consonant, which precedes the case-ending; and that they fall into two great classes according as they retain the consonant or vocalize it into i or a. The characteristic is very often lost in the nominative singular, but it may always be recovered by a careful examination of the oblique cases. § 6. A. First class or purely consonantal Nouns. (a) Labial nouns are limited to some few in b, as plebs (also plebes), scobs (also scobis), scrobs (also scrobis), trabs, urbs (anciently urbis ?), and some few in p, as daps, stips, stirps (anciently stirpis ?), to which must be added compounds in cipfrom capio, as man-ceps, muni-ceps, parti-ceps, prin- ceps. To the same class of compounds we must refer for-ceps, "a pair of pincers," the first syllable referring to the "opening" or "door," which this instrument makes in order to grasp the object. Similarly we have for-fex, "a pair of scissors," from facio, and forpex, "a pair of curling-tongs," from pec-to. = (b) Guttural nouns are a more numerous group, and the tenuis c is a more common characteristic than the medial g. Of the latter class we have only the primitive frux (frūg-), grex (greg-), and strix (strig-): and the verbals lex (root leg-), rex (root reg-), with the compounds il-lex, inter-rex, con-jux (root jug-), remex (root ag.). Supellex is an abridgment of the form in -li- indicated by the genitive supellectilis, and the x does not represent a 9 but cts. The same is the case with senex (= avač), which conversely exhibits a shortened form in its genitive senis cf., however, senectus, seneca, senecio, &c. In nix the a represents gu or qu: cf. ninguo. The genitive něvis may be compared with vivo quiqvo, struo = struquo, &c. The tenuis c is the characteristic of a number of primitive nouns, such as far (făc-), lux (lūc-), codex (codic-), cornix (cornīc-), &c.; it also appears in nouns containing the root of c verbs, as dux (duc-), ju-dex (dic-), and other nouns from dico; pol-lex (lic-), and other nouns from licio; arti-fex (fic-), and other nouns from facio; and we find a great number of feminines in -trix corresponding to real or possible masculines in tor, such as nutrix (nutric-), obstetrix (obstetric-), &c. The last word deserves some special notice, as showing the true meaning of ob in composition. For ob-stetrix must mean "a woman who stands by to assist"-a Beisteherin-and Tapaσтivaι or ovμπapa σTivat is especially used to denote this by-standing or as-sistance in childbirth: so Pind. Οl. VI. 42: πραΰμητίν τ ̓ Ελευθώ συμπαρέστασέν TарéσTаσév Te Moipas. cf. Ol. XI. 54. If then ob-sto may signify "to assist," like Tapioτnu, as well as "to oppose," it can only bear this meaning in consequence of the sense of extension, continuance, and perseverance borne by ob; and thus of-ficium may denote "beneficial aid," though officio signifies harm and hinderance. Compare the two applications of our word pre-vent, which means to go before, either for the purpose of clearing the way, or for the purpose of obstructing the passage. From this explanation of ob-stetr-ix, it is plain that Stator does not imply, actively, "one who causes to stand," but "one who stands by, ready to help"-qui stat opem laturus—of a præsens Divus, according to the proper meaning of that term, as in Cic. Tusc. Disp. I. 12. § 28: "Hercules tantus et tam præsens habetur deus." (c) The most numerous and important class of the purely consonantal nouns are those which have a dental mute for their characteristic; for while the labial and guttural nouns are limited to the masculine and feminine, these exhibit also some neuter nouns of very common occurrence. (a) Masculine and feminine nouns in -d are such as pes (pěd-), frons (frond-), vas (văd-)', and its derivatives præs (= præ-vad-), custos (custo-vad-), and mercēs (merce-vad-); palus (palūd-), &c. Masculine and feminine nouns in -t, are such as dens (dent-), frons (front-), pars (part-), comes (comit-), quies (quiet-), nepos (nepōt-); a very long list of abstract words in -tas (-tat-), as boni-tas, with a smaller number of supplementary forms in -tus (-tut-), as vir-tus; and active participles in -ns (-nt-), which are occasionally used as nouns, as serpens (serpent-), &c. The genitive plural in -ium would lead us to infer that these must have been originally older forms in -tis of those nouns in -t, in which the characteristic is preceded by another consonant; cf. scobs with scobis, and dens, gen. pl. dentium, with sementis. 