Note. The word satrapes (oarрáπηç, ov) is best declined after the first declension; but no example of the genit. sing. being satrapae is known: Nepos (Lysand., 4) uses satrapis. This does not necessarily presuppose the existence of a nominative satraps, which occurs only in later times, but may be the same as Miltiades, genitive Miltiadis. Instances of the dative satrapae, accus. satrapën, and ablat. satrape, occur in other writers, as well as in the correct texts of Q. Curtius. The form satrapem must be rejected; but the Latin form satrapam may be used. The plural is throughout after the first declension, satrapae, satraparum, &c.

CHAPTER X.

GENDER OF THE NOUNS OF THE FIRST DECLENSION.

[§ 47.] NOUNS in a and e are feminine, and those in as and es (being chiefly names of men) are masculine.

Note.-Nouns denoting male beings are of course masculine, though they end in a; as, auriga, collèga, nauta, parricida, poëta, scriba. Names of rivers in a, such as Garumna, Trebia, Sequana, Himera (to be distinguished from the town of the same name), and Hadria (the Adriatic), are masculine, according to the general rule. (See Chap. VI.) The three rivers Allia, Albula, and Matrona, however, are feminine. Cometa and planeta, which are usually mentioned as masculines, do not occur in ancient writers, who always use the Greek forms cometes, planetes; but cometa and planeta would, according to analogy, be masculine.

CHAPTER XI.

SECOND DECLENSION.

[§ 48.] ALL nouns which form the genitive singular in i belong to the second declension. The greater part of them end in the nominative in us, the neuters in um; some in er, and only one in ir, viz., vir, with its compounds, to which we must add the proper name, Trevir. There is only one word ending in ur, viz., the adjective satur, satura, saturum.*

The

*[Originally this declension had but two terminations, us for the masculine and feminine, and um for the neuter. All the forms, therefore, belonging to this declension, which subsequently ended in er, ir, or ur, terminated in early Latin in erus, irus, urus. This is plain from the remains of the early language that have reached us, as well as from other sources. Thus, in Plautus (Men. v., 5, 84) we have socerus instead of socer. same writer, and others also, employ the vocative form puere, which supposes, of course, a nominative puerus. In later Latin we have even Siler and Silerus both occurring, the former in Lucan, ii., 426; the latter in Pomponius Mela, ii., 4, 9; and with these we may compare Vesper and Vesperus. That the Latin vir arose from virus is also highly probable, and is in some degree confirmed by the existence of vira, as a feminine, in earlier Latinity. (Festus, s. v. Querquetulanæ. Serv. ad Virg. Æn., xii.,

The genitive of those in us and um is formed by chan ging these terminations into i. The vocative of words in us ends in ĕ; as, O felix anně, O happy year! In all other cases the vocative is like the nominative.

The neuters in um are declined in the same way; but in the plural they have the termination a, and the nominative, accusative, and vocative are alike in the singular as well as in the plural.

Vir and its compounds, as well as satur, simply add the terminations of the different cases to the nominative.

Some of the words in er are likewise declined by merely adding the terminations to the nominative; as, puer, puĕr-i, puĕr-o, puĕr-um, puĕr-orum, puĕr-is, puĕr-os; others reject the short e in the oblique cases; as, liber (a book), libr-i, libr-ō, libr-um, &c. Those which retain the e are not very numerous, viz., adulter, gener, puer, socer, vesper Liber (the god Bacchus), and liberi (the children, only in 468.) So Iber is only shortened from Ibērus, and satur from saturus, &c. Many Greek words likewise appear in Latin shorn of their appropriate termination; as, ȧypós (agrus, agerus), ager; Evavdpoç, (Euandrus,) Euander, &c. It may be added that, in most words of this class, the e in the middle is syncopated; as, liber, libri; ager, agri. That this is a true syncope is clearly shown by dexter, gen. dexteri and dextri. (Struve, über die Lat. Declin., &c., p. 11.)]—Am. Ed.

the plural); the adjectives asper, lacer, liber (free), miser, prosper, and tener. To these we must add the compounds of ferre and gerere; as, Lucifer, armiger, and the words presbyter, Iber, and Celtiber (plural Celtiberi). The adjec tive dexter has both forms, dextera and dextra, dexterum and dextrum, although the elision of the e is more frequent.

