pecially those of Freund, in an appendix to the preface to his Latin Dictionary) compel us to give up this opinion, especially with regard to the genitive; for it is only in late technical writers that we find, e. g., cornu cervinum and cornu bubulum making the genitive without any termination of the first word: cornucervini, cornububuli. The dative ui is likewise mentioned by an ancient grammarian (Martian. Capella, lib. iii.), but there is no instance except cornu in Livy, xlii., 58, which must be looked on as a contraction of cornui.

[$81.] Note 1.-The genitive of the words in us was originally uis, which was afterward contracted into us. Instances of the ancient form are still found in our authors; as, anuis in Terence. Sometimes, on the other hand, the genitive of words in us was i, after the second declension, which is still found now and then, as well as us, not only in comic writers, but in good prose, e. g., senati and tumulti in Sallust. The dative in u instead of ui is still more frequent, especially in Cæsar, who is said by Gellius (iv., 19) to have sanctioned this form exclusively; e. g., equitatu, magistratu, usu, for equitatui, &c.; it is, however, found also in a few passages of other writers.

[§ 82.] Note 2.-Some words make the dative and ablative plural in ubus instead of ibus. They are contained in the following two hexameters: Arcus, acus, portus, quercus, ficus, lacus, artus,

Et tribus et partus, specus, adde veruque pecuque.

But it must be observed, that instead of ficubus a better form is ficis, from ficus, i (see § 97), and that arcubus and quercubus, though mentioned by ancient grammarians, do not occur in other writers any more than arcibus, or quercibus. Portus has both forms, ubus and ibus, and tonitrus has more commonly tonitribus than tonitrubus.

[§ 83.] Note 3.-Domus takes, in some of its cases, the forms of the second declension; but this is exclusively the case only in the genit. domi in the sense of "at home;"* in the abl. domo in the sense of " from home;" and in the acc. plur. domos in the sense of "home," when several places are alluded to. In the other signification, the forms of the fourth declension prevail, though we find the ablat. domo, genit. plur. domorum, acc. plur. domos, along with domu (see Garatoni on Cic., Philip., ii., 18), do muum, and domūs (see my note on Cic. in Verr., iv., 4); but domo for domui seldom occurs.

GENDER OF WORDS OF THE FOURTH DECLENSION.

[§ 84.] The words in us are masculine. The following only are feminines: acus, domus, manus, porticus, tribus, and the plurals idus, iduum, and quinquatrus, quinquatru

um.

To these must be added colus, which, however, also follows the second declension. (See § 53 and 97.) The words anus, nurus, socrus, and quercus are feminine, according to the general rule, on account of their signification.

Note.-Penus, us (provisions), is feminine; but there are two other

[Domi, "at home," is in fact not a genitive, but an old locative case Compare Anthon's Greek Prosody, p. 227, seq.]—Am. Ed.

Forms of this word, one after the second declension, penum, 1, and the second after the third, penus, oris, both of which are neuter. Specus is most frequently masculine; but in the early language, and in poetry, it is found both as a feminine and as a neuter. In Valer. Maximus, i., 2, we have in quoddam praealtum specus for in quendam specum; but the reading is doubtful. Secus, when used for sexus, is neuter, but occurs only in the nominat. and accus. in the connexion of virile and muliebre secus. (Com pare $89.)

The few words in u are neuter, without exception.

CHAPTER XXI.

FIFTH DECLENSION.

[§ 85.] THE fifth declension, like the fourth, may, with a few changes, be traced to the third. The nominative ends in ēs, and the declension is as follows:

Note 1.-Only the three words dies, res, and species have their plural complete; and Cicero condemned even specierum and speciebus as not being Latin. The words acies, facies, effigies, series, and spes are found in good prose writers only in the nominative plur. (perhaps in the vocative also) and accus. plur.; the others have, from their signification, no plural.

Note 2.-The é in the termination of the genitive and dative singular is long when preceded by a vowel, as in diei, maciei, but short in spei, common in fidei and rei.

