PA UI'79 | Ae Se ae Philological Seminary Library. THE GIFT OF HENRY W. SAGE. CORNELL UNIVERSITY. OUN LARK CBE ULATION wows =| JAN DP 00 GAYLORD PRINTEDINU.S.A Cornell University Library ' PA 4179.M75 | A grammar of the Homeric dialect / olin 3 1924 011 326 190 Clarendon Press Series eps HOMERIC GRAMMAR D. B. MONRO London HENRY FROWDE OXFORD UNIVERSITY PRESS WAREHOUSE 7 PATERNOSTER ROW Clarendon Press Series A GRAMMAR HOMERIC DIALECT BY WE By MONRO, M.A. FELLOW OF ORIEL COLLEGE, OXFORD L’objet de cette science est de rechercher dans l’esprit de l’homme la cause de la transformation des idiomes. M. Breau. Oxford AT THE CLARENDON PRESS 1882 ce) [ All rights reserved } DEDICATED TO THE MEMORY OF THE REV. JAMES RIDDELL LATE FELLOW AND TUTOR OF BALLIOL PREFACE. Ir may be said, without fear of giving offence, that a new Grammar of the Homeric dialect is sorely wanted. The admirable Griechische Formenlehre of the late H. L. Ahrens is now just thirty years old, and is confined, as its title indi- © cates, to the inflexions. Not only has the-course of discovery been going on since Ahrens wrote (and with hardly less rapidity than in the first years of the new science), but the historical method has been carried into the field of syntax. And apart from ‘comparative philology, the researehes of Bekker, Cobet, La Roche, and many other students have brought together a wealth of material that only needs careful analysis and arrangement to make it accessible to the general body of learners. The plan of this book has sufficient novelty to call for some explanation. I have not attempted to write a Comparative Grammar, or even a Grammar that would deserve the epithet ‘historical :’ but I have kept in view two principles of arrange- ment which belong to the historical or genetic method. These are, that grammar should proceed from the simple to the com- plex types of the Sentence, and that the form and the mean- ing should as far as possible be treated together. Now the simplest possible Sentence—apart from mere exclamations— consists of a Verb, or word containing in itself the two ele- ments of all rational utterance, a Subject and a Predicate. We begin, therefore, by analysing the Verb, and classifying (1) the Endings, which express the Person and Number of the Subject (§§ 1-7), and serve also to distinguish the ‘ Middle’ or Reflexive use (§ 8), and (2) the modifications of the Stem which yield the several Tenses and Moods. These modifica- tions, we at once perceive, are more numerous than the mean- ings which they serve to express, and we have therefore to choose between classifying according to formation—i.e. ac- cording to the process by which each Tense-Stem and Mood- Stem is derived from the simple Verb-Stem or Root,—and the vill PREFACE. ordinary classification according to meaning (Present, Future, Perfect, Aorist, &c.). The former course seemed preferable because it answers to the historical order. The problem is to find how pre-existing forms—common to Greek and Sanscrit, and therefore part of an original ‘Indo-European’ grammar— were adapted to the specifically Greek system of Tense-mean- ings. I have therefore taken the different formations in turn, beginning with the simplest (§$ 9-20, 22-27, 29-69, 79-83), and introducing an account of the meaning of each as soon as possible (§§ 21, 28, 70-78). This part of the subject naturally includes the accentuation of the different forms of the Verb (§§ 87-89), The next great division of the subject is concerned with the first enlargement of the Sentence. A word may be added which taken by itself says nothing-—contains no Subject and Predicate —but which combines with and qualifies the primi- tive one-word Sentence. The elements which may gather in this way round the basis or nucleus formed by the Verb are ultimately of two kinds, Nouns and Pronouns; and the relations in which they may stand to the Verb are also two- fold. A Noun or Pronoun may stand as a Subject—limiting or explaining the Subject already contained in the Person- Ending—or may qualify the Predicate given by the Stem of the Verb. These relations are shown by the Ending, which again may be either a Case-Ending or an adverbial Ending. We begin accordingly by an account of the Declensions, sup- plemented by a list of the chief groups of Adverbs (Chapter V). When we pass from the Endings to the Stems of Nouns and Pronouns, we find that they are essentially different. A ‘Nominal Stem’ consists in general of two parts, (1) a predi- cative part, usually identical with a Verb-Stem, and (2) a Suffix. Each of these two elements, again, may be complex. The addition of a further Suffix yields a fresh Stem, with a corresponding derivative meaning; and thus we have the dis- tinction between Primitive or Verbal and Secondary or De- nominative Nouns. The Suffixes employed in these two classes are generally distinct, and deserve a more careful enumeration than is usually given in elementary grammars. The predicative part, again, may be enlarged by a second PREFACE. ix Nominal Stem, prefixed to the other, and qualifying it nearly as a Case-form or Adverb qualifies the Verb. The Compounds thus formed are of especial interest for the poetical dialect of Homer. The analysis which I have given of the chief forms which they present must be taken to be provisional only, as the subject is still full of doubt. With respect to the mean- ing I have attempted no complete classification. It is always unsafe to insist on distinctions which may be clear to us, but only because we mark them by distinct forms of expression. The chapter on the formation of Nouns should perhaps have been followed by one on the formation of Pronouns. The material for such a chapter, however, lies for the most part beyond the scope of a grammar. It is represented in this book by a section on Heteroclite Pronouns (§ 108), which notices some traces of composite Pronominal Stems, and in some degree by another on the Numerals (§ 130). When we come to examine the syntactical use of the Cases, we find ourselves sometimes dealing with sentences which contain at least two members besides the Verb. Along with the constructions which may be called ‘adverbial’ (using the term Adverb in a wide sense, to include all words directly construed with the Verb), we have the constructions in which the governing word is a Noun or Preposition. And in these again we must distinguish between the government of a Case apparently by a Noun or Preposition, really by the combined result of the Noun or Preposition and the Verb, and the true government by a Noun alone, of which the dependent Genitive and the Adjective are the main types. These distinctions, however, though of great importance in reference to the deve- lopment of the use of Cases, cannot well be folléwed exclu- sively in the order of treatment. I have therefore taken the Cases in succession, and along with them the chief points which have to be noticed regarding the ‘concords’ of Gender (§§ 166-168) and Number ($§ 169-173). In the Infinitive and Participle (Chapter X) we have the first step from the simple to the complex Sentence. The pre- dicative element in the Verbal Noun is treated syntactically like the same element in a true or ‘finite’ Verb; that is to say, it takes ‘adverbial’ constructions. Thus while retaining x PREFACE, the character of a Noun it becomes the nucleus of a new imperfect Sentence, without a grammatical Subject properly so called (though the Infinitive in Greek acquired a quasi- Subject in the use of the Accusative before it), and standing to the main Sentence as an adverb or adjective. While the Infinitival and Participial Clauses may thus be described as Nouns which have expanded into dependent Sentences, the true Subordinate Clause shows the opposite process. In many instances, especially in Homeric syntax, we can trace the steps by which originally independent Sentences have come to stand in an adverbial or adjectival relation. The change is generally brought about, as we shall see, by means of Pronouns, or Adverbs formed from Prono- minal stems. Hence it is convenient that the account of the uses of the Pronouns (Chapter XI) should hold the place of an introduction to the part in which we have to do with the relations of Clauses to each other. The next chapter, however, does not treat directly of sub- ordinate Clauses, but of the uses of the Moods in them. It seemed best to bring these uses into immediate connexion with the uses which are found in simple Sentences. In this way the original character of Subordinate Clauses, and espe- cially the cardinal distinction (due to Delbriick) between Conditional and Final Clauses, comes into a clearer light. In the account of the Particles (Chapter XIII) we examine the relations of co-ordinate Sentences, so far at least as these are expressed by grammatical forms. The last chapter contains a discussion of the Metre of Homer (Chapter XIV), and of some points of ‘phonology’ which (for*us at least) are ultimately metrical questions. Chief among these is the famous question of the Digamma. I have endeavoured to state the main issues which have been raised on this subject as fully as possible: but without much hope of bringing them to a satisfactory decision. A book of this kind is necessarily to a great extent a com- pilation, and from sources so numerous that it is scarcely possible to make a sufficient acknowledgment of indebtedness. The earlier chapters are mainly founded on the great work of G. Curtius on the Greek Verb. More recent writers —I will PREFACE. xi not say the ‘new school,’ since they have neither the common tendencies nor the collective importance implied by the word —have cleared up some difficulties, especially in the phono- logy. I have learned very much from M. de Saussure’s Mémoire sur le systéme primitif des voyelles, and from several articles by K. Brugman and Joh. Schmidt, especially the last. I would mention also, as valuable on single points, the papers of J. Paech (Vratisl. 1861) and H. Stier (Curt. Stud. II) on the Subjunctive, B. Mangold on the ‘diectasis’ of Verbs in -do (Curt. Stud. VI), F. D. Allen on the same subject (Zrans. of the American Phil. Assoc. 1873), Leskien on oo in the Fut. and Aor. (Curt. Stud. II), and K. Koch on the Augment (Brunsvici 1868). On the subject of Nominal Composition I may name a paper by W. Clemm in Curt. Stud. VII, which gives references to the earlier literature of the subject, and one by F. Stolz (Klagenfurt 1874). On the forms of the Personal Pronouns there is a valuable dissertation by P. Cauer (Curt. Stud. VII): on the Numerals by Joh. Baunack (K. Z. XXV): on the Com- parative and Superlative by Fr. Weihrich (De Gradibus, &e. Gissae 1869). Going on to the syntax of the Cases, I would place first the dissertation of B. Delbriick, Ablativ Localis Instrumentalis, §c. (Berlin 1867), and next the excellent work of Hibschmann, Zur Casuslehre (Miinchen 1875). On the Accusative I have obtained the greatest help from La Roche, Der Accusativ im Homer (Wien 1861): on the Dual from Bieber, De Duali Numero (Jena 1864). On the Prepositions I have used the papers of C. A. J. Hoffmann (Liineburg 1857- 60, Clausthal 1858-59), T. Mommsen (see p. 147), Giseke, Die allméliche Entstehung der Gesdnge der Ilias (Gottingen 1853), La Roche, especially on tad (Wien 1861) and ézi (in the Z. f. dst. Gymn.), Rau on apd (Curt. Stud. III), and the articles in Ebeling’s Lexicon. On this part of syntax the fourth volume of Delbriick’s Forschungen is especially instructive. Of the literature on the Infinitive I would mention J. Jolly’s Geschichte des Infinitivs im Indogermanischen (Minchen 1873), also a paper by Albrecht (Curt. Stud. IV), and a note in Max Miller's Chips from a German Workshop (IV. p.49 ff.). The use of the Participle has been admirably treated by Classen, in his Beobachtungen twiber den homerischen Sprachgebrauch xii PREFACE. (Frankfurt 1867). A paper by Jolly in the collection of Sprachwissenschaftliche Abhandlungen (Leipzig 1874) is also suggestive. On the subject of the Pronouns the chief source is a dissertation by E. Windisch in Ourt. Stud. Il. On the Article almost everything will be found in H. Foerstemann’s Bemerkungen tuber den Gebrauch des Artikel’s bei Homer (Mag- deburg 1861). The controversy on the Reflexive Pronoun is referred to on p.173. On the Homeric uses of the Moods, besides Delbriick’s great work, I would mention Jolly’s monograph entitled Hin Kapitel vergleichender Syntax (Mun- chen 1872), and L. Lange’s elaborate papers on «/ (Leipzig 1872-73). It is to be regretted that they have not yet been carried to the point of forming a complete book on the Homeric use of ¢é. For the general theory of the subject Prof. Goodwin’s Greek Moods and Tenses is of the very highest value. Regarding the cognate question of the uses of dy and xév the main principles have been laid down by Delbrick. It is worth while to mention that they were clearly stated as long ago as 1832, in a paper in the Philological Museum (Vol. I. p.96), written in opposition to the then reigning method of Hermann. For the other Particles little has been done by Homeric students since Nagelsbach and Hartung. I have cited three valuable papers; on 7é by Wentzel, on 7 (ne) by Praetorius, and on pH by A. R. Vierke. I would add here a paper on the syntax of Causal Sentences in Homer, by E. Pfudel (Liegnitz 1871). On all syntactical matters use has been made of the abundant stores of Kuhner’s Ausfiihrliche Grammatik. And it is impossible to say too much of the guidance and inspiration (as I may almost call it) which I have derived from the Digest of Platonic Idioms left behind by the lamented friend to whose memory I have ventured to dedicate this book. On the collateral subjects of Metre I have profited most by Hartel’s Homerische Studien, La Roche, Homerische Unter- suchungen (Leipzig 1869), Knoés, De digammo Homerico (Upsaliae 1872-79), and Tudeer, De dialectorum Graecarum digammo (Helsingforsiae 1879). Oxronrp, July 18, 1882. Am pW HD 13. 16. 17. 19. 20. 22. 29. 31. TABLE OF CONTENTS. CHAPTER I. The Person-Endings. . Sentences—Subject and Predicate . Stem and Endings . The Person-Endings 7 . Thematic Vowel—Nenlthenintle forms . . Table of Person-Endings . . Influence of the Ending on the Stem 7. The forms of the 3 Plural . Meaning of the Middle . "i CHAPTER II. The Tenses. . Verb-Stem—Tense-Stem . 10, Il. Formation of the Tense-Stems The Simple Non-Thematic Present 12. Variation of the Stem The Simple Non-Thematic Aorist 14. Metathesis 15. Aorists in -«& and -& The Non-Thematic Reduplicated Bioaent The Presents with -vy (-v&) and -vu 18. Thematic forms (sporadic) Non-Thematic Contracted Verbs—Presents 55 ‘i + Aorists 5 21. Meaning of Non-Thematic Tenses i The Perfect % F 2 : . ‘ 23. Reduplication 24. The forms of the 3 Pioral 25. Long and Short Stems 26. The Perfect Participle . 27. Thematic forms . 28. Meaning of the Perfect ‘ ‘3 The Simple Thematic Present . : . 30. With Short Stem 3 d ‘ . The Thematic Aorist 7 e 7 3 . 32. Doubtful forms . ss 7 . 33. Thematic Aorists in tomes PAGE STAM WN A OW Io Io Il Il 12 12 12 13 14 16 16 16 18 19 20 20 21 22 23 24 25 26 27 XiV § 34. 35- . Aorists in -& 39- 40. 41. 42. 43. 45- . T-Class (-Te, -r0) 47- 48. 50. 63. 64. ‘ . Futures from Perfect and Aorist isis 3 67. 70. TABLE OF CONTENTS, The Reduplicated Thematic Present The Reduplicated Aorist (Thematic) 37. Stems with € in the Present wate) 38. The three forms of each Root Tense-Stems formed by a Suffix . The Aorist in -o% The Aorist in -ce, -co (Thematic) Passive Aorists—the Aor. in -y-v 55 a Onv 44. Manning of the Passive Aorists . Suffixes of the Present-Stem Nasal Class (-ve, -vo, &c. ) Iterative Class (-oke, -0Ko) 49. Iterative Tenses I-Class (-ye, -yo) . : 51. Verbs in -ww, &c. 52. Epenthesis 53. Assimilation of y 54. Verbs in -aw, -ew, -ow 55. Assimilation 56. Contraction . 57. Synizesis 58. Meaning of Verbs of the I- Class 59. Desideratives 60, Frequentatives 61. Intensives . 62. Collateral forms of the Proeautt in Hiner The Future in -ow The Future in -cew 66. Future Middle . The Augment (Historical Tenses) 68. The Pluperfect . 69. Loss of Augment Meaning of the Present and Aorist Stans 71. Present-Stem—Relative progress 72. Essentially progressive action 73. Past process (the Imperfect) 74. Descriptive Imperfect 75. Aorist-Stem 76. Aorist of eomtplation 77- Aorist Participle (coincidence) 78. Aorist of present time: Aor, in Similes—gnomic Aor. PAGE TABLE OF CONTENTS. CHAPTER III. The Moods. § 79. The Moods—Infinitive and Participle 80. The Subjunctive—Non-Thematic Tense-Stems 83. 84. 87. go. gl. 93- 96. 105. 106, 107. 108. Tog. IIl. 113. 114, 8x. Contraction . . 82. Thematic Tense-Stems The Optative The Verbal Nouns 85. The Infinitive . 86. The Participle . CHAPTER IV. Accentuation of the Verb. General rule of accentuation of Verbs 88. Accent in Composition . 89. The Infinitive and Participle CHAPTER V. Nouns and Pronouns. Nominal and Pronominal Stems Declensions 92. The Vocative The Case-Endings 5 94. Stems in-.-vand-o . 95- Stems in -& s . Cases—The Nominative Singular 97. The Accusative Singular 98. The Genitive Singular . gg. The Dative Singular 1oo, The Accusative Plural tor. The Genitive Plural . 102. The Dative Plural 103. The Dual 4 104. The Instrumental in -¢(v) Contraction, Synizesis, Hyphaeresis Variation of the Stem Heteroclite Nouns Heteroclite Pronouns 3 Adverbial Endings ; ‘ 110. Case-forms as Adverbs Accentuation of Nouns . 112, The Vocative . CHAPTER VI. Formation of Nouns. Nominal Stems—Primary and Secondary Primary Suffixes 115. Accentuation . 116, Gender PAGE 49 49 50 51 51 53 53 54 yo 71 75 76 XV Xvi TABLE OF CONTENTS. § 117. Secondary Suffixes 118. Compound Suffixes . 11g. Gender 120. Denominative Vaile 121, Comparison of Adjectives 122. Meaning of Comparatives suit Sapettakindy. 123. Composition 124. Form of the Prafixed ‘Stem j 125. Form of the Second Stem 126. Meaning of Compounds 127. Stems compounded with Seapuslllons 128 Accentuation of Compounds. 129. Proper Names 130. Numerals 131, Relation of Nouns and Pronouns to the Verb . 132. The Accusative—Internal and External Object CHAPTER VII. Use of the Cases. 133. Neuter Pronouns. : 134. Neuter Adjectives . 135. Cognate Accusatives . 136. Other Adverbial Accusatives 137. Accusative of the part affected 138. i of Time and Space 139. is with Adjectives . 140. es of the External Object . 141. Double Accusatives . 142. The Dative 146. The 154. The 143. The ‘true’ Dative 144. The Instrumental Dative 145. The Locatival Dative Genitive . ‘ 147. The Genitive aii Nouns 148. ,, 3 in the Predicate 149. 4, ¥5 of Place 150. 4, ¥ of Time 151. The quasi-Partitive Genitive 152. The Ablatival Genitive 153. The Genitive of Price Case-Ending -gu(v) . 155. Instrumental 156. Ablative 157. Locative 158. Dative and Genitive, 159. Forms in -Oev and -ws . 160,-The Ending -ws PAGE 77 78 79 79 80 82 82 82 85 86 88 88 89 go gt 92 93 93 94 95 95 95 97 97 98 99 100 Io2 102 104 104 105 105 108 10g 110 IIo IIo IIl I1I I12 113 § 161. 164. 165 169. TABLE OF CONTENTS. The Nominative—Impersonal Verbs . 162. Nominative in the Predicate 163. Interjectional Nominative The Vocative 2 . Substantive and Adjective 166. Gender of Adjectives 167. fi Pronouns . 168, Implied Predication . CHAPTER VIII. Use of the Numbers, Collective Nouns 170. Distributive use of the Singular 17. Plural of Things is é ; 172. Neuter Plural with Singular Verb . 173. The Dual ‘ CHAPTER IX. The Prepositions. 174. Definition ‘i : 175. Adverbial use of Prepositions . 176, Tmesis . 177. Ellipse of the Verb . 178. Use with Oblique Cases ‘ 179. Use with the Genitive 180, Accentuation—‘ Anastrophe’ of Prepositions . 181 .aupi. ‘ 182. dugi with the Dative 183. », Accusative 184. ,, », Genitive 185. wept ‘ 2 é 186. wepi with the Dative. 187. 4, » Accusative 188. ,, » Genitive 18g. mapa ‘ 3 é 190. apa with the Dative Igl. 4, » Accusative 192. 4 »» Genitive 193. wera. 2 ‘ 194. peré with the Dative 195. 55 » Accusative 196. ,, » Genitive 197. éni 5 . 5 . 2 ‘ 198. émi with the Dative . : ‘ ‘ 199. 5» » Accusative 200. ,, » Genitive é 3 § b xvii PAGE 113 114 IIs 116 117 117 118 11g Ig 11g 120 121 121 123 123 123 124 125 126 126 129 129 130 131 131 131 132 133 133 134 134 135 135 136 136 136 137 137 138 139 Xvili § 201, 205. 209. 211. 214. 217, 220, 221. 222. 223. 224. 225. 226. 225 228. 230. 231. TABLE OF CONTENTS. Pu) - : , ‘ 202, iné with the Dative . 203: 55 », Accusative 204. 45 », Genitive mpoti (mpés), tort F : 206. mpori with the Dative 2073. 45 » Accusative 208. ,, » Genitive dvé—with the Dat.—with the Gen. . 5 210, dvd with the Accusative natal. 5 ‘ 4 212, xaTd with the Accusative 213. 55 » Genitive bid a ‘ , 215. 6&4 with the Accusative 216. ,, » Genitive tnép 4 é : 218. imép with the Accusative 219. 4, 1 Genitive avi (&) . ody, Evy eis eg and mp6 avi Double Prepositions Improper Prepositions . 229. Homeric and Attic uses CHAPTER X. The Verbal Nouns. Nature of the Verbal Nouns The Infinitive—original meaning 232. Infinitive with Nouns 233. 3 with Impersonal Verbs 234. 45 as the Subject 235. a with Relatives 236, #3 with mpiv and mdpos 237. Accusative with the Infinitive 238. Tenses of the Infinitive 239. Dative with the Infinitive 240. Attraction é . 241. Infinitive used as an Imperative 242. Origin and History of the Infinitive . PAGE 139 140 140 141 142 142 142 142 143 143 144 144 145 145 145 146 146 146 146 147 147 147 148 148 £49 149 150 150 151 153 153 155 156 157 158 T58- 158 160 161 161 162 163 TABLE OF CONTENTS. xix PAGE § 243. The Participle—uses . * ‘ . 164 244. Tenses of the Participle é ‘ : - 166 245. Implied Predication . ‘ ‘ : . 166 246. The Genitive Absolute : ‘ 3 . 167 CHAPTER XI. Use of Pronouns. zl 247. Subordinate Clauses. : ‘ . . 168 248. Deictic and Anaphoric Pabinbiens ‘ : F . 168 249. 55e, Toadabe, Todade, We, evade : . 169 250. Kelvos . é i ‘i ‘ : é . 169 251. ovrTos . 3 4 ; : . 170 252. adrés,atras ’ : ‘ j ‘ . 170 253. The Reflexive Pronoun ‘ : ‘ g 2 2 254. The Possessive éds, és : > AGB 255. éds, ds as a general Reflexive ‘ « 193 256. The Article . ‘ ‘ » 5 257. The Sicbstantival rere ; ; . 176 258. The Attributive 5 ‘ . 178 259. With connecting Barlidles ; . 178 260. With Adjectives é : - 179 261. The defining Article. 5 . 181 262. The Article asa Relative . ; F . 182 263. The Article with ré ‘ . 184 264. Homeric and Attic use of the Avttets : » 184 265. 6s7 65 . - , : ; : 185 266. és Te, Os Tis. ‘i , : . 186 267. Correlative Clauses . 2 s : . 187 268. otvexa : : 5 ‘ - 188 269. 6, O71,57e . a ‘ ‘ 5 . 189 270. 6, dr 6 Te as Conjunctions ‘ ; . Igt 271, Form of the Relative Clause i + 192 272. Double Relative Clauses é ‘ + 193 CHAPTER XII. Use of the Moods. 273. Classification of Sentences ‘ ‘i a ‘ - 194 274. The Subjunctive in Principal Clauses . i ‘ - 196 275. Affirmative . . : : ‘ - 197 246. Negative . ‘ ‘ ‘ : - 198 277. Interrogative . : : ‘ . 199 278. Prohibitive . i ‘i : + 199 279. Homeric and Attic uses , 201 280. The Subjunctive in Subordinate CianseiCausi vith pte 201 281. Clauses with yj F ‘ i 3 + 202 282. Relative Clauses—Final .. 3 é . 202 283. ay a Conditional 7 7 + 204 b 2 TABLE OF CONTENTS. PAGE § 284. Relative Adverbs. é : . 206 285, ws, Sms . - : ‘ . 206 286. iva ‘ ‘ : : . 207 287. dppa ss. r é . i . 207 288. ews, eis 6 : : . . 208 289. dre, émére 3 5, 5 208 290. eUTE, HMOs : ‘ , 210 291, Clauses with ei, &c. . : : . 210 292, Conditional Protasis . ‘ ‘ 211 293. Final Clauses with ei F ‘ . 212 294. Object Clauses with ei . a‘ - 212 295. The Subjunctive with ds ei . F a » 213 296. éweé with the Subj. . ‘ . 213 297. mpiv 3 ‘ , ‘ 214 298, The Subjunctive after Secondary Tenses « 214 299. The Optative in Simple Sentences ‘ é . 215 300. With dy or nev 4 5 és * . 217 301. The Optative in Subordinate Clauses . j { . 219 302. Clauses with 7é—ie . : ‘ ‘ » 220 303. Clauses with pq : j . . . 220 304. Relative Clauses—Final j 7 * “221 305. 53 ‘5 Conditional « “B22 306. ws, Oras, iva . ‘ : : a (223 307. Ews, Sppa ‘ 3 : ‘ . 224 308. Ore, 6wére . s : i 225 309. éwet . , : : 226 310. mpiy . ‘ ‘ , . 226 311. ei—Optative of Wish . 226 312. ei—Conditional Protasis ‘ + 22% 313. ef «ev—Conditional Protasis , ; . 228 314. ei—Final and Object Clauses - 228 315. History of the Subjunctive and Optative— Uses in Principal Clauses. 229 316. ,, Subordinate Clauses . 229 317. Original meaning . i ; » azt 318. Conditional Protasis with e . . 232 319. Final Clauses with ei 5 ‘ - 233 320. Order of the Clauses . : ; « 233 321. ld dye 6 , ‘ 234 322. Homeric and Attic Uses. ‘ 234 323. The Indicative—Modal Uses . : 5 : » 235 324. Conditional A podosis 7 ; 5 . 236 325. Past Tense by Assimilation . : = 237 326. Future Indicative . 4 - i » 5337 327. The Imperative ‘ : ‘ 3 ‘ + 239 328. Prohibition . : ‘ * + 239 TABLE OF CONTENTS. CHAPTER XIII. The Particles, § 329. Classification of the Particles . 330. Kat ‘ ‘ 331. Té s s 332. ré in general statement 333. 56 é . 334. 5€ of the Apodosis 335. Enclitic 5€ 336. GAAd, adrdp, drdp 337. av, atre, tums . 338. é . 339. Tin, émein 340. 96, ‘ - ; 341. Dependent Interrogative 342-5. pay, BAY, pey 346. Tot ‘ 347-8. dpa, yap 349-352. otv, 54, vb, Ov 353. mép . 354. “yé 355. ov, jo —Bisltnction of usage 356. odd, undé, ovddels 357. Double negatives 358. Uses of p—Indicative 359. od and yw in Conditional Clauses 360. od with the Infinitive and Participle 361. BH ” ” ” a. 362, dv, xev—Summary of uses : 363. Difference of dy and nev 364. Original meaning of dv 365. xéev—ZMolic dialect . CHAPTER XIV. Metre and Quantity. 366. The Hexameter 367. Diaeresis and Cusaisa 368. Spondaic verses 369. Quantity of Syllables 370. Position 371. Lengthening hetere:p Pr A, pv, 0, 8 372. Origin of the lengthening 373. Final - of the Dat. Sing. 374. Final -a of the Neut, Plur. 375. Short Syllables ee in a Consonant 376. Elision, &c. . : A 377. Crasis 378. Synizesis xxi PAGE 240 241 241 242 245 245 246 247 247 247 248 249 250 251 252 252 255 257 258 259 259 259 260 261 262 263 263 265 268 269 270 270 272 273 373 275 276 278 278 279 281 281 282 xxii TABLE OF CONTENTS. PAGE § 379. Hiatus . : ‘ . ‘ . . « 282 380. Long vowels before Hiatus . , . 283 381. Shortening of diphthongs before Hiatus : . 284 382. Hiatus after short syllables . é . 284 383. Doubtful Syllables i E : : . 285 384. Doubtful vowels. . : . 286 385. Doubling of consonants . : : - 286 386. Metrical licence . ‘i z : . 287 387. Vocatives . 2 : 7 ‘ - 287 388. The Digamma . . , : . 287 389. Nature of the evidence fein Te a ‘ : . 288 390. Words with initial F : 7 3 . 289 391. Words with initial of (‘F) . 5 ; - 296 392. F inferred from metre only . Z . 297 393- Loss of fF, esp. before o,w . . : . 298 394. Initial 5F St : “ : 3 + 300 395. Initial Fp, &e. ‘ ‘ F ® - 300 396. F not initial . e s ; . 301 397- Loss of initial o andy . 5 - : . 301 398. Summary a : . # 3362 399. Theories of the Diana . ‘ : 3 . 302 400. Hypothesis of alternative forms si 2 » 303 401. Explanation from fixed phrases, &c. ¥ » 304 402. Hiatus, &c., asa survival . ¥ + 305 403. Explanation from the nature of the F 3 + 305 404. F in other Greek dialects & 5 . ‘ - 306 405. FinIonic . . : ‘ ‘ . 308 APPENDIX. A. On the Tenses with Stems ending in -&. : é - 310 B. Ono in Verbal Stems . ‘ 7 ‘ ‘i . 314 C. Onyandein Homer . 5 ‘ ‘ . 316 D. On the Assimilated forms (§ 5 eA ‘ ‘ * - 318 E. Order of the Particles and Enclitic Pronouns . g - 319 Other Notes and Corrections . 3 J 4 . 320 Index I.—Homeric Forms a r 3 F s 323 II.—Subjects (including Syntax) . 5 3 » 335 IIT.—Chief passages referred to 5 ‘ : » 342 BOOKS FREQUENTLY REFERRED TO, WITH THE ABBREVIATIONS USED. RiZu siteuiiess Zeitschrift fiir vergleichende Sprachforschung, herausgegeben von A. Kuhn: Berlin, 1852 ff. Curt. Stud.,,. Studien zur griechischen und lateinischen Grammatik: Leipzig 1868-78. (Edited by G. Curtius]. Bezz. Beitr.... Beitriige zur Kunde der indogermanischen Sprachen, herausge- geben von Dr. Adalbert Bezzenberger: Géttingen, 1877 ff. Curtius, Gr... Grundziige der griechischen Etymologie von G. Curtius: 5 Auf, Leipzig, 1879. », Werb. Das Verbum der griechischen Sprache seinem Baue nach darge- stellt, von G. Curtius: Leipzig, 1873-76. Kiihner ...,.. Ausfihrliche Grammatik der griechischen Sprache, von Dr. Raphael Kiihner: 2 Aufl., Hannover, 1870. Synt. Forsch.. Syntaktische Forschungen, von B. Delbriick und E. Windisch. Vols. I—IV. Halle, 1871-79. Bekker, H.B. Homerische Blatter, von Immanuel Bekker: Bonn, 1863-72. Veitch .,,.... Greek Verbs irregular and defective, their forms, meaning, and quantity, by William Veitch: Oxford, 1871. LaRoche,H,.T. Die Homerische Textkritik im Alterthum, von Jacob La Roche: Leipzig, 1866. Roby ......... A Grammar of the Latin Language, from Plautus to Suetonius, by Henry John Roby: London and New York, 1871-74. EDITIONS. Homeri et Homeridarum opera et reliquiae: ex recensione Frid. Aug. Wolfi: Lipsiae, 1804-1807. Carmina Homerica: Immanuel Bekker emendabat et annotabat: Bonnae, 1858. Homeri Ilias ad fidem librorum optimorum: edidit J, La Roche: Lipsiae, 1873-76. Homeri Odyssea ad fidem librorum optimorum: edidit J. La Roche: Lipsiae, 1867-68. Page 2 ” ”» ” ERRATA. 4, line 2, for § 81 read § 82. » ” 10, ”? ” ” II, 5 38, IP 3, for § 82 read § 83. 28, for § 81 read § 82. 5, for f-ara read f-arat 20, for z Dual read 3 Dual. 33, dele hv, Env. 2, dele 1 Sing. énv and the. ”, for § 82 read § 83. 41, last line, for § 75 read § 86. 50, line 39, for § 82 read § 83. 66, ,, 76, 5 81, ,, 86, ” 945 a” I20, 167, ” 186, 4, 197, 1 210, 224, , 14, for épsw read Epws. 20, add venvins (Od. 14. 524). 3, for Heoros read Haoros. 30, for ed-wvopos read eb-dvupos. 44, for 24. 375 read 24. 735. I, for 872 read 782. I, for 13. 16 read 13. 353. 15, for § 259 read § 262. 29, for wé read ke. 17, for 8é read 5é. 28, for drws read Sos. 251, lina 38, for Td0c rend rdde. 255, heading, for TAP read PAP. HOMERIC GRAMMAR. CHAPTER I. Intropuctory.—Tur Prerson-Enpinas. 1.] Att language of which grammar takes cognisance consists of SENTENCES. The simplest Sentence must express the combination of a Supsect—that about which we speak (or think); and a Prept- caTE—that which we say (or think) about the Subject. 2.| In Greek (and generally in languages whose structure resembles that of Greek) every Verb is a complete Sentence, consisting of two parts, the Stem, which expresses the Predicate, and the Ending, which expresses the Subject. Thus éo-ti he (or it) is, pa-0 say thou, 7AOo-pev we came, are Sentences; the Predicates are expressed by the Stems éc-, ¢a-, 7AGo-, and the Subjects by the Endings -7, -6:, -wev. As the Endings of a Verb may always be translated by Personal Pronouns they are called the Person-Endings. Note. It may happen that the ending has been lost by phonetic corruption, as in é\aBe (for éAaBe-r) he took. This however does not form a real exception, because in Greek such words are used exactly as if the lost ending were still sounded. In English the case is different. The form took, for example, can only be used to express a Predicate. The original Subject is lost to the mind as well as to the ear. It may be worth while to notice that the term ‘Verb’ is used in Grammars with a double meaning, sometimes of a single form—as when we say that érimro- pev is ‘a Verb’— sometimes collectively, as when we say that érdmzo-pev is a ‘ part’ of ‘the Verb rumrw.’ Here ‘a Verb’ means a group of forms. 3.| There are three main sets of Person-Endingss :— 1. Those used in the ‘ Principal’ Tenses (the Present, Perfect, and Future Indicative), and in the Subjunctive ; these are called the Primary Endings. 2. Those used in ‘the ‘Historical Tenses’ (the Imperfect, Aorist, and Pluperfect), and in the Optative; these are called the Secondary Endings. 3. Those of the Imperative. 4.] The further modifications which the Endings undergo depend chiefly upon the final letter of the Stem. B a HOMERIC GRAMMAR. [5 In certain Stems the Ending is preceded by O or E: that is to say, O before the nasals p, v, and E before other letters; e.g. réntO-pev, TéntE-re, réntO-vre (older and Dor. form of span We shall call this the Thematic Vowel,* and the Stems whic contain it Thematic Stems. The term will naturally include the corresponding Subjunctives, in which the final letter of the Stem varies in the same way between y and o, as Téntw-pev, rumTy-TE, &e. The 1 Sing. Ending -w probably stands, in the Indie. for -o-p., 1n the Subj. for -w-p.. The forms which do not contain this variable e or o are called Non-Thematic. Among these, again, we have to distinguish a group of Tenses with Stems ending in -d, viz. the Perfect, the First Aorist, and some forms peculiar to the Ionic Dialect, as the Plpf. jdea I knew, the Impf. ja I was, ijia I went, &e. In these Stems the -a changes in the 3 Sing. to -e(v).+ The distinction between Thematic and Non-Thematic applies in strictness only to forms, but may generally be extended to Tenses and Moods. E.g. the Future is Thematic, the Optative is Non-Thematic, &c. But it does not apply to ‘ Verbs’ (in the collective sense of the term), because almost every Verb is made up of forms of both kinds. 5.| In the following Table of the Person-Endings found in Homer the Endings distinguished by larger type are those of the Non-Thematic Tenses. The Endings in smaller type are, first, those of the forms with -d, and, under them again, those of the Thematic forms. In the Dual and Plural (except the 3 Plur.) the Endings are the same throughout. * This vowel has also been termed the ‘Connecting’ or ‘ Auxiliary’ Vowel— names given on the supposition that it is originally euphonic, inserted in order to allow the Stem and the Ending to be distinctly heard in pronunciation. The name ‘Thematic’ implies the opposite theory, viz. that it serves to form a ‘Theme’ from a simpler element or ‘ Root,’ as Aey-e from the Root Aey-; see Curt. Chron. p. 40. On this theory the Stem Aey-e, Aey-o is originally the same as the Theme of the Noun Adyo-s, + The & of these Stems is of course quite different from the final vowel of the Stem in such forms as da-pév, tora-pyat, TéTAa-61, where it is part of the Verb- Stem or ‘Root.’ For the non-radical & has the character of a ‘connecting’ vowel, and so far it resembles the Thematic € or o. That it is distinct from the Thematic Vowel is well shown by Brugman, who points out (in Curt. Stud. ix. p. 293) that there is a consistent set of differences between the ‘connecting’ & and the Thematic € or 0, in the Noun as well as in the Verb ;—that, for example, the & of the Acc. Sing. and Plur. of the consonantal Declension is analogous to the -& of Verbs, just as the -o (Voc. -e) of the Second (or Thematic) Declension answers to the € or o of Verbs. Thus— the a of 7é5-a and 75-is is to the o of Adyo-y and Adyous (Adyo-vs) as the a of ja (j0-a) I was to the o of €-Aeyo-v, and again— as the & of AeAdyx-Gar to the o of Aéyoust. And while € or o is carried regularly through all Numbers and Persons (as wel as all Cases of a Noun), we shall see that the forms of a Tense with -& often interchange with forms which have no ‘ connecting’ or ‘auxiliary’ vowel. Whether the @ is originally auxiliary is a further question, on which see Appendix A. PERSON-ENDINGS. 5.] (a)jom- *g ‘(a))on0- 409 93- aMLno~ oLdo= ao= WL1O=- 'S “p1d0= amgop- amsap- o1Lan= ap- (a)jov- 1A090- AMLA- OLD~ ‘oLd- adavo-‘d- IDLD- 9QDLA- ‘(a)2.00- ‘(a) 1.00 c 390- DL- 399- aL- 39-0- ab~| ot fs — + — | vgar!-‘ng.097"- aan/- | ngan~‘ng.0971- aar- | apg 1 AMg0- AwL- aligo- dao.-‘als- 09.0- d0L- € dog0- aoL- 0g0- Ao0L~ 069-0- doL- z SS = aogar- — | ruqi 9.03- o.13- 013+ (a)3- rsl= "g “1139- qoli- = *g 13- p0p- — @1- o.0- («1)2- (a)3- 90- M1- OL- (2)- 10.L- (a)r0- ‘(a)ob- ‘Surg € 03- = $3- 9- 0>- S3- roli- g “n= sli-° g {s13- on-‘0.00- ao- op- sp- so- 0-‘00- S-‘19- 00- S- 9190-‘I190- Dgo- ‘S- 9.90- | ‘Sug % alirlo- ao- orlo- *g ‘rlo- ao- ‘o- ‘fqng ‘m= alinip- D- D- alr- a- on qrf- | “Butg I “din “LOV “cin “LOV “cine “LOV “‘TALLVAGd M1 “LUVANOOES ‘AMV B2 4 HOMERIC GRAMMAR. [5- Remarks on the Table of Person-Endings. 1 Sing. On the Subjunctives in -w-p. see § 81, and on the Optatives which take -p. in the 1 Sing. see § 82. 2Sing. The original -o. remains only in éo-ot thou art. Eis (or enclitic eds) is read in nine places, but there is only one (viz. Od. 17. 388) in which the metre does not allow goo’ to be read instead. Probably therefore égot is the genuine Homeric form. The Attic ef is not found in Homer. -o04 is the regular ending in oto@a (*oi5-06a) thou knowest (oléa-s, Od. 1. 337 only), 75-060 thou knewest, ho0a thou wast, et-c8a (Il. 10. 450) thou wilt go. It occurs as an alternative form in éy-06a or $f-00a thow saidst ; in some Sub- junctives, as €0é\y-o0a, eiry-o8a; in two Optatives, Bddor-oOa (Il. 15. 571) and KAalou-o6a (Il. 24. 619); and in tl0y-00a, S8o0i-00a (Il. 19. 270). In the Mid. -oat, -co are the usual Non-Thematic Endings: but the Opt. always has -ou-o. We find pépvy-ar (Il. 21. 442), as well as pépyy-oar (Il. 23. 648), Stfq-ar (Od. 11. 100), BeBdn-a (IL. 11. 380, scanned BeBAgar 0v8’), épdpva-o (Od. 22. 228), and Imper. papva-o (Il. 15. 475), waplora-o (Il. 10. 291); contracted pépvg (I. 15. 18, but see § 27), éxpepw (Il. 15. 18), Sapva (Il. 14. 199). Verbs in -ew, which would form -eeat, -eeo, usually drop one e; pW0ear, drroaipeo, &cAeo (from #Aéw, Il, 24. 202); but puvOefa (Od. 8, 180). There is one instance of -ev, viz. diet (Il. 23. 620). In the Imper. -6 is commonly retained in Non-Thematic forms, as xAb-01 hear, rérda-Ou endure (Imper. of rérAnxa), iw-6e (Od. 3. 380). It is dropped in torn (Il. 21. 313), daivo (I. 9. 70). Sometimes it becomes -s, as 0¢-s, 5d-s, mpd-e-s ; also in the Thematic évi-one-s tell, ep. the Attic oyé-s. 8 Sing. The original -r remains only in éo-ré(v). On the Subjunctives with y-ov see § 81. 8 Plur. The Ending -do (for -ovm) is found in é-dou (for *éo-aov) they are and i-aou they go. Stems in a, ¢, 0, uv form -aor, -eror, -over, -dor (for -a-vrr, &c.), as ioraor, TWeior, Sidodo1, Cevyvicr (not T10é-acr, &e., as in Attic). The Perfect Act. has -ao., and (rarely) -déov. The latter occurs twice in Homer, tepvx-aou (Od. 7.114), AeAdyxaow (Od. 11. 304); for other examples in Jonic see Curt. Verb. ii. 166. In these forms the @ belongs to the Ending, since -dov is for -én, which corresponds to the -vr of the Doric ga-vrl, A€yo-vts (as -dta in the Mid. to -vra:). The forms with -aov are of later origin, and belong to two essentially distinct groups; see § 7. -dv is found in jjo-av. The Ending -v is common in Homer, as éa-v, érie-v, though partly displaced by the later -cav. The vowel preceding -v is always short, e.g. éora-v, but gory-cav. prdvOny (Il. 4.146) must be a Dual: perhaps for é-pud-cOyv (cp. pa-pds). In the Middle, the forms -dra, -dro are regular after conso- nants and the vowel c (including the diphthongs e:, n, o, &c.); the forms -vrat, -vro after d, «,o. After v, 7 both forms are found: e.g. elpv-arat, eipv-aro, but AéAv-vTau, Kexv-vTar; BeBAjaras (IL. 11. 656), but peury-vro, EipBAy-v70. 6.] PERSON-ENDINGS, 5 The Imper. Endings -rwcav, -cOwoav are post-Homeric. 1 Dual. -p<@ov occurs only in repiddueOor, Il. 23. 485. 2 and 3 Dual. In the Historical Tenses, according to the ancient grammarians, the regular Endings are— 2 Dual Act. -tov, Mid. -céov, 3 oo» » "THY, 5, -oOny. This scheme, however, is open to some doubt; for— (1) Homer has three instances of the 3 Dual Impf, in -rov, where the metre does not admit of -ryv, viz. &mxe-rov (11. 10. 363), érevyxe-Tov (Il. 13. 345), Aapvave- tov (11.18. 583). Three others in -oOov occur as various readings, where the metre admits of either -o@ov or -oOqv, viz. dpixe-oov, read by some ancient critics (pro- bably Zenodotus) in Il. 13. 613: Owpnace-cOor, the reading of A. (the Cod. Venetus) and Eust. in Il. 16. 218: mére-cOoy, a marginal variant of A. in Il. 23. 506, (2) Three forms of the 2 Dual in -ryv were read in the text of Zenodotus, viz. xapé-ryy (Ll. 8. 448), AaBé-ryv (Il. 10. 545), jOeAé-rqv (Il. 11. 782). Aristarchus read xdpe-rov, AdBe-rov, 7OéAc-rov. The metre gives no help to a decision. (3) In Attic the examples of the 2 Dual in -tyv, -o@qv are so common that Elmsley (on Ar. Ach. 733) held these to be the only correct forms, thus making the Dual of Historical Tenses uniformly end in -nv, as the Dual of the Principal Tenses ends in -ov. Cobet maintains the same view (Misc. Crit. pp. 279 ff.). But the account of the Greek grammarians is strikingly borne out by the forms of the Sanscrit Dual. In Sanscrit we find that in the Historical Tenses the 2 Dual ends in -tam, 3 Dual in -¢dm, answering perfectly to the Greek -roy, -rnv. This therefore is to be regarded as the rule. The exceptions which have been quoted are evidently due to the tendency towards uniformity: and it is to be noticed that this tendency seems to have acted in Homer in the direction of making all Duals end in -rov, -c@ov, whereas in Attic the Endings -rny, -oOyv were extended to the Second Person, Influence of the Ending on the Stem. 6.] In Thematic Stems it is plain that the Ending influ- ences only the final e(o), leaving the rest of the Stem unaffected. Non-Thematic forms, on the other hand, are liable to variations in quantity which affect the main vowel of the Stem. These variations are governed by the general rule that of two forms of the Stem the longer is found with the Endings of the Sing. Indic. Act., the shorter with all other Endings, viz. those of the Dual and Plural, the Imperative, and the Middle. Thus :— (1) a, ¢, o interchange with the corresponding long vowels a (in Ionic n), 9, #3 as py-pl, Edy-v, but 1 Plur. pa-yev, Imper. pa-0l, Mid. é-pa-ro ; riOn-p, Mid. ride-par; déd-m, Mid. di50-ya, (2) t with e« and o: as ef-ys, 1 Plur. t-yev, Imper. i-4; ota, 1 Plur. t6-yev. (3) o with ev and a: as &yeva, Mid. xd-r0 (§ 15); delkvd-mr, i Plur. deixvd-yev. 6 HOMERIC GRAMMAR. [7 The same law governs the interchange in the Perfect of— (4) a with o: as yéyova, 1 Plur. yéya-wev; mémovda, Part. Fem. men dd-via.