1 This word is interesting from its connexion with the Low-German wæd, or wad, a pledge," found in wad-set, wed-ding, &c. Another form was bad, as in the old compact gif bad genumen sy on monnes orfe, “if a pledge be taken from a man's chattels ;" and from this comes out bet. From the Low Latin vad-iare comes the Romance guadiare, guaggiare, and our wager. (See Palgrave, History of the Anglo-Saxons, Pref. p. xxi.) (3) Neuter nouns of this class originally and properly terminated in -t. Although caput, gen. capitis (for which the oldest MSS. of Lucretius give capud), is the only word in which the characteristic is retained unaltered, Greek analogies and many collateral indications enable us to see at once what nouns belonged to this dental declension. Some Greek nouns in μας μaт-=μevт- (New Crat. § 114) have been naturalized in Latin, such as poema, gen. poematis; and lac, gen. lactis, retains more of the termination in the nominative than the corresponding yáλa, gen. yáλaктоs. The T, which is lost in keap, cor, is represented by the medial in kapdía, cordis. And though carmen (cf. carmentis), agmen (cf. armentum), have omitted the characteristic t, not only in the nominative, but also in the oblique cases carminis, agminis, &c., they at all events retain the preceding liquid, which is lost altogether in the Greek neuters in And while corpus, opus, &c. agree with Teixos τεῖχος in softening the into s, they retain some trace of it in the r of the oblique cases, where the Greek, according to the rule (New Crat. § 114), has dropt the s between the two vowels. There is an assimilation of the t in the oblique cases of os, oss-is (cf. õσTcov), mel, mellis (cf. μéλ-T), fel, fellis, and far, farris. The singular forms jecur (also jecinor), iter (also itiner), and jubar, probably ended originally in -rat, like the Greek аp for ἡπρατ, ήπατος. πрат, gеn. TаTоs. The following table will show the gradual degeneration of the forms: -μα, ματος. T β. Y cor os[t] carmen[t] corpus caput lac[t] capit-is lact-is cord-is oss-is carmin-is corpor-is. Here it will be observed that in a the t is preserved intact; that in ẞ, it is lost after another consonant in the nominative, and preserved in the oblique cases; that in B2 it is retained in the medial form which comes nearer to the preceding liquid r (above, p. 256); that in ẞ, it is assimilated to s; that in ẞ, it is altogether dropt after n; and that in y it is softened into s and r. In comparing corpus, corpor-is, with Teixos, Teixe-os, we observe that although the latter has lost the σ, according to the rule, because it is flanked by two vowels, it could retain the neuter characteristic before a consonant: thus we have ópéoβιος from ὄρος, σακεσ-πάλος from σάκος, &c. Similarly, that ther or s which takes the place of t in the Latin nouns, is retained in derivatives, like gener-osus, from genus, generis, robus-tus from robur, and tempes-tas from tempus. (d) Liquid nouns are generally of dental origin, and many of them recal to our recollection the neuter nouns, which have just been mentioned. The only noun in m is the word hiem-s, gen. hiemis, which is probably the corruption of a longer form in mn: cf. χείμων and χείμα= χείμεντ. There are a few nouns in l, as sol, sōl-is, sāl, sălis (which is neuter, as well as masculine, and which, in that use, has lost a final t), nihil (for nihilum), which is neuter and undeclined, and some compounds derived from salio, as con-sul, præ-sul, ex-sul. The great majority of liquid nouns have crude forms in n or r=s. Of the former we have some in -o, -žnis; many in -do, -ēdo, -ido, -tudo, of which the