe, grange, étrange.

13. Sometimes attraction takes place, i being transposed to the radical syllable, especially in French, as huile, bain, témoin, Antoine, gioire, histoire. Before and after r peculiar forms appear—Ital. primaio {-jo); muojo morior; Ital. primiero, cavaliere, from primario, probably assimilated into *primcr(i)o (Or Thomsen). Forms like planer, sorcerus are found as early as the 8th century.

14. Vocabulary.—During this period the Roman world, after being conquered by the Germanic nations, adopted many words from their conquerors. The German influence was strongest in France ; hence we find Germanic expressions for many of the most common words except form-words, though the stock of the language remains Latin. A curious instance, characteristic of the Middle Ages, is rauba, which from "robbery,™ "prey," came to mean "property," Ital. roba (robba), "things in general," Span, ropa, "linen," "stuff," Fr. robe, "gown." In Rumonsch we have la rauba c'auda mi, "the property that belongs to me. " The German h, as in hair, old hadir = "hate," is peculiar to French, and is still sounded in Normandy. Some Germanic sounds had to be modified, as w to gu, for instance, guerra from wcrra, "war" ; guanto, Old Fr. guant, Fr. gant, from want, Dan. Vanlc ; in North French io remained, whence Eng. war, &c. Among the leges barbarac, the LEX SALICA is perhaps the most remarkable as a document of Low Latin. It has sentences and words like—hoe sunt parieulas causas=ce sont pareilles choses ; si in dominica ambactia fuerit occupatus, Ital. ambasciata, Fr. ambassade ; si quis alteram voluerit occidere et colpus praetor fallierit, Ital. se il colpo fallisce; si quis alteram de sagitta toxicata percutere voluerit et praeter selupaverit (see stloppus, p. 665) ; si quis caballum extra consilium domini sui caballicaverit, Ital. caval-care, Fr. chevaucher ; si quis per malo ingenio in curte (Ital. córte, Fr. cour) alterius aut in casa (Ital. casa, "house") aliquid de furtura miserit ; companium, Ital. compagnia ; baro, baronis, "a free man," Old Fr. noni, ber, obj. barun, "hero," "baron"; diffaccrc, Ital. disfare, Fr. défaire ; excorticate, Ital. scorticare, Fr. écorcher. From OTHER LAWS we have—fortia, Ital. forza, Fr. force ; hostis exercitus, Eng. host; vassus (Celtic), "a vassal"; anca, Ital. òca, Fr. oie (from avica) ; troppus, grex (from Germ, thorp), Ital. truppa, Fr. troupe, and also Ital. troppo, "too much," Fr. trop ; forcstis (from foris), " foreign," Ital. forestiere ; marca (Germ.), "border," whence marchensis, Ital. marchese, Fr. marquis ; tornare (Greek ; also Dacian Latin ). GREGORY OF TOURS (6th century) has* pagensis, Ital. paese, Fr. pays. GREGORY THE GREAT has merces, "mercy"; fiasco, Ital. fiasco, ' ' a flask." CORIPRUS (Afr., c. 570) : cara, ' ' face " (Gr. Kdpa, "head"), Span, cara, Old Fr. chicrc, whence bonne chère, "good cheer" (originally the kind, hospitable countenance of the entertainer). VENANTIUS F'ORTUNATUS (C. 580): cofea, "coif"; crema, "cream"; viaticum, Ital. viaggio, Fr. voyage. OTHER SOURCES: caminus, "road" (Spanish Latin, 7th century ; probably Celtic), Ital. cammino, Fr. chemin ; directum, jus (Italian Latin, 551), Ital. diritto, Fr. droit. The late AGRIMENSORES : circare, Ital. cercare, Fr. chercher. OLD GLOSSES (7th and 8th centuries) : aciarium, Ital. acciaio, Fr. acier ; cosinus, consobrinus, Ital. cugino, Fr. cousin ; gamba, Ital. gamba, Fr. jambe (Gloss. Cass.). ISIDORE : trucia, Fr. fruite, Eng. trout ; cama, Span, cama, " bed " ; ficatum, " liver " (properly jceur ficatum, liver of geese fed with figs), Ital. fégato, Span, higado, Fr. foie ; and cusire, consuere, Ital. euscire, cucire, Span, coser, Fr. coudre (Gloss. Isid.); sclvaticas, Ital. sel-vaggio, Fr. sauvage ; fornuitieum, Ital. formaggio, Fr. fromage.