* (5) &p with op: as épOopa, Mid. &pGap-ra.* (6) Sometimes the longer Stem contains an additional con- sonant, viz. in the Perfects and Aorists in -xé, as orn«a, 1 Plur. éord-pev 3 2Onxa, 1 Plur. éOe-wev. The Endings which are found with the long Stem have been called the Light, the others the Heavy Endings. 7.] The 3 Plur. offers some exceptions to the general rule :— (x) The Ending -cot (for -d71, -NT1) is used with the long Stem, as NeAdyx-tiot, Tepix-dot. Cp. reredx-drat, é-reredx-Gro (§ 22, 3). (2) The long Stem is also found in a few forms of the 3 Plur. PE in -aou, as rerolOdor, éotijxaor (§ 24), and the 3 Plur. of the Aor. in -xa, as 20nxav, 2wxay ($15). These are apparently formed on the analogy of the Singular, like ofda-pev, of8acr in Herodotus. ) The Endings -(c)aov, -cav (for -=ANTI, -SANT) are found with the short Stem. The leading examples are :— With Simple Stems: é-a-cav, &-Oe-cav, é-d0-cav, Ke. Presents: 710é-acr, di8d-aor (Att.) ; é-ri0e-cav, é-5i50-cav, &e. Perfects : tact (i8-caot), toav; eigacr (Att. 3 Plur. of éovxa). BeBd-act, yeyd-act, weud-act; BéBa-cav, péna-cav. , éotaor (*éord-acr), teOvacr; Ecta-cav, TéOva-cav. mepv-act, ded{-acr; deldi-cav. The hiatus shows that -aov is for -cdo1, the Primary Ending answering to -civ. The corresponding Mid. -céra: is found in Doric (yeypdyarat, Tab. Heracl. i. 121, in C. I. 5774). In the Plpf. Ending -ecav the -eoa belongs to the Stem (#dea for 78-ecd). The contraction in rao, TeOvaor is evidently due to the impossibility of éord- act, reOvd-aot in the hexameter. Brugman regards them as wrongly accented, and derives them directly from tora-vri, Té0va-vrt (Curt. Stud. ix. 296). This is open to the objection that in all other Stems which take -ka the Endings -vr and -v are confined in Homer to the long forms: thus we find— mepix-dovr and tmepv-dou, but not mépvor éorpxdor, &c. ,, BeBd-acr, » BeBacr (otSaor Hdt.) ,, toacr » B-aore €0yKa-v » e-wav » Bev eSwka-v » 0-cav x €0-v (Hesiod). The short form with -vtt, -v is therefore confined to Verb-Stems ending in a vowel, as in gaol, retort (for pavti, ride-vr:). And even in these it might be maintained that the short vowel is due to the (original) following -NT, as in @-oriv, hyepev, GAd-vtes, Ke. * The interchange of & with ev, and of dp with ep (parallel to that of t and % with e and ev) cannot be exemplified within the Non-Thematie conjugation. 8.] PERSON-ENDINGS, y Meaning of the Middle. 8.] The original force of the Middle Person-Endings is ‘ Re- flexive ;’ that is to say, they denote that the action of the Verb is directed towards the agent. Greek has no Passive Endings distinct from those of the Active and Middle: it is desirable therefore to speak, not of Passive forms, but of the Passive meaning or use of a form. The chief uses of the Middle are— (1) The use to signify that the agent is also the indirect object of the action—that the action is done by some one for or toward himself, or in his own interest: évvv-yar I put (clothes, &c.) on myself; déxo-yat I take to myself; dop df épvocdyevos having drawn him his sharp sword; hpetro réfov took his bow with him; pepéodw let him bear away as his prize. (2) The use in which the agent is the direct object of the action, as Aovo-pat I wash myself. This is comparatively rare. (3) The Intransitive use, in which the reflexive sense is faint, as gaive-tat appears (but daiver Eaurdv he shows himself). So, generally, when the action centres in the agent; as in Verbs of motion (€pxopat, TéTomat, GAAopuat, olxouat, &c.), and in such uses as AaBéoGa to gain a hold of (not to take a thing); also in Verbs of feeling and thinking (aic@dvopat, aidéouat, BovAouat, otopat, peurnpat, emiotapat, pedopar, peupouat, &c.) So in French, ‘je m/apercois’ J perceive, ‘je me doute’ I suspect. (4) The Reciprocal use; dueSdueros taking his turn, héyerOau to tell over (in talk). (5) The Passive use, as éxe-rat 7s possessed. It may be illus- trated from the similar use of some Reflexive Verbs in French, as ‘je me trouve’ I am found; so in Italian, ‘dicesi’ i zs said, &e. The Middle is rather more common in Homer than in later Greek. For example, in the class of Verbs of feeling and thinking we may add the Homeric Zpayae I love, -yavupar, éAdopar, ZATropat, GOopat, dvopat, oTévopat, KexXdpovro, é8vcac0a. And the use is extended to Verbs of seeing and hearing, as épa-pat I see (Aor. idé-c0a), dxovo-par I hear (used as well as 6pd, ideiv, dxovw), Sépxopat, Bocopa, oxérropat, ppacopar; cp. the Attic cxorov-pa I consider, Conversely, Homer has the Act. diw I think, as well as the Mid. dio-yar I harbour the thought, suspect. Sometimes (esp. in Homer) the Middle appears to be used because the Verb implies acting arbitrarily, as a superior, &c.; e.g. otvopa: and dyAgopae to do mischief for one’s pleasure ; épiraro made a favourite of ; Bie-vTac run in a race, SiecOar to chase (but Siov I fled); dediccecOa to terrify; Kékdrero shouted in command, A use intermediate between the Reflexive and the Passive (pointed out by Riddell, Dig. § 88) may be exemplified in dwjyGero incurred hatred. On the Futures only used in the Mid., see § 66. 8 HOMERIC GRAMMAR. [9- CHAPTER II. Tur TENSES. g.] Verb-Stem and Tense-Stem. A comparison of the dif- ferent forms of a Greek Vers usually enables us to see that some ‘ one syllable or group of syllables is present in them all: as tun- in the forms of réntw, or Bovdev- in those of Bovdedw. This we shall call the Verb-Stem. A Verb-Stem not derived from more primitive elements is called a Foot. Again, the different forms belonging to any one TENSE are based upon a common part, which we shall call the Zense-Stem. This part may be the same as the Verb-Stem ; or it may con- tain an additional element, as 8- in 8df-b0-pev, b1-d0-fy-v, &e.; -te, -to In TUn-Te-TE, TUT-TO-pEV, &-TUT-TO-V, TUT-To-I-pL, Ke. The Subjunctive and Optative, again, are distinguished by a Suffix to the Tense-Stem: e.g. b0-iy-v, b:d0-in-v, TUTTO-1-pL, otjoa-i-ut. These may be called Mood-Stems. Finally, the Stems used in the ‘ Historical ’ Tenses—the Impf., Aor., and Plpf.—are formed from the Tense-Stem by prefixing the dugment. The Stems of the augmented forms are therefore parallel to the Mood-Stems, the only difference being that they are formed by a prefix, while the Mood- Stems are formed by a suffix, They may be described as Time-Moods of the several Tenses,—combining the notion of Past Time, which is expressed by the Augment, with the meaning contained in the Tense-Stem. Each Tense-Stem furnishes an Infinitive and a Participle. Thus we have (supplying one or two links by analogy) from the three Tense-Stems Badhe (or -o), Bade (or -0), BeBAnka. PRES. AOR. PERF. Principal Tense Badre-Te wanting BeBAnjxa-te. Historical €-Badde-TE é-Bade-re *e-BeBArjKe-a. Subjunctive BaddAn-Te Badn-Te *BeBAnKy-TE. Optative Baddo-1-Te Bado-t-Te *BeBAnKo-t-Te. Imperative BaddAe-re Bade-te * BeBAnKa-te. Infinitive Badde-pevat Badé-euw BeBaAnx-évat. Participle Baddo-vT0s Badd-vros BeBAnk-dros. It is evident that there might have been a Future ‘Time-Mood’ as well as a Past for each Tense-Stem. In English indeed we can distinguish progressive action in the future as well as in the present and past: I shall be writing as well as I am writing and I was writing. See Goodwin's Moods and Tenses, § 25, N.1 ; 10. | THE TENSES, y Driver’s Use of the Tenses in Hebrew, § 4. Modern Greek has two such Futures : 6a ypapw and 6a ypdyw differ almost precisely as éypapoy and éypaya, 10.] Formation of Tense-Stems. Leaving out of sight the meanings of the several Tenses, and looking to the mode of their formation, we may distinguish the following groups :— (1) With the Verb-Stem serving as Tense-Stem— The Simple Non-Thematic Present, as py-jul. The Simple Non-Thematic Aorist, as é-8n-v. (2) With Reduplication— The Perfect. The Non-Thematic Reduplicated Present, as r/-Oy-. (3) With the Thematic Vowel— The Simple Thematic Present, as A€¢y-«. The Simple Thematic Aorist, as &AaB-o-v. (4) With Reduplication (Thematic)— The Thematic Reduplicated Aorist, as ijy-ay-o-v. The Thematic Reduplicated Present, as yi-yv-o-pas. (5) With other Suffixes (Non-Thematic)— The Present in -vy-pe and -vu-pi, as oxld-vy-pt, delk-vu-pt. It may be that the Suffixes -vy (-vu&) and -vu are properly speaking ‘ Thematic’ (i.e. are suffixes used to form Noun-Themes from ‘ Roots’), but as they vary like Non-Thematic Stems this group of Verbs belongs practically to the Non-Thematic Conjugation. The Aorist in -oa. The Aorist in -y-v (Aor. II. Pass.). The Aorist in -6y-v (Aor. I. Pass.). (6) With other Suffixes (Thematic)— The Present in -rw (T-Class of Curtius). The Present in -vw (Nasal Class). The Present in -oxw, and the Iterative forms. The Present in -w or -yw (I-Class). The Future in -ow. It will be more convenient however to bring the ‘ Verbs in -pu’ together; we shall therefore treat first of the Non-Thematic Present (py-mi, tl-On-pu, oxid-vy-y1), and Simple Aorist (é-8n-v) ; then of the Perfect ; and then of the remaining groups of Tenses in their order, ; 10 HOMERIC GRAMMAR. [rts The Non-Thematic Present and Aorist. 11.] The Simple Non-Thematic Present. The chief Presents in which the Tense-Stem is the same as the Verb-Stem are— ei-pi (for éo-pi) Iam, ei-yu Igo, pn-pi LT say, H he said, Kel-TaL lies, ho-rae sits (3 Plur, ei-arat, properly j-ara, for *io-arau), énl-cra-par I know, dya-pa I wonder, épa-pyat I love, diva-pat IT am able, é-xpépw (for é-kpéua-o) didst hang, déa-r0 seemed, dle-vrat race (ér-dl<-cav tried to scare), dvo-cat dost blame (Sva-ro Il. 17. 25), &n-rov blow, tky-O. be appeased, Kuxi-Thv caught, €-pwevar to eat, épv-ro protected, aret-rau is ready, threatens. Some apparently Non-Thematic forms are due to Syncope, 4s of-naz, for oto-par, T think, dot-cOar, for dove-cOat (or Ade-cbar) to wash, gép-re bring. So Curtius explains o7ed-rat for oreve-rar; so perhaps épv-7o for épte-ro (§ 18, fix.), and dpeva to be satiated (@). On the Non-Thematic forms of Contracted Verbs (such as popy-pevos), see § 19. 12.] Variation of quantity according to the ‘weight’ of the Ending takes place with great regularity in gn-pi J say, pi-s, gn-ot; Impf. é-on-v, &pn-cOa, é-pn; but pa- in all other forms. So ci-ys I go, ei-oOa, ei-o1; Plur. tyev, t-te, t-aor; Impf. 2 Dual i-rnv, 3 Plur. 7-i-cay, t-cav; Imper. i-6, i-ro, i-te.* The vowel of the Stem is long before Heavy Endings in— an-, 3 Dual dy-rov, Inf. dzj-pevar, Mid. &q-r0, Part. daj-pevos. xiyn-, in Keyy-rqv, e-KlXn- MEV, KLXAR-MEVAL, KLXT|-HEVOS. Except before -vr and -1, as dé-vres blowing, xixe-in may find. (On the long vowel in these Stems, see § 19.) Also in keiuar, épt-ro (Inf. pi-cOax), tAn-. ; It is short throughout in eui (for *éo-pt) I am, which is inflected as follows :— Sing. Dual. Plur. 1. eipt eiuev (for éo-pév) Pres. 2, ég-oi, els (§ 5) éo-rév éo-Té 3. éo-ri(v) éo-rdv eigi (Dor. évré), é-dor. I, fa, éa, Av, env, — Huey Them, éo-v Aer 2. hoba, énoba [fo-rov] re (IL 16. 557) 3. Fev, hy, nv, tnv Ho-Tny joay, éoav. The only shorter form of the Stem which analogy leads us to expect is that * The Sing. forms qia, Het-s, Her are difficult of analysis, Comparing the Greek and Sanscrit forms— Sing. dyam ais ait, Dual aitam aitdm, Plur. aima, &e. qa Hes = Her, i-tyy. Delbriick points out that the ai of the Du. Plur. is 4+i (augm.+ short Stem), while the ai of the Sing. may be &+e, i.e. augm.+long Stem (Altind. Verb. p. 100). If so, j-e-s, q-ev answer exactly to ai-s, ai-t. In other words, the augmn. being 4-, the 3 Sing. 4-et is formed precisely as é-py. On this view, again, the 1 Sing. 7-ta may be for an original q-ea, formed like é-xeva (§ 15). Thus the Greek and Sanscr. forms go back independently to a primitive— , Sing. d-ai-am, d-ai-s, d-ai-t, Dual. d-i-tam, &e., Plur. d-i-ma, &c. 14.| NON-THEMATIC TENSES. II which results from dropping the initial e, as in Lat. ’sumus, sunt. The t Sing. ja (Sanser. dsam) is the most primitive.* The 1 Sing. éyv and the 3 Sing. éqv and jnv are anomalous, and (like similar forms in Sanscrit and Latin) seem to require a Stem éoa-. The -v of the 3 Sing. may be due to the influence of fev, which is never found without the -v. 13.] The Simple Non-Thematic Aorist. Examples: @&n-v I went, &orn-v 1 stood, &yve-v I knew, eu grew, é-dv sank under, €-~0n came before, é-0Bn was quenched, Variation of quantity occurs (e.g.) in &By-v, 3 Dual Ba-ryv (also ¢-8y-rnv), 3 Plur. trép-Ba-cav (Il. 12. 469). Generally, however, the Stem is long in all forms of the Active (except the 3 Plur. in -v); e.g. &orn-v, Plur. é-orn-pev, €orn-re 3 €-yvo-pv, 3 xi e-yvw-cav ; &-00-pev ; Kata-mTi-Tyv (Du.) cowered ; «d5-Ot ear |! On the Aorists é7-é-ndw-s didst sail over (Part. ém-dds), Bid-tw let him live (Inf. Bi6-vat), 6AG-vat to be taken (Part. GAots), é-yijpa grew old (Part. ynpds), see § 20. The vowel is also long in the Mid. mAj-ro was jilled, &n-dvy-to profited (Part. dvij-yevos blessed, Opt. 6va-to), du-mvi-ro recovered breath ; but the regular quantity appears in &nrd-ro flew, mpla-ro bought, pOd-yevos coming before, &-pbt-ro perished, Kri-pevos built, érov-ro (av-70) hastened, xv-ro was poured, dtb-70 was loosed (once db-ro, Tl. 24. 1). The vowel is short throughout in é&«ra-v I slew, 3 Sing. é-«rd, 1 Plur. é&«rd-wev, Mid. é-«ra-ro, Part. «rd-yevos; also in otta he wounded, Inf. obrd-pevar, Part. odrd-pevos. On danvpa see § 31, 1.1. Tt will be seen hereafter that «ré- is the regular short form answering to krev-, Hence é-«7a-pev, é-xrd-ro are formed according to the primitive rule. Aorist-Stems ending in a consonant are only found in the Middle: Gd-ro leaped, é-nad-ro was shaken, dp-to started up (Imper. dp-o0, Part. dp-yevos), yev-ro seized, déx-ro received, d€x-To counted, also lay down, jlkx-ro was mixed, xat-€-myx-T0 was fastened into, éhédtx-to + wheeled, and the Participles dp-pevos fitted, do- pevos (for dd-pevos) welcome, tx-uevos ‘coming, favourable, This group of forms is only found in Homer. 14.] Metathesis. Some Aorist-Stems are formed from the Verb-Stem by metathesis ; fuy-BdAr-rnv the two encountered, 3d4-T0 was struck (from Bade-), é-rAn endured (rdda-s), wAR-TO came near (méAa-s), é-orpw-ro was strewed (atope-). The vowel in these Stems is always long. * For the 1 Sing. #v cannot well be derived from the hypothetical qo-y, orig. as-m, which according to all analogy gives Gr. *jo-a, Fa, and Sanscr. as-am ; cp. #ia, Sanscr. 4y-am. . ; ds : : + This form is probably an old misreading for ééAcwro, i.e. Fe-FéAukTo; see Cobet, Mise. Crit. p. 278. 12 HOMERIC GRAMMAR. [15- 15.] Aorists in -xé and -é, The three Aorists, &-OyKa I put, é-yxa I sent forth, é-dwxa I gave, are inflected as follows :— In the Active— 1 Sing. €-Onka i Plur. é-6e-yev 2 4, €Onka-s 2 Dual é-6e-rov 2 4 €-Oe-re 3 55 €-Onxe(v) 3 4 e-O€-THv 3 4» €-Oe-cav and €-Onka-v. In the Middle, é-0é-unv, &c. (with 6e- throughout). Thus @nxa- and 6e- serve as long and short Stems respectively. The only forms in Homer which do not conform to this scheme are év-njca-wev (Od. 12. 401) and O7jKa-To (Il. 10. 31, 14. 187). The four Aorists, érceva (o8-) I urged, é-xeva, éxea (x8-) I poured, éx-na (for *é-«nu-a, short Stem cdv-) I burned, jdeba-ro avoided (Opt. dAga-t-ro, Inf. ddéa-cOar), take -d; &yeva and éoceva standing to é-yv-ro, érav-ro, as é-py-v to €-pa-ro. These forms are not to be explained by loss of o (as if éxeva were for €-xev-ca, &e.). The & is ‘auxiliary’ (§ 4, note), The Mid. é-xeva-ro, found in Homer along with @-yv-70, is a duplicate form, created on the analogy of the Sing. Act., and accordingly related to éxuto nearly as @jKatTo to ero, It will appear that the regular Mid. of the Aor. in -oa was formed by a similar process ; see Append, A. On evra, ijveyxa, see § 36. 16.| The Non-Thematic Reduplicated Present. These Presents are formed by Reduplication, usually of the initial consonant with ¢; 1/0y-o1 puts, diw-m I give, in-or (for *yiyn-or) sends, tora (av-ord-) they set, mywmdaor they fill (the wis euphonic: it is dropped after win éu-mimtdny-61), didn bound, BiBa-s striding; and with Attic Reduplication, évivy-ou (for dv-ovn-) benefits. In these Present Stems the quantity of the vowel in the Stem regularly varies under the rules laid down in § 6 (1). The vowel is long in éy-mémdn-@ (IL. 21. 311), di50-0. (Od. 3. 380),* and the Inf. r16j-wevor (Il. 23. 83, 247) and Part. r161- pevos (I].10. 34). Also in d((y-yar L seek (for *8.-0d7-), the Homeric Verb answering to Attic ()-réw. 17]. Present Stems in -vy (-vé) and -vv. The Tense-Stems of this class—which may be called the Non-Thematic Nasal class —form the Present-Stem from the Verb-Stem by the Suffixes -vq, -vé (which with Heavy Endings regularly become -vd, -vv). The Presents with -vy (va) are nearly all peculiar to Homer, Sdy-vyn-we I subdue, xlp-vn mixed, wép-va-s selling, oKid-va-rar is scattered, mid-varrat comes near, pdp-va-tat fights, Note t for « in kip-, oxi8-, mA-; cp. the later Verbs wir-ve, xriv-vupt. A few Presents with -vv are common to all periods of Greek, * The variation is perhaps less regular in the Imper.; ep. tAy-O1, KAD-Ou. 18.] NON-THEMATIC TENSES. 13 detx-vu-us I show, su-ve-y. I swear, Ceby-vv-ye I join, dddrvpu (for OA-vu-wt) I destroy; but they are mainly Homeric or poetical ; op-vb-01 arouse, dal-vd feasted, &y-vu-rov break, ctop-vica spreading, dn-ondpy-vi wiped away, eépy-vv shut in, pny-viow they break, ya-vu-rar is gladdened, rd-vv-ra1 is stretched, i}-vu-ro was finished, ki-vu-vto were moved, ti-vu-vrar punish, al-vv-rur takes, é-Kal-vv-to surpassed, ap-vi-cOnv won, ax-vu-uar I am vexed, dly-vv-vto were opened, Evvuro (for éo-vu-to) put on, Cévvv-ro (for (wo-ve-) girded himself, dpey-vi-s stretching out. In the Verbs in -vnpt the Verb-Stem is nearly always dissyllabic: ep. Sapd-car (rav-Sapd-rwp, &c.), kepd-car, werd-co, mepd-car, oxeda-cat, wéda-s. So in some Verbs in -vipr; cp. du6-car, dAé-car, oropée-car, The Verb-Stem, it will be seen, has most commonly its short form (note espe- cially ra-vu-rar, Pf. ré-r&-7a1), sometimes the long form, as in Present Stems, Selk-vu-pe, Sevy-vupe, Pyy-vupe. The forms in -avvupt and -evvuss are post-Homeric. 18.] Thematic forms. Some forms of Non-Thematic Tenses follow the conjugation of the corresponding Contracted Verbs in -aw, -ew, -ow ($ 56); especially in the Impf. Indic. and the Imperative. Thus we find :-— edduva (as if from *dayrdw), éxipva (Od. 7. 182, &c), mirva: Imper. «a6-éora (Il. 9. 202). erider, ter (ap-ter, mpo-ter &e.), det (v. 1. dn), blew, wlxeus: Imper. rier, ter (Evv-ier). éd(Sous, edidov: Imper. dédov. Examples occur also in the Pres. Indic. ; daurg (3 Sing. Act.) in Od. 11. 221 (with v. 1. dduvar’); dang (2 Sing. Mid.) in Il. 14. 1993; dv-veis (Il. 5. 880), ped-rets (Il. 6. 523, Od. 4. 372), peO-vet (Il. 10. 121), ret (Il. 13. 732, Od. 1. 192), for which the MSS. usually have davies, &c.: did0%s (Il. 9. 164), dud07 (IL. g. 519, Od. 4.237). So for mpote in Il. 2. 752 we should read zpoxei. Add the Part. B:Bévra (Il. 3. 22, ep. 13. 807., 16. 609), Fem. BiBéoa (Od. 11. 539); for which Bekk. writes PiBdvra, BiBaca. Editors differ in their manner of dealing with these forms. The weight of authority seems to be for the spelling which follows the Thematic conjugation, viz. -e7s, -ofs in the 2. Sing. Pres., and -es, -e, -ovs, -ov inthe Impf. of riOnme, inue, BiBwpr (Cobet, Mise. Crit. p. 281, is extremely positive on this side). But Verbs which have q in the Dual and Plural (éy-rov, xxf-79v) should follow the analogy of the Passive Aorists: hence dn, xixys. And we may leave undisturbed the form 8/57 he bound (Il. 11. 105), for which no one has proposed to read dide. The 1 Sing. mpo-lev (Od. 9. 88., 10. To0., 12. g) stands alone. Porson (in his note on Eur. Or. 141) condemns fumets, T:Oe?s, &c. on the ground that if 7:Oe7s were right we ought also to have 710%, 7iHe?, TiWodpev, Terre, Tt is possible, however, that a form like 7:0e?s may have crept in through the analogy of the Verbs in -ew, although no ‘ Verb’ 7:6éw was in use. It is characteristic of the working of analogy to be partial and gradual. In Homer we find the cor- 14 HOMERIC GRAMMAR [19. responding 3 Sing. Pres. dapv4, 7:0e?, webct, 5:50i—forms which are guaranteed by the metre. The forms so guaranteed are, indeed, few, and perhaps were not found in the oldest text of the poems; but they are supported by similar forms in Herodotus and other Ionic writers.* Similarly, in the Presents formed with -vv there is evidence of a tendency to introduce the Thematic -vve or -vvo. The instances are :— ép-vu-ov (Il. 12. 142), Gyrve (Il. 14. 278), Cedyvvov (Il. 19. 393); Imper. durv-éro (Il. 19.175), ravd-ovor, ravd-ew (Il. 17. 391), dvdw (Il. 4. 56, but may be Fut.), ravi-ovro (occurs four times). Sawvy occurs twice in the Od. as 2 Sing. Subj. In Od. 8. 243, where the metre requires - <“-, we should probably read Saiwv-eat, the Subj. answering to a Non-Thematic Indic. (§ 80). In Od. 19. 328 Sawve’ is equally admissible with dawn. Also, the Verb épvouat (or ptopat) protect, save, is for the most part Non-Thematic, see § 11. The Verb épvw (épvo-) draw is wholly Thematic in Homer. It should be observed that in all the foregoing cases the Thematic form is obtained by combining ¢(0) with the final vowel of the Non-Thematic Stem In other cases the original fina vowel is lost, as «éxe(v) for é-xixn aré-o6a Inf. of é-7a-70, di¢w for Si¢y-yat, and the like. 19.] Non-Thematie Contracted Verbs. The following Homeric forms are usually regarded as instances of ‘irregular Contraction’ of Verbs in -aw, -€w, -ow :— (-aw): ovvavri-tyv met, ovdr-rnv spoiled, mpooavd)-ryv spoke to, goirj-rnv went about, xvi scraped, dpi-yevat to pray, yor-jevar to bewairl, wewvij-pevat to hunger, Oj-cOar to milk. (-ew) : dmewdyj-ryv threatened, dpapri-tnv met, Kadz}-pevar to call, mevorj-wevar to mourn, moOn-pevar to regret, pidrry-pevat to dove, hopr- pevat, opi-vat to carry, ddwTi}-pevos sinning, Tepon-yevar to get dry (§ 42). (-ow): odo 3 Sing. Impf. and also 2 Sing. Imper. of cadw I keep safe. These forms cannot be explained by the ordinary contraction with the Thematic € or 0: e. g. pour#j-rnv cannot come from *dor- taérnv, popij-vat from *opeé-var, dAtTi}-evos from *dured-pevos, * In considering this and similar questions it should be remembered (1) that we do not know when the Homeric poems were first written down; (2) that we do not know of any systematic attention having been paid to spelling, accentuation, &c. before the time of the Alexandrian grammarians; (3) that the tendency of oral recitation must have been to substitute later for earlier forms, unless the metre stood in the way; (4) that the older Ionic alphabet confused e€, et, q and 0, ov, w, 19.] NON-THEMATIC CONTRACTED VERBS. 15 odw from odoe, &e, On the other hand, as Curtius has shown (Stud. iii. 377-401, Verb. i. 352 ff.) they agree exactly with those Non-Thematic forms in which the vowel before the Ending as long except before -vr and -1, such as the Pres. KLX}-pevat, dij-pevat (§ 12), the Aor. or/-pevar, TA-vat, yvd-pevar, &e. and (as we may add by anticipation) the Passive Aorists in -nv and -Onv. Moreover, the same type of inflexion appears in the peculiar ‘Verbs in -p.’ of the Molic dialect, as piry-ps, 1 Plur. irn-pev, 3 Plur. pirevor (for pirc-vru), Part. pidrj-pevos 5 so yeAai-pl, Td@=[L0: and also in the Latin Verbs in -are and -ére, except in the 1 Sing.; e.g. ama-mini is parallel to wewyj-pevar, docemini to poprj-pevat, docemus, doce-nt to pidn-pev, pideuor. Further traces of Non-Thematic formation may be recognised in the ‘irregular contraction’ of Attic—reuf, duff, Ch, Wi, oph, xp7-7Oat, prydv (Inf.), tépGvrv*—and in the Opt. in -ey-v, ou-v (for which however, in the case of Verbs in -ew, We expect -ety-v, as in kixe-(y and Alolic dude-y). The agreement of the Homeric forms given above with the fEolic forms of the same Verbs, and of both, again, with the similar Latin Verbs, seems to show that this formation is older than the corresponding Thematic Verbs in -aw, -ew, -ow. The relation is doubtless, as Curtius maintains, the same as that which we traced between rié@y-s and riOeis, de¢xvu-yur and derxvi-w ($ 18); that is to say, the regular Contracted Verbs represent the extension of the Thematic type beyond its original limits. We shall return to this when we come to treat of the class of Thematic Verbs to which these belong (§ 54). The origin of the unvarying long vowel is not necessarily the same in all cases. Thus in éory-cav, €yvw-cav it is probably due to imitation of the Sing. ;+ in é-$Ayn-r0, é-crpw-ro it is connected with the metathesis (§ 14). In the Homeric forms now in ques- tion (dpy-pevar, ptdy-pevat, cdw, &e.) it is explained by the fact that they are Denominative, i.e. derived from Nouns (apy, pido-s, odo-s) by means of a Suffix which has amalgamated with the final vowel of the Noun-Theme. The vowel in these cases is therefore the result of contraction, and the Verbs so formed may be called the Non-Thematic Contracted Verbs. The theory now stated can hardly be proved without resorting to the parallel Sanscrit conjugation, that which forms the Present Stem in -aya. From this comparison it is evident that the original Suffix was -ya, which in prehistoric * See Veitch. In the Inf. of these Verbs the same shortening has taken place as in the Aolic Inf. edAnv (nddnv-at). In the 3 Sing. the final -ov is similarly lost. The t subscr. of wewh &. may be due to ancient epenthesis—as Curtius explains Aéyn from Aéyn-o1—or merely to analogy. : But see Cait. Verb. i. 195, aud on the other side, De Saussure, Mém p. 146. 16 HOMERIC GRAMMAR. [20. Greek might appear in the forms -yy, -ye, perhaps varying (like -v7, -v@) according to the Endings. To explain the long vowels y, we have only to suppose (with Curtius) that the y of the Suffix, before it ceased to be pronounced in Greek, had the effect of lengthening the preceding vowel, as in the Genitive médn-os (for wéAe-y-os), Thus dA€-yn(e)-, oao-yy(€)- would become $Ay-n(€)-, caw-7(e)- throughout the Pres. and Impf., and by contraction $uy-, caw-. 20.] Aorists. Of the Aorist Stems noticed in § 13, several are probably- derived from Nouns, and do not differ in formation from the Presents discussed in the preceding section: e. g. é-yijpa (yijpa-s), Bid-rw (Bio-s), ém-émAw-s (wAdo-s), GAG-vat, perhaps an-dvn-to. Regarding the Passive Aorists, see §§ 42-44. The forms ri7-yevat (Il. 23. 83, 247), T9j-yevov (Il. 10. 34) are probably due to the analogy of the Non-Thematic Contracted Verbs. The 2 Sing. Imper. tory (Il. 21. 313) is perhaps for *lory-O (like Molic pln for pidy-cr). 21.] Meaning of the Non-Thematic Pres. and Aor. The Presents formed by Reduplication, and by the Suffixes -vy and -vu, are nearly always Transitive or ‘Causative’ in meaning, as torn-pt, oxid-vy-ut, dp-vu-yr; whereas the simpler Verbs, whether Present or Aorist, are usually Intransitive, as éory-v, 087. Regarding the Tense-meaning, it is enough to point out here that the difference of the Present and Aorist is not given by the form of the Tense: thus the Impf. @@y-v is the same in forma- tion as the Aor. é-By-», é-ory-v. The Perfect. 22.| The Perfect-Stem is formed by Reduplication, and is liable to vary with the Person-Endings (§ 6). This variation is the rule in the Homeric Perfect. In Attic it survives in a few forms only ; it is regular in ofda and éoryxa. The short form of the Stem is the same (except for the Re- duplication) as in the Tenses already discussed. The long Stem is often different, with a predilection for o (w, ot, ov), where the Present and Aorist have e (n, eu, ev). The variation appears in the interchange of — (1) 9 (@) and &@: as reOyA-er loomed, Part. Fem. reOcid-via; dipype is fitting, dpip-via, So deAnx-bs, AcAdk-via yelling, weunk-ds, veudx-via Oleating ; Mid. dédracrat (for Aedad-rar) has forgotten, axax-peévos sharpened, népav-rat has appeared ; and the long forms conte, TéerTyKa, TEONTA, &C. 4 and ¢ (doubtful): in ¢3-n8-6s having eaten, wéunde is a care (but pewaddras, Pind.). wand o: in dé50-rar (dw-), éx-néno-rar is drunk up, épwpe is aroused, dhwda, dénwma, 6565-e1. Of the interchange of w and a, as in éppwya (Pres. Piy-vupe, 22, ] THE PERFECT. 17 Aor. bm-eppay-n), and of » and e, as in the Noun €5wd7, there is no clear Homeric instance.* _ juin dédne 18 on fire (for dddnve, dédyfe) answers to cu (aF-) In datw (baF-yw): ep. Kai, Aor. Kya ($ 15). (2) ov and t: as ofda, x Plur. t-yev; mérouda 1 Plur, Plpf. é-né710-yev ; oka, Dual. 2ix-rov, Part. Fem. éix-via. (3) ev and é: as mepevy-ds having escaped, Mid. Tepuy-LEvos ; Terevx-aTar are made, 3 Sing. réruc-rar; Kéxevde hides (Aor. Ge). Other short Stems: Kéx¥-rat, Zoo¥-rar (§ 15), méruo-pat, KEKAV-OL. ov interchanging with % is much less common. It is found in eiAfAovea I ain come, perhaps in BéBovAa (but not if it is connected with Lat. vol-o, vél-im), and the Part. Se50um-dr0s (cp. «rvm-os), & in pépowe (Aor. pire). (4) op (po) and dip (pa) ) as upope has a share, Mid. efuap-ro was od an fated (ep. pép-os). So d:-Ep9opa-s art destroyed, SéSpope runs, rérpode is thickened, SéSopKa, Eodma, €opya and the short forms menap-pévos pierced, térparto (cp. Tpéx-w, Tpém-0s), €-rérah-ro, But ep, eA of the Pres. appears in éep-pévos strung, épy-ara are packed in, and éed-peévos cooped in; ep. § 31, 6. (5) ov and a: as yéyove is born, 1 Plur. yeyd-nev (ep. yév-0s) ; nenovda I suffer, 2 Plur. rénacde (for mena0-re), Part. me7a0-via, So pépova-s art eager, 2 Plur. péud-re; eAdyx-or have as portion (ep. é-Aaxo-v); wépi-rar is slain (cp. pdvo-s); téri-tar is stretched (révo-s). But we find av in xexav5-hs containing (Aor. é-xa5e). wérracGe, not mérooGe, should be read (with Aristarchus) in Il. 3. 99, Od. 10. 465, 23. 53. (6) o and e: as in réroxa (Hesiod). The examples in Homer are doubtful: efw6a (if it is for *é-cFoda) ; dvqjvode mounted up, én-evnvobe is upon; possibly xexon-ds striking. The form dv-hvole supposes a Pres. *avé0-w, perhaps related to dv@-os as dAéyw to day-os: ep. the Attic use of énavGéw. So évfqvobe supposes *éveO-w, and a short form *év0- (Dor. év6-«@v ?). The short Stem with € occurs in débey-uar I await (ep. mpo- dox-ai ambush), &eo-ro was clothed. In these cases loss of € is phonetically impossible: see § 37. (7) -xa& (preceded by a long vowel), interchanging with— (a) A short vowel : as éornxe stands, 1 Plur. éord-pev ; répuxe, 3 Plur. wegd-acr; deidouxa I fear, 1 Plur. deidu-yev 5 so BeBnxa, TéOvnKa, TETANKA, pépune is closed (of a wound) and SéSuce is sunk in are similar, but the short Stems are not actually found. * Unless elw@a is for é-ofwa, short Stem ofe8-; see De Saussure, Mém. p. 168. Cc 18 HOMERIC GRAMMAR. [23. (2) A long vowel: as BeBrrjx-er struck, Mid. BéBAy-rar (BAn- by metathesis, § 14). So kécpyka-s art weary (xdpa-ros), BeBpwk-as having eaten (Bop-), péepBrwoxe is gone (Aor, €-podov), Mid. wenAn-pévos brought near (méda-s), wemvi-rar has his senses, KécAn-pat, E_py-TaL, Pepvy-paL, TETHY-LEVOS, TreTPPw-pLEvOS. A Perfect in -9& occurs in éypyyép-Cacr (Il. 10. 419) keep awake; perhaps in the Opt. BeBpwd-ors (Il. 4. 35)- (8) The Perfects of Verbs in -aw, -ew, -ow, -vw (§ 51, 4) resemble those of the last class, the vowel being invariable. They are chiefly found in the Middle and in the Part. Act.; e.g. kekoTH-OTE, kekopy-dra, teTn-dtes, BeBapy-dra, Kexapy-dTa,, kexadyn-dta—all forms peculiar to Homer. The only examples of the Indic. Act. are— treBapoijxact (Il. 9. 420, 687), BeBin-kev (Il. 10. 145, 172., 16. 22), map-Sxnxev (Il. 10. 2.52); vm-eurnpoxe (Il. 22. 491), Sedanxe (Od. 8. 134, 146), rerdxyxe (Od. 10. 88), dedermv7jx-er (Od. 17. 359). iAhenot (Od. 21. 365) may be a Pf, Subj., see § 8£5 cp. iAq-61, § II. These examples represent the earliest steps taken by the language towards forming the class of regular Perfects in -«é. They are evidently suggested by the Perfects BéByka, oryka &c. which are a well-established group in Homer.* Other Perfects of Verbs of the I-Class (§ 50) are invariable, as Kkexopv0-uévos, mendAuc-T0, Kexovi-péevos. They are only found in Homer in Middle forms. 23.] The Reduplication takes the following forms :— (1) An initial consonant is repeated with «. This is the general rule: we need. only notice the Perfects in which an original consonant has been lost, viz. :— A labial semi-vowel (which we may write F) in é-ed-evos cooped in (for Fe-Fed-p€évos), eiAv-to (Fedv-), é-opya (Fépy-ov), é-odma, €ouxa, Mid. ijix-ro (unless this comes from éioxw). A sibilant {c) in &ornxa (for *oé-ornka), é-ep-uevos strung together (Lat. sero). * A word may be said here on the origin of the Perfects in -xi. They may be regarded as formed in the ordinary way from Stems in which a Root has been lengthened by a suffixed x, as in 6Aé-K-o, épt-K-w (§ 45), TWIAGow (for Try-K-yo, cp. é-mra-x-ov), SeSicocopar (for de-5Fix-yo-par). Thus dA@Aexa is the regular Pf. of éAécw, and mwémtwxa, SeiSouca, answer to the short Stems mré-«-, 5ft-x-. So BeByka, Exryka answer to (possible) Presents *87-xw (cp. Bax-rpov), *oTH-Kw. It is not necessary to suppose an actual Stem in k in each case; a few instances would serve to create the type. The reason for the use of the longer Stems Bn-«-, orn-x, &, was probably that the forms given by the original Stems were too unlike other Perfects. The 1 and 3 Sing. would be *8é8n, *éory, ke. The Aorists in -«& are to be accounted for in the same way. The « may he traced in @qk-y, which points to a Verb-Stem @y-K-. It is worth notice that in Homer the Pf. in -«& is almost as exceptional as the Aor. in -«&. (See Curt. Gr. p. 61, ed. 1879; Verb. ii. 206 ff.) 24.] THE PERFECT. 19 (2) Stems beginning with two consonants (except when the second is a Liquid), or with ¢, usually prefix « only : 8:-é-8opa-s, €-pOi-aro (but me-mreds, mé-nravrat) ; e-Cevy-pévar, é-Krf-obar. The group of has been lost in &-a8-ds (either oe-cF G8-hs or é-of G5-ds) pleasing, and etw8a, The group 8f has the effect of lengthening the vowel of the Reduplication in Beibta (for d¢-3F ra), Sel-Souca, Initial 6 (which generally stands for original Fp) gives épp-, as in éppyx-rar (Feny-), éppifw-rar. Sometimes cip-,as eipy-rat (Fpy-, cp. Lat. ver-bum). One Stem reduplicates p, viz. fepumw-péva, from suTdw, Note also éupope, Mid etpap-rar (probably for €-cpope or c€-cpope, cé-cpap-Tat) ; and éoov-rat (instead of cé-ct-rar, Pres. cev-w), (3) Stems with initial vowel followed by a single consonant usually repeat these two letters (the vowel taking the quantity of the short Stem); as dm-wma J have seen, éd-jda-to was driven, This is called Attic Reduplication. A vowel and ¢wo consonants are repeated in éyp-yyopa I am awake (but see Curt. Verb. ii. 141). (4) Where Attic Reduplication is not admitted, the vowel quantity is increased if possible; e. gi ép-jm-rat (dm-Tw), Kat-7KLo- Tat (aikiw), foKn-rat (dokéw), Hoxuypevos. (5) In a very few cases the Reduplication is lost, viz. :— ota (for Foiéa). Séx-Grar (3 Plur.) await, Plpf. é-5éy-unv, Part. 6éy-evos (with irregular accent). épx-Grat are shut in (Fepy-), Plpf. épyaro and. éépxaro, €o-oar (Feo-) thou hast put on, Plpf. éo-00, go-to and éeo-r0, In the 1 Sing. efuac and Part. efzévos the e. may be due to the loss of the o, (as in efua for 4Féo-pa). For the 3 Sing. which occurs once (Od. 11. 191) the best MSS. give fora (the reading of Zenodotus), others efora: and eira:: an oracle in Hat. 1. 47 gives émi-eora:. émr-x-Gro (Il. 12. 340) had been shut (of gates) may be for é1-edy-aro, (related to éy-w as éw6a to &6-w). The similar form dv-wya I command seems as yet unexplained, In tépev-ro (Il. 24. 125) the short f may be due to the vowel following; so dpo-iax-via (Il 2. 316); moreover the Stem ity- (for Fifax-) is already re- duplicated (Curt. Verb. ii. 148). On dpy-pévos, aBn-Kores see § 26, 5. (6) The Reduplication in de-5¢x-dras they welcome, seems to be that of the ‘Intensive’ forms, as in det-déoxopar: see § 61. The form belongs to defk-vyp, not déx-oar (see Veitch). 24.| In the 8 Plur.— 1. The long Stem with -do. (for -a-NTI) is comparatively rare :— nerolaar (Il. 4. 325), €orixaoe (Il. 4. 434), xara-reOvjxaor (I], 15. 664), Tedaporjxaor (Ll. 9. 420, 683), éypnydp9aon (Il. 10. 419). These forms evidently result from generalising the Stem in -a, So we have otSa-s (Od. 1. 337), ol8a-pev, ofSacr in Herodotus and even in Attic. C2 20 HOMERIC GRAMMAR. [25- 2. The final consonant of the Stem is sometimes aspirated before the -tra: (for -vra) of the Mid.; as émi-rerpdp-arar are entrusted, tetpap-aro were turned, &py-ara (Fepy-) are shut in, dpwpéx-arat (dpéy-w) are stretched out, devdéx-arat (Selx-vupe) welcome. The later aspirated forms of the Act., such as ciAnga, xéxopa, are entirely unknown to Homer. 3. An anomalous e for U appears in det-5€x-arar (delx-vupe, see § 23, 6), ép-npéd-ara (épel-w, ep. jpio-uévos Hesych.), and dk- nxéd-atar (aKa (Co). In Od. 7. 86 ydAneor pev ydp Torxo: éhnAdbar’ (La R.) the reading is uncertain, some good MSS. having épypéSar’. 25.| Confusion of Long and Short Stems. 1. The long Stem is found with a Heavy Ending in cidjAov6-yev we are come (we might read «iAjAv0-uev), awpto (for aopro?) was hung, and the anomalous éyp-iyop-0e (Imper.) keep awake, Inf. éypiyop-Oar. Again, the Stem of the Present appears in ovy-eppyx-tat, MéAeuT-Tat, eCevy-yévat, jpijperc-ro (epeid-w). Note that the irregularity only occurs where the syllable is already long by ‘position.’ It is evidently due, in the latter group, to the influence of the Present. 2. A short Stem occurs with Light Endings in deéé.a (instead of *defdora for d€-dFova: ep. deidi-uev for de-dFi-uev). Also in ava-BéBpoxev (Il. 17. 54), for which Zenodotus had dva-BéBpoxev, the correct Pf. answering to the Pres. Bpéy-w. 3. In Dissyllabic Stems we have to note that— The syllable following Attic Reduplication is often long throughout, as 68-d8ve-rar, €A-jha-rat, ap-npo-pévos, d«-nxe-pévos. It is short in épépim-ro, akaxn-par, dAGAT- par. Homer has ¢iA-qAov8a (Part. éA-ndovd-ds, Il. 15. 81); also (less commonly) €AnATOa: and so kat-epyptre is fallen (Mid. épépim-ro). For é5yS0ra. in Od. 22. 56, we should probably read (with Aristarchus) €55-Grat, the regular 3 Plur. 26.| The Perfect Participle was formed originally from the short Stem, but there are exceptions in Homer, due partly to the fF of the Suffix (-Fés, -via, -Fés), partly to the general tendency to adopt the form of the Sing. Indic. as the Stem. Thus the Homeric Pf. Part. is intermediate between the primi- tive formation with the short Stem (as in Sanscrit), and the nearly uniform long Stem of Attic. In particular— 1. When the Mase. Ending -ds (-éros) follows a vowel, one or both of the concurrent vowels may be made long: peua-dre, peva-Ore (both for peud-Fére). So yeyd-Gras; eu-BeBa-Gras ; 27.| THE PERFECT. 21 Tepu-Gras 5 Kexuy-Gre; TeOvy-dros, TeOvn-GTa; menTe-Gra (for mentn-dra). Both are short in éord-dres. ° In rerpiy-Gras (Il. 2. 314) and kexAny-dres (with v. 1. eexdjyovres Il. 16. 430) the w of the Ending is unexplained. 2. When -ds follows a consonant, the Stem takes the vowel or diphthong of the long form, as dpyp-os, peunk-bs, NeAnK- ds, €0k-@s, mET0LO-ws, Copy-ws. We once find a Fem. eloux-vie.* But oida, ¥ > black, nén-ov (Voc.) tender one, apny-dv-es defenders, ayx-dév, Gen. -@v-os elbow. -VT, -ovt, in Participles, and in a few Substantives, as dpd«-wv a serpent, lit. the ‘staring’ animal (dépx-oyat), ax-wv, rév-ov. -ar, in oblique Cases of Neuter Nouns as (wp), tdarT-os, &e. 114] PRIMARY SUFFIXES. 73 The @ of this Suffix represents the short form of a nasal syllable; see § 38, and the note at the end of this section. -VO, -dvo, -vn, -dvn; as dei-vd-5 fearful, réx-vn art, dx-avo-v handle, dpendvn sickle. Adjectives such as frye-davd-s horrible, jwe-Savd-s gentle, mevee-bavé-s bitter, ovri-avd-s worth nothing, are probably derived from Verbs in -f0; e.g. ob715-avds supposes a Verb ovd7i(w in the transition stage when it was pronounced oiti5-yw, § 53. So frye-5-avd-s implies fryé-5ym, related to fryé-w as -alw (-a-dyw) to -aw; see Curt. Verb. I. p. 326. -VEO ; réue-vos enclosure, iy-vos imprint, -DU; ta-vv- stretched out (Lat. tenuis), found in Compounds, as tavi-memAos : Opy-vu-s a foot-stool. -€P, -wp, -&p; as dip (aF-p) air, aid-rp (al0-e) bright shy ; éd-wp booty, td-wp water ; pdx-ap great (Il, 11. 