prex (which occur only in the plur.), grex, lex, nux, rex, vox, Phryx, and Thrax. Fur and ren have furum, renum; lar, too, has more frequently larum than larium. Of those words which have not been noticed here a genitive cannot be proved to exist; but it is probable that the genit. plur. of vas (vădis) was vadium; and, in like manner, cor, par, and sal probably had cordium, parium, salium, in order to avoid the ambiguity which would arise from vadum, cordum, parum, salum. Cordium occurs in the Vulgate, Jerem., iv., 4.

(d) Substantives of two or more syllables, ending in ns and rs, have ium and um, though the latter occurs more rarely; e. g., cliens, cohors, Picens, Veiens, Camers; and, in like manner, those which, like adolescens, infans, parens, sapiens, serpens, are properly participles, and admit um only because they are substantives (whence we frequently find parentum, from parentes), commonly make their genitive in ium: adolescentium, sapientium, &c. The names of people in as, ātis, such as Arpinas, Fidenas, form their genitive almost exclusively in ium: Arpinatium, Fidenatium. Penates and optimates, which usually occur only in the plural, follow their analogy. Other substantives in as generally have um e. g., aetatum, civitatum; but ium also is correct; and Livy, for example, always uses civitatium. The genit. plur. ium in words with other terminations, if it should occur, must be regarded as an exception. Quiris and Samnis, however, contrary to the rule, generally make Quiritium, Samnitium.

[§ 67.] 7. Names of festivals in alia, which are used only in the plural; as, Bacchanalia, Compitalia, Saturnalia, Sponsalia, make their genitive plural in ium or orum; as, Bacchanalium or Bacchanaliorum. And Horace (Carm., iii., 5, 10), on this principle, makes anciliorum from ancile, plur. ancilia; and Suetonius, in several passages, has vectigaliorum instead of vectigalium.

8. With regard to the dative and ablative plural, it is to be remarked that the Greek words in ma prefer the termination is of the second declension to ibus. Thus, Cicero and other authors use poëmatis, epigrammatis, emblematis, hypomnematis, peripetasmatis, peristromatis, toreumatis; but ibus occurs now and then; as, diplomatibus, in Tacitus and Suetonius; poëmatibus in the Rhetor. ad Herenn., iv., 2; and in Sueton., Tit., 3; strategematibus in Frontinus, Strateg., Præf., lib. iv.

[§ 68.] 9. The accusative plural of words which make the genitive plur. in ium ended, in the best age of the Latin language, in is, which was also written eis, but not

pronounced so; e. g., artis, montis, civis, omnis, similis, mediocris. But the termination es was also in use, and in the course of time became so prevalent that is was preserved only in a few exceptions, such as tris.

Note.-Priscian, towards the end of his seventh book, discusses the accusative plur. in is instead of es, more minutely than any other ancient writer. Among modern works, see especially Norisius, in his Latinitas et Orthographia utriusque Pisanae Tabulae, which is reprinted in Cellarius, Orthographia Latina, vol. ii., p. 233, foll. ed. Harles. There is no doubt, that, until the time of Augustus, those words which form their genitive plural in ium (to which must be added celer, as in all other respects it follows the analogy of the adjectives in er, is, e, although it makes the genit. plur. celerum), had in the accusative plural more commonly the termination is than es; but it must be borne in mind that es was, at the same time, in use with is. Thus we find even in the Columna Rostrata of Duilius, clases, that is, classes, together with claseis; and in the ancient Florentine MS. of Virgil we find urbes, ignes, tres, fines, as well as urbis, ignis, &c., although es, on the whole, is not so frequent as is. (Comp. Gellius, xiii., 20.) In the newly-discovered fragments of Cicero, it is true, we generally find is in words of this kind; but there are instances, also, of es being used in the same words. The ancient grammarians in vain attempted to fix the varying practice by rules and exceptions. Pliny (ap. Charisium, p. 104, ed. Putsch.) denied the accusative funis, and Varro (ibid.) the accusatives falcis, mercis, axis, lintris, ventris, stirpis, corbis, vectis, neptis, and even urbis, and in his work, De Ling. Lat. (viii., 67, ed. Müller), he asserts that gentis alone was used, and, on the other hand, that mentes and dentes were the only correct forms. Valerius Probus (see Orthograph. Noris., p. 242) gives us to understand that the words in es, genit. is, did not form the accusative in is, although they have ium in the genitive plural. Thus much is clear, that the termination is gradually became antiquated, and that the desire of scholars to have an outward distinction of the accusative from the nominative gave way to the general practice. Charisius (p. 122, ed. Putsch.) says: consuetudo traduxit ad nominativi et accusativi formam. And this probably took place about the end of the Augustan age; for in the ancient MS. containing the fragment of the ninety-first book of Livy we no longer find the accus. in is; and in the best MSS. of the complete books, it occurs only in a few isolated passages, and Quintilian does not mention this disputed point at all. Afterward is was still sometimes used by Tacitus and Gellius; but with Tacitus this arose from his desire to revive the ancient power and energy of the language, and with Gellius from his antiquarian studies. This is not the place to inquire in wha manner an editor of ancient authors has to act in the face of this obvious inconsistency of the writers themselves; there are few who faithfully fol low the authority of the MSS.; others, such as Bentley, in his Terence and Horace, everywhere restore the accus. in is (why Bentley, without incon sistency, edited arces and rates in Horace, has not yet been examined) and most of them pay as little attention to the difference in doubtful cases as to the ancient orthography in general, but merely follow the vulgar tra dition. We have noticed here the difference of opinions to caution the student, that, in reading the ancients, he may not confound the short is of the genit. sing. with the long is of the accus. plur.