Note 3-An old termination of the genitive was es (contracted from eïs), but is not found in our authors, except in the word Diespiter = Diei pater. But there are several instances of è and i being used for the ei of the genitive and dative. The è for the genitive occurs very frequently in poetry (Virg., Georg., i., 208, die. Horat., Carm., iii., 7, 4; Ovid, Metam., iii., 341, and vii., 728, fide); and also in some passages of Cicero, Cæsar, and Sal lust; e. g., pernicie causa (some write pernici), in Cic., pro Rosc. Am., 45. In sinistra parte acie in Cæs., Bell. Gall., ii., 23, and several times in Sallust. Instances of the dative ending in e occur in Horace, Serm., i., 3, 95, commissa fide; and in Livy, v., 13, insanabili pernicie nec causa nec finis inveniebatur. The dative in occurs in Nepos, Thrasyb., 2: pernicii fuit; and the genitive in i appears in Livy, ii., 42, in the connexion of tribuni plebi for plebei (plebes plebs).

GENDER OF WORDS OF THE FIFTH DECLENSION.

[§ 86.] The words of the fifth declension are feminine, with the exception of dies, which is mascul. and femin. in the singular, and masculine only in the plural. The compound meridies is masculine only, but does not occur in the plural, as was remarked above.

G

Note.-Good prose writers make the singular of dies much more fre quently masculine than feminine. The latter gender, gei erally speaking is used only when dies denotes duration or length of time, and in the sense of a fixed or appointed day. Thus we find certa, constituta, praestituta, dicta, finita dies, but also stato die.

CHAPTER XXII.

IRREGULAR DECLENSION.-INDECLINABLES.-DEFECTIVES.

[§ 87.] THE irregularities in the declension of substantives may be comprised under two general heads: A. Indeclinables and defectives; B. Heteroclita and heterogenea.

A. Some substantives have a defective declension, inasmuch as they have either no terminations at all to mark the different cases (indeclinables), or want particular cases, or even a whole number (defectives).

I. Indeclinables, or words which retain the same form in all cases, are chiefly the names of the letters of the Greek and Latin alphabets, e. g., alpha, beta, gamma, digamma, delta, iota, a, c, v, &c. It is only late and unclassical authors that decline the Greek names in a. Delta, as a name of a country, is likewise indeclinable; but it is found only in the nomin. and accus. Farther, a number of foreign words, such as git, manna, pascha, and a few Greek substantives in i and y, such as gummi and misy, which, however, occurs also as a declinable word (see § 55); and besides the indeclinable gummi there exist other declinable forms also, e. g., haec gummis, hoc gumma, and hoc gumen. Hebrew proper names, which differ in their terminations from Greek and Latin words, are either not declined at all, as Bethleem, Gabriel, Ruth, or they take a Latin termination in the nominative also, e. g., Abrahamus, Jacobus, Josephus, Juditha. David and Daniel are the only names which, without taking any termination in the nominative,* make the genitive Davidis and Danielis. Others, as Joannes, Moses, Judas, Maria, have already acquired, through the Greek, a declinable termination, and are accordingly declined after the first or third declension. Jesus makes the accusat. Jesum, but in the other cases it remains unchanged, Jesu.

* [Modern writers of Latin verse, however, give Dāvides as a form fo the nominative. Compare Hodgson (Provost of Eton), Sacred History fo Latin Verse, p. 95.]—Am. Ed.

Among the genuine Latin words we must notice pondo, which is used only as a plural, and remains unchanged in all its cases, e. g., auri quinque pondo, five pounds of gold. This peculiarity arose from the omission of the word librae, to which was added the superfluous pondo, an ablative in the sense of "in weight" (in which it still often occurs; see § 428); afterward librae was omitted, aud pondo retained its place. Semis, half an as, has become an indeclinable adjective (one half) from a declinable substantive, gen. semissis, and is used as such in connexion with other numerals.