[$ 49.] Note 1.-The genitive of nouns, both proper and common, in ius and ium, in the best age of the Latin language, was not i, but i; as, fili for filii, and, in like manner, Appi, ingeni, imperi, consili, negoti. So, at least, it was pronounced in the poets before and during the Augustan age, as in Virgil, Horace, and Tibullus. Propertius is the first who, in a few instances, has, which occurs frequently in Ovid; and in the later poets, who preferred regularity of formation to euphony, it is quite common. (See Bentley on Terence, Andr., ii., 1, 20.) With regard to poets, the metre must determine this point; and it was in consequence of the metre that Lucretius (v., 1004), though one of the early poets, wrote nāvīgi, because otherwise the word would not have suited the hexameter. But the orthography of prose writers who lived before the Augustan age is doub ful, on account of the great discrepancy which, on this point as on every thing connected with orthography, prevails in the MSS., even in the most ancient ones of Cicero, which have recently been discovered. It is, however, probable that, although i may have been written, only one i was pronounced, as was always done in the words du and diis. The genitive mancipi for mancipii, which occurs in many legal expressions, is a remnant of the ancient practice, and remained in use in later times. Concerning the accent of these contracted genitives, and of the vocatives of proper names in ius, of which we shall speak hereafter, see above, § 33, and Bentley, 1. c. Note 2.-The following nine adjectives or adjective pronouns, unus, solus, totus, ullus, uter, neuter, alter, nullus, and alius, together with their compounds uterque, utervis, uterlibet, utercunque, and alteruter, form the genitive in all their three genders in ius, and the dative in i; in addition to which, uter and neuter eject the e preceding the r. The i of this genitive is long in prose, but in verse it is sometimes made short. (See § 16.) Alterius alone has the i short both in prose and in verse (with a few exceptions, as in Terence, Andr., iv., 1, 4; see § 850), according to the statement of Priscian, p. 694, 958. It is true that alterius cannot be used in the dactylic hexameter without the i being short, but it is used in the same manner in a trochee by Plautus (Capt., ii., 2, 56). There are only a few instances in which these words follow the regular declension. (See below, $ 140.)

[$ 50.] Note 3.-The vocative of proper names in ius ends in iinstead of ie, e. g., Antôni, Mercuri, Terenti, Tulli, Virgili. In like manner, the prope. names in iius, being sometimes softened down into ius, make the vocative in a simple i; as, Gai, Pompei. But this rule cannot be applied to prope names in ius from the Greek ɛtos, as in Arius, Heraclius; nor to thos names which are in reality adjectives, and are used as proper names only when filius, deus, or heros is understood, such as Laërtius, the son of Laërtes, i. e., Ulysses; Cynthius, Delius, the Cynthian or Delian god, i. e.. Apollo; Tirynthius, the Tirynthian hero, i. e., Hercules. All such words retain ie in the vocative; and in like manner Pius, when used as a proper name, probably formed the vocative Pie; for all common nouns and adjectives, according to the testimony of the ancient grammarians, regu larly formed their vocative in ie; as, nuntie, adversarie, impie, although there are no passages in ancient writers to prove it. But filius and genius make their vocative fili, geni, and meus (though not mea or meum) makes mi. Deus, in the vocative, is like the nominative; as, O deus! mi deus !*

*[The form dee, as a vocative, first occurs at a later period, in the Christian writers; as, for example. Prudentius and Tertullian.]—Am. Ed.

What has here been said of deus alone is applied by poets to other words also: they not unfrequently imitate the Greeks by making the vocative like the nominative, e. g., Terent., Phorm., ii., 2, 10: O vir fortis atque amicus! Horat., de Art. Poët., 292, vos, O Pompilius sanguis! Carm.. i., 2, 43, almae filius Maiae. Ovid, Fast., iv., 731, populus. In Livy, too, it occurs in some ancient formulæ ; as, viii., 9, agedum pontifex publicus populi Rom.; and i., 24, tu populus Albanus; but there is no reason for doubting the form popule, which occurs in other passages.

[§ 51.] Note 4.-The genitive plural of some words, especially those which denote money, measure, and weight, is commonly um* instead of orum, particularly nummum, sestertium, denarium, cadum, medimnum, modium, jugerum, talentum. Nummum is commonly used in this way in connexion with numerals; whereas otherwise, when it merely denotes money in general, nummorum is the usual form, e. g., tantum nummorum, acervi nummorum. There are some other words in which this is the usual form in certain combinations, such as praefectus fabrum, or socium, from faber and socius; so, also, duumvirum, triumvirum, decemvirum. Liberi and deus have both forms, liberorum, deorum, and liberum, deum. Poets indulge in still greater licenses, especially with names of nations; they say, e. g., Argivum, Danaum, Poenum, &c., instead of Argivorum, Danaorum, Poenorum, and in Livy we find Celtiberum, as well as Celtiberorum. We might point out several more isolated peculiarities of this kind; as, ephorum in Corn. Nepos, Agesil., 4. Respecting the genitive of numerals (cardinal, and especially distributive numerals), see below, Chap. XXIX. and XXX.

