Latin Language

and acc. pl. in both classes), and in conjugation, among other innovations, it developed a new set of forms in -b- (e.g., amabam, "I loved," amabo, "I shall love") for the imperfect and future tenses, and in -r for the passive voice, it amalgamated the old perfect and aorist tenses, and the old subjunctive and optative moods. Finally, in syntax, the development of the elaborate and somewhat rigid principles which govern the "sequence" of tenses and the forms of indirect quotation, is perhaps the most note worthy feature.

I. The pronunciation of Latin in Italy in the classical period has been determined with approximate accuracy by the applica tion of strict philological method. Thus in the last century B.C.

the Romans never pronounced C like an English s, or G as j, but always like k and g (in get) respectively, I and V, as conso nants, were sounded like English y and w respectively (not j and v), S was always as -ss in hiss, and the vowels had roughly much the same sounds as in modern Italian.

2.

Accent. In pre-ethnic I.E. the position of the chief accent of a word (or word group) was not limited either by the quanti ties or the number of the syllables. But in an early stage of the development of Latin the I.E. accent, which in character had been predominantly musical, at least in the final period of I.E., gave way to a system of accentuation by which the initial syllable of a word was uttered with more force, and subsequently the Latin accent remained a stress accent, though before the classical period it had come to be restricted to the last three syllables of a word, the precise position being determined by quantity, and the penultimate being accented if long (e.g., amare), the antepenulti mate if the penultimate was short (e.g., amdbimus). The chief difference between this and a somewhat earlier system of accentu ation, which prevailed as late as the days of Plautus, is that in words of f our or more syllables, in which the penultimate and the two preceding syllables were all short, the fourth syllable from the end was accented (e.g., fcicilius). A few typical examples of the changes dependent upon these conditions of accent may serve to indicate their importance (the position of the accent, where indicated, is that not of classical but of pre-Plautine Latin) : (a) Syncope of short medial and final unaccented vowels, e.g., in cirdere, "to be parched," beside aridus, "dry, parched" (un syncopated), mens, "mind," from an older *mentis (*denotes a form which is not recorded, but which can be proved to have ex isted) ; syncope with vocalization of a neighbouring consonant ; e.g., amicio, "throw around," from *am-iacio (the y-sound of

the initial i in iacio "throw," has become the vowel i at the same time that the -a- was lost), sdkros, "holy," became sacer (-ros first passed into -rs, in which r is vocalic, like a French final -re, and this successively into -ers, -err and -er).

(b) Weakening of short medial vowels in unaccented syllables; e.g., Aside°, "beset," beside sedeo, "sit"; efficio, "complete" but fad°, "do, make"—contrast (before two consonants) effectus, "completed," or (before r) peperi, "have borne," but pdrio, "bring forth"; illco, "on the spot" beside locus, "place"—contrast (af ter i) pietas, "sense of duty," but old Latin *pigs (classical pius), "dutiful," or (before two consonants) venustus, "charming," but old Latin Venos, "(the goddess) love"; capitis gen. sg. "head," but caput n. sg. ; weakening of diphthongs in unaccented syllables; e.g., cecidi, "have slain," but caedo (older caid-), "slay," excliido, "shut out," but claudo, "shut." (c) Dependent upon stress accent, too, is the regular shorten ing of a long syllable when preceded by a short syllable and also immediately either preceded or followed by the accent ; thus bene, "well" (order -e), but, e.g., honeste, in which the second syllable is long.

3. In addition to those already mentioned, some of the more important changes of sound are : (A) Of the vowels and diphthongs : e followed by I became o (except before e, i, or a second 1), hence volo, "wish," volebam, volam, but velim, vellem, velle; -e before nc, ng, and gn (pro nounced ngn) became i, e.g., septingenti, "seven hundred," but septem, "seven"; -ei became i, e.g., dico, "say," older deico, and similarly ai became ae, e.g., aedes, "temple" (older ai-), but in rus tic Latin a long open e, as in eclus for haedus, "goat," oi became successively oe and e.g., lotus (older oi-) ; eu became first ou and then e.g., iiimentum, "beast of burden," old Latin pl. iouxmenta (see above), but in Greek 'eiry-vv-iu. "yoke"; and with this ou original ou also became it, e.g., lficus, "grove," lit. "a clearing," in old Latin /oucom (acc. sg.), except in rustic Latin, in which both ou and au (ordinarily preserved in polite Latin) gave d, e.g., losna, "moon" at Praeneste (see above), cognate with lux, "light," lucas, "grove," and &um, "gold" (hence Ital. oro, Fr. or) but polite aurum.