genitive is formed in -dinis, &c.; others in -go, -āgo, -īgo, -ūgo, which have their genitives in -ginis, &c.; others, again, in -o, -io, -mo, -sio, -tio, which form the genitive in -ōnis, &c. It is superfluous to give examples of all these different classes. In comparing caro, gen. carnis, with virgo, gen. virginis, we see that two liquids in the former have coalesced to the exclusion of the short ; and virgo=virgin-s differs from sermo sermón-s, just as δαίμων=δαίμονς differs from χείμων=χείμων-, or as ποίμην= ποιμεν-s differs from σπλήν=σπλήν-ς. In some of the nouns in s=r this characteristic represents the neuter t; such are œs, gen. æris, rus, gen. rūris, os, gen. oris, ver, gen. veris, &c. Other nouns in r really belong to the i declension, as laquear, gen. laqueāris. But we have a large number of masculine and feminine nouns of which r is the genuine characteristic. These are formed in -ěr, or -es, or -us, -ĕris, as mulier, Ceres, Venus; in -ōr or -os, oris, as labor, flos; in -ur, -ŭris, as augur; in -ūr = -us, -uris, as tellus; in -or, -õris, as arbor: we have an important class of nouns denoting agency, and ending in -ter, -tris, as pa-ter, ma-ter, &c., to which must be added u-ter, u-tris, ven-ter, ven-tris, and the compound ac-cipi-ter (-tris) from accipio: cf. capys, the Etruscan word for a falcon (above, p. 155). The instrumental ending in -ter is extended, in a very numerous class of nouns, to -tor, -toris, assibilated to -sor, -sōris; thus we have duc-tor from duc-o, ara-tor from aro, moni-tor from moneo, spon-sor from spondeo, &c. We have seen that the r often appears as s in the nominative; in two nouns an e is changed into in this case; thus we have cinis, cinĕris and pulvis, pulveris. In consonantal derivatives from nouns in r, as in the corresponding neuter-forms, this characteristic is retained as a simple sibilant; thus, from Venus, Veneris we have venus-tas ; from honor, honoris, hones-tas; from arbor, arboris, arbustum ; &c. § 7. B. Second class or semi-consonantal Nouns. (a) Nouns in i exhibit some phenomena of considerable linguistic importance, which have eluded the observation of all previous grammarians. It has been shown elsewhere that the termination i, as a guttural residuum, is derived from the second pronominal element. But it appears as an extension not only of other pronominal affixes, but even of the second pronoun in many of its forms, and especially under the form c-k. Thus. we have not only a large class of Greek adjectives in -kós, and nouns in -κ-s, but we have also the extensions -κ-ɩs, -к-Єos, &c., in which that element is repeated under a softened or vocalized form. Similarly in Latin, although the nouns in x c-s or g-s8 form their genitive plural in -um and are therefore independent of any additional elements, adjectives of the same form show by their ablative in -i and their genitive plural in -ium, that the full ending of the crude form is not c-, but c-i. It has been already remarked that some nouns in b- or p- have by-forms in -bi- or -pi-, and that nouns in -nt- must have been originally formed in -nti-. The last phenomenon connects itself with a very interesting fact-namely, that forms in -nts in Greek and Latin stand beside forms in -ntus and -ndus. From the regular change of -VTS in Greek into -ntus in Latin-as when we have πâs πάντες by the side of quantus, or Τάρας = Τάραντος by the side of Tarentum-no inference can be drawn. But as -d- is generally, if not always, a shortened form of the articulation which appears as the second personal pronoun and the second numeral, and as we have verbal forms in -dus (as cupidus, &c.) by the side of verbals in -TÉOS, -TÚS, -TIS, it is not unreasonable to conclude that if orien-t-s oriu-n-dus, the former is an abridgment of orien-tis analogous to sementis, &c., and this explains the genitive plural in -ium. Although there are some nouns in -i- which retain their characteristic throughout the cases-as sitis, Tiberis, febris, puppis, &c.,-it not unfrequently happens that the shorter = = |