68), gap spring. -P9, -Ao ; generally with the short Stem; mux-pd-s Jitter, ax-po-s point, €-pn seat, écOdds (éa-Ad-s) good: also with an auxiliary a, oOev-apd-s strong, amados tensler. -Pl; in id-pi-s knowing. -/40, -py; with the O- form, as of-yo-s and ol-pn (€i-1) path, mér-po-s (met-) fall, xop-po-s (keipw) a trunk, bd-po-s (Fed-) a roll- ing stone. -fAb; in pij-pi-s report, Siva-pi-s power. -MEV, -pov, -pov; mvO-yjy (Gen. -yév-os) base, diit-yjv breath, de(-uwv (Gen. -pov-os) fearing, On-pav-a (Acc.) a heap. -par; as dei-ya, Gen. -par-os, fear, dvoya name, &e, Of these suffixes -pov and -par go with the E- form of the Stem, -pev with the short form. _ With -o, -n are formed -pevo (in Participles), and -pvo, -pyn, as Bédc-pvo-v a dart, At-pvn a marsh ; -pya (-pv-y4), IN pépi-pva care. -pap, -wwp ; as Téx-wap and réx-ywp @ device ; -pepo, In f-pTP-0S 2. Heteroclite forms occur when different Suffixes are brought into a single declension. In particular— (1) Suffixes ending in -v interchange with Suffixes in p. Thus we find tiwv, Gen. wiov-os fat, but Fem. miepa (ni-Fep-yé), and the Neut. Substantive miap fatness. (Cp. the Lat. femur, femin-is, and jec-ur, jecin-or-is, which is evidently for an older Jecin-is.) (2) Final + is introduced in the Suffix; as in the Gen. #7a-7-0s (for jv-T-0s, cp. the Sanscr. yakrt, Gen. yakn-as, Lat. jecur, jecin-or-is), and other Neuters in -dp, -wp, Gen. -t-os: also in Neuters in -pa, Gen.~padr-os (for -pv-t-os; ep. the corresponding Latin Suffix -men, Gen. -min-is). The syllable before + is shortened, so that the Suffix has three degrees of quantity; e.g -pwv : -pev : -p& (or -pav). Similarly mpdppwv (ppev-), Fem. mpéppacoa (for mpoppa-t-ya). Cp. § 107, 2. 115.] Accentuation. The accent is often connected with the form of the Suffix, and sometimes varies with the meaning. But the rules that can be given on this subject are only partial. 1. Stems in -o are generally oxytone when they denote an agent, barytone-when they denote the thing done; e.g. popo-s bearer, but dpo-s that which is brought; ayd-s' leader, apwyd-s helper, téxo-s offspring. But vopd-s pasture, rovyd-s pestilence (perhaps thought of as an ageitt, ‘ destroyer’). 2. Stems in -y. are-generelly oxytone, but there are many exceptions (as dix-7, 4ax-7). 3. Most Stems in -8, and all in -a8, are oxytone. But those which admit an Acc. in -w are all barytone. 4. Adjectives in -v are oxytone; except O7A-v-s. “Substantives in -v are mostly oxytone; but see § 116, 4. 5. Neuters in -eo-are barytone, but. Adjectives in: -ec, and the Fem. Nouns in -ws, Gen. -oos, are oxytone. 6. Nominatives in -np and -yv are oxytone. Those in -wp and -wy (Gen. -ovos) are barytone, those in -ov, (Gen. -wvos) are oxytone. Exceptions ;. pajrnp, Ovyarnp ; tépnv,-dpnysv, dndév; Nouns in -8dv (Gen. -3dv-0s). 7. Stems in -ro with the O--form are barytone, with the short form oxytone; ¢. g. Koi-Tu-s; vdo-To-s, ‘but ora-7d-s, &e, 76 HOMERIC GRAMMAR. [116. 8. Abstract Nouns in -n, -ov are barytone; in -rd oxytone. It will be seen that, roughly speaking, when the Verbal Stem is in the short form, the Suffix is accented, and vice versd: also that words with an active meaning (applicable to a personal agen) are oxytone, those witha passive meaning (expressing the thing done) are barytone. 116.] Gender. The Gender of Nouns is determined in most cases by the Suffix. The following rules do not apply to Com- pounds, as to which see $ 125. 1. Stems in -o are Mase. or Neut., with some exceptions, as 686s, drapméds, xédevOos, vijvos, hyyds, dmedos, vooos, Tappos, Wipos, orodds, Wapabos, paBdvs, doxds, fuvds, mpd-xoos. In these the change of Gender seems to be due to the meaning. kAvtés is used in agreement with a Fem. in Il. 2. 742; and mixpods in Od. 4. 406, . [1vAos has the two epithets jua0des and jyadey, and is probably therefore of both Genders. 2. Stems in -y (for -d) are mainly Fem. ; but— Stems in -ry denoting an agent-are Mase., as 5€ék-ry-s a beggar, aixuyn-ty-s a warrior. Also, mopxn-s the ring of;a spear, éty-s com- rade, tapyin-s dispenser, perhaps dyyedly-s a messenger ; also. the proper names Bopéa-s, “Epueia-s, Aiveia-s, Adyela-s, Tewpeota-s, *Ayxion-s, ’Aidy-s. Nouns in -r7-s are. probably formed from Feminine. abstract or collective Nouns in -ty. Thus there may have been Feminines 5éx-rn begging, aixun-7h the body of spearmen, &. The formation of a concrete Noun from such words may be illustrated from various usages. The first step is the use of the abstract or collec- tive word as a concrete; cp. Od. 22. 209 dpundixin 5€ pot éoor thou art one of the same age (=6pijré) with me; Il. 12. 213 Spor édvra being one of the common people; Latin magistratus, potestas (Juv. 10. 100), optio; English a relation (=a relative). So g@uy-ds doubtless meant primarily a body of fugitives, vou-ds a pas- turing tribe, &e. With the change of Gender (without a fresh Suffix) we may compare the French wn trompette, meaning a bearer of a trumpet, Italian il podesta the magistrate. So é7n-s is probably from a word ofé-rn kindred, dyyedin-s (if the word exists, see Buttmann Lexil. s. v.) from dyyeAin. The Mase. tayin-s may be formed from the concrete Fem. rayin, the office of household manager being generally filled by a woman (yuv7n Tapin, Od.). See Delbriick, Synt. Forsch. iv. p. 7-13. 3. Stems in -1d (-ya), -18, -c5 are Fem.; also most Stems in -. But pdv-r-s is Masc., and some Adjectives—id-pi-s, tpd@-i-s, edvi-s—are of all Genders. Mase. Nouns in -o sometimes form a Fem. in -t, -18, -a3: as Oodpo-s, Fem, Oodpi-s (Ace. Ootpi-v, Gen. Oovpid-os); édp-ro-s burden, pop-ti-s (Gen. pédprid-os) a ship of burden ; roxo-s, Fem. Toxdd-es; Aevkd-s, Fem. Aevkad-a (érpyv). Originally (as in Sanscrit) the chief Feminine Suffix was -t; and it is this Fem. i that yields the varieties -8, -i5 and -y&. The metre shows that the long + II 7. SECONDARY SUFFIXES., 77 should be restored in vis (Body fuw ebpupérwrov Il. 10. 292, Od. 3. 382), BAo- ovpams (Il. 11. 36), and Bo&ms (Il. 18. 357, where the Ven. A has Boam mérma “Hpn). The t appears also in apid-os, xvnpid-as, 2imAoxapid-es, 4. Adjectives in -v are generally Mase., and form the Fem. in -ed or -ea (for -eFya), as fda, axéa. But OfAv-s as a Fem. is commoner than @jAea; and we also find ods diruy (Od. 12. 369), movdty ep’ yp (IL. 10. 27). On the other hand most Substantives in -v-s are Fem. (and oxytone), and this uv is naturally long, as in i@v-s aim (whereas the Adj. i6v-s straight has 0), tAnOv-s multitude, iAv-s mud, Epwwt-s, and the abstract Nouns in -10-s, as Bpw-rd-s, dpyno-ti-s, KAt-TU-s. But there are a few Masc. Substantives (chiefly barytone) in -v, Opive-s, aTaxu-s, Botpu-s, véxv-s, ixv-s ; also the Neut. dorv, ddpv, your, TU, weOv. It appears that the long -v was originally Fem., like @ and t; cp. the pre- ceding note. 5. The Suffix -eo is almost confined in Homer to Neut. Sub- stantives ; the only examples of Adjectives are dymjs (Il. 8. 524), eheyxé-es (Il. 4. 242., 24. 239), ppade-os vdov (Il. 24. 354). In Il. 4. 235 (ov yap emt evdéoou narijp Zebs eooer’ dpwyds) we may equally well read Wevdeoor (Zeus will not help falsehood). It seems very probable that these words are to be accounted for in much the same way as the Masculines in -rys, viz. as abstract turned into concrete Nouns by a simple change of Gender. The transition may be observed in Wevdos in such uses as Il. 9. 115 ob yap Pevbos éuds dards xarédctas not falsely (lit. not falsehood) hast thou related my folly. So édéyxea reproaches/ passes into édeyxées. 6. The different Feminine Suffixes are chiefly used to express an abstract or a collective meaning; e.g. Kaxd-s coward, Kan cowardice ; pia and gvy-7 flight ; Bovdyn counsel, also the body of counsellors, a council; ppov-i-s understanding ; vip-as (-dd-os) a snow-storm ; mhnyO-b-s multitude (collective and abstract) ; and the Nouns in -ris (-ats), -tus, -n, -Sov. Denominative Nouns. 117.] Secondary Suffixes. The following are the chief Secondary or ‘Denominative’ Suffixes. (Note that final o of the Primitive Stem is elided before Secondary Suffixes beginning with a vowel.) -L0, -1 3 a8 Slka-to-s just, dppov-ty a joining, dpOu-.o-s friendly, aidoto-s (for aidoc-to-s) reverenced, yedoiio-s (probably to be written yerd-to-s) laughable, -et0, -€o (chiefly used to denote material, especially the animal which furnishes the material of a thing); e.g. im7m-e.o-s, tavp-evo-s, aly-e1o-s, Bd-cvo-s and Ad-co-s, Kuy-én, XGAK-e1o-s and 78 HOMERIC GRAMMAR. [x118. xdAk-€o-s, kvdv-€o-s, doupdr-co-s, PAdy-€o-s, Hyd-1(v). I 54.] The Case-Ending -$(v) is found in a number of Homeric forms which appear to be construed indifferently as Datives or Genitives. It will be shown, however, that there is ground for believing these forms to have been used for the Dat. only in the Instrumental and Locatival senses (the latter being comparatively rare), and for the Gen. only in the Ablatival sense. They formed, therefore, a ‘mixed Case,’ composed of the same elements as the Latin Ablative, viz. the original Instr. Abl. and Loc. In respect of usage these forms are archaic: that is to say, they are confined for the most part to lines and phrases of a fixed conventional type. In several instances the survival is evidently due to the influence of the metre: thus 8a- Kpuddgt, ornPeodt take the place of daxpiwv, ornbéwy; daredpiv and ixprodw, of doréwy, do7éo1o1, and ixpiwv, ixptococ—forms impossible in a hexameter. So kart’ Speodhr, tm’ Sxerdu, for nar’ dpéwy, bn’ dx ew. 155.| Instrumental. The forms in $:(v) appear to have been forms of the Instrumental (Sing. and Plur.), and certainly the majority of the Homeric examples may be referred to that Case : érépng. with the other hand (11. 16. 734, &c.), de&vrepnpe (Od. 19. 480); Bindu by force (Il. 16. 826, Od. 1. 403, &c., and in the phrase kpatepigt Binds with might and main), also in strength (Binge géprepos, Od. 6. 6, &e.); dvayxaings dayévras (Il. 20. 143); yeve- apt vedraros (Il. 14. 112, &e.). In the ‘comitative’ use, adroiow dyeoduy chariot and all, tn- mow Kai dxerpuv with horses and chariot (Il. 12. 114, Od. 4. 533); with Prepositions, du’ jot pawoperngw, ovv immo Kal dxerow (often in the Iliad), also map’ éyeo¢w (construed with Verbs of rest, Il. 5. 28, 794., 8. 565., 12. 91., 15. 3)—unless dyeoguy is a Loe. ($ 157); with words expressing agreement, likeness, &e., as mardunpu aprpe fitted his hand, Oedpw pjotwp atddavros (Il. 7. , &e.). With Verbs of trusting; Il. 4. 303 inmootvn te xab hvopénde neToubes ; 80 dyhaingu (Il. 6. 510), Binge (several times). 156.| Ablative. Forms used as Ablatival Genitives are :— Il. 2. 794 vaigw dapopundetev start from the ships. 13. 700 vadouv dpuvydpuevor defending the ships. 3. 368 éx dé prot €yxos HixOn Tarddundiy. 10, 458 amd pev .. kuvenv ceharndw Edovro. Od. 5. 152 daxpudpu répcovro were dried from tears. 8. 279 Kaddmepbe pedrabpdgu efexexurto, With the Prepositions— eg: as ef edvijpi, ek Oedhw, Tmaccaddpi, movrddu, oTnderuy, "EpeBeodur, &e. 158.] FORMS IN -#I(N). III dnd: as dm veuphpu, adrdédiv, yarkddu, o7/becduv, vaddu, &e. mapé when it means from: Il. 12.225 mapa vaddw édevooued” aira xéhevda, and so Od. 14. 498. So probably in the three other places of the Iliad— 18. 305 napa vaipw dvéorn dios "AXtAdeds. 8. 474 mply dpPa mapa vatpr moddxea TInrclwva, 16, 281 éAmdpevor mapa vadgu Todwxea TyAclwva pyvOpov pev dmoppi~a, girdryta 8 éAéoOat, The notion of leaving the ships is implied, so that apd vat = napa vedr. katd down from: kar’ dpeodu (Il. 4. 452., IL. 493). ind from under: im dxeogr (IL. 23. 7), dd Cuydqu (Il. 24. 576). 157.] Locative. This use is found in several clear in- stances, as well as others of an indecisive kind :— Il. 19. 323 BOingr im Phthia ; Tl. 13. 168 kdcoinds AAermT0 was left in the tent; Ovpnqw out of doors, Lat. foris (Od. 9. 238., 22. 220); Kepadipw €Onxe put on the head (Il. 10. 30, 257, 261; ep. 496, Od, 20. 94); dpeogpw in the mountains (Il. 11. 474 os ef Te dadowwol Ojpes dpeopw: Il. 19. 376 7d be Katerar tod? dpecduy : 22.139 nite Kipxos dpecduy k.T.A.3 22. 189 ws 8° Ore veBpdov Gpeoge xvav x.t.d.); Il. 2. 480 Hite Bots ayédndu wey’ eboxos én- Aero tavtov: IL. 16, 487 dyérndu pereAOav coming into the herd. With the Prepositions :—év, as Il. 24. 284 ev yeupl. . detrep7- gw (=Od. 15. 148): mpds, in Od. 5. 432 mpos Kxorvdndovddu (stecking) to the suckers: éppt, in Od. 16.145 Ouida 8 dud’ dotedp. xpos: bwé, in im Gxecgt, Ind Cuyddu (Il. 19. 404, unless the meaning is from under). With én on, at, in the combinations én’ Ixpiddw, ea’ eoyapddup, ént vevpiguy (all in the Odyssey) the Case may be either the Loc. or the Gen. aap’ avrodt occurs four times in the Iliad (12. 302., 13. 42., 20. 140., 23. 640). In three of these places there is a v.]. wap’ adré0t (or mapavTd0r), which generally gives a better sense, and is required by the grammar in 13. 42 éAmovro 8 vias *Axaay alpnoev Krevéay te nap’ adréddt (=mapd vnvoi). It seems probable that the Endings -@: and -gu were confused, possibly at a very early period. 158.] The true Dat. and Gen. There is only one example of the true Dat., viz. I. 2. 363 ds dpitpn ppntpnpw apiyy, podra bé ptrors that phratria may bear aid to phratria, and tribe to tribe. It is not quite clear whether any forms in -$1(v) belong to the true Gen.; the instances in question are :— (1) IL. 21. 295 Kara Idudqu KAvTa Teixea Aadv ééAoat to coop up the army within the famous walls of Llios, (2) Od. 12.45 modds 8 dud’ doredgur Ols dvdpdv mvOopeven there is around a great heap of bones, of men rotting. But this may be an Instrumental of material,=‘a heap (is made) of bones.’ 112 HOMERIC GRAMMAR. [159- (3) IL. 16. 762 xepadrjguy eet AdBev ody! peOle (Gen., § 151, a). II. 350 00d dpdpapte titvoKdpevos Kepadjguy (but the Gen. might be construed with a¢duapre, as an Abl.). (4) The phrase daxpudds tAfoder (Il. 17. 696, &e.). (5) Certain uses with Prepositions; viz. émi in Il. 13. 308 7 emt defidgw .. i) én’ apiotepoduiy towards right or left ; &4 through, in dia 8& orfOecdw edraccen (II. 5. 41, &e.), also 10. 185 épxnrae be dpecdu: mpdoGe in Il, 5. 107 mpdod’ innoiiv cai dyeoguy. The first four of these references evidently do not prove much. The first would be a clear instance of the true Gen. if we could be sure of the text: but there is some probability in favour of IAtoo (§ 98), proposed by Leo Meyer (Decl. p. 55). The form Saxpudde with mApodey &c. may be an Instr. (like the Latin Abl. with similar Verbs) or an Ablatival Gen., § 151, e Again, the uses with émt and mpéoe may be Locatival. The uses with 614 are more important, because they are not isolated, but form a distinct group. It is improbable that 64 through should take an Ablatival Gen. or a Locative. It remains to consider whether the Instrumental is admissible. Now the Sanscrit Instr. is often used of the space or time over which an action extends (Delbriick, Synt. Forsch. iv. p. 58): and so the Abl. in Latin (Roby, Pt. II. 1176, 11~9). This use appears in Greek as the Dat. of the way by which, and perhaps in the phrases mepiidvre 7G Oéper, &c. It is at least possible that 6.’ épeope and da o77- Geog: are fragments of this use. If so, one or two other uses assigned above to the Loc. may be really Instr.; especially dpeogu, Il. 11. 474., 22. 139, 189. On the other hand, if the forms in -v(v) constitute a ‘mixed Case’ ( Locative, Instrumental, and Ablative), there must have been a tendency to extend its sphere from the Loc. and Instr. to the Dat., and from the Abl. to the Gen. Thus the few instances of forms in -dv(v) standing for the true Dat. and Gen. may be first steps towards an amalgamation of five Cases (such as we have in the Greek Dual). Or we may class them with the ‘false archaisms’ which doubtless exist in Homer, though not to the extent supposed by some commentators. Forms in Sev and -ws. 159.] The Ending -@ev expresses the point from which motion takes place; hence it is common in construction with Verbs of motion, and after the Prepositions ég and &mé. Cp. also— Il. 3. 276 Zed mdrep “ldndev pedéeav ruling from Ida. 8. 397 "[dndev émel te when he saw, looking from Ida. 15. 716 "Exrwp 58 mpvurndev émet Md Be when he had got hold Jrom (i.e. in the direction from, beginning with) the stern; so érépwbev on the other side, dudorépwbev on both sides. OF time ; j80ev from (beginning with) dawn, In a metaphorical sense; of an agent (regarded as the source of action), as Il. 15. 489 AidOev BrapOévta Bédreuva: Od. 16. 447 ovd€ rl uy Odvarov tpoucerOat dvwya éx ye pvynoTipwr Oedbev 8° ovdK gor’ ddcacbat, Also, Il. 10. 68 marpddev ex yevefs dvoudcov naming Srom (on the side of ) the father, And in two phrases, Il. 7. 39, 161.] THE ENDING -93, 113 736 oid0ev olos quite alone, and Il. 7. 97 aivdbev alvds quite terribly, —where the force of the Ending is indistinct. It is to be observed that (except in the Personal Pronouns) this form is not found with Verbs meaning to deprive of, free. Jrom, defend, surpass, or with the corresponding Adjectives and Adverbs. Hence it cannot be held to be equivalent to an Ab- lative (§ 152), and probably differed from the Abl. in expressing motion from rather than separation. On the other hand, the Pronominal forms éué6er, oéGer, ev are freely construed — (1) as Ablatives: mpd ev, inép oder, dvev éuédev; and with a Comparative, I]. 1.114 ob €6& éort yepelwv, &c. Cp. also II. 9. 419 wdda yap éGev . . xeipa hy imepeoye. (2) as true Genitives: Il. 4.169 GAAd pou alvov ayxos oébev écoetat I shall have terrible grief for thee ; with Verbs of hearing (Il. 2. 26, &e.), remembering (Od. 4. 592), caring (Il. 1. 180 oéGev 3d €yd ovk ddcyiGo), reaching or touching (dvTidlw, Teipatw, &e.) ; and with docov, mpdcde, dvra, avtiov, évexa, Exntt. 160.] The Ending -ws is generally derived from the -dt of the Sanscrit Ablative of Stems in -2. There seems no reason to doubt this identification, although the transition of meaning which it supposes is not a very easy one, The chiet examples in common use in Homer are— From Pronominal Stems: ds, 76s, 7&s, 66s, adtws, ddAos. From Stems in -o: aivés, donaciws, éxmdyAws, émiotapevos, Oapcadéws, kakOs, KkapTadiuws, Kpaimvds, kpatepGs, étpadéws, TUKI- vos, pyidiws, oTEepe@s, oTVyEepGs, yadeTas. From other Stems: advrws, dtpexéws, dopadréws, adpadéws, Tepippadéws, Sinvexews, evduKéws, Aty€ws, voreuvews, TpoPpovews. It will be seen that comparatively few of these Adverbs come from the short familiar Adjectives. Thus adds, aicxpiis, weyadws, raxéws, pikws are very rare in Homer; and there is no Adverb of the kind from dewds, icos, dp0és, Bapds, devs, ofds. The Nominative. 161.] Impersonal Verbs. It is evident that in a language which distinguishes the Person and Number of the Verb by the Ending, it is not essential that there should be a distinct word as Nominative. The Verb éo-ri (e.g.) stands for he is, she is, i¢ is; the person or thing meant by the Ending may be left to be gathered from the context. In certain cases, however, the Sub- ject meant by an Ending of the Third Person is too indefinite to be expressed by a particular Noun, such as the context could supply to the mind. For instance, in the sentence otTws €o-rh ab is so, the real Subject given by the Ending -1 (in English by I 114 HOMERIC GRAMMAR. [162. the word i#) is not a partieular thing already mentioned or implied, but a vague notion—‘ the case,’ ‘the course of things,’ &e.* Verbs used with a vague unexpressed Subject of this kind are called IMPERSONAL. The vague Subject may be conceived as a Plural, as Il. 16. 128 obdxérs punta méAw@vras the case shall no longer allow of flight, Od. 2. 203 iva éroerat things will be even. It may be observed that a Neuter Pronoun used as the Subject sometimes gives a vague meaning, not far removed from that of an Impersonal Verb ; e.g. Tl. 1. 564 el 8 obtw rotr’ gori if this is so (cp. obrws éori it is so); éaOAdv nal 76 TéruKTAL ait is a good thing too. An Impersonal Verb is often followed by an Infinitive, or dependent Clause, which supplies the want of a Subject. See Chapter X. 162.] Nominative in the Predicate. In certain cases the Predicate of a sentence may be limited or modified by a Nomina- tive in agreement with the Subject. This is especially found— 1. With Adjectives of time; as éorépioe adixovto they came in the evening, évvixios, mpoyorsyv coming forth by night, ebdov mavrixior slept all night, xO.Cos &Bn went yesterday. Such Adjectives seem to answer most nearly to the Gen. of time within which, but may also express duration, as mavnuéptos and tavvix.os. 2. In describing the attitude, manner, position, &e. in which an action is done: as maAlvopsos améotn stood off with a start back- wards, intvos obder epetsOn was dashed face upwards on the ground ; so meds eidjAovda, AaBpds eraryiQwv, mpdppwy TérAnKas (cp. mpo0- gpovews), duerpoenns exodwa, &c. Similarly, d3¢ and xeivos are sometimes used instead of Adverbs of place: Il. 5. 604 kal viv ot mapa xetvos”Apns now, too, yonder is Ares at his side; 10. 434 Opijixes ol8 dmdvevOe here are the Thracians apart. 3. With Verbs meaning Zo de, to become, to appear, to be made, called, thought, &c.; as Kdptioto. tpddev they were nurtured the mightiest, (i.e. to be the mightiest); elownol éyévovto vedy they came to be in front of the ships: Hde dpiorn aiveto Bovdy this ap- peared the best counsel. Tn all such cases the Nominative which goes with the Verb not only qualifies the notion given by the Verb-Stem, but also becomes itself a Predicate (7. ¢. the asser- tion of an attribute). E.g. «dprioro tpapev implies that they were kdpricrov, A Noun so used is called a SeconDARyY Predicate. The use of eipl as the ‘logical copula’ is merely a special or ‘singular’ case of * See Riddell’s Digest, §§ 95-100, 163.] USE OF THE NOMINATIVE. 115 this type of sentence. The Verb has then little or no meaning of its own, but serves to mark the following Noun as a Predicate. The final stage of the develop- ment is reached when the Verb is omitted as being superfluous. 4. With Impersonal or half-Impersonal Verbs meaning ¢o Ze, &e.; the Predicate being— (a) a Neuter Adjective; as pdpoudy ore it is fated; veueo- onroy d€ Kev ely it would be worthy of indignation ; ob tow deuxés it as not unmeet for thee: with a Pronominal Subject, éo@Adv yap 76 réruktat it is a good thing. In the Plural, odkére guxra médovrar there is no more escaping ; cp. Aolyta Epya 748’ Eooerat this will be a pestilent business. In one or two instances the Adverbial form in -ws is used in phrases of this kind: Tl. 9. 547 Koupijrecot xaxas av things went all for the Curetes; Vl. 7. 424 d:ayvdvar yadenGs jv it was hard to distinguish ; Il. 11, 838 ms Kev &o1 tdde Epya; This may be re- garded as older than the Neut. Nominative, since it indicates that the Verb is not a mere ‘copula,’ but has a meaning which the Adverb qualifies. Cp. Il. 6. 131 div qv Lived long (=8yvaids qv); Od. 11. 336 mGs type dvip be paiverar civar; (2) an abstract Noun; as Il.17. 556 col wey 8) Mevédae xary- dein cal dveidos €ooerat ei x.t.d. to thee it will be a humbling and reproach if &e.; od vépeots it 1s no wrong; ei dé por aica but if it is my fate: with a Pronominal Subject, AdBn rdde y éooerar this will be a shame. The use of an abstract Noun instead of an Adjective is a license or boldness of language of which we have already had examples; see § 116 and § 126. The Latin idiom called the Predicative Dative (Roby, Pt. II. pp. xxv-lvi) may be regarded as a less violent mode of expression than this Nom., since the Dat. is a Case which is originally ‘adverbial,’ ¢.e. construed with the Predicate given by the Verb-Stem. In other words, dedecori est is a less bold and probably more primitive way of saying it is disgraceful than dedecus est; just as kakds 4v is more primitive than kakov jv. It is worth while to notice the tendency to import the ideas of obligation, neces- sity, &c. into these phrases: e.g. ob véueots it is not (worthy of, a matter of) indignation, dveidos éccerat it will be (ground of) reproach. So in Latin vestra existimatio est =it is matter for your judgment. 5. The ordinary use of the Participle belongs to this head: as dcacritny épicavte parted after having quarrelled, In this use the Participle qualifies the Verb-stem, and at the same time makes a distinct assertion : see Chapter X. 163.] Interjectional Nominative. The Nom. is not unfre- quently used in Homer without any regular construction, as a kind of exclamation: e. g.— Il. 5. 405 col & éml rotrov dvfxe Od yhavadais *AOjvn, vimios, ovde TO olde «.T.A. fool! he knows not, &e. 12 116 HOMERIC GRAMMAR. [164. Similarly oyérdwos cruel! dsdcpopos the unhappy one! (Od. 20. 194): and so Il. 1. 231 dnpoBdpos Bactdeds devourer of the people ! Of the same kind is the interjectional use of aides shame ! (Il. 5. 787., 13. 95.5 16. 422). : ; A similar account may be given of one or two passages in which commentators generally suppose * anacolouthon’ : viz.— Il. 10. 436 10d 8) KadAiorous tmmous tov 7b€ beyiorous* Aevxdrepot xidvos, Oelew 8 dvéporoww dpotor whiter than snow they are! &c.; and so in the equally abrupt— Il. 10. 547 alvGs axrivesow éorxdtes nediovo. 2. 353 dotpdnrav embdeE’ evaloyia ojpara aivwv (he did so I tell you) by lightning on the right &e. Od. 1. 51 vijoos devdpjecoa, Oca 8 évi Sduara vate. an island (it is) well wooded, and a goddess has her dwelling there! These forms of expression, when we seek to bring them under the general laws of the grammatical Sentence, resolve themselves into Predicates with an unex- pressed Subject. That there may be logical Propositions of this kind has been justly recognised by a recent writer, Dr. Christoph Sigwart (Logik. L. p. 55 ff.). The Predicate, he shows, must always be expressed in a word (or words); but the Subject, when it is of the kind which would be expressed by a Pronoun (?t, this, &c.) may be indicated by a gesture. The simplest examples of the type are the imperfect sentences used by children, such as horse! for this is a horse. When such sentences are introduced into literary language, they give it an abrupt and interjectional character, as in the examples quoted. We might add the phrases such as ov vépeots it is no wrong (§ 162), in which the want of a Verb makes the expression somewhat interjectional. Compare, for instance, od véyeows with aidws, *Apyetor shame on you, Greeks ! The Vocative. 164.] Regarding the use of the Vocative in Homer the chief point to be noticed is the curious one (common to Greek and Sanscrit) that when two persons are addressed, connected by re, the second name is put in the Nominative.* For instance— Tl. 3. 277 Zed warep “ldndev pedéov kvd.oTe péyiote, "Heduos 6’ bs x.t.A. Similarly, the Vocative is not followed by 8é or any similar Conjunction, but the Pronoun od is interposed; as Il. 1. 282 *Atpetdn ob b& matic «.7.A. but, son of Atreus, cease, Se. The Nominative is often used for the Voc., especially, it would seem, in order to avoid the repetition of the Voc.; e.g. Il. 4. 189 pidros ® Mevédae. On this point however it is not always possible to trust to the accuracy of the text. Cobet (Mise. Crit. p- 333) has good grounds in the conditions of the metre for pro- posing to change a great many Vocatives into Nominatives. * Delbriick, Synt. Forsch. iv. p. 28, 166.] ADJECTIVES, 117 Adjectival Use of the Noun. 165.| Substantive and Adjective. This seems a convenient place for one or two remarks on the distinction expressed by these terms. It will be seen from §§ 114 and 117 that there is no general difference in the mode of forming Substantives and Adjectives. Certain Suffixes, however, are chiefly or wholly employed in the formation of adstract and collective Nouns (which are necessarily Substantives): as in the Feminine Nouns in -ri-s, -rv-s, -dor, the Neuters in -ua(r), the Denominatives in -rys (Gen. -ryr-os). Further, the Suffixes which form Nouns in -r7-s, -ryp, -rwp and -evs are practically confined to Substantives. In respect of meaning and use, again, the distinction between concrete Substantives and Adjectives is practical rather than logical. Certain Nouns are mainly used as qualifyg words in agreement with other Nouns; these are classed as Adjectives. In such combinations as Bots raiipos, dvépes Gddnoral, xadkijes dvdpes, Bactdeds Kopos, ’"Ayapyéuvev ’Arpeidys, where the qualifying word is one that is not generally used as an Adjective, we speak of the ‘adjectival use’ of a Substantive. Conversely, when an Adjective stands by itself to denote an individual or group of objects, the use is called ‘substantival’: e.g. xaxds a base fellow, kaka evils, tuxtrov kaxdv a made mischief. This is a use which arises when the objects to which an Adjective applies are such as naturally form a distinct class, Abstract and Collective Nouns, it is evident, are essentially Substantives. Thus there is a clear distinction, both in form and meaning, between Abstract and Concrete Nouns; but not between Substantives and Adjectives. The common definition of an Adjective as 4 word that expresses ‘ quality’ (‘ Adjectives express the notion of QUALITY,’ Jelf, ii. p. 7) is open to the objections (1) that an abstract Substantive may be said to express quality, and (2) that every concrete Noun of which the etymological meaning is clear expresses quality in the same way as an Adjective. .g. the definition does not enable us to distinguish paxntns from paxjpwr. It is evident that the use of a Nominative in the Predicate—as Baotreds éor he is king—is strictly speaking an adjectival use. The corresponding distinction in the Pronouns does not need much explanation. The Personal Pronouns are essentially Sub- stantives (being incapable of serving as limiting or descriptive words) ; the Possessive Pronouns are essentially Adjectives. The others admit of both uses; ¢.g. odros this one, and dvijp otros (in Attic 6 dvip otros) this man. 166.] Gender of Adjectives. In a few cases the Gender of the Adjective is independent of the Substantive with which it is construed. 118 HOMERIC GRAMMAR. [167. 1. When a person is described by a word which properly denotes a thing (viz. a Neuter, as réxvov, téxos, &c., or an abstract Noun, Bin (pido, &c.), the concord of Gender is not always observed. Thus we have ide réxvoy (but pédov réxos, pin kepadr) ; again— Tl. 11. 690 eAOav yap p éxdxwoe Bln “HpaxAnety (= Heracles). Od. 11. 90 7AGe 8 ent oxy OnBaiov Tetpectao Xptoeov oxinrpov xwv. In such cases grammarians speak of a ‘construction according to the meaning’ (xard oiveow). The term is unobjectionable, provided that we remember that constructions according to the meaning are generally older than those in which meaning is overridden by idiom or grammatical analogy. 2. Where an Adjective refers to more than one Noun, it fol- lows the most prominent: or (if this is at all doubtful) the Mase. is used of persons, the Neut. of things: e.g. Il. 2.136 at d€ mov jérepal 7 Gdoxor Kal vim réxva elar’ évt peydpo.s moridéypevat because the wives are chiefly thought of: but— Jl. 18. 514 retyos pév p’ dAoxol te pidar kal vimia réxva pbar’ épeoradres, pera 8° dvépes x.T.d. because the boys and old men are also in the speaker’s mind, Od. 13. 435 dui b€ yw pdxos GAo kaxdy Badrev 73e xiTGva, poyadéa puTdwvta. The Neut. Plur. is especially used of sheep and cattle: Il. rz. 244 7pG0" Exatov Bots dOxev, emerta BF XA dréoTy, alyas éuod Kal dis; Il. 11. 696 ék 8 6 yépay ayédnv te Body Kal TOU péy’ oldy cihero, Kptvdpevos tpinxdov 738 voyfas (three hundred head): ep. also Il. 5. 140., Od. 12. 332. 3. A Noun standing as Predicate may be Neuter, although the Subject is Mase. or Fem.: as ot« dya0ov modvkoipavin. This is a kind of substantival use. 167.] Gender of Pronouns. A substantival Pronoun de- noting a person may retain its proper Gender although the antecedent is a Neuter, or an abstract word; as Il, 22. 87 oirov Oados, dv Tékov adty. Conversely a Neuter Pronoun may be used substantivally of a thing which has been denoted by a Mase. or Fem. word: Il. 2. 873 ds Kal xpdoov exwv médAcudvd? lev Hire Kovpn, vimtos, ovd€ TL ot rd y’ Emnpkece Avypov GA€Opov. Cp. Il. 11, 238., 18, 460, Od. 12. 74 (with the note in Merry and Riddell’s edition), On the other hand, a Pronominal Subject sometimes follows the Gender of a Noun standing as Predicate, as airy dikn éort this is the manner, } Ous @orl as is right. But the Neuter 170.|| THE NUMBERS. 119 is preferred if a distinct object is meant by the Pronoun; as Od. 1. 226 od« &pavos réde y' earl what I see is not a club-feast. _ 168.] Implied Predication. An Adjective (or Substantive in an adjectival use) construed with a Noun in an oblique Case may be so used as to convey a distinct predication ; as otkér’ éuol pida tadr’ dyopevers= this (that you now speak) is not pleasing to me. So after Verbs meaning to make, cause to be, call, think, &e. ; Aaods bE AiPous Toince Kpoviov Zeus made the people (to be) stones. This use is parallel to that of the Nominative in the Predicate (§ 162): ep. the forms of sentence Aaoi éyévovTo AiBo1, Aaovs éxoinae Aldous. In the latter the pre- dicative Noun (Ai@ous) is construed with an oblique Case, instead of with the Subject. A Noun so used is called a TERTIARY PREDICATE: ep. § 162, 3. CHAPTER VIII. Use or tHE NuMBERS. 169.| Collective Nouns. The Subject of a Plural Verb may be expressed by means of a Collective Noun; as ds ¢dcav 7 mAnOds thus they said, the multitude (ep. Il. 15. 305., 23. 157). Conversely, a Participle construed with a Singular Nomina- tive and Verb may be Plural: as I]. 18, 604 mepiicraé’ Susdr0s teptopevot. So too Il. 16. 281 éxivyndev b& padrayyes eAmopevor. In these instances, again, the construction is said to be ‘according to the mean- ing’ (§ 166). The principle is evidently that an abstract or collective word may be used in ‘apposition’ to a concrete word. It may be noticed however that the combinations such as Spiros - reptépuevor are only found when there is some pause between the words; otherwise the Genitive would be used (construed as in Tpwwy karedvaed Spdov, &e.). 170.| Distributive use of the Singular. The word éxaoros is often used in the Sing. with a Plural Verb, as €@av oixdvde éxaotos they went home, each one, deduyyecba Exaotos we are each one obedient. Other words in a clause may follow éxacros in respect of Number:: as Il. 2. 775 immo 58 map’ Gppaow oiow éxactos the horses each beside his chariot ; Il. 9. 656 of b& Exaatos éddv dénas dudixtmeddov oTrelocavtes Tapa vijas toay makw. Even the Verb is made Sing. in I]. 16. 264 of 8 GAktpov Hrop exovtes mpdoow mas néretar Kal dudver olor réxeoou: but this is a slight boldness of expression. On the same principle we may explain the Sing. in Od. 4. 300 ai 3 toav éx peydpoo Sdos pera xepolv éxovoa (= each with a torch 120 HOMERIC GRAMMAR. [irae in her hands). 11,13. 872 tervppévw xara xelpa (each of the two) wounded in the hand. Similarly the Dual is used alone with the Plural of a group which consists of pairs :— Tl. 16. 370 woAdol & ev radpy epvodppates @xées trmot d€avr’ év mpdro pup Altov dppar’ dvaxrav where the Dual &gavre (like the Sing. juu¢) refers to one chariot. Probably, too, we should read dpya dvaxrwy (i.e. Favdxto). The Dual is often used in this way in Aristophanes: cp. Av. 622 dvareivoyres 7 xeipe, and other instances given by Bieber (De duali numero, p. 44). In 11. 5. 487 po] rms ws dio Aivov dAdvre ravd-ypou, the Dual adAdvre is explained by Schol. B bpeis cal ai yuvaixes, If so, it is a distributive use: ‘see that ye be not taken, man and wife in one net.’ In speaking of the characteristics of a group or class it is common to pass from the Plural to the Singular, or vice versd ; e.g. Od. 4. 691 7 7’ eort Bikyn Oelwy Bacidjwr, GAdrov x’ exPatpyse Bpordv x.r.d. it is the way of kings, (a king) will hate one &e. Conversely, Il. 19. 71 6s ke @iynou.. tn éyxeos Hueteporo which- ever one shall have fled before the spear of one of us ; so in deserib- ing a kind of helmet, Il. 10. 259 pverar 8 Kdpn Oadrgcpov al(nav. Cp. also Tpdwv adoxos (Il. 2. 355) =the wife of some Trojan. The Plural is possible because the preceding Singular is distributive, 7.é, takes one instance to represent the class intended. 171.] Plural of Things. The Plural form is not confined in Greek (or indeed in any language) to the expression of ‘ plurality’ in the strict sense (2.¢. to denote a group composed of distinct individuals), but is often used (esp. in Homer) of objects which it is more logical to think of in the Singular. Many words, too, are used both in the Sing. and the Plur., with little or no difference of meaning. Notice especially the uses of the Plural in the case of— (1) Objects consisting of parts: réfov and rééa bow and arrows : éxos and éxea, dpya and appara a chariot: ddua, wéyapov a hall or room, Sepuara, wéyapa a house: Aéxtpov and déxtpa a bed. mihar a gate is only used in the Plur.; 6vpy is used as well as Oupat, but only of the door of a room (@aAapos). (2) Natural objects of undefined extent: ydpabos and Wduador (as we say sands), dAes (once GAs) salt, xovin and xoviar dust, nupés and tupot wheat, jidves, péeOpov and péeOpa, kya (in a collective sense) and xdpara, ddxpv and ddxpva, xpéa (seldom xpéas) meat, cdpxes (once Sing.) flesh. (3) Parts of the body: véroy and réra, orf80s and (more commonly) orca, tpdownoy and apocwna the. countenance, ppiv and ¢péves. 173 .| NEUTER PLURAL—DUAL, 121 (4) Abstract words: Acdaopuévos innocuvdor forgetting horse- manship, noduxeino. metobeas trusting to speed of foot, dvadrxeinar dauévtes overcome by want of prowess, modvidpeinar vooww through cunning of understanding: so atacbaXriat, dppadia, aynvopiat, deor- ppootva, Textoovvat, peOnwootva, &c.; note also mpodoxai ambush, mpoxoal mouth of a river, ddpa gift (xpicos yap éptxaxe, dpa Ocoio). The Plural in such cases is a kind of imperfect abstraction; the particular manifestations of a quality are thought of as units in a group or mass,—not yet as forming a single thing. (5) Pronouns and Adjectives; see the examples of adverbial uses, §$ 133, 134; ep. also § 161. 172.] Neuter Plural. The construction of the Neut. Plur. with a Singular Verb is the commoner one in Homer, in the pro- portion of about three to one. When the Plural is used, it will generally be found that the word is really Plural in meaning (i.e. that it calls up the notion of distinct units), Thus it is used with— Nouns denoting agents; as éOvea applied to the men of the Greek army (Il. 2. 91, 464), to birds (Il. 2. 459), to swine (Od. 14. 73); so with Pd’ avOpdmev (Od. 15. 409). Distinctly plural parts of the body: mrepd, xeiAea, od@ara, pédea: so medida (of Hermes). Numerals: 8éka ordyara (Il. 2. 489), ovata récoapa (Il. 11. 634), téccapa deppara (Od. 4. 437), almddra Evdexa mdvta (Od. 14. 103); so with mavra and moAdd (Il. 11. 574., 15. 714.,17. 760., Od. 4. 437, 794-5 9. 222., 12. 411), and when the context shows that distinct things are meant: as I]. 5. 656 tév pev dovpara (the spears of wo warriors), 13. 135 éyxea .. ATO XELpav. A few instances which occur in fixed phrases may represent an earlier syntax ; Avvto 6é yvia (but also Avro yovvara), durxava épya yévovto, &c. Note especially the lines ending with wéAovrat (ra re mrepa vyvol médovrat, Gre 7 iyara papa méAovTat, puKTa méAovtat, &e.) The exceptions to the use of the Sing. are fewest with Pro- nouns and Adjectives: doubtless on account of their want of a distinct Plural meaning (see the end of last section). 173.] The Dual is chiefly used (1) of two objects thought of as a distinct pair, and (2) when the Numeral dvw is used, 1. Thus we have the natural pairs xeipe, 76de, mixee, TEévorre, Spuw, pnpd, dace, dpOadrus: orabud door-posts ; inma the horses of a chariot, Boe a yoke of oxen, dpve a pair of lambs (for sacrifice) ; 122 HOMERIC GRAMMAR. (173. dvudpe (in Tl. 13. 241., 16.139 of the two spears usually carried, but 8vo dof%pe is more common); qurayd (Ll. 5. 773) of the two rivers of the Troad, and so xpowd (Il. 22.147). So of the two warriors in a chariot (Il. 5. 244, 272, 468), two wrestlers (Il. 23. 707), two dancers (Od. 8. 378), the Sirens (Od. 12. 52, &c.); finally, of the ’Arpeida and Aiavre. The Numeral is generally added in speaking of two wild animals (@fpe dvw, A€ovre Sdw, &e.): xdmpw (Il. 11. 324) and A€ovte (Il. 16. 756) are hardly exceptions, since the context shows that two are meant. It is not easy to account for aierw (Od. 2. 146), used of two eagles sent as an omen, and ytme (Od. 11. 578), of the vultures that devoured Tityos, The Dual in Il. 8. 185-191 (where Hector calls to fowr horses by name) might be defended, because two is the regular number; but probably v. 185 is spurious. In IL. 23. 413, again,—al «’ droxndqoarte pepwpeba xeipov deOAov—the Dual is used because it is the horses that are chiefly in the driver’s mind, although he associates himself with them. In II. 9. 182-195 the Dual refers to the two envoys, Phoenix being overlooked. Again, when two agents have been mentioned together, or are represented as acting together in any way, the Dual may be used: as Il, 1. 531 7h y’ Ss Bovdcdoavte (of Thetis and Achilles), 16. 823 (of a lion and boar fighting), Od. 3. 128., 13. 372, &c. Similarly, of the meeting of two rivers, Il. 4. 453 €s ploydyKeray cvuBadrrcrov 6Bp.yov biwp (cp. 5. 774). The Dual Pronouns vai and oat are used with comparative regularity: see Il. I. 257, 330, 574.5 5. 34, 287, 718, &c. This usage may be a matter of traditional courtesy. Hence perhaps the scrupulous use where the First Person Dual is meant; Il. 4. 