[§ 69.] 10. Juppiter (which was more common than Jupiter) is declined as follows: genit. Jovis, dat. Jovi, accus. Jovem, voc. Juppiter, abl. Jove. In the plural Joves only is found.*

+

[Jupiter, gen. Jovis, is to all appearance very irregular; but there is

F

Bos, bõvis, makes the nominat. and accus. plur. boves, gen. bum, dat. and ablat. būbus, and less frequently bōbus. Sus makes the dat. and ablat. plur. subus, which is a contraction of the less frequent form suibus.

CHAPTER XVI.

GREK FORMS IN WORDS OF THE THIRD DECLENSION.

[§ 70.] A GREAT number of Greek words, especially proper names, belongs to the third declension; and as their genitive terminates in oç (εws, ovs), they follow the third declension in their own language also. Among the terminations of the nominative mentioned above, some belong exclusively to Greek words, viz., ma, i, y, ăn, in, ōn, ÿn, ēr, ÿr, ys, eus, yx, inx, ynx, and the plurals in e; but there are also Greek words with other terminations, most of which, however, are quite treated as Latin words, for which reason the termination on is generally Latinized into o (see above, § 56), and the Greek forms are used by Latin writers, especially the poets, only in some cases.

1. In the genitive singular, the poets frequently use the Greek termination os instead of the Latin is, especially in words in is which usually make their genitive idis, whether simple or derivative (see § 245), e. g., Daphnidos, Phasidos, Atlantidos, Erymanthidos, Nereïdos; so also in nouns in as and ys; as, Pallados, Tethyos; and in eus; as, Pelēõs, Theseos (Ovid, Metam., viii., 268), although the Latin termination ei or contracted ei (according to the second declension), as in Thesei, Terei, is more commonly used. (See above, Chap. XII., 4.)

But in prose the Greek termination of the genitive is seldom used. Substantives in is derived from verbs in particular, such as basis, ellipsis, mathesis, poësis, make their genitive like the nominative, and not baseos, matheseos, &c., which forms are found only in unclassic writers. here in reality a blending of two forms of declension. According to Priscian (6, p. 695, Putsch.), the regular genitive is Jupiteris, or Jupitris. On the other hand, the genitive Jovis, as well as the other oblique cases, are to be traced to a nominative Jovis, which occasionally occurs, and of which Varro makes mention. (L. L., vii., 38.) The stem of this appears to be Jov, or rather Jou, which, with the Latin deus, the Æolic Aɛuç, the common form Zeus, the Oriental Ja, Jao, Jehovah, &c., points to one and the same origin. (Compare Müller, Etrusker, vol. ii., p. 43. Buttmann Mythologus, vol. ii., p. 74.)]—Am. Ed.

(See Vitruv., X., 15. Spartian. Ael. Verus, 3; Sever., 3.) In the few words in y the genit. in yos is used for the sake of euphony, e. g., misyos. Pan, the shepherds' god, admits the Greek genit. Panos in prose, to distinguish the word from panis, bread.*

The feminines in o, however, such as echo, Calypso, Dido, Io, Sappho, have usually the Gresk genitive in ûs; as, echus, Didus, Sapphus, the Latin termination onis being less common. Their dative, accusative, and ablative end in o, and the Latin terminations oni, onem, one, are but rarely used.

