(c) With the last syllable wholly omitted, as in Kάктave, κάσχεθε, καστορνῦσα, καφθίμενος, καβαίνων, κάπετον for κατέπεσον. The important particles av and Kеv ог κа arе аросоpated forms of ανά and κατά.

Ectasis.

141 The second class of vowel-affections is known by the general name of ectasis (ěkтaσis), productio, or "lengthening." It may generally be regarded as a substitute for some lost consonant, and frequently appears as a transposition or hyperthesis. We have already considered this in its connexion with the assimilation of consonants (above, 104). But, for the sake of system, the doctrine must be formally stated here, in its relation to the pathology of vowels.

142 Primary ectasis appears as a direct insertion of or v without transposition. Thus we have delovs by the side of déovs, τιούχαν for τύχην, δεξιάσθω for δεξάσθω, and so forth. So also we have μοῦνος by the side of μόνος, νοῦσος by the side of νόσος, οὐλόμενος by the side of ὀλόμενος, and the like. But there are etymological reasons for these insertions of and v.

し

143 The insertion of (if it can be called an insertion, for strictly speaking it indicates the primitive form), very commonly represents itself under the form e, with that palatal pronunciation which so often yields to synizesis. In Boeotian inscriptions we have the forms ἀγωνοθετίοντος, χοραγίοντος, &c,, for which the Ionians wrote, probably with synizesis, ἀγωνοθετέοντος, χοραγέοντος, &c., and the Attics the contractions ἀγωνοθετοῦντος, χοραγοῦντος, &c. Compare also ἠερέθομαι with ἀείρω = ἀέργω, ἠΰτε with εὖτε, ἀληθηίη (pronounced ἀληθήγη) with ἀλήθεια, &c.

144 As we have already seen, an apparent ectasis with is often nothing more than an hyperthesis of that letter. The same occasionally happens with v. Thus ἐλαύνω (root έλα-) is to be explained by a transposition in the formative adjunct vv- (110, b).

145 This hyperthesis must be carefully distinguished from the strengthening of or v in the root, by the prefix e or o. Thus πείθω and πέποιθα exhibit modifications of the root πιθ-, found in

ἔ-πιθ-ον; σπεύδω and σπουδή, κέλευθος and ἀκόλουθος, point to lost roots in which v alone appeared; whereas μέλαινα and ἐλαύνω exhibit transpositions of the and v. It is possible, indeed probable, that the e, o, prefixed to the root-vowels, v, may have originated in hyperthesis, but, even then, this, as a transference into the root, must be distinguished from the other transferences which are more distinctly consonantal, and more formally terminational. In comparative grammar the strengthening of a root by prefixing e, o is known by the Sanscrit name guna, i.e. "corroboration.' The substitution of n (involving a vocalized guttural) for a is also a kind of guna, which stands half-way between the prefix e, o, and the hyperthesis of . By a principle of compensation ão or io may always be represented by ew; thus we have λaós by the side of λεώς, the old particle afos by the side of ἕως, the genitives ἱκέταο and ἱκέτεω, ναυτάων and ναύτεων, πόλι-οs and πόλεως, βασιλέος and βασιλέως, &c.

Examples of Ectasis.

(α) At the beginning of a word: αἰετός, αἰεί, ἠγαθέος, ἠλασκάζω, ἠμαθοείς, ἠνεμοείς, ἠΰς, εἰαρινός, εἰλάτινος, εἱλίσσω, εἶνακοσίοι, εἰρωτάω.

(β) Within the word: βασιλήιος ; κληίς; ἀληθηΐη; πατρῴιος; εὐρωείς; Διώνυσος; ἐλαία by the side of ἐλάα; ἀητός by the side of ἀετός; καίω and κλαίω by the side of κάω and κλάω ; 'Αχαιικός, Πλαταιικός, Αθηναιικός; ἀγνοιέω; ἀλοιάω; ποίη; ῥοίη; στοιά; εὔνοια, εὔπλοια, πνοίη, εὔροια, διάῤῥοια, χροίη, χλοίη by the side οἱ νόος, πλόος, πνόος, ῥόος, χρόος, χλόος; γούνατα, δούρατα, κοῦρος, μοῦνος, οὐδός, οὐλος, Οὔλυμπος, πουλύς, δουλιχοδείρων by the side of γόνατα, &c. ; κεινός, στεινός by the side of κενός, στενός, δε,

(c) At the end of a word, chiefly the prepositions: ἀπαί, καταί, ὑπαί, διά, παραί, ὑπείρ.

PART III.

INFLEXION.

CHAPTER I.

DECLENSION OF NOUNS AND PRONOUNS.

§ I. General Remarks.

146 IT has been already shown that every noun and verb in the Greek language may be reduced, by stripping off an affix, prefix, or both, to some single syllable, which constitutes its meaning, and which is found also in other words of cognate signification. This ultimate element we call the root, stem, or basis of the word. The prefix may generally be stript off at once, but the removal of the affix is often a double operation. To confine ourselves for the present to the noun, we find that almost every word of this kind ends with a short termination, often a single letter, which marks its immediate relation to the other objects in connexion with it, and which we call the case-ending. But in the majority of nouns, we find, between this and the root, an affix, consisting of one or more pronominal stems, which marks the definite class and quality of the noun, and points out the restriction with which the general force of the root is applied in the particular instance. When the case-ending alone is removed, the remaining part of the word is called its crude or uninflected form, whether it has another pronominal affix or not. The affections of the uninflected form, as such, belong to a different subject-that of the formation of words by derivation or otherwise (Part IV.). At present we are concerned only with the crude forms and cases of the noun.