407 dyaydv6’ (* Diomede and 1’); 8. 109 Oepamovre our attendants ; II. 313 Ti maddvre AcAdopeba 4.7.A.3 12. 323 & wéwov ci. . puydvTe; Od. 3. 128 Eva Ovpov éxovre (‘Ulysses and I’). Cp. the Second Person, Il. 1. 216 (Athene and Here), 322 (the heralds), 3. 279 (Hades and Persephone ?), 7. 279 Taide pidw. 2. Of the use with the Numeral the most significant examples are Od. 8. 35, 48 kovpw S& kpwwOévte S0w Kal TevtiKovTa ByTHY: where the Dual is used by a kind of attraction to the word dv. The Dual is never obligatory in Homer, since the Plural may always be used instead of it. Hence we often have a Dual Noun or Pronoun with a Plural Verb or Adjective, and vice versd. The Neut. Dual (like the Neut. Plur.) may go with a Sing. Verb: thus we have doce with all three Numbers. Certain of the ancient grammarians—Zenodotus among them—supposed that Homer sometimes used the Dual for the Plural. But Aristarchus showed that in all the passages on which this belief was founded the Dual either had its proper force, or was a false reading. The use of the Dual in Attic is nearly the same as in Homer: whereas in other dialects it appears to have become obsolete. This was one of the reasons that led some grammarians to maintain that Homer was an Athenian, 176.] PREPOSITIONS, 123 CHAPTER IX. Tue Prevositions. Introductory. 174.| Prepositions are words expressing some local relation, and capable of being used as prefixes in forming Compound Verbs. The Greek Prepositions are also used in construction with oblique Cases of Nouns and Pronouns. The Adverbs that are construed with oblique Cases, but do not enter into composition with Verbs, are called Improper Prepositions. The list of Homeric Prepositions is the same (with perhaps one exception, see § 226) as that of later classical Greek. In the use of Prepositions, however, there are some marked differ- ences between the two periods. There are no ‘Inseparable’ Prepositions in Greek: see however the note to § 221. 175.] Adverbial use. In post-Homeric Greek it is a rule ee to a few exceptions only) that a Preposition must either 1) enter into Composition or (2) be followed immediately by and ‘govern’ a Noun or Pronoun in an oblique Case. But in the Homeric language the limitation of the Prepositions to these two uses is still far from being established. A Preposition may not only be separated from the Case-form which it governs (a licence sometimes found in later writers), but may stand as a distinct word without governing any Case. In other words, it may be placed in the sentence with the freedom of an Adverb : e.g. as éyy’s is used either with or without a Genitive of the point to which something is near, so dupé may mean either on both sides (of an object expressed by an oblique Case) or simply on both sides ; év may mean in (taking a Dat.), or simply zuside ; and so of the others, ¢.9.— yédaoce 8 Taca mepi xOdv all the earth smiled round about. imal d€ re kduros dddvTwv yiyvero beneath arose rattling of teeth. These uses, in which the Preposition is treated as an ordinary ‘ Adverb of place,’ may be called in general the adverbial uses, 176.] Tmesis. The term Twzsts is sometimes applied gener- ally to denote that a Preposition is ‘separated’ from the Verb which it qualifies (thus including all ‘adverbial’ uses), but is 124 HOMERIC GRAMMAR. [22% more properly restricted to a particular group of these uses, viz. those in which the resulting meaning is the same as the Prepo- sition and Verb have (or might have) in Composition: e.g.— ot Kara Bods “Yreplovos "Hedloro fodiov who eat up (xarjoOovr) the oxen of the sun. ots mor’ dn’ Aivelav EAsunv which I took from (apercunv) Aeneas. tnd 8 eryxero pucddv and promised (bwécxero) hire, This at least is the sense in which the word tp4jots was employed by the Greek grammarians, who looked at the peculiarities of Homer as deviations from the later established usage, and accordingly regarded the independent place of the Preposition as the result of a ‘severance’ of the Compound Verb. We may retain the term, provided that we understand it to mean no more than the fact that the two elements which formed a single word in later Greek were still separable in the language of Homer. The distinction between Tmesis (in the strict sense) and other ‘adverbial’ uses cannot always be drawn with certainty. The question is not so much whether a given Preposition and Verb are actually found in Composition, as whether they coalesce in meaning, so as to form a new compound notion; e.g. weTa vora Baroy turning his back, yeipas and Elpei tunas cutting off his hands by the sword. The clearest cases are those in which the construction of other words in the sentence is affected by the Preposition ; ¢.g. in da Alvefay EAdunv, txd 8 xxeto picddv. On the other hand, the use is simply adverbial in— mept ppevas tyepos alpet desire seizes his heart all round (because the Compound mepiaipéw means to strip off, to take away from round a thing). &s rods fyeudves Stexcopueor . . wera be Kpetwy "Ayayeuvor and in the midst the king Agamemnon, Os Tpdes mpd pev Gdrdot dpypdres, adrap én’ addou the Trojans, arrayed some in front, others behind. 177.| Ellipse of the Verb. In certain cases, viz. when the Verb is understood, a Preposition may represent the whole Pre- dicate of a clause :-— olwvol b& méps mAges HE yuvatkes about (him) are more birds of prey than women, év0’ vu pev pidrdrns therein is love. od tot éme b€0s there is no fear for thee. mapa 8 aunp the man is at hand. adn’ ava but up ! Where the Preposition takes the place of a Compound (as é& for éveott, ému for éreots), Tmesis may be recognised: and cer- tainly in such an instance as— map’ euorye Kat GAdot others are at my command (not are beside me, but = mdpe.or in its derived sense). 178.] PREPOSITIONS, 125 So when a Verb is to be repeated from a preceding clause; as Tl. 24. 229-233 ev dddexa pev mepixaddéas efere mémAovs .. é« 5& bY aidwvas tpimodas: Il. 3. 267 Sprvro & adrik’ émeta dvak avopGv ’Ayauéuvor, dv 8 ’Odvceds (se. Spvvto). 178.] Use with oblique Cases. Prepositions are frequently used in Greek with the Accusative, the Locatival and Instru- mental Dative, and the Ablatival Genitive ; less commonly with the true Genitive; rarely (if at all) with the true Dative. The phrases formed by the combination of a Preposition with a Case- form are equivalent in construction to single Adverbs. It may be shown (chiefly by comparison with Sanserit) that the government of Cases by Prepositions belongs to a later stage of the language than the use of Prepositions with Verbs. Origi- nally the Case was always construed directly with the Verb, and the Preposition did no more than qualify the Verbal meaning. F.g. in such a sentence as els Tpoinv Ge the Acc. Tpotnv originally went with Ade. If however the construction Tpoiny 7AGe (that is to say, the Accusative with the simple Verb) ceased to be usual, it is evident that the Preposition would become necessary, and the combination e’s Tpoiny would be soon stereotyped, so as to be equivalent to a single Adverb. In Homer we find many instances of a transitional character, in which a Case-form which appears to be governed by a Prepo- sition may equally well be construed directly with the Verb,— modified, it may be, in meaning by the Preposition. Thus we have époi with the Dat. in the recurring form— dug 8 dp wpoow Barero fidos, but the Preposition is not necessary for the Case, as we see from its absence in— dpos ipOiporor Bar’ aiyisa, and again from the transitional form— dupl 52 xarrar | duos diccovrar, where the Preposition is best taken in the adverbial use—on each side his mane streams out on his shoulders. Cp. IL 17. 523 év 5€ of éyxos | vndvioror par’ ov xpadavdpevov Ave yuia, where év is adverbial. The transition from év in Tmesis to év with the Dat. may be seen in év yain énayn (=yaln evenayy) was fixed in the earth; ep. obSe. émaoxippén, &c. Again, we seem to have api governing the Accusative in— Il. 11. 482 ws fa 767 dup’ ’OSvaqa . . TpWes Error. But dpi must be taken with €mov, as in— IL. 11. 776 op@i pev audi Bods Enerov xpéa, So in— ind (uydv iyaye brought under the yoke the supposition of Tmesis is borne out by the form— bnaye (uydv dxéas inmous. Again, the Ablatival Genitive in— HAO’ ef GAds came out from the sea may be explained like reixeos éfeAeiv, &c.; and— yds dnd mpiurns xapddis wéoe by vos droOpwoxwy, and numerous similar forms. 126 HOMERIC GRAMMAR. [179. A transitional instance might be found in— Il. 6. 100 bvmep pact beds ééupevar’ GAN’ bbe Alqv, where however the rhythm of the verse is rather against writing Oeds 2€ Eupevar. Thus the history of the usage of Prepositions confirms the general principle laid down in a previous chapter ($ 131), that the oblique Cases, with the exception of the (true) Genitive, are primarily construed with Verbs, and that consequently the con- struction of these Cases with Nouns and (we may now add) Prepositions is always of a derivative kind. 179.] Use with the Genitive. Where the Genitive with a Preposition is not Ablatival, it may usually be explained in two ways, between which it is not always easy to choose :— (1) It may be derived from one of the uses with Verbs dis- cussed in §$ 149-151. .g. the Genitive in— bs 7 elow bia dSovpds which goes through the wood is probably the Genitive of the space within which motion takes place. For efow 61a d0vpds has the same relation to medio10 didkew and wedioo Stampyjocew, that jAdev eis Tpolyy has to Tpotny hAGev and Tpotny ciondrGev. (2) It may be of the same kind as the Genitive with a Noun: e.g. the construction with dvri may be the same as with the Adverbs dra, dvtiov, dvtia, &e., and the Adjectives dvrfos, évav- tios, &e., and this is evidently not akin to any of the construc- tions with Verbs, but falls under the general rule that a Noun or Pronoun qualifying a Noun is put in the Genitive ($ 147). It is held by Curtius (Elucidations, v.17) that the Genitive with avrt, mpd, 514, Saép, 6, and other Prepositions, when they do not necessarily imply motion JSrom, is of the same kind as the ordinary Genitive with Adjectives and Adverbs. This view is supported by the fact that in Greek the Improper Prepositions nearly all govern the Genitive, whatever their meaning: e.g. éyyvs and éxas, évrés and éxrés, dvra, wexpt, évexa, &c.; for the explanation of this fact can only be that the construction does not depend upon the local relation involved, but is of the same kind as in S¢uas tupés, xdpww Tpadov, &e. On the other hand, it is argued by Delbriick (Synt. Forsch. iv. p. 134) that such a construction of the Genitive with Prepositions is unknown in Saunscrit, and therefore is not likely to be the original construction, at least in the case of the Prepositions common to Greek and Sanscrit—a list which includes mp6, t76, and imép. He would allow the supposition however in the case of dvti (the Sanscrit anti being an Adverb), and perhaps 814; regarding these words as having become true Prepositions more recently than the others. 180.] Accentuation. The rules for the accentuation of Com- pound Verbs have been already given in § 88. They proceed on the general principle that (except in the augmented forms) the accent falls if possible on the Preposition; either on the last syllable (as dmé-dos), or, if that is elided, then on the first (as im-aye). 180.] ACCENT OF PREPOSITIONS. 127 In regard to the other uses, and in particular the use with Cases, the general assumption made by the Greek grammarians is that all Prepositions are oxytone. They do not recognise the modern distinction according to which éy, eis, and ég are unac- cented—a distinction which rests entirely on the practice of the manuscripts (Chandler, p. 254). Dissyllabic Prepositions, however, are lable in certain cases to become barytone. The exact determination of these cases was a matter of much difficulty and controversy with the ancient writers, and unfortunately we cannot now determine how far their dicta rest upon actual observation of usage, and how far upon theoretical and fanciful considerations. The chief points of the generally accepted doctrine are :— (1) The dissyllabie Prepositions, except doi, dvri, avd, and dua (except also the dialectical forms xaral, tral, wapat, anai, tmelp, mpori), are liable to ‘ Anastrophe ;’ that is to say, when placed immediately after the Verb or the Case-form to which they belong, they throw back the acute on to the first syllable, as oven ato (=damodovey), exev Kata, @ emu (=ep’ ©), waxy en, Zeqvpov imo, &e. (2) Also (according to some, see Schol. A on Il. 5. 283) if they stand at the end of a verse, or before a full stop. (3) Also, when they are equivalent to Compound Verbs ($177); as én, ém, mapa (for év-eor, &e.). So dva (for dvd- o7n6t); although dvd according to most authorities was not liable to Anastrophe. (4) Two Prepositions are barytone in the adverbial use,— amo when it is =dmo6ev at a distance, and mépt when it is=Tepicods eaceedingly. To which some added zo (as rpopeer 8’ to yvia, &c.). (5) Monosyllabic Prepositions when placed after the governed word take the acute accent (as an equivalent for Anastrophe) ; but only when they come at the end of the line. Some however accented Od. 3. 137— Kakecoapevw dyopyy és mdvtas ’Axatovs. The general effect of these rules being that Prepositions when they have the character of independent words retract the accent, we are led at once to the inference that they are properly bary- tone (as they are in Sanscrit), and that the grave accent which they take in the use before Case-forms only means that they lose their accent altogether. In other words, just as there is an ‘orthotone’ Zor: and an enclitic éor, sometimes written éo7i, so there is (e.y.) an original orthotone mép., and a ‘ proclitic’ mepr, written 7epi, but in reality without accent (except in the case of a following enclitie, as mepf re). 1. On this view the accent of Prepositions is originally the same in Greek and 128 HOMERIC GRAMMAR. [180. in Sanscrit. Moreover, it will serve to explain one or two minor peculiarities of Greek usage. Thus (1) it is the rule that when the last syllable of a Preposition is elided before a Case-form, the accent is not thrown back. This is intelligible on the ground that the Preposition is in fact without accent; and the same account will apply to the same peculiarity i in the case of dAAd and rivd, On the other hand, (2) the accent is retracted in the case of elision before a Verb (as Um-aye), because the Preposition is then the accented word*. Again, (3) the general rule of the Aeolic dialect, that all oxytones become barytone, does not extend to Prepositions; doubtless because they are not real oxytones. The word ért (Sanscr. dt?) is a Preposition which happens to have survived (with the original accent) in the Adverbial use only: cp. the use of apes = besides. It will occur as an objection that on this view all Prepositions in Tmesis ought to be barytone, not only those which follow the Verb or governed Case. The answer may be that this was in fact the Homeric accentuation, but was forgotten before the time of the grammarians, whose teaching on the subject of accents was entirely derived from the usage of later Greek. Hence, while they knew some- thing of the accent in the case of ‘ Anastrophe,’ they were ignorant of it in the wider group of adverbial uses to which Anastrophe belonged. For Anastrophe survived in later usage, whereas Tmesis in general is exclusively Homeric. It is worth observing, however, that some grammarians carried the doctrine of Anastrophe further than others. Ptolemaeus Ascalonites wrote ace 5° dao fuvdv tpnxds Aidos (11. 5. 308), holding that the insertion of 5€ did not prevent Ana- strophe. Some wrote mépa yap Geot eiou wat Hpi (Il. 3. 440), for the singular reason (founded on the literal sense of Tmesis) that the Compound is mdp-eor. A better ground would be that the Preposition is as emphatic as in wap’ Zyovye xat a@ddot (Il. 1.174), Thpa 8 avqp, &c. There was also much doubt about the accent of a Preposition placed between a Substantive and an Adjective construed with it; as BdvOou dao Siwjevtos, morayod dao SeAAnevros. Apollonius and Herodian adopted the view which recognised Anastrophe in every such case: and Tyrannio even wrote Ados tro furs, on the ground that the order in prose would be id Alou puTrjs. 2, One or two suggestions may be added in reference to the Prepositions which are generally said to be incapable of Anastrophe :— dvé was thought by some to be capable of Anastrophe, and this view is sup- ported by the adverbial use dva up! &poi is probably a real oxytone, since the Adverb duis is so. The correspond- ing Sanscrit Preposition abhi is oxytone, contrary to the general rule. The assertion that brat, wapal, mporl, &c. are not liable to Anastrophe is difficult of interpretation. It may mean no more than that later usage furnished the grammarians with no examples. * See Wackernagel in K. Z, xxiii. p. 457 ff. On this view, however, the original accent would be dmo-5os, évi-orres, Tapa-oxes, &e. It may perhaps be preserved in the Indic. évt-omes and Imper. évt-ore (see § 88, where a different explanation of these forms was suggested). The Imperatives in -s may be paroxytone because they stand for dn6-506t, évi-omeb, &c. Since the passage in the text was in type I have found that a similar account of these matters is given by Prof. Benfey in his Vedica und Linguistica, pp. go ff. He goes so far however as to hold that the Du. and Plur. forms of ecip{ and pypt were all originally barytone (p. 98). But, admitting an original orthotone éopt, 7pt, &e. in the Sing. (and probably in the 3 Plur.), surely the change of accent in dap, Plur, dapev, is original, and related to the change of quantity ; as also in émi, Plur. imds, véda, Plur. vidmds, &e. The difficulties found in applying the theory to the Imperative are not conclusive, 182. | "AMEI, 129 3. There are many places in Homer where it is uncertain whether a Preposition is part of a Compound or retains its character as a separate word, either ‘adverbial’ or governing a preceding Noun. Thus we find the readings — Il. 4. 538 moAAol 2 wept xrelvovro Kal dAdo (Wolf). 16. 497 éped mépt papvao xadKn@ (wept Ven. A.). 18. IgI oretro yap ‘Hpaicroio map’ oicépev évrea xadd (so Aristarchus). I. 269 wat pey roto eyo ped’ dpireor (Ar.). 4. 423 Lepupov tro xvhoavros (Bekker, &c.). with the variants repxreivoyro, mepiudpvao, &c, And the existing texts contain a good many Compounds which we might write divisim without loss to the sense ; as Il. 18. 7 vnvoly émdovéovra, Od, 8.14 mévrov émimrayxGeis, Od 16. 466 doru KaTaBAworovTa, In reference to such forms it is natural to argue that the tendency of gram- marians and copyists unfamiliar with the free adverbial use of the Prepositions, would be always towards forming Compounds; and hence that modern critics ought to lean rather to the side of writing the words separately. : api, 181.] The Preposition di means on both sides, or (if the notion of two sides is not’ prominent) a// round. It is doubtless connected with dude both. The adverbial use is common; ¢.g. with a Verb understood, Od. 6. 292 év 8% Kpyun vader, dupt b& Aepdv and around is a meadow. It is especially used in reference to the two sides of the body : Tl. 5. 310 dudi 8% doce xedawy vie exddrvwe black night covered his eyes on both sides (i.e. both eyes); Il. 10. 535 dui xrimos otara BddAer; Tl. 18. 414 ondyyy & dui mpdcwna kcal dupa xeip’ amopdpyvy; Od. 2.153 maperas audi re depds; Od. 9. 389 mavta d€ of BrAéhap audi Kai dfpdtas k.7.A. So Il. 6. 117 dul 8€ pv opupa rinre Kal abxéva Béppa kedrauvdy the shield smote him on the ankles on both sides and on the neck. Here dpdi is generally taken to mean above and beneath ; wrongly, as the passages quoted above show. This use of dpdt is extended to the internal organs, esp. the midriff (ppéves) regarded as the seat of feeling: e.g. Il. 3. 442 ob yap Th more p’ GBe Epws ppévas dupexaduype (so 14. 294). 6. 355 érel oé pddrora mévos ppévas aur BéBnne. 16. 481 0’ dpa Te ppéves Epxara: dpe’ ddivdv Kip (so Od. 19. 516). Od. 8. 541 pdda rod piv dxos ppevas duprBeBhee. So Hesiod, Theog. 554 xwoaro 58 ppévas dui; Hom. h. Apoll. 273, bh. Ven. 243 5 Mimnerm. 1. 7 fpévas dpi xaxal reipovor péptpvac. Hence read— IL. 1. 103 péveos 88 péya ppéves dpol pédauvar wipmdavt’, and similarly in Il. 17. 83, 499, 573- 182.] The Dative with éyoi is a natural extension of the ordinary Locatival Dative—the Preposition being adverbial, and not always needed to govern the Case. Compare (e. g.)— K 130 HOMERIC GRAMMAR. [183. Tl. 1. 45 dé’ Sporow eww (Loc. Dat., § 145, 3)- 20. 150 audi 8 dp’ dppykrov vepedrny @uorow Evavro. IT. 524 dud’ Suorow exer cdxos has a shield on both sides on his shoulders, 2. e. across his shoulders. In a metaphorical sense éoi is applied to the object about which two parties contend: as Il. 3. 70 dud’ “EAévy kal xrjpace mao. pdxerOa. So of a negociation, I. 13. 382 cvvdueba apdi yduwo we shall agree about the marriage. Here the locatival sense is lost, and the Preposition is indispensable for the syntax. So too Il. 7. 408 dydi 8% vexpotow as to the question of the dead ; Il. 16. 647 dud pove THarpdéxdou pepynpicov. It is a further extension of this use when dpi with the Dat. is construed with Verbs meaning to speak, think, &c., as Od. 4. 151 dy’ ’Odvoqi pvdedunv. This last variety (in which the notion of two sides disappears) is confined to the Odyssey: ep. 5. 287., 14. 338, 364. A true Dative may follow apot, but cannot be said to be governed by it; e.g. in Il. 14. 420 dui 5€ of Bpdxe redyea his arms rattled about him, the Dat. is ‘ethical,’ as in Il. 13. 439 pitey 5é of dul x:rGva. So in IL. 4. 431 appt 8 rar Tevxea ToKid’ €AapTe, the Dat. is not locatival, but the true Dat. The two kinds of Dat. may be combined, as Il. 18, 205 duel 5é of eeparg védos Ecrede. The construction of ép¢t with the Dat. is not found in Attic prose. It survives in the poetical style, and in Herodotus. 183.| The Accusative with dpi is used when the Verb ex- presses motion, as— Il. 5. 314 dydi 8 dv Pirov vidv éxevato mye AcvKa. Also to express extent, diffusion over a space, &c. (ideas naturally conveyed by terms denoting motion) :— Od. 11. 419 as dul Kpytijpa tpamé(as te tAnOovcas KeiueOa as we lay (scattered) about, Sc. Accordingly it is especially used in Homer— 1) of dwellers about a place, as Il. 2. 499, 751, &e. 2) of attendants or followers; as Il. 2. 445 of 3’ dud’ ’Atpetwva . » Oivov they bustled about Agamemnon. The description about (a person) need not be taken to exclude the person who is the centre of the group; e.g. in Il. 4. 294 (Agamemnon found Nestor) ods érdpous oréAdovTa .. dup péyav TeAdyovra AAdoropd Te Xpopiov re, where Pelagon &c. are included under the word érapor. This is an approach to the later idiom, of dypt TAdrava = Plato and his school. It should be observed that the motion expressed by the Verb when dpi takes an Acc. is not motion zo @ point, but motion over a space. Hence this Acc. is not to be classed with Accusa- tives of the terminus ad quem, but with the Accusatives of Space (§ 138). This remark will be confirmed by similar uses of other Prepositions. 186. ] "AMI, TEP. 131 184.] The Genitive with éy¢i is found in two instances, — Tl. 16. 825 pdxecOov midaxos aud’ ddlyns fight over a small spring of water, Od. 8. 267 detdeww dd’ ”Apeos guddrytos k.T.A. Another example may perhaps lurk in— I. 2. 384 eb 5€ ris dpparos duis iduv «.7.d. if we read dug Fiddv (having looked over, seen to his chariot). With this mean- ing compare Il. 18. 254 dui para ppdeode; and for the construction the Attic use of mepropipar with a Gen.=to look round after, take thought about (Thue. 4. 124): also the Gen. with dugipdyeodar Il. 16. 496., 18. 20., 15. 391, , TEpL, 185.] The Preposition wepi (or épt, § 180) has in Homer the two meanings around and beyond. Both these meanings are common in the adverbial use; the second often yields the derivative meaning leyond measure, ex- ceedingly, as— Il. 16. 186 mépt peév Belew raxdy exceeding swift to run. 18. 549 76 34 mépt Oadua rérvkto which was an exceeding wonder. Od. 4. 722 mépt ydp pot edwxe for he has given to me beyond measure. The meaning Jeyond is found in Tmesis, I]. 12. 322 aédeuov mept tévde guydvtes escaping this war: Il. 19. 230 ToA€uoto Tept atvyepoio Almwvtat shall remain over from war: and in Composi- tion, wepleyu I excel, neprylyvoua I get beyond, surpass, meptowda I know exceeding well (Il. 13. 728 Bovdrj mepiidpevar dAdrAwv to be knowing in counsel beyond others ; cp. Od. 3. 244., 17. 317). The Gen, in such constructions is ablatival (§ 152). 186.] The Dative with rept (as with audi) is usually Locatival ; as Il. 1. 303 épwrjoes mepl doupl will gush over (lit. rownd upon) the spear ; 2. 416 yiTGva wept arnOecor datga to tear the chiton about (round on) the breast. ; Hence, when used of an object of contention, it means over ; as I]. 16. 568 epi maidi.. mévos ein the toil (of battle) might be over his son, ep. Il. 17. 4, 133, Od. 5. 310: and in a derivative sense, Od. 2. 245 paynoacbat rept dari to fight about a feast. 1, It is a question which meaning is to be given to mepi in— Il. 5. 566 wept ydp die morpere Aady (SO g. 433-) II. 566). 10. 240 gb5ercev 5 wepi favOG Meveddw. 17. 242 Sooo épA Kepadry mépe Seibsa (or mepidet6.a). and in the Compound (Il. 11. 508 7@ fa mepideccay, 15.123 nepideloaca Geoit, 21, 328., 23. 822). Most commentators here take arepl = exceedingly, and the Dat. of the person as a Dativus ethicus: wept ydp dic monet for he feared exceedingly Sor the shepherd, &. But it is difficult to find Homeric analogies for such a use K2 132 HOMERIC GRAMMAR. [187. of the Dative, and the meaning over, on behalf of, is supported by the Homeric Hymns (h. Mere. 236 ywdpevoy wep Boval, h. Cer. 77 dxvupévny rept madi), and later writers (Hdt. 3. 35 mepl éwu7d Seuaivovra, Thuc. 1. 60 Sebidres wept x.7.A.) ; also by the use of dpi with the Dat. (§ 182) in nearly the same meaning. 2. Much difficulty has been felt about the use of wept in a group of phrases of which the following are the chief instances :— Il. 4. 53 Tas damépoa, br’ dy Tor dwéxOwvran wept Ape (cp. 4. 46, &c.). Od. 6. 158 Keivos 8 ad rept «fjpt paxdpratos éfoxov ddAwy. Il. 21. 65 wept 8 40eA€ Oud (so 24. 236). 22. 70 dddvogovtes mept Oupd. Od. 14. 146 wept yap pw epider ral endero bvpd. Il. 16. 157 rotoiv re wept ppecty domeros GAKn. Od. 14. 433 mept yap ppeolv atoipa 76. In all these places the Dative may be construed as a Locative (although k7jpt without twepi is only found in Il. 9.117): the only question is whether the Prepo- sition is to be taken in the literal local sense round, all over, or in the derivative sense exceedingly. In favour of the latter it may be said that the same combina- tions of Preposition and Verb are found without a word such as kfjpt or up, where accordingly wept must mean exceedingly ; compare— IL. 13. 430 Tv Tept KApt pine maTIhp Od. 8. 63 Tov wept Moda’ épidnae Od. 14. 433 mept yap ppeaty aioipa 7'5y 2 88 wep! xépdea olde Il. 16. 157 wept ppeoty dann Od. 12. 279 mepi Tot pevos Od. 5. 36 mepl «fpr Gedy ds tipnoover } IL. 8. 161 epi pév oe Tiov Aavaoi. Again, in Il. 4. 46 rdwv por mepl «fjpe Te€oxero the meaning beyond is required by the Gen. rdwy; ep. 4. 257 wept wey oe Tiw Aavady TaxuTwrwr, On the other side, it must be considered that the representation of a feeling as something surrounding or covering the heart, midriff, &c. is common in Homer. Thus we have— Il. 11. 89 cirov re yAviepoio wept ppévas inepos aipe. Od. 9. 362 émel KixAwna rep) ppévas HAvOev olvos. So of a sound, Il. 10. 139 wept ppévas HAvO’ io (cp. Od. 17. 261), And more fre- quently with éudt; cp. Od. 19. 516 mumvat d€¢ por dud’ adivov mip dtetar percdaves dd5upopévny épéover ; and the other passages quoted at the end of § 181. Similarly mepi Kipt, wept pect, may have been meant in the literal sense,—the feeling (fear, anger, etc.) being thought of as filling or covering the heart. On the whole, how- ever, the evidence is against this view ;—unless indeed we explain wept kfjpt as » traditional phrase, used without a distinct sense of its original meaning. The occasional use of the Dat. with wept in Attic is probably due to familiarity with Homer. 187.] The Accusative with wepi is used (as with api) when motion ox extent im space is expressed: as Il. 1. 448 éxarduBnv €ornoav mepl Bwoudv placed the hecatomb round the altar; 2.750 tept Awdduny otk’ Oevto made their dwellings round Dodona. Generally speaking the Accusative implies surrounding in a less exact or complete way than the Dative. It makes us think of 189.| TIEPI, MAPA, 133 the space about an object rather than of its actual circumference. Occasionally, of course, the circumference is the space over which motion takes place, or extent is measured: as Il. 12. 297 pave paBdo.rr dinvexéow mepl kb«Aov round in a circle; Tl. 18. 274 éordpevat mep) totxov to stand along the wall all round it. 188.] The Genitive with wept is used in three distinct ways:— 1, With wep’ meaning deyond (in the figurative sense, =excel- ling) it expresses the object of comparison: Il. 1. 287 wept mdv- Twy éupevat to surpass all, Od. 1. 235 diotoy émoinoay Tepl ndvTwy have made him unseen more than all men, 4. 231 éntotduevos rept navtwv. This use is distinctively Homeric. 2. With wepi=round, over (in the local sense) the Gen. is very rare; the instances are— Od. 5. 68 173° adrod rerdvvato mept onelous yAadupoto Hpepts nBawca, 130 Tov pev eyov eodwoa Tept tTpdmlos BeBadra. The Gen. may be akin to the (partitive) Gen. of place (§ 149): the vine e.g. grew round in or over (but not covering) the cave. 3. With wepi=over (the object of a contest), as I]. 16.1 as of Mev Trepl vnds eVooéApoto pdyxovTo, 12.142 dudverOar Tept vndv to defend the ships ; sometimes also in the figurative sense, about, as Il, 11. 700 wept tpimodos yap euedAov OedoecOat, Od. 9. 423 date wept woxijs as when life is at stake; and of doubt, Il. 20. 17 7 te mept Tpdwv cat Axatay pepunpicers. The use with Verbs of anger and fear is closely akin; Il. 9. 449 maddAakidos mépt xdoato; 17. 240 véxvos Tépt deidia (unless we read repiydoaro, mepideldia). The weapons of the contest are said to be fought over in Od. 8. 225 eéplCecxov mept ré€wv; so Il. 15. 284 drmdre Kotpou epioceray wept pvOay. And this is also applied to the quarrel itself, Il. 16. 476 ovvirny Ep.dos mépt OvpoBdporo (cp. 20. 253). By a not unnatural extension, wept with the Gen. follows Verbs meaning to speak, know, &c., but only in the Odyssey; viz. 1. 135 (= 3.77) Wa piv mept matpds amouxouévoto porto; 15.347 el dye pot wept pntpos K.T.A.3 17. 563 olda yap ed epi Keivov; also 1. 405.. 7: 191., 16. 234., 17. 371., 19. 270. Note that the cor- responding use of dot with the Dat. is similarly peculiar to the Odyssey (§ 182). The origin of this group of constructions is not quite clear. It may be noted, however, that they answer for the most part to constructions of the Gen. without a Preposition; cp. dpdvec@a: mepl vndy and dpiveoba vndv; and again eimé wept pntpés, of6a wept xetvov, &c. with the examples given in § 151, d. Tapd. 189.] The Preposition mapé (rapai, by Apocope mép) means alongside. It is common in the adverbial use (see § 177), 134 HOMERIC GRAMMAR. [190. and also in Tmesis and Composition. Note the derivative meanings— (1) at hand, hence at command ; as Il. 9. 43 mdp tor ddds the way is open to you; Od. 9.125 od yap Kuxddrecot vees mapa. (2) aside; as Il. 11. 233 mapal bé of érpdmer’ éyxos the spear was turned to his side (instead of striking him). (3) hence figuratively, wapé wy’ ijrage cozened me ‘aside, away from my aim: and so sapremOdy changing the mind by per- suasion, napendv talking over, &c.; also, with a different metaphor, wrongly. (4) past, with Verbs of motion, as épyoua, erative, &c. 190.] With the Dative mapdé means Jeside, in the company of, near. It is applied in Homer to both persons and things (whereas in later Greek the Dat. with mapé is almost wholly confined to persons); thus we have mapa vyi, mapa vnvoi (very frequently), zap’ appact, mapa Boyd, Tap Toot, wapa crabue, &c. This Dat. is either Locatival or Instrumental: see § 144. It may be used after a Verb of motion (e. g. Il. 13. 617), see § 145, 4. 191.] The Accusative with mapé is commonly used— (1) when motion ends beside or near a person or thing: as II. 3. 406 ijoo map’ airov lodoa go and sit by him; Il. 7. 190 Tov pev map 70d’ édv yapddis Bare. Hence the use of the Acc. often implies motion: as Il. 11. 314 map ew’ israco place yourself beside me; Od. 1. 333 orf pa mapa orabu.ov came and stood beside the pillar; Il. 6. 433 Aadv be otjcov Tap epived. (2) of motion or extent alongside of a thing (esp. a coast, a river, a wall, &c.); Il. 1. 34 87 8 dxéwy mapa Oiva went along the shore; Od. 9. 46 7oAAG BF pra Eohaloy mapa Giva sacri- jiced many sheep along the shore; Il. 2. 522 map motapov évaov dwelt by the side of the river; Il. 3. 272 map Eipeos KovAedy dwpto hung beside the sword-scabbard. (3) of motion past a place; as Il. 11. 166, 167 of 88 map “IAov ona .. map’ épwwedv eaoetovro they sped past the tomb of lus, past the fig-tree; Il. 6. 42 mapa tpdxov e€exvdlsOn rolled out past the wheel. The derivative meaning beyond (=in excess of) is only found in Homer in the phrases rap dvvapw (II. 13. 787) and mapa poipay (Od. 14. 509): but cp. the Adj. mapatc.os against fate. Note that wapé is often found with an Acc. of the place near which a weapon has passed: e.g. Il. 5.146 kAnida map Gpyov trike struck the coltar-bone near the shoulder: Il. 16. 312 otra Odavra orépvov yuuvadérta nap’ donida passing the shield). 193.] TIAPA, META, 135 Comparing these different uses we see that the same Acc. may imply motion, either ending near an object or going past it, as the context may require. In the latter case the Acc. is one of space: just as wepi takes an Acc, of the space round which action takes place. 192.| With the Genitive mapé properly means sideways or aside from. As with the Dative, it is used of things as well as persons (whereas in later Greek it is practically restricted to persons). On the other hand, it is confined in Homer to the local sense ; thus it is found with Verbs meaning to go, bring, take, &e. not (as afterwards) with dxovw, wavOdve, oida, or the like. An apparent exception is— Tl. 11. 794 ef € teva gpeoiy jot Oeomponiny ddeeiver, kal rivd of map Znvds enéppade mérvia pyrnp, where however the notion of bringing a message is sufficiently prominent to explain the use. So Il. 11. 603 Pdcy£auevos mapa nds sending his voice from the ship, The original meaning sideways or at the side from is visible in some of the uses with a Gen. denoting a thing: as Il. 4. 468 wap’ domidos éfepadvOn appeared beyond (outside the shelter of) the shield: so probably Il. 4. 500 vidy TIpedpoo vdov Bade .. nap into dkedwy struck him (aiming) past the chariot. So too a sword is drawn rapa pnpot sideways from the thigh. The same meaning lies at the root of the frequent use of mapa in reference to the act of passing from one person to another (as in mapadidw and napadéxopa), hence of gifts, messages, &ec. It is usual to regard mapd with the Gen. as meaning from the side of, from beside, de chez. But this is contrary to the nature of a prepositional phrase. The Case-ending and the Stem must form a single notion, which the Preposition then modifies ; hence (e. g.) mapa pnpod means beside from-the-thigh, not from beside- the-thigh. This is especially clear where the Preposition is joined to a Verb; e.g. Od. 19. 187 maparAdyfaca Madedv driving-aside from-Maleae: and in— Il. 4. 97 Tod Kev 82 wdunpwra map’ dyAad Sapa Péporo the rhythm connects rapé with pépoo rather than with ro}—thou wilt bring-aside (= transfer) from-him. So with other Prepositions: dad Tpoins off from-Troy, not from off-Troy: kar’ obpavot down from-heaven, not from under-heaven. As to ia6 with the Gen.=from under, see § 204. peta. 193.] The Preposition perd in the adverbial use means mud- way, in the middle; e.g. with a Verb understood, Il. 2. 446 pera 88 «.7.A. and among them &c. Hence alternately, as Od. 15. 460 xptoeov Spuov éxav, peta d pr€xtpoow eepro strung with electrum between (the gold); so in succession, afterwards, as Od. 21. 231 mpatos éyd, peta 8 types I first and you in turn; Od. 15. 400 pera ydp Te kal ddyeoe répmerar dvijp=a man has his turn of being pleased even in the course of his sufferings. The notion of alternation appears in Compounds with perd, as peTapdddgw, peraorpépw: in Tmesis, Od. 12. 312 pera 8 dotpa 136 HOMERIC GRAMMAR. [194. BéBnxe the stars have changed their place. So peranavépevor (Il. 17. 373) means with turns or intervals of rest. 194.] With the Dative perd means between or (less exactly) among. The meaning Jetween is found in phrases such as pera xepoi, pera moocl, pera pect (on the double character of the dpeves cp. § 181); also, of two parties, per’ dudoréporor. The use in reference to several objects (among) is mostly restricted to persons, since it conveys the idea of association of units forming a group, &c. (whereas év and avy are more loca‘). Hence per’ dorpdcn (Il. 22. 28, 317) is said of a star among other stars (with a touch of personification): and in Il, 21.122 xeioo per’ ixOvor there is perhaps a sarcastic force—tlie there with the fish for company. Cp. also the phrase Od. 5. 224 mera kal rdd¢ Toiot yevérOw let this be as one among them. The expression in Il. 15. 118 pe6” aipare wal xovinar is equivalent to a Collective Noun, nearly =‘the wounded and the fallen’ So Il. 21. 503 perd orpoparryye xovins, a somewhat bolder phrase of the same kind. The construction of perd with the Dat. is in the main Ho- meric. It is occasionally imitated in later poetry. 195.| With the Accusative perd has the two meanings among and after. The meaning among is found after Verbs of motion with Plurals, and also with Collective Nouns, as ped’ éunyupiv, pel? dptoy ; SO pera detmvov to (join the company at) a feast, pera 7 70a kal vouov tnnav=to the pasture ground where other horses are. It occurs without a Verb of motion in Il. 2.143 maox pera TAnOvy to all among the multitude ; Il. 9. 54 werd mavras éphdiwas émdev dpiotos (so Od. 16. 419). And with a Singular in IL. 18. 552 Spdypara per’ dypor mirror the handfuls of corn fell in the middle of the furrow (between the ridges). Of the other meaning we may distinguish the varieties— (1) after, following ; Tl. 13. 513 énatgar pei? édy Bédos follow- ing his weapon, Od. 2. 406 yer txvia Baive Oevio. (2) after, in order to find (with a Verb of motion), as per’ ép’ navdes has come in search of me, Od. 1.184 és Tepéony pera XaAkov. (3) in succession to, neat to; rov d pera KT.r. and after him fe.; and in the derived sense of time, Il. 8. 289 mpére roe per Que to thee after myself; of rank, Il. 7. 228 ofoe. . per- €act kat per’ “AxtdArja even (in the second rank) after Achilles. 196.] With the Genitive perd occurs in five places (with a Plural Noun), in the meaning among or with— Tl. 13. 700 peta Bowrév eudxovto. 198. ] *ENIL. 137 Il. 21. 458 088% preO? udev meipa «.7.A. 24. 400 Tv péra TadAcuevos KANpw AdXOV. Od. to. 320 per’ GAdwv A€Eo Eralpw. 16. 140 peta dudav 7’ vt oixw ive K.T.A. Of these instances the first is in a passage probably inserted afterwards to glorify the Athenians; the second is in the Qcdv paxy, and therefore doubtful ; in the third we should perhaps write peraraAAépevos and construe of them cast- ing lots in turn I was chosen. But the last two indicate that the use had crept into colloquial language as early as the Odyssey, taking the place of ovv or &pa with the Dative. See § 221. emi. 197.] The Preposition émi means over, upon ; sometimes after (as we speak of following upon); with, at (i.e. close upon); in addition, besides, esp. of an addition made to correspond with or complete something else; also, attached to, as an inseparable in- cident or condition of a person or thing; and conversely, on the condition, in the circumstances, &c. Examples of these meanings in the adverbial use are :— Il. 1. 462 éni & aldora oivov AciBe poured wine over (the meat). 13. 799 mpd pev 7’ GAN’, adbrap én dAda in front—behind. Od. 1. 273 Oeoi & én pdprupor éotwy the gods be witnesses thereto. 5. 443 él oxénas jv dvéeuoro there was thereto (the place was furnished with) a shelter from the wind. Il. 18. 529 xreivov 8 énl pndroBoripas killed the shepherds with the sheep. I. 233 ent wéyav Spxoy duoducn I will swear in confirmation. With a Verb understood, ému=is present, is in the case, as Od. 2. 58 ob yap én’ duip there is no man here (for the purpose); Il. 1. 515 ov rou én d€os there is no fear with or for you (as part of your circumstances); I]. 21. 110 ému Tow Kal uot Oavaros death is my tot too (ep. 6. 357 olow ent Zebs Oijxe kaxdv pydpov). It is very much used in Composition. Note the meaning over in émi-mAéw to sail over, also én-olyouat to go over, review, ém- Twrdopat, et-addouat; besides, in emi-didmju, &e.; to (of bringing aid, joining, &c.) in én-apryw, én-adétw, ém-apapioxe, é7-adrdoow, &e. ; for, in émi-kAdIw to spin for (so as to attach to); hence of assent, émi-vetw, émi-TAqvat, émi-e(xw (with a general affirmative meaning, on as opposed to off, for as opposed to against). 