Note 5.-Deus has three forms in the nom. and ablat. plur., viz., dei, dii, and di, and deis, diis, and dis. The forms in i are the most usual, and in reality only one of them, since di and dis were pronounced as monosyllables (Priscian, p. 737), and are most frequently found thus spelled in the ancient MSS.

The following words may serve as exercises of declen sion: Annus, year; corvus, raven; hortus, garden; lectus, bed; medicus, physician; morbus, illness; nuntius, messenger; populus, people; rivus, brook; taurus, bull; ventus, wind. Neuters in um: Astrum, star; bellum, war; collum, neck; dolium, cask; donum, present; membrum, limb; negotium, business; ovum, egg; poculum, cup; proelium, battle; sepulcrum, sepulchre; signum, sign; tergum, back; vinculum, fetter. Those in er, genit. eri, have been mentioned above. The following are the most common among those which reject the e before the r: Ager, field; aper, boar; arbiter, arbitrator; auster, south wind; cancer, cancer, or crab; coluber, snake; culter, knife; faber, workman; liber, book; magister, teacher; minister, servant. To these must be added the proper names in er, e. g., Alexander, genit. Alexandri. The adjectives which reject

* We do not write ûm, as is done in most editions, for several reasons: 1. Because it is doubtful whether this form arose from contraction; 2. Because, according to the testimony of the ancient grammarians, no final syllable in m with a vowel before it is long (which would be implied in the circumflex), whence no one would be able to distinguish by his ear such a genitive as nummum from the accus. sing., as Quintilian, i., 6, 17, attests; and, 3. Because no accents are used in Latin.

the e are aeger, ater, creber, glaber, macer, niger, piger, impiger, pulcher, ruber, sacer, scaber, sinister, taeter, vafer.

CHAPTER XII.

GREEK WORDS OF THE SECOND DECLENSION.

[§ 52.] GREEK words in os and neuters in ov, which make ov in the genitive, are commonly Latinized in the nominative by the terminations us and um, such as the common nouns taurus, antrum, theatrum, and the proper names Homerus, Pyrrhus, Corinthus. Other common nouns, which are more rarely used, admit of both terminations in the nominative; as, arctos and arctus, barbitos and barbitus, scorpios and scorpius; and this is still more frequently the case in proper names, so that, e. g., Paros, Delos, Isthmos, and Ilion are used along with Parus, Delus, Isthmus, and Ilium. Generally speaking, however, the Greek forms belong more particularly to poets and the later prose writers. Greek names in poç, with a consonant before it, sometimes become Latinized by the termination er, and sometimes they change pos into rus, and make their vocative in ĕ. The former takes place in by far the greater number of cases, e. g., Alexander, Maeander, Teucer; the only instances in which the termination rus is found are, Codrus, Hebrus, Locrus, Petrus.* In the compounds of μéтpov, and a few others, both forms are used, as hexameter and hexametrus, though the latter occurs more frequently. Words ending in os in the nominative may make the accusative in on instead of um; as, Delon, Bosporon, Tarson. The nominative plural sometimes ends in oe (the Greek diphthong ot), as in canephoroe, Cicero, in Verr., iv., 3, 8; Locroe, Quintil., x., 1, 70.1

* [To these Schneider subjoins Myriandrus, Antandrus, hydrus, amphi macrus, diametrus, and perimetrus. (L. G., vol. iii., p. 75.)]-Am. Ed.

+[This ending belongs properly to the earlier state of the language. Thus, in Plautus (Cas. prol., 31) we have clerumena, and also (Poen., i., 1, 9) lerœ. In many MSS., too, the Comedy of Terence which we entitle Adelphi, is called Adelpho. Besides Cicero and Quintilian, however, we find it in Nepos (Miltiad., iv., 3), hemerodromo; in Pliny (H. N., 37, 10), bola; and most frequently in the names of nations and cities; as, Selle (Lucan, iii., 180); Holma (Plin., H. N., v., 27); Arimaspa (Pomp. Mela, ii., ì, 2), &c. (Consult Schneider, L. G., vol. iii., p. 82, seq.)]-Am. Ed.