VI. Sixth (Last) Period. For the sixth and last period—that is, for the history and
distinctive traits of the great modern Romance languages —the reader is referred to the separate articles. lan"

1. Italian is distinguished by its harmonious form, its vocalic endings, and the rich fulness of its tones.

2. Spanish is distinguished by its regularity, by its short, distinct sounds and its fixed tones, and by many Arabic words. Certain " thick " sounds, as the j (like Dutch and South German ch, though in the south of Spain much weaker, almost h) and the lisping c, z seem to be rather modern developments than due to direct Arabic influence.

3. Portuguese is, with Gallego (the dialect of Galicia), the western dialect of Spanish, and has almost the same words, but a very different pronunciation ; in sound it approaches somewhat to French, as in the nasal vowels (which, however, are less purely vocalic than in French) and the voiced sounds of s, z, and,/. It has partly retained the Old Spanish form, as in filho for Span, hi jo, and partly it has a character of its own owing to its many obscured vowels and contractions, as boa for bona, dôr for dolor.

4. Provençal in many respects represents the earliest form of French ; in others it has peculiar developments (see PROVENÇAL). Catalan is the southern dialect of Provençal.

5. French makes up for the want of the full forms and tones of Italian by its grace and delicacy. It has more of a history than the other Bomance languages, Old French being very different from Modern.

6. Ladino (Rumonsch, Germ. Churivälsch, from the town of Chur) or Central Romance extends from the Grisons to Friuli on the Adriatic. It is not uniform, being only an agglomeration of cognate dialects ; and it is scarcely more Latin than any other Bomance language. It has chiefly been elucidated by Ascoli.

7. Roumanian has probably not survived from the old Roman colonists of Dacia, but been imported from Istria (which has a. cognate dialect) or Northern Italy. It has been greatly mixed up with Slavonic words and sounds (such as the " mixed " vowels), and has some distinctive marks, such as the post-positive article, Romunul = Romanus ille; compare the similar phenomenon in the Slavonic dialect of Bulgaria and in the Albanian language.

Literature.—The real founder of scientific Romance philology and linguistics is Friedrich Diez, in his Grammatik der romanischen Sprachen, 3 vols., Bonn, 1836-42, and Etymologisches Wörterbuch der romanischen Sprachen, 2 vols., 1852. He also published Altro- manische Sprachdenkmale}; 1846 ; Zwei altromanische Gedichte, 1852 ; and Altromanische Glossare, 1865. Pott contributed several articles on the Low Latin of the leges barbarae in the Zeitschriften of Höfer and Kuhn. Other authorities on various branches of the subject are—Ducange, Glossarium médite et infirme latinitatis, 7 vols. ; Marini, Papiri diplomatici ; Muratori, Anliquitates Italics, ; Schuchardt, Der Vocalismus des Vulgärlateins, 3 vols., 1866-68 (a valuable collection of materials), also several minor works by the same author ; and Gaston Paris, Étude sur le Pôle de l'Accent Latin dans la Langue Française, Paris, 1862, and La Vie de St Alexis (a poem of 11th century), 1872. The principal magazines devoted to the subject are—Jahrbuch für romanische und englische Literatur (ed. IWolff, Ebert, and Lembcke), later only für romanische Liter- atur ; Archiv für das Studium der neueren Sprachen (ed. Herrig), of a more popular character ; Romania (a quarterly, ed. Gaston Paris and Paul Meyer, since 1872), contains articles of the most eminent Romanists ; Revue des Langues Romanes (Montpellier, from 1870 onwards), chiefly devoted to Provençal ç Romanische Studien (ed. Boehmer) ; Zeitschrift für romanische Philologie (ed. Gröber, since 1877) ; and Romanische Forschungen (ed. Vollmöller, since 1884). Mussafia has written many articles and treatises, chiefly in the Transactions of the Vienna Academy of Sciences. Ascoli, author of Studi Gritici, has edited since 1873 the Archivio Glottologico, which has articles by Flecchia. The Rivista di Filologia Romanza (ed. Manzoni, Monaci, and Stengel, 1873) in 1878 became Giornale di Filologia Romanza (ed. Monaci). For the etymological dictionaries of the separate languages see the special articles. (J. ST.*)



The above article was written by: Johan Storm, LL.D. Edin., Professor of Philology, University of Christiana.