198.] With the Dative én has the same group of meanings; note éspecially— (1) émt vnvot by the ships, én’ decor with the sheep (of a shep- herd), éml xredreoor with (in charge of) the possessions ; Il. 4. 235 émt wevdeoow ecoer’ dpwyds will be a helper with (on the side of ) falsehood (or fulse men, reading wevdéoor). 138 HOMERIC GRAMMAR. [199. (2) Il. 4. 258 dAdolo em) epy im (engaged upon) other work, So dredeuritw ent epyo with a work unfinished: so Il. 4. 178 én maor in all cases dealt with. (3) Od. 17. 454 od« dpa col y emi cldei Kal ppéves joav with JSorm thou hast not understanding too; Tl. 13. 485 Tao? ent Ove with this spirit (toc). (4) Od. 11. 548 10188 én’ d0Am with such a prize (when such a thing is prize); pcdG em pyro for fixed hire (given the hire, hence iz view of it). (5) én’ qpuare for the day, i.e. as the day’s work, in a single day. Note also that éwt meaning wpon very often takes the Dat. after Verbs of motion, as xaréxevev én’ ovde poured on to the ground: hence against, as éw ddAjAowow idvres, papvacOae én’ dv- dpaat, &e. 199.| With the Accusative éni implies (1) motion directed to a place, seldom (2) to a person; or (3) motion or (4) diffusion, extent, &c. over a space. 1. After Verbs of motion the Acc. does not (like the Dat.) distinctly express that the motion ¢erminates on the place: e.g. él xOdva-is merely ¢o or towards the ground, but éni x@ovi implies alighting on it, Cp. Tl. 18. 565 drapmrés tev én’ adtiy there was a path leading to it; Il. 9. 218 emt orqOos cvvoywxdre bent im over the chest. Hence the phrases expressing attitude, as émi ordua, él yodva, &e. Two forms, él defia and én dpiorepd, are used even when motion is not expressed; as Il. 5. 355 etpev éweita pdyns én dpiotepa Oodpov “Apna juevov. Note however that én’ dpicrepots and én’ dpiorepGv are metrically impossible. 2. The use with persons in the meaning towards, in quest of, is rare, and almost confined to the lliad: as 2.18 Ba 8 ap’ én’ ’Arpeidny “Ayapeuvova, tov 8 éxiyavey: also 5. 590., 10. 18, 54, 85, 150.) II. 343, 805., 12. 342., 13. 91, 459., 14. 24., 16. 535-5 a1. 348, Od. 5. 149. 3. The meaning over, with Verbs of motion, is very common ; én mévrov (idv, TA€wv, pedywv, &e.), emt yatav, emt yOdva, em Kd- para, &e. Also with Verbs of looking, as Il. 1. 350 épdwv én’ amelpova mévTov. Hence such phrases as émt otiyas, of troops &ce. moving in ranks, i.e. over or along certain lines: as Il. 3.113 tmmous épugay ént orixas: and so Od. 5, 248 ém o7dOpunv tOuve straightened along (hence by) the rule. So with Plural Nouns, Il. 14. 381 olydevor ent mdvtas going over them all, Od. 15. 492 TwokAd BporGv én) dore’ GAduevos ; and of a distribution, Od. 16. 385 daccduevor Kata poipay ed’ jyéas 2.é, equally, so as to go round, 201.] EI, ‘YIIO. 139 4. The instances in which extent (without motion) is implied are chiefly found in the Odyssey. Examples from the Iliad are: 9. 506 POdve dé re Tacay én aiav she is beforehand all the world over (sO 23. 742): 10. 213 KAéos ein mavtas én’ avOpdtous, 24. 202, 535. It will be seen that they are from books 9, 10, 24. Notice also the use with Neuters expressing quantity; as Il. 5.772 Téccov ém Ophoxover to such a distance they bound ; also én modddy a long way, ént toa to an equal extent ; and esp. the common phrase dcoy 7’ émi, see Il, 2. 616, &c. 5. OF time: Tl. 2. 299 petvar’ emi xpdvov wait for (lit. over) a time ; Od. 7, 288 eddov mavvdxtos Kal én’ 7G Kal peécov ijpap slept all night and on through morning and midday. 200.] The Genitive with émi is used in nearly the same sense as the Dative, but usually with less definitely local force; in particular— (1) with words expressing the great divisions of space, esp. when a contrast is involved (land and sea, &c.); as émt Xepaov, én Hmeipov, én dypoti; Od. 12. 27 7 adds 7 el ys adyjoere (cp. Il. 13. 565). This is evidently a Gen. of place, § 149. For the difference of Gen. and Dat. ep. II. 1. 485 én’ jrelpowo epvocay tot em Wauddos. (2) where the local relation is a familiar one; as él uns, én’ dajuys, ep tnmov, ent Opdvov, én’ obd05, emt mipyou, én’ dykGvos, etl pedins (epeoOe(s). Thus él vnvot means on or beside ships, énl yndv on board ships. (3) with Verbs of motion, upon (of the terminus ad quem), as Il. 3. 293 xaréOyxev emt xOovds; so bearing down on, as Il, 3. 6 mérovrat em’ "Qxeavoio podwr: Il. 5. 700 mporpémovto pe- Aavdoy emt vydv: Od. 3. 171 veoivefa vycvov ent Wupins taking the course by the island Psyria. So perhaps Il. 7.195 (ebxeabe) oryy ep’ tpelwv (keeping the words) to yourselves, (4) of time; én’ eipyyns (Il. 2. 797, &e.); ent mporépwv avOpa- mov (Il. 5. 637, &e.) Cp. the Gen. of Time, § 150. In later prose the Gen. is very common, and the uses become indistinguishable from those of the Dat. es uTro, 201.] The Preposition éné (also Swat) usually means beneath, as in Il. 2. 95 tnd 88 crevaxiCero yaia the earth groaned beneath (their tread). The original sense, however, seems to have been upwards, as in the Superlative ta-aros uppermost (cp. tyr aloft, in-tws facing upwards). On this view we can understand why éné is not applied (like xaré) to express downward motion. Hence, too, it is especially used of supporting a thing, as Il, 1. 140 HOMERIC GRAMMAR. [202. 486 ind 8 Epuara paxpa tTdyvocay: and on the same principle it expresses resistance to a motion (whereas xard implies yielding, going with the stream &c.); as Il. 5. 505 tnd 8 Eorpepov jvioyijes the drivers wheeled them up, i.e. to face (the Trojans): and so bm-avtidoas meeting face to face, ino-péve to stand against (as we say, up to); and with the derived notion of answering, in-acida to sing im correspondence, tro-xplvopar (= Att. droxplyvouat), t70- Badrho to take up (a speaker), in-axotw=to show that one hears (by answering or obeying). So too the Compounds ¢-op&, ta-dypros, imd-8pa, &c. do not express looking down, but looking upwards from under ; even in Il. 3. 217 ordoney imal dé iSeone Kara x9oves dppara mhtas it is the face that is bent downwards: cp. Il. 19. 17. From the notion of being immediately under is derived that of being moved Jy, i.e. of agency or cause. The transition may be seen in trro-elxw to give way (before), bno-tpéw &e.; so Il. 16. 333 breOepudvon was warmed by (the blood), 202.] With the Dative iwé is very common in the simple local meaning, wader. It is sometimes found with Verbs of motion, as Od. 4. 297 ddunt ta’ aidovon Oguevac; and even when motion rom is intended, in Il. 18. 244 @dvoav tp’ Gpuacw @kéas inmovs. In this case however we have to consider that dpydrwv is metri- cally impossible, The derived sense wnder the charge or power is found in such uses as II, 5. 231 td jvidx@ (of horses), 6.139 Zeds yap of tm oKintpy eddpacce, 6.171 OcGv tn’ dpdpor mouT7: also, with the notion of an effect produced (where the Gen. would therefore be rather more natural), i7d xepot (dapmfvar, Oavéew, &e.), bd doupi (rumels, &c.); Il. 13. 667 votow tn’ dpyadrten POicba, Od. 4. 295 tmvo tro yAukep@ tapnoépeOa: and often of persons, as Il. 5. 93 b1d Tvdeidy muKival Kkdoveovto Padayyes. 203.] The Accusative is used with 6wé (1) of motion to a point under, e. g.— Il. 2. 216 bd “IAtov HAGe came under (the walls of) Troy. 17. 309 Tov BaN b76 KAnida péonv : often with Verbs of striking, &e.). Also (2) of motion passing under, and hence of extent under: Od. 15. 349 et mov ert GSovow bm’ abyas jedloro i.e. anywhere that the sun shines (ep. tm’ 7@ 7 HéAtdy TeE—an equivalent phrase). Il. 2. 603 of & exov Apxadiny ind Kvddjvns épos. 3. 371 ayxe d€ piv ToAdKEoTOS twas Gradny bro Setpyy (i.e. passing under the throat). In one or two places it is applied to time: Il. 16. 202 mav6’ ind pyviOudr all the time that my anger lasted ; so perhaps Il. 22. 102 204. ] . “CIIO. I41 vbx@ tro rHvd ddoqv (but night is often regarded as a space of darkness). 204.] The Genitive with éné is found in two or three distinct uses :— (1) with the force of separation from: as Il. 17. 235 vexpov tr Aiavtos éptew from under Ajax; Od. 9. 463 tm’ dpvewod Avéunv: so Il. 19. 17 doce devov ind Prchdpov as ef oédas eLepaavber. In this use the Gen. stands for an Ablative, ep. § 152. Note however that originally dw6 with an Abl. probably meant upwards from : see § 192. (2) of place under, with contact (especially of a surface); as— Il. 8. 14 t26 yOovds eats BépeOpov. Od. 5. 346 ré8e kpidepvov tnd orépvoto tavtooat, Tl. 1. 501 dekirepn 8 dp’ bn’ dvOepedvos édodoa taking hold of him under the chin. : 4. 106 10 arépvoto Tuxjoas. 16, 375 ty 8 dedda oxidvad’ tnd vepéwy, i.e. seeming to reach the clouds (cp. 15. 625., 23. 874). These uses of the Gen. are evidently parallel to some of those discussed in § 149 and § 151; compare (¢.g.) b1d vepéwy with the Gen. of space within which (mediovo didnev, &e.), and im dvOepedvos Edodca with xdpns ere (§ 151 a) took by the hair. They are doubtless to be regarded (like the Gen. with éql, § 200) as varieties or developments of the Genitive of Place. As with the Dative, the notion wnder passes into— (3) the metaphorical (or half metaphorical) meaning wader the influence of, by the power of ; as Il. 3. 61 6s 7° etowy d1a dSoupds im davépos under the man’s hand ; Od..19.114 dperGau S€ Aaol im avrod under his rule ; and many similar uses. Cases may be noted in which the agency intended is indirect (where later writers would rather use 5:4 with an Acc.) :— Tl. 16. 590 fv par’ aviip aden mepdperos 7) ev d€0ro, He kal ey TorAdum dylwv tro Ovpopaicréwy, where it is=wnder the stress of an enemy (so 18. 220); Il. 23. 86 ctré pe... Hyayer tyerepdvd’ dvdpoxracins v0 duypijs by reason of a homicide (committed by me). As asound is said to be over or about (nept, dud) the person hearing, so he is wnder the sound: hence (e.g.) with a half meta- phorical meaning II. 15. 275 rév d€ 9 d76 laxfjs epdvn dis. So of other accompaniments, as Il. 18. 492 daldwv tm0 Aapmopevdov- in the light of blazing torches. Tt is not quite clear whether the Gen. with tné expressing a cause or agent is to be regarded as Ablatival or not. It is natural that the effect should be thought 142 HOMERIC GRAMMAR. [205. as proceeding from the agent: but on the other hand we have seen that a locatival Dat. with éw6 may express the same notion. Probably this use of the Gen. with té (which has no parallel in Latin) was developed when the Genitive and Abla- tive had ceased to be distinct, Tport., 205.] The Preposition mpoti (pds, wort) expresses attitude or direction towards an object. It is found in the adverbial use ; Od. 5. 255 mpos 8 dpa mdddtov toujcato he made a rudder to be put to (the raft); hence commonly im addition, besides—a use which remained in later Greek. It isa question whether mport and mori are originally the same word. The present text of Homer does not indicate any differ- ence of usage. 206.| With the Dative mpori means resting on, against, beside a thing: as Il. 4. 112 wort yaln dykdivas resting (the bow) against the ground: Od. 5. 329 mpds GAAnAnow Exovrar hold on to one another. With Verbs of motion it implies that the motion ends on or beside the object ; Od. 9. 459 Oewopevov mpos ovdei. The later meaning besides, in addition, is only found in Od. to. 68 dacdy wv erapot re Kaxol mpos Totot Te Umvos. 207.] With the Accusative mpori is very common, meaning towards: as mpos méAw towards the city (not necessarily reaching it), Il. 8. 364 kAateoxe mpds otpavdy cried out to heaven ; hence to, on to (mostly with Verbs of motion), as Od. 4. 42 é&kdwav mpods evra leaned against the walls: against (persons), as mpds daipova im opposition to a god ; also addressing (persons), with Verbs of speaking, &c.; in one place of time, Od. 17. 191 mori éo7epa to- wards evening. Note that the literal local sense appears in all the Homeric uses of mpott with the Acc.: the metaphorical uses, viz. in respect of, for the purpose of, in propor- tion to, according to, &c., are later. 208.| With the Genitive mpori expresses direction without the idea of motion towards or rest on the object: as Od. 13. 110 ai pev mpos Bopéao.. at 8 ad mpds vdrov t.e. not at or facing the north and south, but more generally, ix the direction fixed by north and south; Il. 10. 428-430 mpos pév GAds .. tpds OvuBpns ; Tl. 22. 198 mort ardédtos in the direction of Troy ; Od. 8. 29 7% mpos jolwy i) éoneplwv dvOpdrov (=from east or west). Among derived senses we may distinguish— (1) at the hand of, from (persons), as Il. 1. 160 rTipqv adpvdpevor mpos Tpdwv, 11. 831 ra oe mpori pac AyiAdjjos dediddy Oar. (2) on the part of, by the will of, as Tl. 1. 239 of re Oduoras 210.] IIPOTI, ’ANA, 143 mpos Ards eipvarar who uphold judgments on behalf of Zeus ; Tl. 6. 456 mpos GdAns iordv ipaivors at another's bidding. (3) Gefore, by (in oaths and entreaties); as Il. 13. 324 mpos matpos youvdcouat I entreat in the name of thy father. The Preposition here implies that the god or person sworn by is made a party to the act; cp. Od. 11. 66 viv d€ ve TOV dmEv youvacopar od Tapedvtwy, mpds 7’ GAdxou Kal Tarpds KTA. on the part of the absent ones I entreat &c. It will be seen that mpori with a Gen. is seldom used in the strictly local sense except when there is a contrast between two directions. Hence the use approaches closely to that of the Gen. of Place given in § 149 (2); compare (e.g.) mpds Bopéao —npos vérou with Od. 1. 24 of piv ducopevou “Lrepiovos of 8 dvidvros. The Case is accordingly ‘ quasi-partitive’ (i.e. true) Genitive, and has no ablatival character. ava, 209.] The Preposition dvd (dv) means up, upwards, up through. It is rarely used as a pure Adverb (the form dvw being preferred) except in the elliptical ava wy / But it has a derivative adverbial sense in I]. 18. 562 uédaves 8 Gva Bdrpves foav there were dark grapes throughout. Tmesis may be seen in Il. 2. 278 ava 8 6 mroAtropbos ‘Odvoceds orn, and in dva 8 éryero (aveoxero), &e. In Tmesis and Composition it sometimes expresses reverse action, as dva-Ajw. So dva-Baddo to put off. ava is seldom used with the Dative; the meaning is up on (a height of some kind), as Il. 1.15 xpucéw ava oxijntpy raised on a golden staff ; 15.152 ava Tapydp@: so 8. 441., 14. 352., 18.177. Od. 11. 128., 23. 275., 24. 8. This use is confined to Homer. With the Genitive it is only used in three places in the Odyssey (2. 416., 9.177., 15. 284), and only of going on board a ship (ava vyds Balvw). The meaning wp from is only found in Composition: dvébv modujs adds, &e. 210.] With the Accusative dvd means up along, up through, of motion or extent: ava dorv, dp mediov, ava dduara, dv’ 6ddv, dv ‘EAAdda, &e.; Il. 5. 74 av’ ddovras tnd yhdooay tdye xadxds the spear cut its way up through the teeth and wnder the tongue ; so éva. ordua, used literally (11. 16. 349., 22. 452, &e.), and also of words uttered, Il. 2. 250 Baordijas ava orgy’ exov having the kings passing through your mouth (i.e, talking freely of them); similarly ava Ovudv of thoughts rising in the mind, Note also the applica- tion to miwving, as Od. 4. 41 map 8 EBadov Cerds, ava € kpt hevKov Zuéav; ep. Od. 9. 209 (with the note in Merry and Riddell’s edition). This Accusative is evidently one of Space (§ 138). The use with collective Nouns, as dy’ Su:hov through the press, paxyv avd, ap gévov dy véxvas, &c, seems to be peculiar to the Iliad. 144 HOMERIC GRAMMAR. [2z1. The use in Il. 14. 80 dvd vi«ra may be explained either of ¢ime or of space: cp. tm6 vinta (§ 203) Sid vinra (§ 215). The meaning wp on, up to (of motion) may be traced in I]. ro. 466 Oey dvd prpixny; Od. 22.176 iov” dv’ driv éptoat draw (the cord) up to a high pillar; perhaps in the phrase dvd 0’ dppara toxin’ éBawov (Od. 3. 492, &c.). Kara. 211.|] The Preposition «ard (by Apocope «d8, &c.) means down, and is parallel in most uses to dvd, It is never purely adverbial (karw being used instead, cp. dvw), but is common in Tmesis, as Il. 1. 436 card 8& mpuprijor eyoav, 19. 334 Kara Tap- mav TeOvapev, &c., and in Composition. Besides the primary sense (seen in xat-dye to bring down, xata-vebw to nod downwards, le. 2 assent, &c.) it often has the meaning a// over, as kata-ewdw to clothe, xataxéw to pour over ; hence completely, as kata mavta gayeiv to eat all up, xata-ktelvw to kill outright: also in the place, as before, as xatadeinw to leave where it was, &e. xaté is not used with the Dative. If such a use ever existed it was superseded by éwé (just as dvé with the Dat. gave way to ént), The possibility of the combination may be seen from the phrases car’ aird@1, car’ atc. 212.| With the Accusative card means down along, down through, as kata poov down stream; ep. 1. 16. 349 ava ardpa Kat kata pivas (of blood). But it is very often used (like dvd) of motion that is not upward or downward, except from some arbitrary point of view; as xa’ é6ddv along the way, kata mTdAw through the city, &e.: again, cata dpéva cal cata Oupdv in mind and spirit. Other varieties of use are :— (1) with collective Nouns (chiefly in the Iliad), as xara orpardv through the camp, méheyov Kara, kata kAdvov, &e. (2) with Plurals (less common), as kar’ attovs going among them, kar’ avOpamovs aAdAnoOat. (3) of the character or general description of an action, as cata mpngw on a piece of business, kara xpéos, kata Anida (all in the Odyssey). (4) to express place; esp. of wounds, e.g. kar’ Gpov about (some- where on) the shoulder, Cp. IL 1. 484 tkovto kara otpatdév arrived opposite (within the space adjoining) the camp (so Od. 5: 441). (5) to express agreement (from the notion of falling in with), in the phrases xara Oupdy, xara kdopov, kata poipay, kar aicay. 215.| KATA, AIA, 145 (6) distributively : as Il. 2. 99 éprjrudev 88 Kad? &8pas in their several seats ; and so, in 2. 362 xara pida kara ppirpas. (7) xara oféas (udxera) by themselves (to the extent consti- tuted by themselves): so Il. 1. 271 kar’ éw airdv. 213.] With the Genitive xaré has two chief meanings: (1) down Jrom ; as kat’ obpavod down from heaven, xa® trmwv adro leaped from the chariot. This Genitive is clearly Abla- tival in origin. (2) down om (in, over, &e.): as Il. 3. 217 xar&d xOovds Supara mas fiaing his eyes on the ground; xara 8 dpBadpav Kéxvr’ axhis a mist was shed over his eyes ; kata yains down in the earth, Comparing the similar uses of él (§ 200), tm (§ 204, 2), and mport (§ 208), we can hardly doubt that the Genitive in this latter group is originally akin to the Genitives of Place (§ 149). Sud. 214.] The Preposition 8:4 seems to mean properly apart, in twain. It is not used freely as an Adverb; but the original sense appears in the combinations diampd, diaymepés, and in Tmesis and Composition, as d:a-orfvat to stand apart, dia-Téyve to cut asunder ; dia xtijow dSatéovro divided the possession, From the notion of going through it means thoroughly, as in b.a-TépOw to sack utterly. In several Compounds, as dra-rdpve, dt-aipéw, b1a-ddrT@, the notion of division is given by the Preposition to the Verb; e.g. dia-Tduve to separate by cutting, &e. 215.| The Accusative with 84 is often used to denote the space through which motion takes place: as— Il. 1. 600 814 Sépata Tomvdorra bustling through the palace (so dua orréos, Sia BHooas, dia pwmijia, &c.). 14. 91 pdOov dv ob Kev dnp ye dia oTdéua mdymay ayouTo (=with which a man would not sully his mouth: cp. ava oropa, § 210). Od. 9. 400 dxeov ev omijecor b0 dxpias dwelt im caves about (scattered through) the headlands. So Il. 2. 40 bid xparepds topivas lasting through hard fights: and 6d véxra (chiefly in the Odyssey, and books 10 and 24 of the Iliad). This use is distinctively Homeric. Sometimes also 84 with the Ace. is used in Homer to express cause or agency; as Il. 3. 73 i dud pavtootynv (Calchas led the army) by virtue of his soothsaying ; Od. 8. 520 5:4 peydOvpov ’AOjvny (to conquer) by the help of L 146 HOMERIC GRAMMAR. [216. Athene; so Il. 10. 497., 15. 41, 71., Od. 8. 82., 11. 276, 282, 437, 13. 121., 19. 154, 523. These places do not show the later distinction between by means of and by reason of. 216.] The Genitive with 8 implies passing through some- thing in order to get beyond it; esp. getting through an 0- stacle: as— . Tl. 4.135 did per dp Cworipos éAjdaro. So of a gate, Il. 3. 263 dua Skadv exov tmmovs: and of lower and upper air, &c. 8c’ aépos aldép’ tkavev, bv’ aidépos ovpavov tke, me- diovde Sia vepéwy. Again did mpopdyor, 8: duldov &c. of making way through the press. The Acc. is used where we expect this Gen. in Il. 7. 247 & 8& Sid mrvyas HAGE went through six folds: but this may be partly due to the metrical impossibility of wruxav. Conversely, in Il. 10.185 8s re nad vAnv Epxnrar & dpeogi the Acc. would be right, and dpeodr is probably a false archaism: cp. § 158. itp, 217.] The Preposition éwép (or Smeip) means higher, hence over, beyond. It is not found in the adverbial use, or in Tmesis, or with a Dative. In Composition éwép expresses going across or beyond, hence excess, violation of limits, &c. 218.] With the Accusative énép is used— (x) of motion or extent over a space, as Il. 23. 227 dmelp Gada kiévatat jas. This use is not common; II. 12. 289., 24. 13, Od. 3. 68., 4.172., 9. 254, 260, (2) of motion passing over an object: as Il. 5.16 tntp doy Gptorepoy HAVO dxwny; Od. 7.135 vaéep oddv éBijcero. (3) metaphorically, ix eacess of, in violation of: tntp atcav, vmép potpav, trép Spxia: also, somewhat differently, Il. 17. 327 dTép Oedy in spite of God. 219.| With the Genitive énép is used both of position and of motion over an object, esp. at some distance from it; as orf 8’ dp tnép ceadijs; Il. 15. 382 vnds daép rolywy (of a wave com- ing’) over the sides of a ship: Il. 23. 327 dcov 7 épyw’ vnép ains a Sathom’s length above ground. Metaphorically it means over so as to protect, hence in defence of, on behalf of; as Il. 7. 449 Telxos eretyfocavto vedv trep; Tl. 1. 444 ExarduBnv peta: dmep Aavady. So IL. 6. 524 86° saép aébev aloxe akovw when I listen to reproaches on your account (of which I bear the brunt). In respect of form éaép (for bép:, Sanser. updri) is a Comparative of éé; cp. 222, | "ENI, SYN, EIS. 147 the Superlative traros, and the Lat. superus, summus. Bence the Gen. is pro- bably Ablatival, like the Gen. with words of comparison; see § 152. évi, _ 220.) The Preposition évi (also «ivi, etv, év) means within, in 5 it is used adverbially (as Il. 5. 740 év 8? épis, ev 8 GAxr} &e.), in Tmesis (as év 7’ dpa of $8 xepl), and with a (locatival) Dative. Notice, as departures from the strict local sense, the uses— (1) with Plurals denoting persons (=yerd among), as év bpiv (Il. 9. 121, 528., 10. 445), ev maou (Od. 2. 194., 16. 378), evi opiat (Il. 23. 703). (2) with abstract words (rare in the Iliad); év rdvrecot wévoict (Il. 10. 245, 279), év mdvreco’ épyowor (1). 23. 671), év ddyeou (Il. 24. 568); Oadiy gu (Ll. 9. 143, 285), év vymen (Il. 9. 491); €v giddrnte; ev poipy aright (Tl. 19. 186), aton év dp- yaden (Il. 22, 61), év Kapds aion (Il. 9. 378); ev d& if typi (11. 9. 319). These two uses are nearly confined in the Iliad to books 9, 10, 23, 24. ou. 221.] The Preposition ov (or &v) means in company, with. It is not used as a pure Adverb, but is found in Tmesis, as I]. 1. 579 ov & hiv daira tapagy and disturb (cvvrapdccw) our feast. It is used with an Instrumental Dative (§ 144). To express equally with, or at the same time as, Homer uses épa with a Dat.; while ov» commonly means attended by, with the help of, &e. Hence ovv évreor with armour on, abv vynval in ships, ay dpx@ on oath, abv’ AOnvy aided by Athene: so Il. 4. 161 avy te Beyddw américay they pay with a great price. The use of ovv with the Dative has been recently shown by Tycho Mommsen to be confined, generally speaking, to poetry. The Attic prose writers (with the singular exception of Xenophon) use pet&; while the practice of the poets varies, from Homer, who hardly ever uses pera with the Gen., down to Euripides, who uses it about half as often as owv. It is evident that in post-Homeric times peta with the Gen. became established in the ordinary colloquial language, while ovv with the Dat. was retained as a piece of poetical style, chiefly through the influ- ence of Homer, but gradually gave way to living usage. Thus otv became an ‘Inseparable Preposition’ (like dis- and ambi- in Latin). See T. Mommsen’s dissertation Mera, avy und dpa bei den Epikern (Frankfurt am Main, 1874). eis, 2.2.2.] The Preposition «is (or és) expresses motion Zo or into. It is not used adverbially (the Adverb being efow), and seldom in Tmesis. The motion is sometimes implied: as Tl. 15. 275 ebdvn dis jiyéveos els 6ddv: 16. 574 és IInaiy’ ixérevoe (came as suppliant). L2 148 HOMERIC GRAMMAR. [223. Of time ; és }édvov karaddvra to sun-set; so és ri how long? «is 6 until: Od. 14. 384 és Opos 7) és dmdpyy as late as summer or autumn. Metaphorical uses: Il. 2. 379 ei 5€ mor’ és ye pilav Bovrcdoouen if we take counsel to one purpose; Il, 9. 102 elmety els dyaddy to speak to good effect (so 11. 789., 23. 305). éé. 223.| The Preposition ég (or éx) usually expresses motion owt Jrom an object. It is not used purely adverbially, but there are many examples of Tmesis: as é€ épov évro, éx d€ of Hvioxos mANYN gpévas his charioteer lost (lit. was struck out of) his wits. With the idea of motion implied :— Il. 13. 301 x Opyxns Equpous peta Oupyocecbov armed them- selves to come from Thrace after the Ephyri. 14. 129 évOa 8° énest’ adrol pev exducda dyioriros éx Bedgov hold back from fighting (going) out of range: cp. 16. 122, 678,, 18. 152. And with an abstract word, Il, 10. 107 éx ydAouv dpyadéoto pera- oTpéyyn idov irop. So of direction; Tl. 14.153 “Hpn 8 eloeibde .. atac’ é€ OvdAdy- novo standing (looking) from Olympus; Od. 21. 420 (drew the bow) adrddev éx dippoto Kabhevos from the chair as he sat; Il. 19. 375 br dy é« mévrow cédas vavtnor pavinn when a meteor appears to sailors at sea (seeing it from the sea): of choosing out of, Il. 15. 680 éx moAddwv miocvpas tuvaelpera: inmovs, and similarly, Il. 18. 431 800° éuol é« macéwy Kpovidns Zeds adye’ ebwxe to me (taken JSrom, hence) more than all. eé is also used of an agent as the source of action; as Il. 5. 384 trier. . €€ dvdpdv have endured at the hands of men; cp. Il. 22. 280., Od. 7. 70., 9. 512. The meaning in consequence of (a thing) occurs in Il. 9. 566 é& dpéwy pntpos xexoAwpévos, and in the Odyssey (3. 135., 5. 468, &c.). Of t2me: ex toto from that time, é& apyis from the first (Od. 1. 188, &c.), é« vedrnros (Il. 14. 86). Note also: Il. 10. 68 marpddev ex yevefs dvopdtwy calling them by the father's name according to family ; Tl. 9. 343 (486) ex Ovyod Jrom the heart, heartily (but Il. 23. 595 é« Ovpod mecgew to fall away from a person's favour). and. 224.] The Preposition dé means off, away, at a distance from. It is not used adverbially, but is common in Tmesis; as I], 8. 108 ots mor’ dm Aivelay éAdpnv which I took from Aineas. In Composition it generally gives the Verb the notion of separating ; 226.] TIPO, ’ANTI. 149 €. g. dzo-Kdnrw is not to hew at a distance, but to separate by hew- ing: so dnodtw, atoBdAAw, dnorAovo, anoppHyvums, anoxantw, and several others—all used in Tmesis. Hence we must explain II. 19. 254 amd rplyas dpfdyevos cutting hair as an anwapyxy, or first offering ; cp. Od. 3. 446., 14. 422. Sometimes it has the force of restoration or return, as in d70-diSaput, ato-vorréw (ep. ay backwards). So dmo-evneiv means either to speak out or to forbid, refuse. With the Genitive ané generally expresses motion away from, not implying previous place within the object (whereas é« means proceeding from). It is also used of position, as Il. 8. 16 décor ovpavds éor’ dno yains as far as heaven is from earth ; Od. 1. 49 Pirwy dro miyata tacxes suffers woes far from his friends ; meta- phorically, Tl. 1. 562 dad Ovjod paddov euol evear you will be the more out of favour with me; and ddgys away from expectation. The Gen, is clearly Ablatival. mpd. 225.] The Preposition mpé means forward, in front. It is seldom used as an Adverb ; Il. 13. 799 mp0 pév 7’ GAN’, KrA.; TIL. 16. 188 éfdyaye mpd pdwode brought forth to the light: and of time, Il. 1. 70 apd 7° édvra the past. In one or two other instances we may recognise either the free adverbial use or Tmesis: I]. 1. 195 TMpo yap HKe, I. 442 mpd pw emeuwe, Od. 1. 37 apd of elromev. Traces of a use of mpé with the Locative may be seen in thé phrases otpavd6. mpd in the face of heaven, \A1d6. mpd in front of Troy, and (perhaps in the temporal sense) 746: mpd before dawn. In these cases the meaning is to the front in, hence immediately before. With a Genitive, on the other hand, mpd means ix front with respect to, in advance of ; hence, in a more or less metaphorical sense, in defence of, as Il. 8.57 apd te maidwv Kai mpd yuvatkav. The Case is here the Ablatival Gen. (as with dxép and words of comparison), But in Il. 4. 382 mpd 6500 éyévovro the Gen. is partitive, got forward on the way ; and so perhaps Il, 16. 667 mp0 poBoro for- ward in the flight, i.e. having betaken themselves to flight (so Diintzer a. /.). The temporal sense is rare in Homer; Od. 15. 524., 17. 476 mpd yduouo before marriage ; Il. 10. 224 Kal te mpo 6 Tob évdnoe one thinks of a thing before another. , avtt. 226.| The word dvri can hardly be shown to be a true Prepo- sition in Homer. The only Compound appears to be dvri-pépeodar 150 HOMERIC GRAMMAR. [227. to oppose (Il. 1. 589., 5. 701., 22. 482., Od. 16. 238): for the Verbs dvriBodéw to meet and dvtitopéw to pierce are probably derived from the Nouns dvri-Bodos, avti-ropos: also in Il. 8.163 we should read yuvaixos dp’ dvri térv§o, not dvrerervgo (ep. Od. 8. 546 dvtl xaovyvijrov Leivds 0 ixérns te réruatar), and for dvri- oxea0e hold up against (Od. 22.74) avr’ toxeade (2. €. dvta icxeode, ep. Od. 1. 334 dvta maperdwy cxouevn Armapa kpydepva). avzi also resembles the Improper Prepositions (esp. the Adverbs dvra, dvtiov, &c.) in being used with the Gen., but not with the Dat. or Acc, It means in place of, hence im the character of, equivalent to: as Il, 21. 75 dvti rot eiu’ ixérao, Double Prepositions. 227.] It is characteristic of Homer to form a species of com- pound by combining two Prepositions. We have— dpi wept, like our rownd about: also mepl 7’ dui te round and about: used adverbially, as Il. 22. 10 dydau 8’ dudl mepl peydr’ taxov ; in Composition, dugureprotpaga (Il. 8. 348), &e. wapeé out besides, out along, out past: adverbial in Od. 14. 168 &dAa wape€ pepvdpueda: with the Acc., mapeé dra alongside the sea, nape tiv vicov past the island ; mapéx vdov beyond (=contrary to) reason: with the Gen., mapeé dd00 aside from the way. ine, with a Gen. away from under, as Il. 13. 89 pevEecOa trex KQKOU. S.€é, with a Gen. right through, as d1ex mpoOdpov, dik eydpoto. dwonpé quite away, used adverbially and with a Gen. Siampd right through, adverbially and with a Gen. mepumpd round about; Il. 11.180 mepimpo yap eyxeé Ove. In all these instances the meaning and construction are mainly determined by the first of the two Prepositions (so that e.g. wapéé is used nearly as wapd, Si€§ and Siampé as 8d, &c.). The second does little more than add some emphasis. The treble Preposition trexmpé is found in Composition: iaexmpobéw, imex- mpopéw, &c. The sense is represented by dividing the words trex-mpobéw, &c. A curious variety is found in the Compound mpo-mpoxvarvddpevos rolling forward before, where a second awpé is added to give emphasis to the first. Improper Prepositions. 228.] The term ‘Improper Preposition’ may be applied to any Adverb used to govern a Case. The following are some of the most important words of the kind :— Used with a Genitive: dyy: near, close to, éyy'6., éyyts near, dvra, dvriov, &e. facing, mpdobe(v) before, maporde(v) in front of, Omoe(v) behind, weconyds between, évtds, évtocde, évdodev within, 229. | ATTIC USES OF PREPOSITIONS. I51 &w out, exrds, exrob., éxroabe(v) outside, evepOe beneath, aver, dvevde(v). apart from, without, drep without, vdogr away From, éxds, éxdrepOe(v) apart from, peopa until, wépnv beyond, madw back from, avrixpt straight to, ids straight towards, thre, thdO. Sar off, iTaba under, etvexa (évexa) on account of, exnri by the favour of. The Gen. with some of these words may be Ablatival (§ 152). In general, however, it appears to be used with little or no refer- ence to the meaning of the governing Adverb, and merely in order to connect the two words. Hence these constructions are best brought under the general rule that a Noun governs the Genitive (§ 147). With a Dative: Gua together with, plyda in company with, spas in like manner, duis takes a Gen. in the meaning aside from (Il. 8. 444., 23. 393, Od. 14. 352). It is also found with the Acc. in the same sense as dud, in the phrase Oeol Kpdvov dypis édvres (see also Il. 11. 633, 747, Od. 6. 266); and once with a Dat., viz. in Il. 5. 723 cvdnpéw Aor dudis. eiow generally takes an Accusative, as "IAtov elow to Ilium: but a Gen. in Od. 8. 290 6 8 elow depatos fe went inside the house (not merely ¢o the house). The word ds is used to govern an Accusative in one place in Homer, viz. Od. 17. 218 os aiel tov Guotov dyes Oeds ws Tov duotov God ever brings like to like. There is no evidence that as was used in Composition: hence it cannot be counted as a proper Preposition. Note the frequency of Compounds formed by one of these words following a Preposition: év-ayra, elo-avta, dv-avra, xaT-avra, nép-ayra, év-aytiov, kaT-ev-ayTiov : Ep-mpogbev, mpo-mdpode, pet-dmiabev, an-dvevOev, dn-arepbev, and-voogi, in-évepbe, Kat-avtinpd, Op. dv-dvya, dt-apmepés, xat-avTd0, &c. These are not true Com- pounds (otvOera), but are formed by wapdbeots, or mere juxta-position : i.e. they do not consist of two members, of which the first is wholly employed in limiting or qualifying the second, but of two adverbial words qualifying the same Verb. Thus they are essentially akin to the combinations formed by a Preposition and its Case; see § 178. Homeric and Attic uses of Prepositions. 229.| The development of the language between the Homeric and the Attic period is especially shown in the uses of Preposi- tions. It may be convenient here to bring together some of the chief points. 1. Most of the Prepositions,—but esp. dpi, wept, wapd, emt, bn6, mpori, évi—are used in Homer adverbially (7. ¢. as distinct words). Afterwards they become mere unaccented words or prefixes. 2. A variety of the same process shows itself in the disuse of 15% HOMERIC GRAMMAR. [229. Tmesis. Besides the Prepositions already mentioned, this applies to petd, dvd, katd, Sid, ex, dad, eis. In these processes of development we have seen that the loss of independent meaning is accompanied by a change (which is in all probability simply a oss) of accent. 3. The construction with the Dative (which is mostly loca- tival) is the one in which the Preposition retains most nearly its own ‘adverbial’ meaning—so much so that it is often doubtful whether the Preposition can be said to ‘govern’ the Case at all. Accordingly we find that this construction is comparatively rare in Attic. It is virtually lost (except as a poetical survival) with audi, wepi, werd, dvd, and ouv. 4. On the other hand the Genitive is more frequent in Attic, and not confined (as it generally is in Homer) to uses in which it has either an ablatival or a quasi-partitive sense. Thus it is used with dpdoi, wepi, and perd: also with 8d of motion through. In such uses as these the Case ceases to have a distinct meaning : it merely serves (as with the Improper Prepositions) to show that the Noun is governed by the Preposition. 5. The development of meaning is chiefly seen in the exten- sion from the literal sense of place to various derivative or metaphorical senses. Some of these senses are beginning to be used in the Homeric language: ¢.9. époi with the Dat.=about, concerning ; wept with the Gen. (probably also the Dat.) in the same meaning ; mapé with the Acc.=7 excess of, in violation of ; petd with the Acc.=after; émi with the Acc.=towards (a person): Sid with the Acc. =owing to: ék=in consequence of. Others may safely be counted as post-Homeric; note in partieular— nepi with the Acc.=adout, nearly (of time and number); also= concerning, in relation to: napdé with the Dat.=in the opinion of; with the Ace.=during the continuance of ; also compared with: katé with the Acc.=azswering to; also during the time of: with the Gen.=about, against : ént with the Dat.=iz the power of: with many phrases such as 8 dpyijs, ava xpdros, mpds Blay, ex Tod eudavois, &e. 6. There are slight but perceptible differences between the usage of the Iliad and that of the Odyssey (§ 182, 188, 196, 199, 215). Some uses, again, are peculiar to one or two books of the Thad, esp. 9, 10, 23, 24: see §§ 199 (4), 220, 223 (jin.). 231.] INFINITIVE AND PARTICIPLE. 153 CHAPTER X. Tur Versat Nouns. Introductory. 230.] The preceding chapters deal with the Simple Sentence : that is to say, the Sentence which consists of a single Verb, and the subordinate or qualifying words (Case-forms, Adverbs, Pre- positions) construed with it (§ 131). We have now to consider how this type is enlarged by means of the Verbal Nouns. The Infinitive and Participle, as has been explained (§ 84), are in fact Nouns: the Infinitive is an abstract Noun denoting the action of the Verb, the Participle a concrete Noun expressing that action as an attribute. They are termed ‘ Verbal’ because they suggest or imply a predication, such as a finite Verb ex- presses (¢. 9. épxerat c&ywy adtovs implies the assertion aye. abrovs), and because the words which depend upon or qualify them are construed with them as with Verbs (dywy adrovs, not dywv abrav bringer of them). Thus they have the character of subordinate Verbs, ‘governed’ by the finite Verb of the sentence, and serving at the same time as centres of dependent Clauses. The distinction between Infinitives and other abstract Substantives, and again between Participles and other primitive Adjectives, was probably not always so clearly drawn as it is in Greek. The Infinitives of the oldest Sanscrit hardly form a distinct group of words; they are abstract Nouns of various formation, used in several different Cases, and would hardly have been classed apart from other Case-forms if they had not been recognised as the precursors of the later more developed Infinitive. The Participles, too, are variously formed in Sanscrit, and moreover they are not the only Nouns with which the construction is ‘ad- verbial’ instead of being ‘adnominal.’ The peculiarity of the Verbal Nouns in point of meaning may be said to consist in the temporary and accidental character of the actions or attributes which they express. Thus mpatrew and mpata suggest a particular doing, momentary or progressive, at or during a time fixed by the context; whereas mpafis means doing, irrespective of time; mpdxrwp one who does, generally or permanently, a doer ; and so in other cases. It may be added that the distinction is especially important for Homer. In the later language there are uses of the Infinitive and Participle in which they lose the Verbal element, and have the character of ordinary Nouns; e.g. 7d mpdrrew is nearly equivalent to mpgs, of mparrovres to mpaxTtopes, &c. The Infinitive. 231.] Form and original meaning. In form the Greek In- finitive is the Dative of an abstract Verbal Noun (omen actionis): 154 HOMERIC GRAMMAR. [231. hence it properly expresses an action ¢o which that of the govern- ing Verb is directed—viz. a purpose, effect, bearing, &c. of the main action. Thus Sépev-o1 fo give, being the Dative of a Stem So-pev- giving, means ‘to or for giving,’ hence in order to give, so as to give, &e. This meaning evidently accounts for the common construc- tions of the Infinitive with Verbs expressing wish, command, power, expectation, beginning, and the like: as é0€Ao ddpevar lit. Lam willing for giving, dévapa idew I have power for seeing, &e. In Homer it may be said to be the usual meaning of the Infini- tive. It is found in a great many simple phrases, such as Luvénce padxerOa urged together to fight (so that they fought), dos dyew give for leading away (to be led away), oid vojjrar knows (has sense) to perceive, Bi 8 idvar stepped to go (=took his way, ep. yotvar’ évoua pevyéuevat) 5 mpoenxe TvdécOa, Téune véecOat, OpTo mérecOat, dvepou txavov ajuevat, &e. Cp. also— Tl. 1. 22 éemevdrjpnoay ’Axatol, aideioOar xrd. the Greeks uttered approving cries for (to the effect of) respecting, Sc. 3 80 2. 290 ddvpovrat oikdvde veetOa. %. 107 ’Ayapueuvorr Actme Hopfvat, ToAATow vycowwt Kai “Apyet navtt dvdcoew left (the sceptre) to Agamemnon to bear, thére- with to rule over, §e. Od. 4. 634 eue 8% xped yiyverar adrijs "HALd’ és edpdxopov diaB7- pevar L have need of it for crossing over to Elis. The notion of purpose often passes into that of adaptation, possibility, necessity, &e.; e.g.— Il. 6. 227 woAdol pev yap euot Tpdes .. krelvew there are many Trojans for me to kill (whom I may hill); ep. 9. 688 «ict cat olde Tad’ elneuev these too are here to tell this, 11. 342 éyyvs écav mpopvyeiy were near for escaping, to escape with, 13. 98 eiSerar jHuap tnd Tp&ecot dapqra the day is come for being subdued (when we must be subdued) by the Trojans ; ep. Od. 2. 