changes, which may be inconsistent with the development of its separate parts. The most common of these changes is an ectasis at the point of junction or in the first syllable of the final word, which is perhaps due to an euphony of the same kind as that which is observable in the reduplicated perfect and aorist (307). The doubling of p at the point of junction in such words as ισό-ροπος, κατάῤῥυτος, &c. is invariable (above, 105). Synæresis often takes place when the last vowel of the former word is not elided before the initial vowel of the latter part of the compound; thus we have ῥαβδούχος for ῥαβδό-εχος; κακοῦργος for κακό-εργος; θεωρός for θεά-δρος; πυλωρός for πυλά-δρος; φροῦδος for πρόοδος, &c. But independently of these cases we often have an arbitrary ectasis at the beginning of the latter word, when its initial letter is a, e or o, the two former vowels being changed into η, and o becoming ω. The words which most regularly admit of this ectasis in composition are the following: ἄγω, στρατηγός, λοχαγός; ἀγορεύω, κατήγορος; ἄκη, ἀμφηκής; ἀκέομαι, ἀνήκεστος ; ἀκούω, ὑπήκοος; ἄνεμος, εὐήνεμος; ἀνήρ, εὐήνωρ; ἀνύω, ἀνήνυτος; ἀραρίσκω, τριήρης, ποδήρης; ἐλαύνω, χαλκήλατος; ἐλθεῖν, ἔπηλυς; ἐρέσσω, ἐρετμός, ὑπηρέτης, εὐήρετμος; ἐρέφω, κατηρεφής; οβολός, τριώβολον; ὀδύνη, ἀνώδυνος; ὄζω, δυσώδης; ὄλλυμι, πανώλης; ὁμαλός, ἀνώμαλος; ὄμνυμι, ἀνώμοτος; ὄνομα, συνώνυμος; ὄπωπα, δυσωπία; ὀρύσσω, τοιχωρύχος; ὄρος, ὑπωρεία; ὄφελος, ἀνωφελής. These are found in common Greek; the following are poetical only: ἀμείβω, ἐξημοιβός and ἐπημοιβός; ἀριθμός both ἀνάριθμος and ἀνήριθμος; ἀρόω, ἀνήροτος; ἐράω, πολυήρατος; ἐρίζω, ἀμφήριστος, but ἀνέριστος; ὀδούς, ἀμφώδων, but ἀμφόδους, χαυλιόδους, &c. The cases of ἀπολλήγειν, Ἱππόδαμμος, Παρθεννόπαιος, &c., Ιππομμέδων, ̓Αλφεσσίβοια are peculiar results of the constitution of the liquids.

Obs. The ectasis does not take place in the parathetic compound of preposition and verb; thus we have ὑπάγω, ὑπακούω, &c. This needs no explanation.

(3) Formation or inflexions of the whole Compound. 376 A synthetic compound, viewed as an undivided whole, appears either (a) as a noun, or (b) as a verb. the main subdivision in this class of words.

And this suggests

But when the word

as a whole performs the functions of a noun, we have still to

consider (a) whether the included word represented by the last part of the compound is an ordinary noun, or (B) of verbal origin. And when it is an ordinary noun, we must inquire (aa) whether it is a substantive, or (bb) an adjective. With an examination of these different cases we shall conclude the subject of compound words.

(a) The whole compound constitutes a Noun.

(a) The latter part represents a Noun.

(aa) The latter part is derived from a Substantive.

377 Even in this case we must seek a further distinction; for although the last part of the compound may be derived from a substantive, it may not only (aa) retain its substantival value, but may also (88) perform the functions of an adjective, and the latter is much more common than the former.

(aa) The substantival value is retained.

The first part of the word is generally an uninflected prefix; but there are cases in which a noun preceded by its epithet or a dependent case becomes a synthetic compound; thus, while we have ὁμό-δουλος, σύνδουλος, ἡμί-ονος, ἐπί-μετρον, we have ἀκρόπολις, μεγαλό-πολις, ἱππόδρομος, ἡλιό-πολις, &c.