[§ 71.] 2. The Greek accusative of the third declersion in a is very often used by the Latin poets instead of em. Thus, Horace uses only heroa, Cyclopa, Memnona, Agamemnona, Helicona, Chremeta, and not Cyclonem, Agamemnonem, &c. Among the prose writers, Cicero most studiously avoids the Greek termination, except in aër, aether, and Pan, of which he makes the accusative aëra, aethera, and Pana (for the reason mentioned above). In all other instances the Greek accusative in a must be looked upon, in Cicero, as an exception. It occurs much more frequently in Nepos, Livy, Curaus, and the authors of what is called the Silver Age, though principally in proper names and along with the common Latin termination em, e. g., Babylona, Eleusina, Lacedaemona, Marathona, Parmeniona, Sidona, Timoleonta, Troezena, also Periclea, Stratoclea, and similar names ending in the nominative in cles. In like manner, words in is and ys admit, even in prose, the Greek forms in and yn, together with the Latin im and ym, but Cicero uses them only by way of exception; Livy and Curtius have them more frequently, e. g., Nabin, Agin, Halyn, Tigrin. The accus. Eleusin, instead of Eleusinem (a), must be traced to the form Eleusis, gen. is, which, however, is not well attested. For the accusative of words in eus, which later writers usually make ea; as, Persea, Demetrium Phalerea, see above, Chap. XII., 4.

* [But by no means to the exclusion of Panis. (Consult Schneider, L. G., vol. iii., p. 285.)]-Am. Ed.

[These two accusative forms, aëra and æthera, appear the more remarkable in Cicero, when we compare them with his own language on another occasion: "aër; Græcum illud quidem, sed receptum jam tamen usu a nostris: tritum est enim pro Latino.... æther; mutuemur hoc quoque verbum, dicaturque tam æther Latine quam dicitur aer." (Cic., N. D., îi‹, 36, 91, seqq.)]-Am. Ed.

Proper names in es, which in Greek follow the first declension (gen. ov), and in Latin the third (gen. is) (see Chap. IX., 3), have in the accusative the termination ēn along with that in em, e. g., Aeschinen, Achillen, and Ulixen (inasmuch as these names are not formed from Αχιλλεύς and Οδυσσεύς, but from the less common Αχίλλης and Οδύσσης, ου), and especially barbarian names, such as Mithridaten, Phraaten, Xerxen, Araxen, Euphraten. The termination en for em is, moreover, found in those compounds which in Greek follow the third declension, but in the accusative admit of ŋv and ʼn (contracted from ɛa); but en is used much less frequently. Instances of this kind are, Sophoclen, in Cic., De Off., i., 40; Hippocraten and Epicyden, in Livy. Some words are in Greek declined in two ways, either after the first or after the third declension, such Θαλῆς, Χρέμης, gen. ov and ητος; in Latin they may have the shorter form and yet follow the third declension (e. g., the ablat. Thale), and in the accusative they admit also of the termination en, e. g., Chremetem and Chremen, Thalem or Thaletem and Thalen.

[§ 72.] 3. The vocative singular is in most Greek words like the nominative; but those ending in s form a distinct vocative by rejecting that consonant, both in Greek and Latin. Thus, the vocative of words in is, ys, eus: Daphni, Phylli, Thai, Coty, Tiphy, Orpheu, Perseu. Words in is, idis, however, make the vocative just as often like the nominative; as, Bacchis, Mysis, Thais. Nouns in as, antis, make their vocative in Greek ăv and ā, but the latter only is used in Latin, e. g., Atla, Calcha.

Proper names in es, gen. is, have the vocative of the first declension in e, together with the regular one. This is the case with those which in Greek follow the first declension (e. g., Carneade, Simonide, and Achille, see above), and with those which, although they follow the third in all other respects, yet admit of the accusative in nv. Thus, we sometimes find Damocle, Pericle, Sophocle,. Socrate.

[§ 73.] 4. The plural of those Greek proper names, which by the forms of their accusative and vocative sing. show their tendency to follow the first declension, is sometimes formed after that declension. Thus, we find in Cicero, De Orat., ii., 23, the rom. Naucratae; and Orat., 9, the accus. Thucydidas.