284. Again, from the notion of direction or effect the Infinitive shades off into that of reference, sphere of action, &c.; as Il. 5. 6o1 olov 5) Oavyddouev “Extopa Siov alxpytiy 7 ewevar Ktd, for being a warrior ; dpotedecke mdxeoOar was best for (i.e. in) fight- ing, ebxerat iva boasts for (of) being. In the passages quoted the Infinitive is so far an abstract Noun that the action which it denotes is not predicated of an agent. The agent, if there is one in the speaker’s mind, is not given by the form of the sentence; e.g. éyyts éoav mpopu- yelv (were near for escaping) might mean were near so as to escape or (as the context of Il, 11. 342 requires) were near so that he could escape; Sivat émetys- Hevos would usually mean eager to set, but in Od. 13. 30 it means eager for (the sun’s) setting. Hence the apparently harsh change of subject in such a case as— 232.| THE INFINITIVE. 155 Od. 2. 226 Kat of idv év vnvoly émérperev olkoy &mavra meiGeaOai re yépovTi kal Euneda wavra pudragoev to the intent that it should obey the old man and he should guard all surely (lit. Sor obeying—for guarding). And so in Il. g. 230 év dof 88 cawodpev 7) dmodoOar vijas, where vfjas is first Object, then Subject. The harshness disappears when we understand that the abstract use is the prevailing one in Homer. It may also be noticed here that— (1) With Verbs of privative meaning, the Infinitive may be used as with the corresponding affirmative words: as éppuy’ avtu- Bodjoau shudders as to (from) meeting ; Od. 9. 468 ava 8 ddpior vetov éxdotw Kralew I nodded backwards to each for weeping (=for- bidding him to weep). But the proper use also appears, as in II. 22. 5 avtod petvar endédnoe fettered so that he remained. Here the context must determine the meaning. (2) With gpovéw, diw, &e. the Infinitive may express the effect or conclusion: I think to the effect—, hence I think fit ; as Tl. 9. 608 dpovéw 8% terypHoae I claim to be honoured ; 13. 263 od yap dlw . . moAcuitew I have no mind to &c. So eineiv to speak to the intent that, to bid, as Od. 3. 427 einare 8 dow dpywjow .. méverOat. In this use, as was observed by Mr. Riddell (Dig. § 83), the ‘dictative force ’— the notion of thinking right, advising, &¢.—comes through the Infinitive to the governing Verb, not vice versa. The same remark holds of the use with gov it is possible, 232.| Infinitive with Nouns, &c. It will be useful to bring together instances in which the Infinitive depends upon some qualifying word—Preposition, Adverb, Adjective, &c.—construed with the Verb :— IL. 1. 258 ot rept pv Bovddyy Aavady mept & éore pdxecOar excel in fighting. 1. 589 dpyad€os yap Oddvpmios dvtipéperOas is hard to set oneself against ; cp. 20. 131. 4. 510 émel ob ot AlOos xpws od8e oldnpos Xadkdv dvaryxéoBat since their flesh is not stone or iron for withstanding (so as to be able to withstand) bronze. 8. 2.23 if p’ ev pecodre éoxe yeywrepev Gudorépuce for shouting =so that one could shout) both ways. 13. 775 énet rot Ovpids dvairioy airidacdar since your mind is for blaming (is such that you must blame) the innocent. Od. 17. 20 ob yap énl orabpoion pévery ere TyArtwos eiul I am not of an age to remain. ; 17. 347 aldas 8 odk dyabh Kexpnuer dvdpi mapeivar shame is not good to be at the side of a needy man (is not a good ‘ backer’ for &c.). : 21, 195 moto «’ eft’ "Odvore dzuveper ef roBev Edo; i.e. how would you behave in regard to fighting for Ulysses ? 156 HOMERIC GRAMMAR. [233. Od. 2. 60 huets 8° of vd tr rotor duvveuev may be either we are not like him, so as to defend, or simply we are not fit to defend. The construction of the Inf. is the same in either case: the difference is whether rofo: means ‘of the kind’ with reference to olos ’Odvoceds Zoxe or to the Inf. duvvéuer. The latter may be defended by Od. 17. 20 (quoted above). This construction is extended to some Nouns that are not used as predicates; as Oelew rayds swift to run, Batya idécOa a wonder to behold (cp. the use of the Accusative with Adjectives, § 131 fin.). 233.] Impersonal Verbs. The Infinitive is used with €or there is (means, room, occasion, Sc.), tore it is fit, wémpwtar it is determined, eipapto it was fated. For €or. cp.— Tl. 14. 313 xeloe ev gore cal borepov dpynOjvar. Od. 15. 392% aide de vires AOéoparou ort pev ede, gor. 5& repropevoiow axovewv there is (enough) for sleeping and for listening. It is very common with a negative: otk éo71, ob} tws gort, &e. meaning there is no way, it may not be that Se. The Impersonal use is also found in phrases of the two kinds noticed in § 162, 43 viz.— (a) With a Neuter Adjective; as dpyadéov b¢ pol eort OécOar kta, it is difficult for me to make Sc. ; podpiyov d€ ot ear’ ad€éacBar it is fated for him to escape ; so with aioxpdv, venerontov, atciovr, dpx.ov, BéATepov, and the like. (2) With an abstract Noun: as— IL. 14. 80 od ydp tis vépeois puydey Kaxdv there is no wrong am escaping il. Od. 5. 345 001 tot poip’ early ddvEa it is thy fate to Se. II. 330 GdAG Kal apy ebdeuv there is a time for Sec. So with aica, pdpos, Oduis, xped, dvdynn, aldds, d€éos, eAmoph, &e. followed by an Infinitive to express what the fate, need, shame, &c. brings about, or in what it consists. These examples throw light on two much-debated passages : Il. 2. 291 9 phy Kal révos éoriv dvinbévta véeo@a verily there is toil for a man to return in vexation, i.e. ‘I admit that the toil is enough to provoke any one to return.” Thus understood, the expression is a slightly bold use of the form of sentence that we have in apn éotly ebSev, potpa éoriv ddvat, Oupds éoriv dvaitiov altidacba, &c. The other interpretation, ‘it is toil to return vexed,’ though apparently easier, is not really more Homeric; and it certainly does not fit the context so well. Il. 7. 238 of8° emt defid, of8 én’ dpiorepa vapjoa Rov dfadény, 76 pou ort Tadavpivov trode pice I know how to turn my shield of seasoned ox-hide to the right and to the left, wherefore I have that wherewith to war in stout-shielded fashion (=I have a good 234.] THE INFINITIVE. 157 claim to the title of raAatpwos moAcurorijs, elsewhere an epithet of Ares). Here €or is used as in Zorw ebbew, &e. 234.] Infinitive as apparent Subject, &c. In the Imper- sonal uses the Infinitive appears to stand as Subject to the Verb ; dpyadgov éort Oér0ar=to make is hard: ob pkv yap te Kaxdv Bact- Aevewev to be a king is not a bad thing. This construction how- ever is not consistent with the original character either of the Impersonal Verbs (§ 161), or of the Infinitive. I¢ is plain, too, that éorw etdeu can never have meant ‘sleeping is,’ but ‘there is (room &e.) for sleeping’: and so dpyadéov éort Oé00ai is originally, and in Homer, it (the case, state of things, 8c.) is hard as to making. It is only in later Greek that we have the form dpyadéov éart 76 Oéo0at, in which 6éc6a is an indeclinable Neuter Noun. The process by which the Infinitive, from being a word of Jimitation, comes to be in sense the Subject of the principal Clause, admits of being illustrated from sentences of various forms :— (1) With a personal Subject; ¢.y. in— I. 5. 750 rhs emrérpanrat péyas otpavds OtAvytds Te Huey dvaxdivas muKivoy védos 7d’ emOeivar : the meaning ‘to them is entrusted the opening and shutting of the thick cloud of heaven,’ is expressed by saying ‘to them heaven is entrusted for opening and shutting the cloud,’ Il. 1. 107 aie rot ra Kan’ eort pita pect pavreverOa. 4. 345 év0a pir’ dnradréa xpéa euevar. The meaning (aimed at rather than expressed) is, you love to prophesy evils (to eat roast flesh, §c.). (2) The Impersonal form (dpyaAgov éort) only differs from the other in the vagueness of the Subject, which makes it easier for the Infinitive to become the Subject in sense, while it is gram- matically a word limiting the vague unexpressed Subject. It follows from what has been said that the form dpyadéos éorly dvtipépecOar is quite as natural and primitive as the Im- personal form, and is not to be derived from it ‘by Attraction,’ The use of a Neuter Pronoun as Subject (e. g. 76 ye xaddv dxovépev the thing is good, to listen) may be regarded as a link between the personal and impersonal forms of expression: cp. § 161 note. Similarly (3) an Infinitive following the Odject of a Verb may become the logical Object ; as— IL. 4. 247 7} pévere Todas oxeddv eXOeyev; do ye wait for the Trojans for their coming on? i.e. ‘for the coming on of the Trojans.’ 14. 342 “Hon, pjré Oey 7d ye deldiOe pajre Tw’ avdpav oerbat do not fear any one of gods or of men for thew being about to see, i.e. ‘that any one will see’: ep. Od. 22. 39. 158 HOMERIC GRAMMAR. [235. A further development of this use leads, as we shall see, to the ‘ Accusative with the Infinitive.’ As a parallel to the Infinitive used with Impersonal Verbs it may be shown (4) that the Infinitive sometimes takes the place of a vague unexpressed Object. Thus oid vofica: means knows (enough) to perceive: the full construction being such as we have in Il. 2. 213 ds fp’ rea pecty jow deoopd Te TOAAG TE HON . . Ept- Cépevat who knew (had a store of) words wherewith to wrangle. So too didwpt with an Infinitive is originally construed as Od. 8. 44 7 ydp fa Ocds mép dOxev Godin répmev: Il, 11, 20 rév more of Kivipyns d@xe fevniov efvar; thence it comes to mean ‘to give (such a state of things) that some event shall happen,’ z. e. to grant the happening ; as 80s ricac@a: grant that I may punish. In such a pas- sage as Il. 3. 322 rdv dds dwopOiuevoy Siva: “TA. we may take 7éy with 6dés or as an Acc. with the Inf. diva. A Neuter Pronoun, too, may serve as a vague Object, explained by an Infini- tive; e.g. ll. 5. 665-6 76 péy ob Ts éweppaoar’. . éteptoat. The Infinitive may also (5) be equivalent in sense to the Genitive depending on a Noun; as— I. 7. 409 ob yap ms peda verdov natateOvnwTav ylyver’ émet xe Odvewor Tupds perdtooépev Gra é.e. ‘there is no grudging about the appeasing of the dead.’ Hence is developed an idiomatic use of the Genitive parallel to the Accusativus de quo: see Shilleto on Thue. 1. 61, 1, 2.35.| With Relatives. It is remarkable that the use of the Infinitive with ds, dare, otos, doos, &c. is rare in Homer. The familiar construction of écre only occurs twice: Il. 9. 42 énéo- acuta. woTe véecOar is eager to return, and Od. 17. 20 ov yap ént otabuotor pevew ére tHAIkos elul, Gor emTerdapevw . . TLOécOaL. The other instances are: Od. 21.173 rotov—oidv re uevar such a one as to be; Od. 5. 484 Sccov te . . Epvada so far as to shelter ; Od. 19. 160 dvijp ofds Te wddtota olxov KiSecOat, 21. 117 olds 7’. . avereaOa, 236.] With mpiv and mépos. This use is common in Homer: as Il. 1. 98 apiv y ano rarpl Pirw ddueva before they give back to her father &e.; 11. 573 mwdpos xpda Aevxdv emavpeiv before touching the white flesh. The origin of this singularly isolated construction must evidently be sought in the period when the Infinitive was an abstract Noun; so that (e.g.) mply dépuevae meant before the giving. The difficulty is that a word like piv would be con- strued with the Ablative, not the Dative: as in fact we find Ablatives used as Infinitives in Sanscrit with purd ‘before’ (Whitney, § 983). It may be conjec- tured that the Dative Infinitive in Greek was substituted in this construction for an Ablative. Such a substitution might take place when the character of the Infinitive as a Case-form had become obscured. 237.| Accusative with the Infinitive. Along with the use of the Infinitive as an abstract Noun, we find in Homer the 237.] ACCUSATIVE WITH INFINITIVE. 159 later use by which it is in sense the Verb of a dependent Clause, the Subject of the Clause being in the Accusative. _In the examples of the Acc. with the Infinitive we may dis- tinguish the following varieties or stages of the idiom :— _ 1. The Ace. has a grammatical construction with the govern- ing Verb: e.g.— Il. 1. 313 Aaods 8 ’Arpeldys dtodvpalvecbar dvoye Agamemnon ee the people to purify themselves (=that they should purify). 5: 601 otov 87) Oavuddouev “Extopa diov alyunriy r Euevat «rr. (for being a warrior, how he was a warrior). This might be called the natural Acc. with the Infinitive. 2. The Acc. has not a sufficient construction with the Verb alone, but may be used if it is accompanied by an Infinitive of the thing or fact: e.g.— Bovdow’ eyo adv cGv eupevar I wish the people to be safe (the safety of the people). otvex’ Gxovoe teiperdar TpGas because he heard of the Trojans being hard pressed. TO ov veneriCow’ "Axarods doyadday wherefore I do not think it a shame in the Greeks to chafe. In this construction the logical Object is the fact or action given by the Infinitive, to which the Acc. furnishes a Subject or agent, and thus turns it from an abstract Noun to a virtual pre- dication (so that e.g. reipecOar Tpéas is virtually =r érefpovro Tpées). It is found with Verbs that usually take only a ‘Cog- nate Acc.’ (Neuter Pronoun, &c.), as npi, eimov, dxovw, muvOdvo- pat, oida, dio, ppovéw, eOé\w, BovAopa, ~rTopar, veweo(Couat, pbovéw, &e. The construction is the result of two distinct influences : 1. The tendency to shift the weight of meaning from the grammatical Object to an epexegetic or limiting word. This has been fully illustrated in § 234. 2. The principle of the Accusativus de quo, which seems to be that an Acc. of the object spoken, thought or felt about may be used when the thing said, &c, is also expressed in any way, viz.— (a) by an Adverb: ¢% eimety to say good of. (6) by ‘Apposition,’ with implied predication (§ 168); Weddds nev paiper we should call it false: émiOovto pera Tpweoow “Apna heard of Ares as among the Trojans.* * There is no need here to suppose an ellipse of etvar. On the contrary, the Infinitive has sometimes the appearance of being added epexegetically to an implied predicate of this kind. #.g.— Il. 5. 639 Gan’ oféy tevd. pact Biny “Hpaxdneinv éupevat (what they call him as to being). ep. 2. 249., 17. 27.5 21. 463, §70. Od. 1. 233, 377-5 6. 43., 17. 416. 160 HOMERIC GRAMMAR. [238. (c) by a Participle: as Il. 7,129 rods viv ef mr@acovras ip’ "Exropt mayras dxovoa if he were to hear of their shrinking ; see § 245. (d) by a distinct clause (the common Ace. de quo); as Il. 3. 192 etm’ dye por nal révde .. Satis 05 eae tell me of him—who he is. 3. The Acc. has no construction except as the Subject of the Infinitive. This Ace. is found— (1) after Impersonal Verbs (§ 162, 4): as— Il. 18. 329 duo yap rémpwrat suoiny yatav épeticas at is fated for both to Se. 19. 182 od pev ydp Te veuecontoy Bactdja dvdp’ dmapéocacba it is no shame that a king should Se. (2) after mpiv and mdpos; as mplv éAdeiv vias Axadv before the Greeks came, ndpos tdde eépya yevecOar before these things came to pass. This may be called the purely idiomatic Acc. with the Infini- tive. It has evidently been formed on the analogy of the older varieties. 238.] Tenses of the Infinitive. So long as the Infinitive is merely a Verbal Noun, it does not express anything about the time of the action as Past, Present, or Future. But when it is virtually a predication, the idea of time comes in; ¢.g.— Tl. 5. 639 Gd ofdy tuvd pact Binv “HpaxAneinv éupevar (=‘ what they say he was’). 14. 454 od pay adr’ diw . . Gdtoy mndHoat AKovta, GAG tis "Apyeiov Kduuce xpot. The Future Infinitive has the Tense-meaning of the corre- sponding Indicative. It is generally used with étw, édwopo, and other Verbs denoting expectation; also with pé\o, when it means to de about to. The notion of Future time, however, may also be given by an Aorist or Present Infinitive :-— Il. 12. 407 émeé of Ovpds eéArero Kddos apéoOat. Od. 2. 373 dpocov pi) pntpl pidrn Tade pvdjoacat, It is true that the reading is not always certain; e.g. in the last quoted line there was an ancient v.l. puOqoerOar. Accordingly Cobet (Misc. Crit. p. 328 ff.) corrects most of the places where a Future can be substituted for an Aorist or Present: reading in Il. 3. 112 €Amopevor mavocoOar, 3. 366 F 7 ebdpnv ticerOar, and so in IL. 22. 120, 235., 23. 773, Od. 13. 173 (where however the Future mean- ing is not needed), 20, 121: also in Il. 16. 830 kepatkéuev for kepatfewev. But no similar correction can be made in Il, 13. 666-8, Od. 15. 214. The only example of an Inf. representing an Optative is— IL. 9. 684 wai 8 av rots dAAoww py TapapvOjcacda which is the report of the speech (v. 417) kal 8 av . . mapayvdn- oaluny. 240.) CASES WITH THE INFINITIVE. 16f 239.] Dative with the Infinitive. The construction in— aicxpov yap téde y éotl Kal éocopévoror TvdécOat is idiomatic (as compared with opdiv dds dyew, &e.), because the meaning is, not ‘is shameful for future men,’ but ‘is shameful for the hearing of future men’; that is, mvéc0a is the only word which is logically a Dative. The principle is evidently the same as has been pointed out in the case of the Subject and direct Object of the Verb (§ 234). Because the action of the Infini- tive stands in a Dative relation to the governing Verb, the agent or Subject of the action is put in the Dative. A similar ‘ Attraction’ is seen in the Latin double Datives; indeed the Homeric écoopévoror muéoOar can be translated literally into Latin (erit posteris auditut), 240.] Predicative Nouns—‘ Attraction.’ Corresponding to the Nominative in the Predicate (§ 162), an Infinitival Clause may have a Predicative Accusative, in agreement with the (expressed or understood) Subject of the Clause: as Il. 4. 341 opaiv pév 7 eréoixe era mpdroow edvtas Eatduev it becomes you that you should stand among the foremost ; Il. 8. 192 rhs viv KA€os otpavov tke. macav xpuoeinv evevar whose fame reaches heaven (to the effect) that it is all of gold. Or there may be ‘ Attraction’; é.¢. the Subject of the Infini- tive may follow the construction which it has in the principal Clause: as— Il. 1. 76 Kal pou duoocoy 7 pév por tpddpav . . apigev. 12. 337 od Tos of Env Badcartt yeywvelv, The general difference between these two constructions is that a Noun or Par- ticiple in the Accusative is closely connected with the Infinitive, so as to become emphatically part of the predication; whereas a Participle in the Dative may denote something prior to the Infinitive (a condition or reason). Thus— IL. 1. 543 alet rot pidoy éorly éped dnd véoguw édvta npunrabia ppovéovra diralepuer, ‘you like to decide apart from me,’—whereas éév7: would give the meaning ‘ you like when you are apart from me to decide.’ IL. 15.57. €A6n nal efnpor Mooeddwm dvaxre Tavadpevoy ToAguoo TA & mpds SHpal? inécOae ‘to cease from war and come’; not ‘ when he has ceased to come.’ Od. 4. 210 &s viv Nécrop: b&xe dapmepes jpara wévra, abrév piv Aurapais ynpacnépev KTA, But with a Dat.— Tl. 6. 410 épol 8€ xe népdiov ein ced dpapaprovcy xOdva Svpevat ‘it were better for me, if (or when) I lose thee, to go down to the grave.’ 15. 496 ov of derxts duvvopevy mepl marpys Tebvdpev to die when fighting for his country. In other places the Dat. is perhaps used merely as the easier construction: e.g. Tl. 8. 218 émt ppeot 07K’ "Ayapéuvon réta "Hpy aitO momvicayte . . drpivat. When the Subject of the Infinitive is also Subject of the governing Verb the M # 162 HOMERIC GRAMMAR. [241. Nominative is generally used: as Il. 1. 76 (quoted above), I. 415., 4. 101-3., 8. 498, &c. An exception is— Od. 9. 224 év8 éud udv mpiotia6’ rapo: AisoovT’ énéeoot, Tupay alvupévous iévar aA that they might take of the cheeses and so go back. The so-called ‘Attraction, it should be observed, is in reality the simpler and older construction. The idiomatic Acc. with the Infinitive is evidently used mainly to give the Infinitive its later character as a predication. 241.] Infinitive as an Imperative. This use is very fre- quent in Homer, but is chiefly found after an Imperative, so that the Infinitive serves to carry on the command already given :— Tl. 1. 32% epxecOov xdtolnv ’Ayapeuvovos ’Atpeidao xelpos Eddvr’ ayduev Boronida. 2. 8-10 Bdor’ 10... dyopevewev Os eTITEAAM. 3. 459 exdore, Kal Typ dmotiveuev. Od. 4. 415 kal rér’ ere? tiv pedérw kdpros te Bly Te, avo. 8 yew xra. (cp. v. 419, 422 ff.). Or after a Future, to express what the person addressed is to do as his part in a set of acts :— IL. 22. 259 vexpov Ayatoiow dow madi, ds dé od peCew. Od. 4. 408 eivdow E€elns' ob 8 ev Kpivacbat Eraipous. So after a clause which leads up to a command; IlL.-r11. 788 GAN ot pdoda (Achilles is the mightier) but do you advise him well: 17. 691., 20. 335. Compare also, Il. 4. 53 ras dianépoa destroy them (as you say you wish, V. 40): 10.65 ab0. weve (answer to the question am I to remain here?): 5.124 Oapoéov viv.. payeoOa (in answer to a prayer) without fear now you may fight. The use for the Third Person is rare: in a command, Il. 6. 86-92 eine 8 Emerta pnrépe of Kal Quy 7 88 .. Oetvar xrA.3 7. 79 cpa be otkad’ eudv dduevar wddw (let him take my arms) but give back my body ; 80 17.155. 23. 247: ina prayer, with a Subject in the Accusative,— Tl, 2. 412 Zed xddiore, peyote, xedawwedées, alps vaiwy, pay mply én’ jédAcov Sbvat Ktdr. (cp. 3. 285., 7. 179). Od. 17. 354 Zeb ava, Tyrcuaxdv por ev avdpacow ddBoy evar. An Infinitive of wish is used with the Subject in the Nom., once of the Second Person, and once of the First Person : Od. 7. 311 al yap Zed re marep kab AOnvaty xat”AmoAdov Tolos éwy olds eoot, Ta TE Hpovewy AT eye TEP, matod 7 euny éxéuev kal Guds yauBpos xadrgecOat, 24. 376 al yap . . ofos Nijpixov efAov . . Toios ewy . . epeota- pevat kal audveuy, The force of the Infinitive in all these uses seems to be that of an indirect 2.42. | ORIGIN OF THE INFINITIVE. I 63 Imperative. The command is given as something following on an expressed or implied state of things. Thus we may connect the idiom with the use of the Infinitive to imply fitness, obligation, &c. (§ 231); compare eict Kal olde rdd° cimépev these are here to say this, with nal 88 od eiwéwevar it is your part to say. There is a similar use of the Infinitive in Sanscrit (Whitney, § 982, ¢, d; compare the predicative use of the Sanscrit Dative, ibid. § 287). 242. | Origin and history of the Infinitive. That the Greek Infinitive was originally the Dative of an abstract Noun is proved by comparison with Sanscrit. ‘In the Veda and Brahmana a number of verbal nouns, nomina ac- tionis, in various of their cases, are used in constructions which assimilate them to the infinitive of other languages—although, were it not for these other later and more developed and pronounced infinitives, the constructions in question might pass as ordinary case-constructions of a somewhat peculiar kind’ (Whitney’s Sanscrit Grammar, § 969). In the Veda these Infinitives, or Case-forms on the way to become Infinitives (werdende Infinitive, Delbr.), are mostly Datives, ex- pressing end or purpose, and several of them are identical in formation with Greek Infinitives; as ddvane Sobvar (Sof evar), vidmane Fibpevar, -dhyai -cOar,* -ase -oat. In Greek however,—the Dative Ending -a: not being otherwise preserved, and the ‘true Dative’ construction not applied to things (§ 143),—these forms stand quite apart from the Case-system, and consequently have ceased to be felt as real Case-forms. In other words, the Greek Infinitive is a survival from a period when the Dative of purpose or consequence was one of the ordinary con- structions of the language. In Latin, again, this Dative (which includes the Predicative Dative) is common enough, and often answers in meaning to the Greek Infinitive; compare (e.g.) dpy éorly etdev with munitioni tempus relin- quere (Roby, § 1156), duuvew eiol nal dAdAoe with auwxilio esse, &c. The limitations which Mr. Roby points out in the use of the ‘ Predicative Dative’ (Pref. p. xxvi ff.) indicate that it had been employed more freely at an earlier period. The reten- tion of the construction in Latin is connected, on the one hand with the fact that the Latin Dative is a ‘true Dative,’ on the other hand with the comparatively small use that is made in Latin of the Infinitive of purpose. Similarly it is curious that in classical Sanscrit the Dative of purpose &c. is extremely common, but the Dative Infinitives have gone entirely out of use (Whitney, § 287 and § 986)—a result of the ‘struggle for existence’ which precisely reverses the state of things in Greek. The development of the original abstract Infinitive into the Infinitival Clause which we find in Greek and Latin may be traced chiefly under two heads; (1) the construction of the ‘ Accusative with the Infinitive,’ by which the predication of the Infinitive is provided with an expressed Subject (§ 237): and (2) the system of Tenses of the Infinitive, which is gradually completed by the creation of new forms,—esp. the form of the Future Infinitive, peculiar to Greek,—and by the use of the Present Infinitive as equivalent in meaning to the Present and Im- perfect Indicative. Other forms appear to have been added in the post-Homeric language (§ 85,1), and the Infinitive came to be used as an equivalent, not only for the Indicative, but also for other Moods. The use of the Infinitive as an indeclinable Noun is subsequent to Homer; it became possible with the later use of the Article. Some of the conditions, however, out of which it grew may be traced in Homeric language. The first of these was * So Delbriick and others; but see Max Miiller’s Chips, Vol. IV. p. 58. M2 164 HOMERIC GRAMMAR. [243. the complete separation of the Infinitive from the Case-system; the use of a single Case only as an Infinitive, and that in a construction (Dat. of consequence) which had become obsolete with other Nouns: to which we may add the unlikeness of the Infinitive form to any Datives retained in the Declensions. Thus the Infini- tive ceased to be felt as a Case-form, and could be used in parallel construction to the Nom. or Acc.: as— Il. 2. 453 rotor 8 dap wédAEpos yAviion yéver’ He veecOat. 7. 203 8és vinny Alavi kal dyAady ebyxos dpéoba. Again, an Infinitive following a Neuter Pronoun (Nom. or Acc.), and expressing the logical Subject or Object, might easily come to be regarded as in ‘ Apposition’ to the Pronoun: as— Od. 1. 370 éwel 76 ye naddy drovéper early dodod. II, 358 wat xe 7d Bovdoipny, nat Kev TOAD Képdiov etn, mAeorepn adv xept pidny és narpid’ ixéoOar where the Article is as independent as in Od. 4.655 dAAd 76 Oavydw tov err. The only instance which looks like the Attic use is Od. 20. 52 dvin al 7d puddo- aew, where however the juxtaposition of the Article and Infinitive may be acci- dental. The use of the Infinitive with an Article in the Gen. or Dat. is wholly post-Homeric., The Participle. 243.] Uses of the Participle. Following out the view of the Participle as a Verbal Adjective, we may distinguish the follow- ing uses: ; 1. The Participle is often used as an ordinary Adjective quali- fying a Noun; as Oeol aléy édvres, Bporol otrov edovres, 7801 Toth Toixov apnpdres, odkos Tervypéevov, and the like. In one or two cases it is Substantival: as rd yap yepas éotl Oavdvtwy, Woyal eldwdra KapdvTov. In this use—which may he called attributive—the Participle is a simple quali- fying word in relation to the Noun with which it agrees, while it is construed with dependent words like a Verb (aity édvres, otrov gdovres, &c.), A few Par- ticiples have lost their Verbal character altogether: esp. odAdpevos, dvfpevos, txpevos, dopevos (§ 86); also the Substantival pédovres rulers, révovres muscles, dyeiBovres rafters, aidovea a portico; perhaps yépwy, Sepdnwv, potoa, The word xpetww retains a trace of the Verb in edpd xpeiwy widely ruling. 2. Much more frequently, the Participle qualifies or forms part of the predication (§ 162): e.g. in such combinations as— dvacryitny eptaavre parted having quarrelled eippovéwy ayopijcato spoke with good thought the Participle has the same construction as the Adjective in maXtvopsos anéorn, or mpdppwv rérynxas (§ 162, 2). Thus it serves to express a predication which the speaker wishes to sub- ordinate in some way to that of the governing Verb. 3. Finally, a Participle construed in ‘ Apposition’ to a Noun im an oblique Case may imply a predication (§ 168); as xamvéy 243.| THE PARTICIPLE. 165 aroOpdoKovta vonoat to desery the smoke rising (i.e. when it rises, or that it rises, &e.). The two last uses are generally comprehended under the term predicative. The first of them, since it connects the Participle with the Verb of the sentence, is a variety of the Secondary Predicate (§ 162); the second answers to the Tertiary Predicate (§ 168). / Note also that— 1, A Participle of this kind often has the character of a distinct Clause, coming at the end of a sentence, and after a metrical pause: as— I. 4. 420 devdv 8 eBpaxe yarxds émt oriOeoow dvaxros épvupevon (as he roused himself). Od, 23. 205 ds pazo, ris 3 abrod Avo youvara Kat pidov Frop onpar’ dvarwotons (when she recognised the token). 2, Not unfrequently the Noun or Pronoun which should stand as subject to the Participle is understood : I. 2. 153 diy 8 obpavdy Ixev | otxade iepevew: 80 12. 339, 13. 291, 498, 15. 689. 5. 162 méprios 72 Bods EvAoxov nara Bockopevdwy a heifer or cow (of those) that are feeding in a thicket. 5. 665 76 per of Tis eweppdaar’ od’ évdnoe | pnpod etepvoa ddpu peldwor, bpp’ émBain, | omevddvrwv no one... (of them) in their haste: cp.15. 450 76 of ob Tis eéptxakey iepevwy Tep. 18, 246 dp0av 3 écradraw dyoph yéver’ an assembly was held upstanding (of them standing up). Od. 17. 489 TyAépaxos 8 év wey npadin péya wévOos dege | BXnuévou for his having been wounded ; so Il. 13. 417, &e. So with the Dative; Il. 12, 374 émevyouévosor & txovro came as a relief (to them) when they were hard. pressed: Od, 5.152 wareiBero && yAunis aiwy véaror ddupo- pera. 3. The understood Subject may be indefinite ; as IL. 2. 291 mévos early dvinbevta véecOat for a man to return vexed: cp. the phrase baov Te yéywve Bojoas. 4. The Participle is sometimes found in a different Case from a preceding Pronoun with which it might have been construed. Thus we have a Gen. after a preceding Dat.— i Tl. 14. 25 Ade 5é oge meph xpot xadxds areipns vuocopévew (governed by xpot instead of agreeing with ou). 16. 531 Srre of Gk’ Heovoe péyas beds cigapévoro (with jkovce instead of ot). Od. 9. 256 ds epaé’, juiy 8 adre narexrdaOn pidrov Hrop beccavrwy. So after an Acc., Il. 20. 413 Tdv Bade... vra napatocovros wounded him... in the back as he darted past; Od. 4. 646 7 o¢ Bin déxovtos dmnupa, Again, a Dat. is used after another Case, in— IL. 10. 187 ds T&v vndvpos tnvos dnd Brepdpoiiv dAwWAE vUKTA prdagcopevotgt KAKY : 80 14. 141-3. Od. 17. 555 PeTaAAToal Ti E Ovpos apo réce KéAerae Kal Knded wep wenabvip. We need not consider these as instances of ‘ Anacolouthon’ or change of the construction. The Participle, as we saw from note 2, does not need a preceding Pronoun: it may therefore have a construction independent of such a Pronoun, 166 HOMERIC GRAMMAR. [244. It is clear, in most cases at least, that the poet has chosen the simpler or more convenient of two equally admissible constructions of the Participle. 244.| Tenses of the Participle. The distinction between the Present and Aorist Participle has already been touched upon in §§ 76-77, and the meaning of the Perfect Participle in § 28. It may be remarked here, as a point of difference between the two kinds of Verbal Noun, that the Aorist Participle generally represents an action as past at the time given by the Verb: e.g. ds eimdv kar’ dp’ Eero having thus spoken he sat down. This however is not from the Participle itself conveying any notion of past time, but from its relation to the Verb (as a subordinate predication). When a Participle is not predicative, the notion of time is usually absent: e.g. in the phrase és #éAvov xatadvyra to sun-set ; so Il. 13. 37 Opp’ Eumedov abOe pévotev vooTn- cayra dvaxra that they might await the return of their lord. The Future Participle is used predicatively with Verbs of motion; 7\@e Avoduevos came to ransom, kadéovo’ te went to call, jy’ €mukoupjoovta, énédpape Tevxea ovdjowy, &c. The exceptions to this rule are— (1) €coduevos future, in Il. 1. 70 ra 7 écodueva mpd 7” edvta things future and past ; 2.119 Kal €ocopevotor TuOécOat. (2) émBnoduevos, in Il. 5. 46 (16. 343) Ww trmwy emPyocduevor, 23.379 alel yap dippou EmPnoouevotow etxtyv., But see § 41. (3) Od. 11. 608 aiel Badéovre oixds. (4) Il. 18. 309 kal re xravéovra Karéxra, see § 86. 245.| Implied Predication. Where the Participle is pre- dicative, we often find the Noun or Pronoun taking the place in the construction of the whole Participial Clause: as Od. 5. 6 pede ydp of édy év Sdpact vipdns tt troubled her that he was Se. : Tl. 6. 191 yiyvacke 908 yévov jiv édvta knew him for the offspring of a god: Od. 10. 419 col pev vootnoavts éxdpnuev we were glad- dened by thy return: Il. 13. 417 dxos yever’ edfauevoro there was vexation at his boasting. We have here the idiom already observed in the use of the Infinitive (§ 237) by which the weight of the meaning is shifted from the grammatical Subject, Object, &e. to a limiting or qualifying word. With Verbs of saying, hearing, knowing, &c., also of rejoicing and grieving, we sometimes find an Acc. with a Participle follow- ing (analogous to the Acc. with the Infinitive, § 237, 2): Il. 7,129 rods viv ef ardccovras i’ “Extope mdvras axovoat if he were to hear of their shrinking. Od. 4. 732 ef yap ey muddunv radrny 6ddv Sppalvovra. 23. 2 decrolvy epéovea lrov méow evdov édvra. Tl. 8. 378 i vai. . ynOjoer mpopaveioa will rejoice at our ap- pearing. 2.46. ] GENITIVE ABSOLUTE. 167 Il, 13. 16 7x Oro ydp fa Tpwolv dauvapévous he was vexed at their being subdued by the Trojans. A further extension, analogous to the Acc. with the Inf. after Impersonal Verbs, may perhaps be seen in Od. 6.193 dv éwéoty’ ixérny Tadaneipioy dvtidoavta which it is fit that a suppliant should meet with, 246.| Genitive Absolute. This is a form of ‘implied predi- cation,’ in which the Noun or Pronoun has no regular construc- tion with the governing Verb. The Participial Clause expresses the time or circumstances in which the action of the Verb takes place : Tl. 1. 88 ob tis éued CGvros rd. no one, while I am living, shall &e. %. 55L TeptTeAAouevan eviavTay as years go round, 5. 203 dvdpGv eihouevar where men are crowded ; so dvdpdv Atkudvrav, dvopav tpecodvrwy, ToAAGY EAkdvTwY, &e. Od. 1. 390 Kal Kev roir’ €0€douu Ards ye diddvT0s dpécbat that too I would be willing to obtain if Zeus gave it. The Subject is understood in Od. 4.19 porns €Edpxovros when the singer began the music. The Aorist Participle is less common in Homer than the Pre- sent, especially in the Odyssey: the instances are, Il. 8. 164, 468., 9.426., 10. 246, 356., 11. 509., 13. 409., 14. 522., 16. 306., 19. 62, 75 21. 290, 437-5 22. 47, 288, 383, Od. 14. 475., 24. 88, 535. An approach to a ‘ Dative Absolute’ may be seen in such uses as— Il. 8. 487 Tpwotv per p’ déxovaw edu pdos. 12. 374 émevyouévorot & txovro. Od. 21. 115 ob xé por dxvupévw rade Shpara mérvia pATHp Acinor (=it would be no distress to me if &e.) which are extensions or free applications, by the help of the Participle, of the true Dat. (Dativus ethicus). Similarly the ‘Genitive Absolute’ must have begun as an extension of one of the ordinary uses of the Gen.; most probably (if a conjecture is to be ventured) of the Gen. of Time (§ 150). For jeAtou dudyros within the time of the sun’s rising is a Gen. like jods in the morning, vuxrds by night, &c., and answers, as a phrase denoting time, to Gy’ jeriy Karadvvr at sun-set, és j€dvoy xatadivra up to sun-set, &c. So we may compare rov8’ abrod AveaBavros édevcera he will come within this year with % oédev évOdd edvros eredoera he will come within your being here; and again weprteAdopévay énravtay in the years as they go round, with Tay mporépav éréay in the former years. The transition may be seen in éapos véov iorapévoro in the spring when it is beginning. Compare also the phrases émet- yopévav dvépav, Bopéao weadyros, &c. with ynvepins in calm weather, &e. The circumstance that the Ablative is the ‘absolute’ Case in Latin is far from proving that the Greek Gen. in this use is Ablatival. In Sanscrit the Case used * This list is given by Classen, Beob. p. 180 ff.; but he includes meperAdpevos (Od. 1. 16., 11. 248), as to which see Curt. Verb. II. 9. 168 HOMERIC GRAMMAR. [247. in this way is the Locative: and the Latin Abl. Absolute may represent a Loca- tive of time at which, or even an Instrumental of cirewmstance (§ 144). The hypothesis that such Participial Clauses in Greek expressed space of time within which (rather than point of time, or circumstance) is borne out by the interesting fact, noticed above, that in Homer this construction is chiefly found with the Participle which implies continuance, viz. the Present: whereas in Latin the Abl. Abs. is commonest with the Perfect Participle. CHAPTER XI. Uses or THE Pronouns. Introductory. 247.] The preceding chapter has dealt with the two gram- matical forms under which a Noun, by acquiring a verbal or predicative character, is developed into a kind of subordinate Clause. We have now to consider the Subordinate Clause pro- perly so called: that is to say, the Clause which contains a true finite) Verb, but stands to another Clause in the relation of a ependent word. £.. in the Sentence Acvaoere yap Td ye TavTEs & por yépas épxetar dAAn ye see that my prize goes elsewhere, the Clause 6 ou yépas pxerar dAAy stands in the relation of Odject to the Verb of the principal Clause. As the grammatical structure of Subordinate Clauses is shown in general by means of Pronouns, or Conjunctions formed from Pronominal Stems, it will be proper to begin with an account of the meaning and use of the different words of this class. 248.|] Deictic and Anaphoric Pronouns. The Greek Gram- marians divided the Pronouns (d4vrwvupia) into detxtixal * point- ing,’ and dvagopicai ‘referring’ or ‘repeating.’ These words have given us, through the Roman grammarians, the modern terms Demonstrative and Relative: but the meaning, as often happens in such cases, has undergone a considerable change. A Deictic Pronoun—it will be convenient to adopt the Greek words —is one that marks an object by its position in respect to the speaker: J, thou, this (here), yonder, &e.; an Anaphoric Pronoun is one that denotes an object already mentioned or otherwise known,—the term thus including many ‘ Demonstratives’ (that same man, ¢Ae man, he, &c.), as well as the ‘Relative.’ In all, therefore, we may distinguish three kinds of Pronouns : 1. Deictic, in the original sense. 250. ] THE PRONOUNS. 169 2. Anaphoric, i.e. referring to a Noun, but Demonstrative (in the modern sense). 3. Relative, in the modern sense. This however, it should be observed, is a classification of the uses of Pronouns, not of the words or Stems themselves: for the same Pronoun may be Deictic or Anaphoric, Demonstrative or Relative, according to the context. It is probable, indeed, that all Pronouns are originally Deictic, and become Anaphoric in the course of usage. or nw 88e, Kelvos, obTos. 249.] The Pronoun ae is almost purely Deictic. It marks an object as near the speaker,—this here, this on my side, &c.; as vat pa 70d¢ oxijntpov by this sceptre (in my hand) ; “Exropos ijd¢ yur this is the wife of Hector ; Od. 1. 76 pets olde mepippatipeba let us here consider (§ 162, 2); 1. 226 ovdk epavos tdde y’ éori what I see here is not a club-feast. It is especially applied to a person or thing to which the speaker turns for the first time, as— Il. 3. 192 ein’ dye por cab révde, didrov rékos, bots 88 cori. Hence the use to denote what is about to be mentioned—the new as opposed to the known. This is an approach to an Anaphoric use, in so far as it expresses not doca/ nearness, but the place of an object in the speaker’s thought. So in— Tl. 7. 358 otcOa wat GAdrov 000v dyelvova rodd« vonoat the speech is the present one, opposed to a better one which should have been made. The derivatives rtoodcd«, todode, Sd, evOdse, are similarly Deictic: as Il. 6. 463 xnret rows avipds from want of a man such as Iam now. 250.| The Pronoun xeivos is sometimes used in the Deictic sense, pointing to an object as distant : Il. 3. 391 Keivos 6 y év Oadrdu yonder he is in the chamber. 5 604 kal viv of mdpa xeivos ”Apns there is Ares at his side. So of an absent object: as Od. 2. 351 xeivoy diouévn Tov Kdypopov thinking of that (absent) one, the unhappy. Hence in an Anaphoric use, ketvos distinguishes what is past or done with, in contrast to a new object or state of things : Il. 2. 330 keivos ras dydpeve he (on that former occasion), Fe. 3. 440 viv pev yap Mevédaos evixnoen ody ’AOHvD, ketvoy 8 adris éyo. Od. 1. 46 Kat Alyy ketvds ye eouxdre Keira ddACOpw GANG pot dud’ "Odvo7i «.T.d. Here xetvos marks the contrast with which the speaker turns to a new case. The literal sense of local distance is transferred to remoteness in time, or in the order of thought. 170 HOMERIC GRAMMAR. [251. 251.] The Pronoun ofros is not unfrequently Deictic in Homer, expressing an object that is present to the speaker, but not near him, or connected with him. Hence it is chiefly used (like zs¢e in Latin) in reference to the person spoken to, or else in a hostile or contemptuous tone. Instances of the former use are :— Tl. 7. 110 ddpatvers, Mevédae diorpedés, odde tl ce xpi} tavtns appoovyns. 10, 82 ris & otros Kara vijas ava orpardy Epyeat otos ; Od. 2. 