Obs. It is doubtful whether such words as συγγραφή, σύγγραμμα, ἔκπεμψις, ἐκπομπή, and the like, are themselves parathetic compounds or are derived from verbal parathetics; most probably the latter. Such words as συγγραφεύς, εἰσαγωγεύς, &c. are manifestly derivatives.

(88) An adjectival value is assumed.

1. The substantive retains its form, as in a-rais, "childless;" Svo-épws, "ill-starred in love;" μaxрó-xep, "long-handed;" TOλÚ-πOUS, “many-footed;" ev-Oeos, "possessed by God;" μóπονος, “labour-hating;” φιλόπατρις, “country-loving;” δεισιδαίpov, "fearing the lower deities," &c. 2. An adjectival ending is appended, as in σύνδειπνος, “ dining together;” ἀ-χρήματος and a-xpýμwv, “money-lacking;" ǎ-σTOμOS, “without a mouth;" ἄ-σωμος and ἀ-σώματος, “incorporeal;” παραλληλό-γραμμος,

nded by parallel lines;" XETTÓ-Yews, "having a light soil;' "having a good or simple character;" ev-μýns, “of a v-axis, "cowardly;" ǎ-daxpus, “tearless," &c.

ἄναλκις,

Obs. Nouns in -ny and -np generally form the compound in -evos, -Epos, as a-λíuevos, "without harbours;" ev-áσrepos, "bright with stars." But compounds of opv, and some nouns in -p, merely change into ω; thus we have ἄφρων, ἔμφρων, σώφρων, ἀπάτωρ, εὐπάτωρ, εὐήνωρ, ἀμήτωρ, προγάστωρ.

(bb) The latter part is derived from an Adjective.

In this case the adjective retains its form and meaning, except so far as the latter is modified by the prefix. Thus, from toos, "equal," we have av-toos, "un-equal;" from Siaßarós, "passable," δυσ-διάβατος, “ hard to cross;” from γυμνός, “ naked,” ἡμί-γυμνος, "half-naked;" from Xeurós, "white," UTO-λEUKós, "rather white;' from σοφός, “ wise,” πάν-σοφος, “ all-wise,” &c.

(B) The latter part is of verbal origin.

In interpreting a compound, of which the latter part includes the meaning of a verb, we have always to inquire whether the verb involved is to be taken transitively or intransitively. This is sometimes shown by the ending, sometimes by the accent, and sometimes left indeterminate. The endings are as follows:

1 -os, -ov. This is the most numerous class of compounds ending with a verbal, and the meaning of the included verb is generally determined by the accent, according to the following rule: When the meaning is transitive, and the first part of the compound is any noun excepting wâs and woλús, the compound is oxytone if the penultima is long, and paroxytone if the penultima is short; but all transitive compounds are proparoxy tone when the first part of the compound is a preposition, an adverb, or one of the nouns was and Toλús; and the same is the case with all intransitive compounds. Hence, although words compounded with prepositions, adverbs, Tâs and Toλús, do not indicate their meaning by the accentuation, we may tell by the position of the accent whether the other class of compounds represents a transitive or intransitive construction; thus, λιθοβόλος from λίθων βολή, “ a finging of stones,” means "a person who pelts with stones," but Meó-Boλos is "one who is pelted;" μηтро-Kтóvоs would refer to Orestes the matricide, but μητρό-κτονοι to the children of Medea, who were slain by their mother. The following examples will illustrate the different classes:

66

Obs. There are some exceptions to this rule: (1) The epic compounds ἱππόδαμος, ἐγχέσπαλος, σακέσπαλος, πτολίπορθοs are proparoxytone, though the meaning involved is undoubtedly transitive. (2) Compounds with ἄρχω and συλάω are proparoxytone, as ἵππαρχος, ἱερόσυλος. (3) Compounds with ἔχω are proparoxytone or properispone, as ἡνίοχος, δαδούχος, &c. (4) By a peculiar refinement, words compounded with ἐργάζομαι are oxytone when they signify a bodily or material action, but proparoxytone, or by contraction properispome, when they denote a moral action, or an operation and habit of the mind; thus we have λιθουργός, “ a worker in stone;” γεωργός, “ a husbandman ;” ἀμπελουργός, “ a vine-dresser;” but πανούργος, " an unscrupulous rogue;” κακούργος, “ a criminal;” περίεργος, " a busy-body,” &c.