147 The Greek noun (ovoμa, nomen) is either substantive (ὄνομα προσηγορικόν) or adjective (ὄνομα ἐπίθετον oι ἐπιθετικόν) :

the former expresses a person or thing; the latter expresses the quality of a person or thing; or, to speak more strictly, the former expresses an appropriated, the latter an unappropriated quality.

The Greek noun has five cases (TTσeis, casus): three numbers (ἀριθμοί, numeri), singular, dual, and plural (ἑνικός, πληθυντικός, Svïkós): and three genders (yévn, genera),—masculine, feminine, and neuter (αρσενικόν, θηλυκόν, οὐδέτερον).

(a) Cases.

148 The five cases are the nominative, genitive, dative, accusative, and vocative. Properly speaking, the vocative (îλŋτɩý) is not a separate inflexion, but is either the crude form, or the same as the nominative. And the nominative εὐθεῖα or ὀρθή, casus rectus) stands opposed to the genitive (γενική, κτητική, πατρική), dative (δοτική, ἐπισταλτική), and accusative αιτιατική), which are called Tλáɣiai πтúσeis, casus obliqui. The Stoics and Peripatetics differed as to the reason for calling these inflexions TTσeis, casus, "fallings."

The genitive may be translated by "of, from, out of, by:" the dative by "to, with, at, for:" the accusative corresponds to our simple objective case, or implies "motion towards a place or object." The use of the cases must be learned from the syntax. When a noun is used immutably in an oblique case it is called an adverb (eπippnua, adverbium).

Besides these five cases, which alone appear in the ordinary declensions of nouns, there are remains of other inflexions, which are partly absorbed in the existing cases, or still represented by distinct forms of certain words.

(a) The locative or case of rest, which is generally absorbed in the dative, appears as a distinct form in some a and o nouns, chiefly proper names of places; thus we have in a nouns, 'A¤ĥvai, ̓Αθήνησι ; ̓Αλωπεκή, ̓Αλωπεκήσι; Θεσπιαί, Θεσπιᾶσι; Πλαταιαί, Πλαταιάσι; Δεκέλεια, Δεκελείασι; Ὀλυμπία, Ὀλυμπίασι; Ἐρχία, Ἐρχίασι; also θύρασιν, foras, and ὥρᾶσιν, opportune, and in one Attic inscription (Böckh, C. I. no. 140) we have тoîs Taμíari for τοῖς ταμίαις. In o nouns we have Ἰσθμός, Ἰσθμοῖ, τὰ Μέγαρα, Μεγαροί, Πειραιός, Πειραιοί, and the Attic demes Σφηττοῖ, Σουνιο,

&c.; also the common words oikoi, medoî, péσooi. In consonantal nouns we have Πυθώ, Πυθοί, Μαραθών, Μαραθῶνι, and Δωδῶνι from the obsolete Δωδῶν for Δωδώνη.

(b) The form in -p is used by the epic poets, and sometimes also by the lyric poets, to denote a dative, genitive, or even an accusative with or without a preposition: thus we have forms like κεφαλῆφι, κλισιῆφι, θεόφιν, ἰκριόφιν, ἔχεσφιν, στήθεσφιν, ναῦφιν, ἐσχαρόφιν (by a metrical affection for ἐσχαραφιν), κοτυληδονόφι for κοτυληδόμφι), &c. Ιη νόσφι, λικριφίς, &c. the form is merely adverbial.

(c) The form in -Oev or -0e generally indicates derivation or motion from a place, as Evevida тáτρale Σwyeves (Pind. Nem. VII. 70), but is also used as a common genitive, especially in the pronouns ἐμέθεν, σέθεν, ὅθεν.

(b) Numbers.

149 The dual, though a very old form, is never regarded in Greek syntax as a necessary expression for things considered as pairs, and is constantly interchanged with the plural, of which it is merely a corruption. Homer uses it very frequently to denote things taken in couples (see Il. vIII, 41, 45), and it is common in the Attic dialect; but the plural gradually superseded it; after the time of Alexander it became nearly obsolete; and it is not to be found in the New Testament. It is one of the most remarkable coincidences between the Æolic dialect and the Latin, that neither of them has dual forms (see Anecd. Bekk. 1184, 21). But they are found in Sanscrit.

(c) Genders.

150 The main rules with regard to the genders of nouns are the following:

(1) Masculine: names of male persons and animals; as Σwκράτης, ἀνήρ, προφήτης, αλέκτωρ, λέων, and of months, winds, rivers, and hills, as ὁ Γαμηλίων, ὁ Ζέφυρος, ὁ Ἰλισσός, ὁ Παρνασός.

(2) Feminine: names of female persons and animals; as Ασπασία, γυνή, θυγάτηρ, κομμωτρία, αλεκτορίς, λέαινα, and even diminutives of proper names; as Aeóvτɩov; also the proper names