40 odx Exds otros dvijp the man you want is not far off. Again, obros is regularly used of one of the enemy ; as— Il. 5. 257 rovrw 8 ob mddw adris dtotceroy dkées tarot. 22. 38 pp pos piuve, pidrov Tékos, avepa TovTOV. Similarly, with a tone of contempt,— Il. 5. 761 dgppova robrov dvévres (ep. 831, 879). Od. 1. 159 rovrovow pev tatra pede (of the Suitors). More commonly, however, ofros is Anaphoric, denoting an object already mentioned or known. In later Greek it is often employed where Homer (as we shall see) would use the Article. It is a curious proof of the limited use of otros in Homer that it is never found after a Preposition (Kriiger, Dial. 50, 2, 9). attés. 252.| The Pronoun atrés is purely Anaphoric: its proper use seems to be to emphasise an object as the one mentioned or implied,—the very one, that and no other, It conveys no local sense, and is used of the speaker, or the person addressed, as well as of a third person. Specific uses are— (1) To distinguish a person from his surroundings or ad- juncts: as— Il. 3. 195 revxea pev of xeirar emt xovt movduBoreipn, avros dé K.T.A. Q. 301 adrés Kal rod dépa he and his gifts. 14. 47 mpl mupt vijas évumppoat, xrelvat be Kal adrods. 17.152 Os tot TdAX’ Gedos yevero mrdrAet Te Kal abtS to thy city and thyself. So of the dody, as the actual person, in contradistinction to the sovl or life (Wvx7), Il. 1. 4, Od. 11. 602, &e. _ ifence, too, aitds=dy himse/f (without the usual surround- ings) :— Il. 8. 99 Tudetins 8 adrds mep eav mpoudyorow eulyOn. So Achilles in his complaint of Agamemnon, Il. 1. 356 éAav yap 253-| THE REFLEXIVE PRONOUN. 171 exel yépas adrds dmovpas, i.e. at his own will, without the usual sanction : cp. 17. 254, 23. 591. This meaning appears also in aires = merely,— Od. 14. 151 GAN éyd odk adrws prOjoopa GAA ow Spko. Cp. Il. 1. 520 4 8% kal atrws .. vee? as it is (without such pro- vocation) she reproaches me. The Gen. airoé, &c. is used to strengthen the Possessives: as + a an a NY a re w Od. 2.45 uov abrod xpeios: Il. 6. 490 7a o” abrijs épya: Il. ro. 204 @ abtod Ouys (suo tpsius animo). Hence in I. 9. 342 tiv abrod giAéec—where the use of the Art. is not Homeric —we should probably read fv atrod. (2) To express without change, the same as before : Il. 12. 225 ob kéopm mapa vaigw edrevoduel? abra xédevba. Od. 8. 107 Hpye 8% 7 abriy Sddv AvTep of BAdor K.T.A. Hence the use with a Dat., noticed in $144; as Od. 8. 186 aire pape with his cloak as it was (without putting it off); and in the Adverbs aizd., aitod in the place, without moving ; and abrws without doing more, hence without effect, idly: as— Il. 2. 342 atrws ydp p’ énéeco’ épidalvoyev K.T.d. (3) The unemphatic use, as it may be called, in which it is an ordinary Anaphoric Pronoun of the Third Person (Eng. he, she, at). In this use the Pronoun cannot stand at the beginning of a Clause (the emphatic position), or in the Nominative—an unemphasised Subject being sufficiently expressed by the Person- Ending of the Verb. The use is derived from that of the emphatic aérés in the same way that in old-fashioned English ‘the same’ often denotes merely the person or thing just men- tioned: and as in German derselbe and der nédmliche are used without any emphasis on the idea of sameness. The Reflexive use of airds is very rare: Od. 4. 247 GAAw 8 adrév pati xara- apinroy qioxe. On Il. 9. 342 THy adrod pirger see above (1). The Refleaive Pronoun. 253-] The Pronoun éo (7.¢. the Personal Pronoun declined from the Stems ée- or é- and ode-) is sometimes Reflexive (i. e. denotes the Subject of the Sentence or Clause), sometimes a simple Anaphoric Pronoun. In the latter use it is always un- emphatic. The Reflexive sense is chiefly found either (1) after a Preposi- tion, as dudl & mamrjvas looking round him, and so amd é0, ént of, mpott ol, peta oiot, kara opéas, &e.; or (2) when it is reinforced by adrés, as I], 20, 171 €@ 8 atrov émorpiver paxéoacOa stirs him- self up to fight, Other examples are few in number ; 172 HOMERIC GRAMMAR. [254- Tl. 2. 239 ds kal viv "Aytrja, €o péy’ duelvova para k.T.d. 5. 800 7 dAtyov of maida éouxdra yelvaro Tudeds. So II. 4. 400., 5. 56., 24. 134, Od. 11. 433., 19. 446, 481. We should add however such Infinitival Clauses as— Tl. 9. 305 ézet ob tid pyow spotov of euevat x.T. A. where the reference is to the Subject of the governing Verb: so Il. 17. 407, Od. 7. 217, &e. Compare also the similar use in Subordinate Clauses, as— I. 11. 439 yv@ & ’Odvceis 8 of ob Tu TeAos Kataxalpioy 7jdOEV. The strictly Reflexive use is commoner in the Iliad than in the Odyssey. Ex- cluding Infinitival and Subordinate Clauses, there are 43 examples in the Iliad, against 18 in the Odyssey. Note that the use is mainly preserved in fixed com- binations (am6 0, mpori ot, &e.). The Anaphoric (non-Reflexive) use is very much commoner. In this use—which is doubtless derived from the other by loss of the original emphasis—the Pronoun is enclitic: whereas in the Reflexive use it is orthotone. Accentuation. According to the ancient grammarians this Pronoun is ortho- tone (1) when used in a reflexive sense, (2) when preceded by a Preposition, and (3) when followed by a Case-form of airvés in agreement with it. The first and second rules, as we have seen, practically coincide: and the third is not borne out by the usage of Homer. In such places as Od. 2. 33 cl0e of ait@ Zebs dyabdy Tehécese, 11.6. gt Kat of moAd Pidraros abTH, Od. 8. 396 Evpdados 8é é abrdy (‘Odva- céa) dpecodcdw,—add Il. 24. 292, Od. 4. 66, 667., 6. 277—the Pronoun is evidently unemphatic, and is accordingly allowed to be enclitie by good ancient authorities, This is amply confirmed by the instances of pw avrov (Ll. 21. 245, 318, Od. 3. 19, 237, &c.). In one instance, viz.— Od. 4. 24 Prirdy pw TAnyhow dekedino Sapaccas it would seem that pw has a refi mse. The reading, however, is not certain, some ancient authorities giving avtov pev or atrov pév (La Roche a. l. and H. U. p. 138). 254.| The Possessive éés, 8s is nearly always Reflexive. Oc- casionally it refers to a prominent word in the same Sentence which is not grammatically the Subject : as— IL. 6. 500 ai peév ert Cwdv yoov “Extopa @ évt oike, Od. 9. 369 Odriv eye miparov Copa peta ofs Erdpoucr, Cp. Il. 16. 800., 22. 404, Od. 4. 643., 11. 282., 23. 153. And it is occasionally used in a Subordinate Clause to refer to the Subject, or a prominent word, of the Principal Clause : Od. 4. 618 mépev d€ E Paldyyos Hows Subovlwy Bartheds, 60’ ds Sduos duexddupe kelo€é pe vooticavta (ep. 4. 741). 255-] THE REFLEXIVE PRONOUN. 173 Il. 10. 256 Tvdeldy piv 8dxe pevenrdreuos Opacvuhdns pdoyavov dpdnykes, 7d 8 édv Tapa vyt A€AerTTO. 16. 753 €BAnro mpds ariOos, é) ré www Sdrecev GAK?. It will be seen that where és does not refer to the grammatical Subject it is generally emphatic: ¢.y. in the line last quoted, é adky his own prowess, not that of an enemy. This indicates the original force of the Pronoun, which was to confine the reference emphatically to a person or thing just mentioned, 2 55+| Use of és, 8s as a general Reflexive Pronoun. It has been a matter of dispute with Homeric scholars, both ancient and modern, whether éés (6s) was confined to the Third Person Singular (Ais own), or could be used as a Reflexive of any Number and Person (own in general—my own, thy own, their own, &c.).* The question is principally one of textual criticism, and depends in the last resort on the comparative weight to be assigned to the authority of the two great Alex- andrian grammarians, Zenodotus and Aristarchus. It is connected with another question, of less importance for Homer, viz. whether the forms o, ot, ¢ are confined to the Singular, and those beginning with o- to the Plural. In regard to the latter of these questions there is no room for doubt. The only instance in dispute is Il. 2.197, where the reading of Zenodotus was— Oupds be péyas earl dioTpepéwy Bactdyov, Tip 8 ex Ards or, pret 8é € unriera Zevs, and so the first line is quoted by Aristotle (Rhet. II. 2). Aristarchus read 5:orpe- péos BaotAfos, possibly on account of é However, admitting Zenodotus to be right, € need not be a Plural. The change from the Plural to the Singular is not unusual in passages of a gnomic character, e.g. Od. 4. 691 BT eat Bien Ociav Bacirjwv Gddov ’ éxOalpnot Bpotay, ddAov Ke gudoin. Again, the ‘general’ Reflexive use, if it exists in Homer, is confined to the ad- jective éés, 8s. Our texts indeed contain an instance of op(or in this use. In Il. 10. 398 Dolon tells Ulysses that he has been sent by Hector to find out— He puddooovra vies Goal ws 76 wapos meEp, hy 75n xElpecow bp’ jhperepyar bapevtes préw Bovdedovar pera opiow, ov8 e6édovet k.T.A. With this reading Dolon repeats the exact words of Hector (vv. 309-311), and so Aristarchus read; but the best MSS. have BovAevorre pera adiow (consult among yourselves), and é0édo:7e, The Optative, however, is not defensible (esp. after the Indic. puAdocovra:), and was probably introduced by some one who thought the Second Person necessary for the sense. But the Third Person is natural enough: for Ulysses, to whom Dolon is speaking, is not one of the Greeks who can be supposed to be ‘ consulting among themselves.’ The form @ is found as a Plural in Hom. h. Ven. 267. In later Epic poets the * The question was first scientifically discussed by Miklosich, in a paper read to the Vienna Academy (I, 1848, p. 119 ff.), who is followed on the same side by Dr. Karl Brugman (Hin Problem der homerischen Textkritik und der ver- gleichenden Sprachwissenschaft, Leipzig, 1876). The other side of the discus- sion has been chiefly sustained by Dr. E. Kammer in the Neue Jahrbiicher (1877, p. 649 ff.). 174 HOMERIC GRAMMAR. [255- Substantival ¢fo, &c. are used as Reflexives of any Person or Number: see Theo- critus, 27. 44, Apollonius Rhodius, 1. 893., 2. 635, 1278., 3. 99 (Brugman, Prodi. p. 80). But the use is exclusively post-Homeric. The case is different with the adjective és, 5s. Zenodotus appears to have regarded it as a Reflexive of any Person or Number (meaning own in general), while Aristarchus confined it strictly to the Third Person Singular. Accordingly we find forms of és (8s) read by Zenodotus in a number of places in which our editions—following the authority of Aristarchus—have substituted other words. Thus in— Il. 1. 393 dAAd ad, el Sivacal ye, wEpicxeo maL5ds Ejjos, and in similar passages (Il. 15. 138., 19. 342., 24. 550), it is known from the Scholia that Aristarchus read éfjos, Zenodotus éoto (= thine own), Again, in— IL. 3. 244 ds paro, rods 8 H5n Karexev pvaifoos aia, év Aanedaiporr adh, pian ev marpid: yan’ for gidn Zenodotus read &j (their own). So, again, in— Il. 11. 142 viv pév 3) Tod warpos derkéa Ticere AWByV Zenodotus read ot marpés (your own father). It is probable that he read of in the similar places Il. 19. 322, Od. 16.149, &. Besides the instances of un- doubtedly ancient difference of reading, there are several places where one or more MSS. offer forms of é6s in place of éés and oés, Thus— Il. 14. 221 8 re pect ofa pevowds (Gor La Roche’s D). 19.174 ov de ppeot ofow iavO7js (jow in several MSS.). Similar variations (with ¢peo?) are found in Od. 5. 206., 6. 180., 13. 362., 15. II1., 24.357. Again— Od, 1. 402 Shpacr coiow dvdoous (oiow ten MSS.). Similarly in Od. 8. 242., 15. 89 (fot for éuotcr): also— Od. 7.77 Kal ohv és marpida yaiay (jv és in one MS.). 13. 61 od be répreo 7G’ evt oixw (@ evi one MS.). Another instance of variation is detected by Dr. Brugman in— Il. 9. 414 i 5€ xev oixad txwpe pian és warpiba yaiay, where the MSS. (except Ven. A) have txwpar, pointing to the reading éjv (my own).* The existing text of the Odyssey contains three passages which Brugman claims as instances of a general Reflexive sense, viz. Od. 4.192 (as to which see Merry and Riddell’s note), Od. 13. 320 (where there is some reason to suspect an interpolation), and— Od. 9. 28 ob Tot eyw ye Hs yains Sivapar yAucepwrepov dAdo ldécOat, But there is no reason to take 7s otherwise than in v. 34 ds odSey yAvKLov Fs TaTpi- Sos 052 roxhaw yiyverat nothing is sweeter than a man’s own country, &c. The reference of the Pronoun is vague, as in Od. 1. 392 alfa re of 5@ dpvedy médXeTaL a man’s house (when he is a king) quickly grows rich. We have seen that post-Homeric poets (who, even if they do not represent living usage, may be regarded at least as embodying early traditions about the * Dr. Brugman carries his theory into other passages where he supposes Aristarchus to have corrected the text in order to get rid of the use of és for the First or Second Person: but the examples quoted above will suffice to give an idea of the strength of his argument. 2.56. | THE ARTICLE. 175 text of Homer) use the Substantival %o, &c. in the sense in question. The cor- responding use of the adjective 6s, Ss is still more common, as Brugman shows. It is found in Hesiod (Op. 58, Theog. 71), and in Callimachus, Apollonius Rhodius, and Quintus Smyrnzus (Problem, pp. 28, 78-83). In attempting to arrive at a conclusion on this matter we must begin by under- standing that the issue does not lie between supposing on the one hand that Aristarchus had good ground for correcting the text of Zenodotus in all the places in question, and on the other hand that he introduced a strange form like é7jos on his own authority, and merely to satisfy a theory. The latter is impro- bable, not only from the respect for manuscript authority which is expressly attri- buted to Aristarchus, but also because the various readings are not all capable of being explained on this supposition. Thus, (1) the word és is proved to exist by Od. 14. 505., 15. 450, and oto, though excluded by the sense, is found as a variant in the latter place. Also (2) éfos is found for éofo meaning his own in Il. 14. 11., 18. 71,138. It cannot therefore be regarded as certain that éos was systematically introduced to get rid of éofo=my own, &c. Again, (3) the use of the Article in rod rarpés, ris untpés, Tod madds, is not so clearly un-Homeric as to require the change to ot marpés, &c, in every place (see § 258). And if in Il. rr. 763 otos Tijs dperis dwovnoerar Bentley was right in reading 7s (cp. 17. 25), this indicates that the Article might creep in for od, fs, &c. apart from the intention of carrying out a grammatical theory. On the other side it must be conceded that the generalised Reflexive use of és, és, —if not of the substantival forms ¢0, &c.—is of high antiquity, so that sporadic instances of it may have occurred in the genuine text of Homer, If so, the error of Aristarchus will consist in a somewhat undue purism. Brugman holds that the general Reflexive sense is the primary one, belonging to the Stem sva (Greek ‘Fe-) in the original Indo-European language, and surviving in the Homeric use of és, 6s. But even if the readings of Zenodotus which give this sense are right, it does not follow that they represent the oldest use of the Pronoun. Brugman has himself given excellent instances of the extension to the First and Second Person of a Reflexive Pronoun originally confined to the Third Person (Prodlem, pp. 119 ff.). In the present case it is significant that the gene- ralised use of the substantival forms 0, &c. is clearly post-Homeric. If és (8s) is sometimes used in Homer, as well as afterwards, of the First and Second Persons, or of the Plural, it is natural to see in this the result of an extension of usage. And the conclusion is strengthened by the character of the instances. For we see that és (Ss) is so used, according to Brugman, (1) with words of relationship, as in the combinations of marpés, matdds éoio, &c., (2) with words meaning home or country, as fs warpisos, Swpaow oiow, &e., and (3) with ¢péves, esp. in the phrase ¢pecly jow. These are precisely the most familiar combinations into which such a Pronoun enters,—the combinations accordingly in which an exten- sion of use is most likely to be found. 6 t TO. 256.| The Article 6 4 7é may be defined as a purely Ana- phoric Pronoun, conveying some degree of emphasis. It differs from 83¢ ofros and éxetvos in the absence of Deictic meaning: for while it usually marks some contrast between objects, it does not distinguish them as near or far, present or absent, &e, On 176 HOMERIC GRAMMAR. [257- the other hand it is distinguished from the non-Reflexive use of aités and éo by greater emphasis. Three chief uses of 6 4 76 may be distinguished :— 1. The use as an independent Pronoun; 6 # t6=he she it. This may be called the SussrantivaL use: it embraces the great majority of the instances in Homer. 2. The use as an ‘Article’ in the later sense of the term, ¢.e. with a Noun following. This may be called the AtTRI- BUTIVE use. 3. The use as a Relative. 257.] The Substantival Article. This use of the Article is very much the commonest in Homer, and it is also the use from which the others may be easily derived. The Substantival Article either (1) is simply ‘resumptive,’ recalling a person or thing already mentioned, as 6 yap for he, tov pa him I say, avtés cal tod dépa the man and his gifts: or (2) marks a contrast, as 6 5€ but the other. The following points of usage are to be noticed :— 1. The most frequent—we may almost say the regular—place of the Article is at the beginning of a Clause, followed by peév, 84, ydp, dpa, or preceded by adtdp, &Add, 7 Tor, or an equivalent Particle. Hence the familiar combinations 6 peév, 6 dé, 6 ydp, Kat yap 6, adrap 6, 7) Tou 6, Tév pa, GAAG tov, &e. of which it is needless to give instances. It is worth notice that the later Substantival use with pév and 6€ is a surviving fragment of this group of uses. A few others are found in Attic poets, as 6 ydp (isch. Sept. 17, Soph. El. 45, O. T. 1082). The use of the Article with an adversative Particle (8¢, adrdp, adAd) generally marks a change of Subject: 6 5¢ Gut the other, &c. But this is not always the case: e.g. Il. 4. 491 Tod peév auapd’, 6 6€ Aeixov ... BeBrrner hum he missed, but smote Leucus (so Il. 8. 119, 126, 302., 11. 80, &e.); Tl. 1. 496 Odris & od Ader’ epetucwy | maidds éob, GAN Hy’ aveddcero x,T.A.: ep. Il. 5. 321., 6. 168, Od. 1. 4, &e. The Article in all such cases evidently expresses a contrast: not however between two persons, but be- tween two characters in which the same person is thought of. This last use—in which the Article is pleonastic, according to Attic notions—occurs in Herodotus, as 5.120 7a peév mpdrepov ot Kapes €BovAevorro perijKar, of dé adris ToAEpeiv ef dpyxijs apreovro. ‘We may compare it with the pleonastic use of the Pronoun in— Il. 11. 131 Gypet Arpéos vid, od 8 déia d€€ar drowa, where the effect of inserting ov is to oppose the two acts denoted by (éype and dé£au damowa. 257-] THE ARTICLE. 177 2. The Article with yé is frequent in Disjunctive sentences : Tl. 12. 240 ef 7’ em) deEL? Lwor mpds HS 7’ HedLSy Te, el 7’ én’ dpiorepa rol ye «ra. (or else to left). Od. 2.132 Ge 8 y } réOvnxer. Here also the force of the Article is to contrast two things said about the same Subject. 3. The principle of contrast often leads to the placing of two Articles together: Il, 21. 602 efos 6 rdv medi{o1o SidKero, 10. 224 kal re mpd 6 Tod évdnoev. So an Article and a Personal Pronoun, éy 8% od roto (Il. 13. 829, &e.); ep— Tl. 8. 532 eioouar et xe p’ 6 Tudeldns xparepds Aroujdns Tap vndv Mpds TEetxos dmdceETaL, 7} KEV ey@ TOD. Note that when the second of the two is in the Nom., it usually takes yé: hence rod 8 ye, ri p’ of ye, &e. 4. The Article often stands for the object defined by a follow- ing Relative Clause, e. g.— Tl. 9. 615 kaddv ror ody euol rov Kidew bs x” ue Kfdn. 1. 272, Tap ot viv Bporol eiot &e. The use is to be classed as Anaphoric ; the intention of saying something about the object is equivalent to a previous mention. So in Latin the Anaphoric és is used to introduce gui. The Neuter of the Article is similarly used to introduce Clauses beginning with ére, as, and the like :— IL. 15. 207 écOAdv kal 76 réruxtas Or’ dyyedos aloe €id7. Od. 9. 442 76 5& wpmos odk evdyoer | ds of x.7.A. IL. 3. 308 Zeds pev mov 7d ye olde. . . Ommorépy K.T.A. and even independent sentences, as— Od. 4. 655 adda 75 Card’ tov évOdde Mevropa diov. 5. The uses in which the Article is least emphatic (i.e. does not begin the Clause, or express a contrast) appear to be— (a) after Prepositions: esp. in the Dat. Plur, after perd, mapd, mpotl, ovv, év, Spa: as Il. 1. 348 % 8 déxovo’ Gua rotor yovy klev. This is to be connected with the fact that the forms éo, of, opiot, &c. are not used with Preposi- tions in the simple Anaphoric sense (§ 253); so that the Art. is the only Pronoun available. (2) when the Neuter Article is used for a fact or set of facts; as Il. 4. 353 drpeat iy eOeAnoOa, Kal at Kév to. Ta peuirn. Here again the want of a corresponding form of o makes itself felt. This use is chiefly found in the Nom. and Acc.; but also in robvexa therefore, éx roio from that time, &e. N 178 HOMERIC GRAMMAR. [258. 258.] The Attributive Article. The Attributive Article is found in Homer in a limited range of cases, and has evidently grown out of the use of the Substantival Article followed by a Noun in ‘Apposition;’ e.g. Il. 4. 20 ds épaé’, at & enéuvéav ’AOnvain te Kai “Hpn thus he spoke, but they murmured, Athene and Here: Tl. 1. 348 4 8 dékove’ Gua rotor yur) xiev. So with pw, Il. 21.249 Wa pw mavoere mévoro | dtov “AytAAHa, ep. Od. 11. 570. In such cases the Pronoun is still substantival, the Noun beng added by way of afterthought. It is a step towards an Attributive use when the Article needs the addition of the Noun to explain it; ¢. 9. IL. 4. 501 rev p’ "Odvceds Erdporo yoAwodpevos Bare Sovpt Kopony’ % 8° Erepoo dia Kporddoto mépyoev aixuy Xadkein. Here 4 8€ would not be clear without aixyy. So in— Od, 15. 54 Tod yap te £etvos pipvijoKera juata mdvra avdpos Eervoddxov. So too with Proper Names—when a new person is about to be mentioned the Art. anticipates the Noun: e.g. Il. 2. 402 atrap 6 Boty idpevoev vat avipav ’Ayaueuvav. And where the Neut. 1é is followed by an epexegetic Infinitive : Od. 1. 370 émel td ye Kadov dxoveuev early do.dod. In all these cases the combination of Article and Noun is not sufficiently close to constitute an Attributive use; but they serve to show how such a use is developed. The Attributive uses in Homer may be classified as follows :— 1. Uses with connecting Particles, where some contrast is made in passing to the new sentence or clause. 2. Uses with certain Adjectives that imply contrast. 3. Uses to mark a person or thing as definite. 259.] With connecting Particles. The uses that fall under this head, though not very numerous, are characteristic of Homer. The following are the chief :— 1. The Article with an Adversative 8¢, adrép, &c. is not un- frequently used to bring out the contrast in which the Noun stands to something already mentioned: e.g. Il. 2. 217 Godrkds env, xwdds 8 Erepov adda, Ta dé of Suw but then his shoulders—contrasted with the parts of his body mentioned before; so 7& 8¢ of dove (Il. 13. 616), &e. Il. 22. 405 ds rod wey Kexdviro Kdpn day, S€ vu patnp but on the other hand his mother, &c. Il. 4. 399 rotos env Tudeds AlrdAtos’ GAAG Tov viov K.T.A. 260. | THE ATTRIBUTIVE ARTICLE. 179 1. 382 fixe 3° én” ’Apyelours Kaxdv Bédos, of 5é vv Aaol OvijcKoy énacctrepot, Ta & em@xeTo Kida Oeoto. So we should explain the Article in Il. 1. 20 aida 8€ jor Ad- cate plany, ta 8 drowa ddxecOar release my daughter, and on the other side accept ransom. Cp. Tl. 2. 105, 107, 278., 5. 308., 7. 84, &e. The usage is perceptibly rarer in the Odyssey. 2. The use of the Art. with pév—to contrast the Noun with something that follows—is rare: 11. 267 abrap éwel TO pev Edxos érépoero: ep. 8. 73., 9. I., 13. 640., 19. 21., 20. 75, Od. 3. 270 (seemingly the only instance in the Odyssey). There is a similar use however with the Art, following the Noun: Il. 6.147 @vAAa Ta pep T wTA.; Od. 1. 116 pvnotipwy TGv pev KTA. This use should be carefully distinguished from the later Definite Article. For instance, in Il. 1. 20 7a daowwa does not mean this or the ransom, in contradistinc- tion to other ransoms. It means the other, the ransom, in contrast to the person ransomed. To give another example, the 4th book of the Iliad begins of 5 dei, which we naturally take to mean simply du¢ the gods. This however is incorrect. Taking in the last line of the 3rd book, we have— ds par’ ’Arpelins, ént 8 jveov dAdAot ’Axatol: of 88 Geol map Znvi xabqpevor jryopswrro, Clearly the Article marks the turning from the one scene to the other,—from the battlefield to Olympus. Thus the Attic of (@eo/) distinguishes the gods from other beings: the Homeric of (62 Geo?) marks, not this permanent distinction, but the contrast arising out of the particular context. The difference appears also in the use with Proper Names. In Attic the Article shows that a particular known person is spoken of ; in Homer it marks the turn- ing of attention to a person—ushers in the name, as it were: e.g. Il. 5. 509 Tod & expaiavey éperuds | SoiBov ’AndAdAwvos. In short, the Homeric Article contrasts, the Attic Article defines. 3. The corresponding use with copulative and illative Par- ticles, cal, ré, }5€, kal ydp, is much less common: cp.— Il. 1. 339 apds re OeGv paxdpwr mpds te OvnTGv avOpdror, kal mpos TOU BactAjos amnveos. 15. 36 iotw viv réd€ yaia Kal obpavos eipds tmepber, Kat rd KaTeiBopevov Stvyds Ddwp (cp. 18. 486). Od. 22. 103 ddc d& cvBorn | Kal TG BoveddAw GAdAa. Tl. 14. 503 0088 yap 7 TIpopdxoto ddyap KTA. The Article singles out its Noun as the special object intended, or turns to it with fresh emphasis. The only instance of an Art. with an Infinitive falls under this head, viz. Od. 20. 52 dvin cal rd GudAdooev. But we need not take 16 guAdooeww closely together. 260.] With Adjectives. The Article is used before adjectival words that imply a contrast or distinction, especially between definite or well-known alternatives: in particular— (a) dddos and étepos, passim: also avtés=same, N2 180 HOMERIC GRAMMAR. [260. (2) Comparatives and Superlatives ; of wAéoves, of dpiorot, &e. So in the adverbial expressions 7d mpiv, rd mdpos, Ta apora, and the like. In these phrases the Neut. Article is used ‘adverbially’ (§ 113), and has the force of an Adverb of Time (cp. é 700, eis 8). Hence there is a kind of ‘ concord’ between the Pronoun and the Adverb (76 mdpos=then formerly). It is quite dif- ferent when a Mase. or Fem. Article is used with an Adverb, as of évep0e Geoi—a use which is extremely rare in Homer, (c) Ordinal Numerals: as rj dexdrn: similarly 76 fysov, 70 xO.¢év. Also Cardinal Numerals, when a division is made; as Il. 5. 271 tovs pév Técoapas adros éxwv ari- TAAN’ ent pdrvy, To de dv’ Aivelg ddxev four he kept, and the (other) two he gave to Aimeas: Il. 11.174 mdcas* ti d¢ 7 if K.t.A. (the lion chases) all, but to one &e. (d) Possessives; rév éuov xddov, Ta o& kjda, &c. (e) A few words expressing the standing contrasts (as they may be called), of great and small, many and few, good and evil, and the like, esp. when the contrast is brought out by the context: Il. 1. 106 pdure kaxdv, ob md org pot TO Kpiyvoy etmas* aiel ro. Ta Kak’ Corl hidra pect pavreverOar. 3. 138 T@ BE Ke viKjoavte Plan Kexdjon dkoures (the conqueror being one of two definite persons). So 4 aAnOds (Il. 2. 278., 15. 305) the many (in contrast to a single man, or to the few): Alas 6 yéyas the greater Ajax: dvaxtes of véow (Od. 14. 61) masters of the younger generation : ixOvor toils dAlyours (Od. 12. 252) the smaller kinds of fish: rov mpovxovra (Il. 23. 325) the one in advance. So too— Il. 1. 70 ds 78y 1a 7° Cdvra Ta 7” eoodueva mpd 7’ edvta. The use to contrast indefinite individuals (one—another) is rare in Homer: IL. 9. 320 kdrOav’ duds 6 7 depyos avip 6 Te TOAAG €opyos: Od. 17. 218 ds det roy duotoy dyer Oeds os TOY dpoiov. (7) Patronymics and geographical epithets: e.g. Il. 11. 613 Maxdom mdvra éouxe TS “AokAnmiddy (ep. 13. 698., 14. 460., 23. 295, 303, 525): Ll. 2. 595 Oduvpw tov Opijixa: Il. 6. 207 mediov 75’ AAriov, cp. 2. 681., 10.11: and so perhaps I. 21. 252 aletod . . rod Onpyripos an eagle, the hunting kind. This use is rare. (g) In a very few places a Genitive takes the place of the Adjective: Il. 20. 181 rips ris Tpiduov: Od. 24. 497 viets of Aodlovo: Il. 9. 342., 10. 408., 23. 348, 376, Od. 3. 145. When an Adjective with the Article prefixed to it qualifies a Noun, the general rule in Homer is that the Noun comes first. This is also the order which we saw 261.] THE ATTRIBUTIVE ARTICLE. 181 to prevail in the use before a Relative: compare (¢.g.) Il. 14. 279 @eods . . | rods bnoraprapious with 5. 331 Gedwr | Tdwy air’ dvdpdy méAepoy xara Koipavéovow. In both cases the Article is properly ‘resumptive,’ that is to say, it repeats the Noun in order to add the qualifying words, The other order—in which the Noun comes last—is common with Comparatives and Superlatives (as 7d veiara meipara), with the Possessives, as Tov éudv xédAov (but we also have warjp obyds), and with dAdos (tay dAAwy Aavady &e.; but also, Tl, 5.130 Oe0ts .. . Trois ddAows, 9. 219 Tolxov Tob Erépov). With the other groups enumerated above it is exceptional: Il. 14. 274 of évepde Geol; 2, 681 7d TleAaoyindv “Apyos. Commonly the Article begins a new line, as in dvdpdy | Trav rére, dvakres | of véor, Maxdov... | TH AoxAnmaéy; or it follows the principal Cesura, as in Oduup rév Opyixa; or begins the second foot, as in ixOvot Tots dAtyost, TivAs THs TIpiduou, Alas 8 6 péyas, waides rot perdmobe Ac- Actupevot. 261.] The defining Article. The few and somewhat isolated uses which fall under this description may be grouped as follows : 1. The use before a Relative Clause may be combined with ‘ Apposition’ to a Noun: as— Il. 5. 319 008 vids Kamavijos edjdeto ovvectawy Tawv ds eméreAA€ KT.A. This is the primitive order, the Article being ‘ resumptive’—the injunctions, those namely which, &c. So ipat. tH bre &e, The later order—that in which the Noun follows the Article—ap- pears in a few places of the Iliad :— 5. 265 Ths yap Tor yevers hs Tpwt mep «rd. (ep. v. 268). also 6. 292., 8. 186., 19. 105. It is commoner in the Odyssey. 2. Occasionally the Article is prefixed to an epithet used in a hostile or contemptuous tone: thus we have II. 2. 275 rov AwByrTipa, 13. 53 6 Avoosdys, 21. 421 H Kvvdpuia, 22. 59 roy dvatnvov : Od. 2. 351 TOV Kdupopoy 5 12.113 THY dAonD ; 18. 26 6 podoBpds; 18. 333 Tov adjrnv; 19. 372 at Kives aise. So perhaps in Il. 3. 55 4 te xdun td Te eldos. In Od. 18. 114 rodrov rdv Gvadtov does not mean (as it would in Attic) ‘this &vadros,’ but ‘this man—dvadros that he is.’ Cp. 11.13.5396 Oy 6 Avoosdns xrd., where 6 Avooddns—the mad- man—is used as a single term, in Apposition to 6 ye, This use—which is characteristic of Homer—may be regarded as a relic of the Deictic force of 6 4 76. It answers to the later use of otros, Latin iste. 3. The use of the Article to show that the Noun denotes a known person or thing—the defining Article of later Greek—is rare in Homer. It is found in the Hiad— (1) with yépor, yepaids, ava€, ijpos: where however the Pronoun is the important word, the Noun being sub- joined as a kind of title: roto dvaxros= ‘of his lordship’ (cp. the German allerhichst derselbe). Accordingly, 182 HOMERIC GRAMMAR. [262. when the name is added the Art. is generally not used; as yépwy inanddra [Inheds (not 6 yépwr). (2) with ézos and 060s, in certain phrases, as mofoy Tov pddov Zevmes; In these cases the Noun is of vague meaning, adding little to the Article: cp, émei rov pidov dxovce with érel td y’ dkovoe. So in the formula duocey re reredtyaév Te Tov Spxov, perhaps with a touch of cere- monial verbiage. In the Odyssey it occurs with several other Nouns: 6 eivos, passim ; vioos, Od. 5. 55., 9.146., 12. 201, 276, 403, &e.; 7a para Od. 9. 464., 11.4, 20: 6 pdydos Od. 9. 375, 378: 16 ré£ov Od. 21.113, 305. The other examples in the Iliad are chiefly found in one or two books: see Il. 2. 80., 7. 412., 10.97, 277, 321, 322, 330, 408, 497., 20. 147., 23. 75, 257, 405., 24. 388, 801. We may perhaps add a few uses with words of relationship : Il. 11. 142 viv pev 6 rod matpos deukéa ticere AGByv. But here the Art. is resumptive with emphasis: (if ye are sons of Antimachus) ye shall now pay for his, your father’s, outrage. Il. 19. 322 od8 ef Kev rod marpos dmopOuevoro mvdoluny not even if I heard of him, my father, being dead: and so Il. 4. 399., 21. 412, Od. 2.134., 16.149. See however § 255. Possessive Article. It has been a question whether the Article is ever equi- valent to a Possessive Pronoun. If so it would be a kind of defining Article— defining a thing as belonging to a known person. In most of the instances, how- ever, the reference to a person is given by a distinct Pronoun: II. 19. 331 as dy pot Tov maida x.7.A.: Od. 11. 492 GAN’ aye por Too maubds x.7.A,: Od. 8.195 Kal x’ Ghads Tor... 7d ofpa: Od. 18. 380 od8 ay pou tiv yaorép #.7.A.: Od. 19. 535 GAN’ Gye por Tov bverpoy x.7.A,; Il. 1.167 ool 76 yépas Todd peifov: Il, 10, 321 GAA’ dye pot 70 aximrpoy dvdoxeo, Hence the Art. in these places has much the same function as with a Possessive (uo: roy naiéa=rdv éuoy maida); it reinforces the Pronoun which conveys the idea of possession. This account does not apply to rHs etvfs (Il. 9. 133, 275., 19. 176), and Tis dperfs (Od. z. 206). But the Art. is most probably substantival: rs eivy her couch, Tijs dpern her perfection, oe The Article as a Relative. The Article at the begin- ning of a clause may often be translated either as a Demonstra- tive or as a Relative. It has the character of a Relative when it introduces a distinctly subordinate or parenthetical Clause: as— I. 1. 36 ’ArddAor dvaxri, rov HKowos tTéxe Anré. The use of 6 4 16 as a Relative is less common in Homer than that of és 4 8, and is restricted in several ways : 1. It follows the Noun to which it refers; whereas a Clause introduced by és often precedes. 2, The Clause which it introduces does not define,—i. e. show who or what is denoted by the antecedent. E.g. rév 262..] THE ARTICLE AS A RELATIVE. 183 9 ok ys « nixopos téke Anté is meant merely to say something about Apollo, not to show who is meant by the name. But in Il. 1. 78 dtopwa dvdpa yodwoguev bs x.7.A. the Relative Clause is necessary to the meaning,—‘ dhe man who &e,’—the antecedent dvdpa being indefinite. It evidently follows from this that the Art. cannot stand as correlative to a Demonstrative Pronoun (183—38 that—which, not 7é—18): also that a Clause expressing condition must have 8s. The only exceptions to the first rule-— Tl. 1.125 adda ra pev modtwy eLempddouer, ra Sédacrat. Od. 4. 349 (=17. 140) AAAa 7a pey por erme . . TOY K.T.A, are also exceptions to this second rule. It seems to follow that the original reading in these places must have been dAAd @ & per. The other exceptions to the second rule are, in some instances at least, capable of explanation: e.g. in— : IL. 1. 319 ARy Epidos THy mpBrov emnmeiAne ’AxiAHi the meaning is not ‘the same quarrel which he had threatened,’ but his guarrel— which he had &e. And so perhaps in Il, 13. 220, Od. 13.126. Again— Il. 7.452 rod 8 émAnoovra 7d éyw nal PoiBos ’AnéAAwY KTA. and will forget the other—(a wall) which &e. But some MS. have & +” éya, which is free from difficulty. Od. 22. 392 Sopa eros etmwps 76 por KaTaddpudy éorrt that I may utter a word (not the word), which is in my mind, Cp. the recurring bgp’ efma 74 pe Ovpds x.7.A., and Il. 19. 339 mynodpevor Ta ExacTos evi peyaporow éXerrov. Real exceptions (if the text is right) are to be seen in Il. 9. 592 #n5e’ da’ dvOpd- mow TéAe Tov dotu ddd: Od. 1. 444 dd0v Tiyv méppad’ ’AOhvy: and perhaps a few other places (Od. 1. 17., 9. 334+, 13. 263.) 14. 227.) 19. 573+) 23- 355)+ 3. On the same principle, the Art. is little used where the Clause expresses a constant or essential characteristic of the Antecedent. In particular, it is not found in epexe- getic Clauses: as vymdxois, ofs ob TL méAet K.T.A.2 GpXer Kdkovs, al maou kakoyv K.t.X.: Aaprerldns, dv Adumos éyel- vato «.t.A. This however does not apply to the Art. with ré (6 re, 7} re, Td Te), as to which see § 263. If the Clause adds some new fact about a definite Antecedent, the Article is preferred. And sometimes even it has the effect of representing a fact as unexpected: as Il. 1.160 pds Tpder, rév ob tT perarpénn the Trojans—to whom all the time you give no heed: 1. 39% tiv por ddcav vies ’Axaidy (Briseis) whom the Greeks gave me (=although the Greeks had given her to me). Note especially the adverbial use of 16 = wherefore (§ 133): as— Il. 3. 176 GAda Ta y’ odk @yévovto* Td Kal Kdalovoa TéTNKa. 19. 213 TO poe ov TL meTa pect Tabra pépnhev. So Il. 7. 239., 12. 9., 17. 404., 23. 547. There is only one in- stance in the Odyssey, viz. 8. 332 (in the Song of Demodocus). 184 HOMERIC GRAMMAR. [263. 263.] The Article with ré serves as a Relative. In use it is intermediate beween the simple Article and the Relative 8s, for it expresses a constant or general characteristic, but only of a definite Antecedent: as— Il. 7. 112 “Exrope Tpuapldy, rov re orvyéover kal dAdo. 15.621 xdpatd te Tpoddevra, Td TE TpoTEpEvyETat adTHY. Od. 18. 273 ovrdouevys Cuber, Ths Te Zets CABov amndipa. It is especially used in similes (where a ¢ypical case is described), as Il, 13. 390 mirus BAwOpH tHv 7 otpeot KTA.: see § 266 fin. 264.| Homeric and Attic Article. After the account given in the preceding §§ of the Homeric uses of the Article it is hardly necessary to show in detail where they differ from the corresponding uses in Attic Greek. What we have chiefly to observe is that the difference is often greater in reality than it appears to be at first sight. Familiar as we are with the de- fining Article of modern languages, and of Attic Greek, we naturally import it into Homer whenever it is not made impos- sible by the context: and accordingly it is often assumed that the use in Homer is not consistent—that 6 is sometimes a De- monstrative, sometimes an ‘Article’ properly so called. In Homer the Article indicates, not that a person or thing is a known or definite one, but that it is presented to us in an antithesis or con- trast,—which need not be more than the contrast involved in turning our attention from one person or thing to another. Objects so contrasted are usually definite, in the sense that they are already known or suggested by the context: and hence the readiness with which the later defining sense can be applied to passages in Homer. Thus adrap 6 y’ jjpws can usually be trans- lated but the hero (before mentioned), as though 6 distinguished him from other heroes. But when we find that atrap 6 in Homer means but he, or but the other, and that it may be followed by an epexegetic Noun (as atraip 6 Boty tépevoev avag avdpdv ’Aya- peuver), we see that fpws is a kind of epithet—dut he a hero. This point has been explained in connexion with the use of the Attributive Article, § 258, 1 (with the note). It may be further illustrated from instances in which the Article marks contrast, but not definition, and where consequently it cannot be translated by the English the. Such are:— Tl. 15. 66 mod€as dd€cavr’ aicnods Tovs GAAous, pera 8’ vidv éudv Sapmnddva Siov not the others, but others as well, certain others. Il. 5.672 i) mporgpw Ards vidv éprydSodmoto didKor, 7} 8 ye TGv TAcdvwY Avkiwy amd Bvpodv Edo.To or should take the lives of more Lycians instead. Here ot mdéoves 265.| THE RELATIVE. 185 does not mean ‘the greater number,’ but ‘a greater number,’ in contrast to the one person mentioned. Tl. 22. 162 ds & or’ deOAropdpor wept répyara pdvoxes trmoe pupa pada tpwxGou 76 de péya Kelrar deOAov and there a great prize lies ready. So Od. 20. 242 airip 6.. opus but a bird. The same thing is shown by the position of the Article in punoThipwov Tov mev «.T.Ar. (§ 259, 2). It is evident that ray is used, not because the suitors are definite persons, but because a contrast is made by means of pév. The use with Cardinal Numerals (§ 260, ¢) is to be similarly explained. It is not peculiar to Homer, but is regular in Attic also, where it may be regarded as a survival of the Homeric use of the Article. It is a further question, and one that cannot be fully discussed here, whether any uses of the Article found in our texts of the Iliad and Odyssey are post- Homeric, and evidence of a later origin of the books or passages where they occur. It will be seen that in the case of the uses which have been noticed as rare or exceptional most of the examples come from books ix, x, xxiii, and xxiv. See especially the uses treated of in § 260 f, g, and § 261,3. Others again seem to belong to the Odyssey; see § 261, 3, and cp. § 259,1. The use of the Article in the tenth book of the Iliad seems clearly later than in any other part of Homer: e.g. Il. 10. 97 dedp’ és rods pvAakas KaraPetoper. 277 Xaipe 5t 7H dpe ’OSuceds. 322 % pev Tous inmous Te Kal Gppara k.T.r. (SO 330). 408 ms 8 ai Tay GAAwY Tpdwy pudanal #.7.A, Also rediov 76 Tpwirdy (v. 11), 6 TAHpwY ’Odvceds (Vv. 231, 498), THY vUKTA (Vv. 497). So in the Catalogue of the Ships we have @dpupw rév Opnixa (IL 2. 595), and 7d TleAaoycdv “Apyos (2. 681). fed & 265.] The Pronoun 3s # 8, and the Adverbs formed from the same Stem (ds, Ste, éws, &c.) are occasionally used in Homer in a Demonstrative sense; viz.