66

2 -ης, -ες. These generally express the intransitive meaning of the verb, or at least a condition resulting from it; as θεο-φιλής, " beloved of God;” αλουργής, “ made of purple ;" ἀπρεπής, “unbecoming;” εὐμαθής, " easily taught, docile;” αὐταρκής, “ selfsufficing, complete. But ίππομανής, “ horse-maddened" of mares, and "horse-maddening" of a luxuriant meadow, seems to contain both meanings of the verb.

66

3 -ns or -as, gen. -ov. Generally substantives denoting the agent, and therefore transitive; as νομοθέτης, “ the law-giver;" εὐεργέτης, “ the benefactor;” οἶνο-πότης, “the wine-drinker;" ὀρνιθοθήρας, “ the bird-catcher;” πατραλοίας (ἀλοιάω), “ the parricide."

66

4 -. This is rare and generally poetical. If the first part is a substantive, the verb included is generally understood in a transitive sense; otherwise it is passive; thus we have voμo-þúλağ, “a guardian of the laws;" Bov-Tλng, "striking the oxen;" кvаμóκυαμόTpw, "eating beans;" but aπop-pwę, "broken off;" veo-opáž, "newly slaughtered."

(b) The whole compound constitutes a Verb.

66 a

378 It is a fixed law of the Greek language that a synthetic compound never constitutes a verb except as a derivative from one of the synthetic compounds which we have just discussed. In other words: "Verba non possunt nisi per flexuram quandam cum aliis orationis partibus præter præpositiones consociari" (Lobeck, ad Phrynichum, p. 560; see above, 369). To express by a single word the combination of a noun or adverb with a verb, it is generally necessary to pass through a series of derivations; thus, from ἵππον τρέφω, " I keep a horse,” we have ἵππου τροφή, “ the keeping of a horse," from this the synthetic adjective iππоτρóоs, "keeping a horse," and from this the derivative verb iππоτроpéw; from λίθον βάλλω, " I throw a stone,” we have λίθου βολή, α throwing of a stone," from this oßóλos, "throwing stones," and from this again λιθοβολέω, “ I throw stones; similarly with an adverb, we may have Tλe ẞάλλw as two separate words, "I throw once from afar," from this Tλe Boλn, "a throwing from afar," τηλέβολος, “ throwing from afar or thrown from afar,” τηλεβολέω, "I habitually throw from afar;" or with the fixed case of a noun, a3 πὺξ μάχεσθαι, “ to fight with the fists,” πὺξ μάχη, " a fighting with fists," vypaxos, "habitually fighting with fists, a boxer," Tyμaxéw, "I am a boxer," vyuaxía, "a boxing match;" or πυγμαχέω, πυγμαχία, with the instrumental case of the noun, as xeipì ypάow, "I write with my hand,” ἡ χειρὶ γραφή οι χειρὸς γραφή, “ the writing with the hand," Xepoypapos, "writing with the hand" (which is assumed from the meaning of the verb), and xepóypados, "written with the hand" (which occurs in the word Tò xepóypapov), hence Xeipoypapéw, “I write habitually with the hand, I am a copyist,' and finally xeɩpoypápnua, "a thing written with the hand" (in the same sense as тò xeiρоyρapov). The unity of the compound is generally and regularly indicated by the place of the augment and reduplication (above, 309).

πύγμαχος,

66

66