[§ 88.] II. Defectives in case* are those substantives which want one or more cases. There are many words of which the nominative singular cannot be proved to have existed; as, for instance, of the genitives dapis, dicionis, feminis (for which the nominat. femur is used), frugis, internecionis, opis, pollinis, vicis, and of the plurals preces and verbera (for which we use as a nom. sing. plaga or ictus). The genitive neminis, from nemo, occurs very rarely, and its place is supplied by nullius. (See § 676.) The vocative is wanting in a great many words, from their signification. The genitive plural is wanting, that is, does not occur in our authorities, in several monosyllabic words: as, ōs, oris; vas, vadis; glos, pax, and others. (See § 66.) The genit. and dat. sing. of vis are very rare, but the plural vires, virium, &c., is complete.

[§ 89.] With regard to words which want several cases, it most frequently happens that only those cases exist which are alike (i. e., especially the nominat. and accusat.), all the others being wanting. This is the case (a) with Greek neuters in es (properly adjectives) and in os in the singular, and with those in e in the plural, e. g., cacoëthes, chaos, epos, melos, cetos (which make the plural mele, cete, as in Greek), and Tempe. Some of these words, however, have a declinable Latin form in us, i, or um, i, viz., chaus, cetus, melus (mascul.), and melum, from which the ablatives chao, melo are derived; and besides (Tò) Argos, there is a declinable Latin form Argi, Argorum, Argis. (b) With the Latin neuters fas, nefas, nihil, parum (too little), and instar, which was originally a suo*[Consult, on this subject, the following passages of the marians: Charis., p. 22, seq. Id., p. 72, seq. Diom., p. 288. 924. Phoc., p. 1708, seq. Asper, p. 1729. Donat., p. 1749. Cledon., p. 190. Consent., p. 2734.]—Am. Ed.

*

ancient grain Pris., p. 672 Serg., p. 1845

stantive signifying "an image," or "resemblance," and was then used as an adjective in the sense of "like," but only in such connexions as admit of its being explained as a nominative or accusative. Secus, sex, is likewise used only in cases that are alike, especially as an accusative absolute, virile secus, muliebre secus, e. g., canis, muliebre secus; in other phrases, sexus, us, is the ordinary word. (c) With the plural of many monosyllabic words; as, neces, kinds of death; paces, treaties of peace especially neuters; as, aera, brazen images; jura, rights; rura, fields; tura, incense; and others, the plural of which generally occurs only in poetical language; as, farra, corn; mella, honey; fella, bile. To these we must add the poetical plurals flamina, murmura, silentia, colla. The following plurals, grates, munia, munera, likewise occur only in the nom. and accus., and the ablatives gratibus and munibus are rarely used. Metus, which is complete in the singular, and astus, of which the ablat. singular is used, have, in the plural, those cases only which are alike.

The following must be remembered separately: fors occurs only in the nom. and abl. singular (forte, by chance); lues, in the nom., acc., and ablat. singular; mane, in the nom., acc., and abl. singular, and is alike in all of them, but it is used also as an adverb. Satias, for satietas, does not occur, in good prose, in any other form. There are several words which are frequently used in the plural (see § 94), but which in the singular have only one or other case, more especially the ablative; e. g., prece, from preces, occurs in prose also; but the ablative singular of ambages, compedes, fauces, obices, and verbera is used only in verse, and not in ordinary prose.

[§ 90] Some words occur only in particular combinations, and in a particular case: dicis, with causa and gratia; nauci, in the phrase non nauci facere, or esse; diu noctuque, or die et noctu, old ablatives, for which, however, nocte et interdiu are more commonly used; derisui, despicatui, divisui, ostentui, in combination with duci or esse; infitias, with ire; suppetias, with ferre; pessum and venum, with ire and dare; whence venire and vendere, for which Tacitus, in the same sense, uses veno ponere, exercere; foris and foras (from forae fores); gratis (for gratis), ingratiis; sponte, with a pronoun; as, mea, tua,