— (1) After cai, od8€, pySé: as Tl. 21. 198 GANG Kal ds delSorKe even he fears: Il. 6. 59 pnd’ os pdbyo. may not even he escape: and often in the combinations kal &s even so, 008° @s not even so. (2) With pév and 8, to express a contrast between indefinite objects : Il. 11. 64 ds “Extwp ore pév re pera mparoist paveoker, dAdore & ev mupdroise xtA. (so 18. 599., 20. 49). 12. 141 ob 8 Fj tou elws pev Krr. up to a certain time. 17.178 bre 8 adrés énorpiver but sometimes fc. (3) &s thus is common ; especially és—és=as—so, 186 HOMERIC GRAMMAR. [266. These idioms may be regarded as the remains of an earlier use of 8s in the simple Anaphoric sense. The growth of a Relative out of a Demonstrative has been already exemplified in the Article (§ 262). The Demonstrative meaning need not be given to 6s in— Od. 4. 388 rév 7 ef mus ad SUvai0 Aoxnodpevos AcAaBEobat bs nev To elo Oddy H.T.A. him, if you can, lay in wait for and seize, in the hope that he will tell you: cp. Od. 5. 17., 10. 539. In the combination és ydp it makes little difference whether 8s is rendered as a Demonstrative (for he), or as a Relative (ds y’ up who as you suppose, who there- fore): the latter is more natural in parenthetical Clauses, such as 6 ydp yépas éort vepévrav which of course is the privilege of elders. 266.] 8s te, 85 ms. The simple és may be used in any kind of Relative Clause, although in certain cases (§ 259) the Article is preferred, Thus we have— Il. 4. 196 dv ris diotedcas €Badev (a particular fact). 1. 403 ov a kaAgovot (a constant or characteristic act). 2.205 els BactAeds, @ edwxe (a characteristic fact, defining). 1. 218 ds xe Oeots émumelOnra: (definition of a class). If the Relative is meant to refer to an indefinite number of individuals falling under a common description, 8s ts is gene- rally used, =who being any one, whoever. If, again, the Relative Clause generalises by making us think, not so much of all possible zxdividuals in a class, as of different times and circumstances,—in other words, if it lays stress on the general and permanent element in facts—és te is used: @. g.— Il. 1. 279 oxnnrodyos Bactheds, @ Te Zeds KBSos Edwxe to whom as king, to whom in any case. 5. 545 Addetod ds 7’ edpd peer x.7.d. (cp. 5. 876). 4. 361 Ta yap gpovées & 7’ yd Trep (such things as, &e.). Q. 117 dvnp bv te Zeds Khpr pdjon. Od. 1. 348 Zeds atrios, bs Te SiSwaw KTA.: 80 4. 105., 5. 52, &e. 7.74 ototy 7’ &b ppovénar they to whom she is well inclined. Thus és re is constantly used in comparisons: as IL. 3. 61 (wédexvs) és 7° elowy dia Sovpds tm dvépos Os fa TE K.T.A.: 5. 5 GoTEp’ dmapwd evadlyktoy Os TE MaALOTO, K.T. AL So ds te, 60. te, Sev te, Ste re: EvOa Te, ta Te: Saog Te, olds TE. Od. 12. 22 dicOavées, Gre 7 GAdot mak Ovijcxove’ d&vOpwror. 19.178 Kvwods, peyddy mors, Oa te Mivos x.T.d. We have seen above (§ 263) that 8 te is also used in com- parisons, and generally when the Relative Clause expresses a permanent characteristic, or treats the particular Antecedent 267.] RELATIVE ADVERBS, 187 as typical. If any rule is to be laid down for distinguishing the two uses, it must be that 8 te is properly confined to the use with a definite Antecedent ; whereas és te is free from this restriction. Thus 6 te in a comparison either makes us think of an instance as a single definite one, or refers (when the Plural is used) to the whole class. Thus in II. 5. 783 avol Kamporow Trav Te obévos odx ddAanadydv the Relative Clause applies to all boars. This distinction comes out in Il. 16.157 of 82 Avior ds wpopdryor, Totaly TE wept ppecivy daoneros dduh, of 7 Edapov KEepady .. Sdwrovew wtrA, Here toiolv te refers to all wolves, of re to the wolves of the particular simile. Thus Homer has jive Relatives, viz. 6, 8 te, ds, 5s Te, 8s 1s, each with a distinct shade of meaning: Attic retains only és and ds tis. 267.| Correlative Clauses. It is important to distinguish between the simple structure in which the Relative and Ante- cedent merely denote the same object, as— dvdpa Baroy bs x.7.A, having struck a man who &e. T&v ol viv Bporol eiat of those who are now living. év medi 601 mep K.7.A. on the plain where &e. and the yparailel structure, in which the Relative must be an Adverb with the same Ending as the Antecedent ; as— Tas b€ a” dnexOnpw as viv exmayN eplryoa and hate you in the manner in which I have loved you—where the manner is the point of comparison, and must qualify both Clauses. So— ropa & ém Tpdeccr Tiber xpdtos, dpp’ dv ’Axanol K.7.d. Th tuev xev by ov, Kedaivedés, jyemovedys. The difference between the two kinds of Clauses is that in the former the Relative affects a Noun or Pronoun in the prin- cipal Clause ; whereas in the latter it qualifies (or at least helps in qualifying) the whole Clause. The Relative Clause in fact serves as an Adverb (of manner, time, way, &c. as the Suffix may determine), construed, like the Antecedent Adverb, with the principal Verb. It follows that the Demonstrative Antecedent may be omitted without injury to the sense: so that és viv épidnoa may stand for tas—és viv épidnoa, spp’ dv for réppa—épp’ dv, and so on. In sentences of this type the fuller or Corre/ative structure is the older, and may be presumed to have preceded the other in every case. In this way we see how ds (lit. in which manner) came to mean in the manner in which; and so dppa to the time when, § by the way by which, 80 at the place where, sre at the time when, and so on. There is however a further stage of the use of Correlatives, viz. that in which the Relative has no proper construction in its own Clause. 188 HOMERIC GRAMMAR. [268. This may be most clearly seen in the use of odvexa (i.e. of Evexa) for which reason: e. g.— IL. 1. 110 &s 8% Todd vend oduv ExnBdros Gryea TevXeEL, obvek ey . . ovK eOedov K.T.A. Apollo causes sorrow for this reason, that I would not §c. Here we cannot translate oivexa for which reason: the reason does not precede, but is given by the Relative Clause. Hence the first évexa only is logical; the second (in otvexa) is due to the paral- lelism of the Clauses. Thus the correlatives rotvexa—oivexa come to mean for the reason—that. Or,—since ofvexa implies an antecedent rotvexa—the Relative otvexa by itself expresses for the reason that, because. To put the matter in a different way, the correlative structure is rational only when the relation between the clauses can be inverted: e.g. ws idov ds éudvnv I saw and was maddened in the same degree. But in (rotvera) ddyea Tevet Olver’ éyu ove eedoy the second clause gives the ground of the first, not vice versd. Therefore the totvexa is rational, the ovvexa is irrational; and when we translate ofvexa for the reason that, because, we really supply the necessary rovvexa, and render otvexa itself by a mere connecting Particle (¢hat). The same process may be traced more or less distinctly in all Relative Adverbs. Thus és (in which manner) comes to mean in such manner that, &pa for so long that, eis 6 to the point that, iva (lit. where) in order that ; also, as will be shown presently, 6, 67 and é te are adverbial Accusatives, meaning literally iz which respect, hence im respect that, because, that. By an inverse process the suppressed Antecedent may be without a construction. This is found in the use of &s =r ots, as Il. 4.157 ds a EBadror Tpwes since the Trojans have thus shot at you: and so otov dovce at hearing such a thing, oi’ dyopeves judging from the quality of your speech. Here &s comes to mean seeing the manner in which, oiov because of the thing such as, &. The looseness of structure here is due to the ellipse. The Antecedent being omitted, the want of a construction is not felt, 268.] odvexa, This Conjunction (which may be treated as a single word) is used in two ways: (a) to assign a cause or reason: (4) to connect the fact expressed in the Relative Clause with a Verb of saying, knowing, &e, The second of these uses is evidently derived from the first by a kind of degeneration, or loss of meaning. The fact told or known is originally given as the ground of the saying or know- ing. The transition may be seen in— Od. 7. 299 Ect’, 7} To. ev Todrd y’ évatoysov odx evdnoe tats éui) obverd o° ob Tt wer’ Gyimddotoe yuvatéw yey és tmerepov 269.] ‘O, ‘OTI, ‘O TE. 189 my daughter did not judge aright in this, because she did not &e., more simply, im this, that she did not §c, Again— Od. 5. 215 otda kal adrds ndvta mad’, obvexa oelo tepippav TIqveddmera xrA. I know all, inasmuch as Penelope is oo i.e. LT know that she is &c. This use is found with Verbs of saying in Od. 13. 309., 15. 42., 16. 330, 379. In the Iliad it occurs only once, viz. Il. 11. 21 weddeTo . . wéya Kdéos, otver ’Axatol KrA. Note that (except in Od. 13. 309., 16. 379) the Verb is followed by an Acc. of the thing; so that the Relative Clause does not directly take the place of the Object. Thus (e.g.) wevOero KAéos otvexa is literally heard a rumour the ground of which was that &c. 269.| 8, 8,8 re. The Acc. Neut. of the Relative, when used adverbially ($ 133), yields the three ‘Conjunctions’ 8, ém and 6 te, which mean properly iz respect that, hence usually (a) because, or (4) that (after a Verb of saying, knowing, &c.). (1) 6 im respect that, because may be exemplified by— IL. 20. 283 rapByoas & of dyxt mayn Bédos dreading because the dart stuck near him. We have here the phenomenon already noticed in otvexa, viz. the Relative has no construction in its own Clause, but reflects the construction of the Demonstrative in the principal Clause: rapByjoas (rd) 6 mayn Bé\os. The Cognate or adverbial Accusative with rtapByjcas would express the nature or ground of dread (as in ré ye deldi6t, 7éde xéeo, &c.); hence the meaning dreading in respect of (or because of) this, that the dart stuck, Accordingly we find 6= because chiefly with Verbs of feeling, which regularly take a Neuter Pronoun of the ground of feeling :— Od. 1. 382 TnrA€uaxov OatuaCov d Oapraréws dydpeve. So Il. 9. 534 (xaoapévn), Od. 19. 543., 21. 289 (odk dyanGgs 8). A peculiar use to state a consequence which is made a ground of inference may be seen in— Od. 4. 206 rotov yap kat marpds, 5 kal menvupéva Racers Sor you are of a wise father, (as I know) because you speak wisely : so Od. 18. 392, and probably also— Tl. 21. 150 rls wédev eis dvdpGv, & ev errAns avtlos edOetv ; who are you that (i.e. I ask because) you dare &c. The transition to the use of ’=¢hat may be seen in— Od. 2. 44 otre re djpuov GAdo mipatexouat odd’ dyopedw GAN eudv adrod xpeios, 6 wor Kaxov EuTeren oiKw what I tell is my own case (which consists in the fact) that evil has fallen on my house). It is common with oida (Il. 8. 32, Od. 4. 771, &e.), yyvdone (Il. 5. 433-, 16. 120, &e.) dtw (Il. 15. 248): 190 HOMERIC GRAMMAR. [269. and found with Verbs of seeing, as Il. 1.120 Aevooere yap Td ye mdvres & yor yépas épxerar GAA ye see this, that my prize goes else- where (ep. Il. 19. 144., 22. 445, Od. 17. 545)- (2) 8n because is common after Verbs of feeling. We need only stop to notice some instances (parallel to those of 6 just quoted) in which it is=as I know because : Il. 16. 33 vnreés, odk dpa coi ye narip iy inndra TInrets, odds Oéris pairnp, yAavky b€ ve Tikre Oddagaa, mérpa 7 7ALBaror, Ste Tor vdos early aanvijs meaning now I know that you are no child of Peleus §c., because your mind is relentless. Il. 21. 410 vyrire, otdé vi 7d Tep enedpdow Socov dpelwv ebyop’ éyov epevar, Ott por pevos dvtupeplcets. Od. 5. 339 Kdypope, tinre ro Gde Tloceddwv évorixbav @dvoar’ éxadyAws, Ste ToL Kaka TOAAG HuTedver why is Poseidon so enraged against you (as he seems to be) since he causes you many evils? So Il. 10. 142., 21. 488., 24. 240, Od. 14. 367., 22. 36. The transition to the meaning that may be seen in— TL. 2. 255 joa dverdicwy ore of pdda ToAAG bid00cr reproaching him in respect that, with the fact that Se. It is the regular meaning with Verbs of knowing: Tl. 8.175 yrypdoxno 3 dre por mpddpwv katévevoe Kpovlwy I know that &e. The use of $r.=¢hat is commoner in the Iliad than in the Odyssey (where os and otvera partly supply the place, see § 268). (3) The form 8 te (so written by Bekker to distinguish it from ére when) is found in Homer with the same varieties of meaning as 6 and én, Thus we have— Tl. 1. 244 xwdpevos 6 1° dpioroy ’"Axardy oddev Ericas angry because Sc.; so Il. 6.126., 16. 509., Od. 8. 78. Il. 4. 31 daimovly, rl vt ce Tplayos Ipiduoid re aides Téooa Kaa peCovat, 6 T° domepxes peveatvers how do Priam and his sons do you such evil, (as they must do) since you are furiously enraged? So Od. 14. 89 olde b€ ror toaor.. 6 tT’ ovk eOédovar they know something (as is plain) because they are not willing: Od. 21.254 tooodvde Bins émidevées eluev ..6 7’ ov duvduerOa we are so wanting in strength, as appears by the fact that we are not able, And with the meaning that— Il. 1. 411 yuo 8& Kal Arpeldns, edpuxpetov ’Ayapeuver, jy arnv, 67 apicrov ’Axady ovdev ericev may know his folly, in that he failed to honour Sc. Od. 14. 365 eye 8 &b ofda Kal atrés voorov éueio dvaxtos, 6 7 #xOero maou Oeotas 270. ] ‘0, ‘OTI, ‘0 TE. 191 I know of the return of my lord, that (as it showed) he was hated by all the gods. So Ul. 8. 251 etdov’ 67’ dp’ rd. saw that Se.; and with yiyvdoxe, Il. 5. 331, &e. The existence of a distinct 6 re with the meaning because or that depends upon its being shown .that in places such as those now quoted the word cannot be either bru that or Ste when. The latter supposition, though often admissible in parti- cular passages, seems excluded by several examples,—e.g. Il. §. 331 yeyvhonxov 6 7° dvadsms qv Oeds,—and generally by the complete correspondence of meaning between 6, 87, and 6 te. On the other hand it is extremely improbable that the v of Stu was ever capable of elision. In this respect drt that stands on the same footing as tt and 6 71. It has to be considered, too, that in Homer the adverbial use of these words, which gives them the character of Conjunctions, is hardly more than a particular use of the Accusative Neuter. If then the Neut. of ds and $s tts is so used, it is difficult to see any reason why the Neut. of the equally familiar 3s te should be excluded. The only objection is that in most of the instances (perhaps all the clear instances) which support 6 te the final vowel is elided. Why do we find 6 7’ and 6 , but not 6 re, if it existed? The answer must be that 6 te would be liable to be altered into the familiar 61. wherever sense and metre alike permitted. The MSS. vary in some places between Ste and 8tv (as in Il. 14. 71, 72, Od. 13. 129), and on such a point we cannot be said to have any good external authority. In Il. 16. 35 Ar. read Sve. And there are one or two places where all the MSS. give 6 re or Ste (unelided) with the meaning because ; IL. 15. 467 & wéror F 5 mdyxu paxns emt ppdea xeiper Saivow huerépys & Té por Bidy exBare yetpds surely God is frustrating us, (as I judge from this) that he has thrown the bow Srom my hands. So— Od, 5. 356 @& por eye, wh tis por bpatvnow SdAov abre dbavatww, 6 Té pe oxebins dmoBva dvdrye i.e. ‘there is another snare in this bidding me to get off the raft.? So perhaps Od. 13.129 6 Té pe Bporol ob rm Tlovor (I shall not be honoured by the gods, as I judge) because mortals do not honowr me: and even (with La Roche) Il. 1. 518 # dn Aoiyio. epy’, 6 TE pw “TA, it is a pestilent thing that you &c., Il. 16. 433 & poe éydv, 8 re alas for me that &e., and ll. 19. 57. 270.| 6, 61, 6 re as Conjunctions. In a few instances it is impossible to explain these Relatives by supplying an Accusative +6 in the principal Clause. Thus— Tl. 5. 349 7} odx GAts Orre yuvaikas avddxidas rmepoTrevets 5 Here the principal Clause is Impersonal, and the Antecedent might be a Nom. (is 7¢ not enough) or Gen. (is there not enough im this), but hardly an Accusative. Again in— Tl. 8. 362 od8€ re rOv méeuvnrat, 6 of wdda ToAAdKLS KTA, 17. 207 TGv Towny, 6 Tou KTr. (as amends for the fact that) the Relative Clause serves as a Genitive: ep. Od. 11. 540 yn@o- ovvn & of KTh., 12. 374 dyyedos jAGev..8 of krA. Further, in— Od. 20. 333 viv & Hdn Téd€ SHAov, 6 7 odKérs vdoTIWds eore 192 HOMERIC GRAMMAR. [271. it is in Apposition to a Nom., and so in Od, 24.182. Add II. 9. 493 Ta ppovéwy & pot KTA., 23. 54.5 TH Ppovéwy Gru of KrA,: and also Od. 2. 116 ra dpovéova’ ava Ovudv & of xrd., where the v. 1. 8 for & has good MS. authority. In these instances, then, the forms 8, &c. have ceased to be felt as Case-forms, and may properly be termed Conjunctions. It is worth while pointing out the parallel between this exceptional construction of the Clause with & &c. and the extension of use which has been observed in the Infinitival Clause (§ 234). At first the Acc. is allowed because the Infinitive serves as logical Object: and accordingly the construction is found with Verbs that take a Neut. Pronoun as Acc. of the thing, viz. Verbs of saying, knowing, feeling, &c. (§ 237, 2). The next stage consists of the use with Impersonal phrases (as dpyadéov éort), which is also Homeric, and forms the transition to the use of the Infinitival Clause as a Subject, and so to the (post-Homeric) use of the Infinitive as an indeclinable Noun, Similarly the Clause with 6 or tt is primarily equivalent to an Accusative of the reason (§ 133), and is extended (but in a few Homeric examples only) to the relation of a Nom. or a Gen. The three forms 6, 6 re, dt do not differ perceptibly in meaning. Hence the reduction in Attic to the single 671 is no real loss. 241.] Form of the Relative Clause. It is characteristic of the Relative Clause that the Verb fo de is often omitted: as— IL. 8. 524 pidos & ds pey viv dyujs, eipnuevos éora, and so éccot ’Axatol, of rep dptorot, 4 Tus dplotn, Os 7 alrios 8s Te kat od«l, &e. Hence we should write in Il. 11. 535., 20. 500 dvruyes at wept dipov, in Il. 21. 353 ixOves ot xara divas. So with Relative Adverbs; as Od. 10. 176 6dpp” ev vyt don BpGais re méots te 80 long as there is food and drink in the ship. This ellipse leads to a peculiar ‘ Attraction’ of the Relative to the Case of the Antecedent, found chiefly with soos re in the Odyssey, as— Od. 10. 113 tiv 8& yuvaika | etpov Sony 7’ dpeos Kopudyr, which is equivalent to tréonv don eo) kopudi; and so écov Te, Od. 9. 322, 325, 10. 167, 517., 11.253 also ofdy re, Od. 19. 233. The only instance in the Iliad is somewhat different :— I. 1. 262 0d ydp mw rolovs tdov . . oloy TetpiOoov x.7.d. The later Attraction of the Relative into the Case of the Antecedent is not found in Homer: but there are some instances which may be assigned to ‘Inverse Attraction,’ i.e. in which a Noun is put in the Case of a following Relative: as— I. 14.75 vijes Soar mpGra eipdarar dyxs Oaddoons Ekopev K.T.A. (SO 6. 396., 10. 416., 14. 371). Kihner gives, as an example of Attraction, Il. 5. 265 ris ydp rot yevefs As Tpwt wep eipudra Zeds dxe. But there the Gen. may be partitive: ‘the brood from which Zeus gave’ (§ 151 e). So Il. 23. 649 (§ 153). 272,| RELATIVE IN SUCCESSIVE CLAUSES. 193 272.] Double Relative Clauses. When two or more Clauses connected by «ai or 8é are introduced by a single Relative, it need not be construed with any Clause after the first: ¢e.g.— I. 3. 235 obs kev eb yvotnv cal 7’ obvoua prOnoalunv and (whose) names I could tell. So, with change of Subject,— I. 1. 162 @ émt T6AN eudynoa, ddcav b€ por vies Ayadv for which I toiled, and which the sons of the Greeks gave me. Od. 2. 114 76 bred Te maThp Kéderau Kal dvddver adtH and who is pleasing to herself. The Relative is not repeated in any Clause of this form; but its place is often taken by another Pronoun (usually an enclitic, or an unemphatic avrds) :— IL. 1. 78 7 yap diowar dvdpa yohwoguev, ds péya tavrev *"Apyelwy kparées al ot meiOovrar ’Ayanol, Od. 9. 19 ei’ ’Oduceds Aaepriddns, ds tact SdAoLow dvOpadrovot peA@, kat peu KA€os otpavon tke. This idiom, it should be observed, is not peculiar to Homer, but prevails in all periods of Greek (Kiihner, IT. p. 936). Successive Relative Clauses not connected by a Conjunction are frequent in Homer. In one or two places the Relative seems to be repeated for the sake of emphasis: Od. 2. 130 ddyov déxovoay dndcat 7 pw’ rey’ 7} w EOpewe. Sometimes the second of two such Clauses is epexegetic of the first : as— Il. 5. 403 cxérAtos, dBpmuoepyds, ds ote d0er alcvra péCav, bs réLowow exnde Oeods (so 6. 131., 17.674, &c.). Or it marks the return to the main thread of the narrative: as— Od. 14. 288 67 tore Botwé HAGev avijp, amarjrsa €ldds, , a S x ae) Z wa TP@KTYS, Os 67 TOAAG Kak’ GvOpaeTrOLoLY eapyel, bs wv dye mapremOov «rd. (cp. Il. 15. 461-3). Where different Pronouns are used as Relatives in successive Clauses, the reason of the variety may often be traced. Thus in— Il. 16. 152 év 5& mapnopinoww dpopova MfASdacov ver, tov pd tor’ "Hetiwvos éhov méduy Hryay ’AxiAAcUs, bs wal Ovnros ea Exe6’ inmors AOavaro.st, 20, 233 ds 51) KaAALoTOS yéveTo OvnTay dvOpuTwY, Tov Kal dvnpelpayTo Ooi KTA. the Clause with rév gives a single fact or event, that with 6s a characteristic or permanent circumstance. Again, we find és expressing a single fact, while 6s te introduces a general assertion: as LL. 5. 545 “AAgeiod, bs T° edpd feet TvAlwy bid yains, 8s réxer’ "Opaidoxov. The difference between és tis and és te appears in Od. 6, 286 Kal 8 dAAn vepeod H Ts Toad7a ye péCor, 7” Géuenre pirwy watpos kal pntpods edyvrwv dvdpact ploynta. Here 4 ts insists on the inclusion of all members of the class (any one who —), 4 Te prepares us for the class characteristics (one of the hind that —). 194 HOMERIC GRAMMAR. [273. CHAPTER XII. Usrs oF tae Moons. Introductory. 273.] Classification of Sentences. Before entering upon an examination of the Homeric uses of the Moods, it will be con- venient to give some account of the different kinds of Sentences and Clauses with which we shall have to deal. A Simple Sentence—or the principal Clause in a Complex Sentence—may be purely Affirmative. Or, the affirmation may be turned (either by the use of a suitable Pronoun or Particle, or by the tone and manner in which it is uttered) into a ques- tion: ¢.e. the Sentence may be Interrogative. Or, a predication may be framed in order to be denied: in which case a Particle is added to make the Sentence Negative. Or, the Sentence may express Wish, Purpose, or Command ; and any of these may again be combined with a Negative, so as to express some variety of Prohibition. Or, once more, the Sentence may be Conditional, z,e. may assert, deny, command, &c. subject to a hypothesis ; and this hypothesis or condition may be expressed by a subordinate Clause, or by an Adverb or adverbial phrase (then, in that case, or the like): or the condition need not be expressed at all, but conveyed by the drift of the context. A subordinate Clause may be so loosely connected with the principal Clause as to be virtually an independent sentence. We have seen that this is generally the case (for example) with Clauses introduced by the Article (§ 262). The Clauses which chiefly concern us now are— 1. Dependent Interrogative Clauses. 2. Prohibitive Clauses (p= Zest). 3. Relative Clauses proper (introduced by és). 4. Clauses introduced by a Relative Adverb (as, 561, 86ev, dre, ws, dppa, &c.; also évOa, va, and éme/). 5. Clauses introduced by et 7/. This classification is based upon the grammatical form of the Clause. If we look to the relation in point of meaning between the two Clauses of a Complex Sentence, we find that subordinate Clauses fall into-a wholly different set of groups. Thus there are— (1) Clauses expressing cause or reason: as— Il. 2. 274 viv 8€ rdde péy’ Gpiorov ev ’Apyelouowy epeter, ds TOY AwBnTijpa émecBdAov eo’ ayopdwv. 273.| THE MOODS. 195 And so in clauses like Il. 4.157 és o° ¢Badrov Tpdes since the Trojans have thus shot at you ; 6.166 olov dxovoe at hearing such a thing (§ 267 fin.): as well as in the regular Causal use of 8, drt, 8 te ($ 269), and odvexa, (2) Clauses expressing the Odject of Verbs of saying, knowing, thinking, &e. (i.e. the fuct or thing said, &c.): as— Tl. 2. 365 yudon ered bs 0 iyeudver Kaxds, Os Té vu AaSv. Od. 6. 141 6 0& pepunprEev "Odvoceds | i}. . H KTA. Jl. 18.125 yvotev & as 8) Synpov ey worA€poto wéTavpat. 601 Teipyoerat at Ke Oénow (tries if it will run), (3) Clauses expressing condition or limitation ; which may be introduced— By és: as tay ot viv Bpotoi claw of the mortals now living: 85 « émdeuns he who is in want: bs Ke Oeots emumetOnrar he who shall obey the gods: 8 tt ob eloatto whatever seemed to him. By a Relative Adverb: of manner, as os émiré\dw as I bid, ds dv éywv elnw as I shall speak ; of time, éwet, dre, &c., also Ews and édpa when they mean so long as; of place, as émnd0c mdratoy medlov where is the richest of the plain, By ei—the common form of Conditional protasis. It will be convenient to term all these Clauses ‘ Conditional ’"— the word being taken in a wide sense, so as to include every Clause of the nature of a definition or limitation, as well as those in which actual priority in time is implied, (4) Final Clauses, expressing end or purpose: ntroduced— By és; as Il. 4.190 émiOjoe pdppay’ & kev navonor will apply drugs which shall stay: Il. 14. 107 viv & ely bs .. évionot may there be one who may tell. By as, énws, tva—the ordinary forms. By éws and é$pa, when they mean ¢id/ such time that: to which we must add cis 8 wntil, which (like otvexa) is practically a single word. By «i: as Il. 1. 420 efw adr. . al xe wlOnrar I go in the hope that he will listen. By prj lest (=tva pi). It is important to observe that the several groups of Clauses now pointed out are generally indistinguishable in respect of grammatical form; so that Clauses of the same form (introduced by the same Pronoun or Particle, and with a Verb of the same Tense and Mood) often bear entirely different meanings. This will be shown in detail in the course of the present chapter; meanwhile a few instances may be noted as illustrations. 1. Final Clauses introduced by 6s are in the same form as the Conditional or limiting Clauses such as &s we réxy, S77 Kev etmps, &e, 02 196 HOMERIC GRAMMAR. [274. 2. The regular Final Clauses with 4s and Smws are in the same form as the limiting ds dy éydv eirw as I shall speak, énus @0édnow as he pleases, &c. 3. Clauses with p may either be Final (when ph=iva pi), or Object-Clauses after a Verb of fearing (Sel5w un). 4. Clauses with €ws and éppa may either be Conditional (when the Conjunction means so long as), or Final (when it means wnftil). : 5. The Final Clause with ei is indistinguishable in form from the ordinary Con- ditional Protasis: compare al xe mi@nra: to see if he will listen with Il. 24. 592 un pot Tar poxre oxvdpawéeper al xe TUOnat be not angry in case you hear. From these examples it is evident that in this as in so many parts of Greek grammar the most important differences of meaning are not expressed by corre- sponding distinctions of form. The Pronoun or Conjunction which connects the subordinate with the principal Clause generally leaves the real relation between the two Clauses to be gathered from the context. These different kinds of Sentence are distinguished to some extent by means of Particles which it will be convenient to mention here, reserving a fuller account of them for a later chapter : (1) Strong Affirmation is often expressed by 4, and the same Particle is also employed in Luterrogation (especially with ironical force). (2) Negation is expressed by oki (otk, ot), Prohibition by py. (3) The Particle ei, in its ordinary use, marks a Conditional Protasis, 2. e. a Clause stating a condition or supposition. (4) The Particles é and xé(v) mark a predication as Condi- tional, or dependent upon a condition. It is not necessary that the condition should be expressed by a Clause with é¢i, or even that it should be distinctly stated in the context. It is enough (as we shall see) if the predication (assertion, denial, purpose, command, &c.) is made in view of some /imitation to particular conditions or circumstances. The Subjunetive—in Principal Clauses. 244.] The Subjunctive in a Simple Sentence, or in the Prin- cipal Clause of a Complex Sentence, may be said in general to express either the wid/ of the speaker or his sense of the neces- sity of a future event. Like the English must and shall, by which it may usually be rendered, it is intermediate in meaning between the Imperative and the Future, Sometimes (as in the ‘Hortatory’ touev det ws go, or in Prohibitions with p4) it is virtually an Imperative; sometimes it is an emphatic or pas- sionate Future. These varieties of use will be best understood if treated with reference to the different kinds of sentence— Affirmative, Interrogative, Negative, Prohibitive, &c.—in which they occur, 275. | THE SUBJUNCTIVE. 197 275.| In Affirmative sentences the force of the Subj. depends in great measure on the Person used. (z) In the First Person the Subj. supplies the place of an Imperative, so far as such a thing is conceivable: that is to say, it expresses what the speaker resolves or insists upon doing ; ¢. 9. Tl. 9. 121 tyty & év mdvresoe mepixduta dGp’ dvoyjnve (where the list of gifts immediately follows). Od. 2. 222 ofjua ré of xed kal ent xrépea xrepetw TOAAG par doca éorke, Kal dvépe pntépa ddow (the Subj. expresses the decisive action to be taken by Telemachus, viz. to acknowledge his father’s death: the Fut. dé0 expresses what would follow as a matter of course). 12. 383 dvcouat ets Aldao kal ev vexverot hacivw (said by way of a threat), Hence after a Clause containing an Imperative the Subj. is used to show what the speaker will do as his part ; as— Il. 6. 340 GAN dye viv éninevor, dpjia rebxea Siw wait, and I will put on my armour. 22.416 axéade pido katy’ ofoy edoate knddpevov TEp efeAOdvra moAnos tkéoO emt vias ’Axarov" Nicoww avépa TobTov KTA. 450 Setre, dvw pot emecOor, op Sr’ epya TérvKTat. So after the phrases adn’ dye, ei 8 dye, as Od. 6.126 add ay’ evar abros Teipjooma de Wopar: 9. 37 «i 8 dye Toe Kal véorov emov modvkndé’ évionw. On the phrase ei 8 dye see § 321. To show that a purpose is conditional upon something else being done, the Subj. may be qualified by the Particle kév : Il. 1. 137 ef 8€ ké pH ddwow, eyo d€ Kev adtos EXopat —f they do not give her, I wilt (in that case) Se. 14. 235 melOev, yo d€ Ke Tor eldew yap obey, and I witl be.thankful. 16. 129 dvcceo Tedxea Oaccon, ey dé Ke Aady dyelpw. Od. 17. 417 76 ve xpy Sdpevat Kal Adiov Hé wep GAAOL aitou éyw b€ Ké we KAElw KTA, So too Il. 1. 183 ri pev .. meno, eyo 8 « dyw Boonida I wilt send her (as required) and then I will take Briseis—the Subj. ex- pressing the speaker’s own threatened action, and xév marking that it is the counterpart to what is imposed upon him. This use of xév with the Subj. is not found except in Homer. It appears to be the rule that «év is used when the Clause with the Subj. is introduced by 8é€, but not when it follows without a connecting Particle. The First Person Plural is similarly used, as Od. 3.17 ddd’ aye 198 HOMERIC GRAMMAR. [276. viv lOds xle Néoropos immoddoto’ efSomev xtA. And so in the common ‘ Hortatory’ Subj., as pevywuen let us fly. (2) A Subj. of the Second and Third Person in an Affirmative sentence is usually an emphatic Future, sometimes approaching the force of an Imperative. The only example of a pure Subj. (i.e, without dv or xev) in this use appears to be the phrase kai moré tis eimnor and men shall say (Il. 6.459, 479., 7.87). With éy we find— Il. 1. 205 jis tmepordigot tay’ ay more Ovpoy SAnrat (in effect a threat of what the speaker will do). 22, 505 viv 8 dv woAAG TAONo1 Pidrov amo TaTpds dwapTay now he must suffer much Sc. With xéy the examples are rather more numerous :— Od. 1. 396 rap Kev Tis Tod’ exnow, emel Odve dios ’Odvaceds let one of them have this (emphatic assent). 4. 80 dvdpav 8 i Kév ris por epicoerat He Kal odki. 10. 507 Hoda, rHv d€ KE ToL TOW Bopéao pepyot sit still, and her (the ship) the breath of Boreas shall bear along (solemn promise). IL. 9. Jol GAN 7 rot Keivov pev edooper, 7 Kev tnow 7 Ke evy(et him go or let him stay): ep. Od. 14.183. Note that where two alternatives are not expressed by the same Mood, the Subj. gives the alternative on which the stress is laid : Il. 11. 431 onpepov 7) Sovoiow emevgea .. 7 kev ud Und doupt TuTEls d7d Ovpoy dr€oogs. 18. 308 orjooua, i Ke Pépnot péya Kpdros # Ke hepoluny I shall stand firm, let him gain the victory (= though he shalt gain) or I may gain it. Od. 4. 692 dAAov x ex daipnot Bpordy, ddAdov Ke didroin a king will (is sure to) hate one, he may love another, A curious combination of Opt. and Subj. is found in— Tl. 24.654 adrix’ dy e&etmor Ayapeuvov, rome Aadv, kat Kev dvdBAnots Avovos vEeKpoio yévnras he would straightway tell Agamemnon, and then there must be a delay in the vansoming of the dead. The Subj. is used to express the certainty of the further consequence, as though the hypo- thetical case (avrik’ dv é£elmor) had actually occurred, 276.] In Negative Clauses properly so called (¢. ¢. distinguished from Prohibitions) the Subj. is an emphatic Future. We find— (a) The pure Subj. (expressing a general denial) : I. 1. 262 od ydp mw rolovs (ov avépas, odd wpat L have not seen—TI never shalt see. 278.) THE SUBJUNCTIVE. 199 Il. 7.197 od ydp ris we Bin ye éxdv déxovta dinra no man shall chase me against my will. 15. 349 otd€é vu tév ye yvoorol re yuwral te mupds AcAdXwot OavdvTa. Od. 16. 437 odx &00’ ofros duijp ob8 ~roerar ovde yévnrat there 1s not, there never will or can be, the man who &e. (so 6. 201). 24.29 poip’ ddon, Tv ob Tus GAeverat (Cp. 14, 400). (8) The Subj. with av: Il. 3. 54 ob« dy rot xpatopn KiBapis KTr. be sure that then your lyre will not avail you. 11. 386 ef yey d7 dvTiBiov obv Tedyeor TEipnOetys, ovK ay ToL xpalopyoe Bids KTA, The reason for av in these places is obvious: in the following instances it seems to be used because there is a contrast :— Il. 2. 488 wAnddv 8 otk dv eyo pvOjoopar 003’ dvopjve but the multitude I cannot declare or tell by name Od. 6. 221 dvrnv & odk ay éywye ogccopa (here dvryv is an emphatic limiting word: ep. Od. 4. 240., 11. 328, 517). 294.) In Interrogative sentences the Subj. generally expresses necessity, submission to some command or power; as Il. 10. 62 adOe pévw . . He Oéw tr. am I to remain here, or am I to run &e. ; Od. 15. 509 7H yap éyd, pire réxvov, iw; Ted SGpa0’ Tkwpas KTA, where am I to go? to whose house, &c.: Od. 5. 465 & pou yd, té Td0w; Te vd por pjciota yevyta ; what am Ito suffer? what is to become of me? And rhetorically, with an implied negation— Il. 18. 188 m&s 7’ dp’ tw pera pGAov; exovar dé Tevxe’ exetvor how can I go into the battle? They have my arms, Il. 1. 150 w&s ris rot mpddpwv emeoi welOnrar Ayadv ; One or two passages given by Delbriick under this head should be classed as examples of Subordinate Clauses, A transitional instance may be seen in Od. 22.166 ov 5é poor vnpeprés éviomes, 7 puv amoxTeivw .. He col evOad ayw utr. tell me,am I to kill him, or bring him here? Here the Clause may be treated as a distinct sentence; but this can hardly be said of Il. 9, 618 Gua & jot pawwo- perngy ppaccdpued’ 4 xe vedped’ KTA., because the Clause does not express an actual but an intended future deliberation. Still less Od. 16. 73 pytpt 8 éupn diya Oupds évt ppect pepynpiCe 7) adrod map’ éuol Te wéevp KTA., where the form of expression is changed from the First to the Third Person (as in oratio obliqua). These are therefore examples of the Dependent Deliberative Subj. (§ 280), and rank with the Subordinate Clauses that furnish an Object to Verbs of saying, knowing, thinking, &e. 278.| With the Prohiditive Particle p4 the Subj. has the cha- racter of an Imperative. We may distinguish however— 200 HOMERIC GRAMMAR. [278. (2) Direct forbidding, usually with the First Personal Plural (answering to the Hortatory Subj.), and the Second Person Sing. ; sometimes also with the Third Person, as— Il. 4. 37 &p£ov drs eOéreis* pt Tobrd ye velkos dnicow cot kat éuot yey’ propa per’ auorépoicr yévnrar I do not want this to become a quarrel. Od. 22. 213 Mévrop, pa} o° éméecor mapareniOnow ’Odvaceds see that Ulysses does not persuade you. And with the First Person, as Il. 1. 26 pn ce xtyelw let me not catch you; Il. 21.475 wh cev axovow. (4) Fear, warning, suggestion of danger, &e.; e.g.— Tl. 2.195 pj te xoAwodpevos pen (I fear he will Se.). 5. 487 py Tews as ayior Aivov GAdvre Tavadypov avdpdor Svopevéccow EXwp Kal kbpya yévnode see that you do not become a prey Sc. 22.123 pa} pay eyo wey tkapar idv, 6 5€ pw odK ehenjoel. Od. 5. 356 & por eyd, un tis wot bpalyynow dddov adre adavdtwv (I hope some god is not weaving &c.). 18. 334 pay tis Tor Taxa “Ipou dyelvav GAAos avaoth see that a better than Irus does not rise up. The construction is the same in principle when a Clause of this kind follows a Verb of fearing ; and it is sometimes a question whether the Clause is subordi- nate or not. Thus the older editors (including Wolf) punctuated IL. 11. 470 deidw, pn tT 740no1.—as though Sel8 were parenthetical. It is more probable, however, that in such cases the Clause with p has acquired a subordinate character, serving as Object to the Verb (thing feared) ; see § 281. On the other hand, Clauses with pf are often explained by supposing an ellipse of a Verb of fearing: ph féén for Sel8m pu féey. This is open to the general objection that it explains a Simple sentence by supposing a Complex one: and moreover it gives a different account of the formation of Clauses which are essentially similar. For the Clause wu) péép I will not have him do (hence I fear he may do) is identical in form with pi pégns I will not have you do. Similar questions may arise regarding the Final Clauses which are formed in the same way as the Prohibitive Clauses now in question. Thus in Il. 1. 586-7 Térrabt, uTep éuy, .. wh oe... iwpor we may translate endure, mother ; let me not see you &c., or (bringing the two Clauses more closely together) endure, lest I see you de, No clear line can be drawn betweeu independent and subordinate Clauses: for in this as in other cases the complex Sentence has been formed gradually, by the agglutination of simple Clauses. The combination ph oJ—prohibition of a negative—is ex- tremely rare in Homer. In II. 5. 233 pi to pev deloavre pary- cetov ovd edédynrov, and I]. 16.128 pi dy vias EAwou Kal odKeére ura wéAwvrat, the Particles are in distinct Clauses. There is no clear instance of pi} ov in an independent sentence. It occurs ima Final Clause, Il. 1. 28 jan vd toe od xpaioun xrr., Tl. 24. 569: and after Sei8w in Il. 10, 39 Seid yu} od Tis Tou KTA, 280.] DEPENDENT DELIBERATIVE USE. 201 It is well known that the Present Subj. is not used as an Im- perative of Prohibition (with p4). The rule is absolute in Homer for the Second Person. The Third Person is occasionally used when fear (not command) is expressed ; the instances are,— Od. 5. 356 (quoted above); 15.19 prj vd te. . pépytar; 16.87 yuh Juv Keptouewowv. The restriction does not apply to the ‘ Horta- tory’ use of the First Person Plur., as Il. 13. 292 pnxéri tadra -heydueba. We shall return to this subject in connexion with the corresponding rule which forbids or restricts the use of pi with the Aorist Imperative (§ 327). 279.| Homeric and Attic uses. In Attic the use of the Subj. in independent Clauses is either Hortatory, or Deliberative, or Prohibitive. Thus the use with ay (§ 275, a), the use in Affirma- tion (§ 275, 0), and the Negative uses (§ 276) do not survive, The Subjunetive in Subordinate Clauses. 280.] Clauses with 4¢—fe. Doubt or deliberation between alternative courses of action is expressed by Clauses of the form HE (4)—Ye (f) with the Subj., dependent on a Verb such as ¢pa- (ouat, peppnpitw, &c., or an equivalent phrase: ¢.g.— Il. 4. 14 tets 8 ppatdued Stas Cora rade epya, Hp adris méAeudy TE KaKOy Kal pvAomLY aiviy Opoopev, 4 iAdTyTA per’ Gudoreporot PddAwper. Od. 19. 524 @s Kal euol dlya Ovuds dpwperar evOa Kal evOa, He wevo .. 7 On Gy? Em@pat KTA. (cp. 22. 167). So, of doubt as to which of two possible results of the speaker’s action will be realised :— Il. 13. 327 eiSoper, Hé Te ebyos dpéLopev, HE Tis Hutv. This form is also found (but rarely) expressing, not the speaker’s own deliberation, but that of a third person : Od. 16. 73 pnrpl & eujj dixa Ovpos evi ppect pepunpicer, h avrod map’ éuol re wévy Kal d6ua couiGy, KTr. The speaker (Telemachus) here expresses himself from his mother’s point of view, only putting the Third Person for the First. So of mere doubt as to a result— IL. 16. 243 eicerae fa Kal otos éxlorynrar modeulCew HueTepos Oepatrov, 7 of KTA. where énfornrat (will prove to know) is used nearly as the Latin Subj. in Indirect Questions. An example after a Past Tense is found in Il, 16. 646 ff.: see § 298 jin, The accentuation #e (7) in the second of the two Clauses is supported by the unanimous testimony of the ancient grammarians, and is now generally adopted. 202 HOMERIC GRAMMAR. [28r. The MSS, nearly always have %é or # in both places: and so the older editors wrote. 281.] Clauses with pj. These are mainly of two kinds: (1) Clauses following a Verb that expresses the fear of the speaker, as defo pj te 7dOnor I fear that he will suffer. Here the Clause with p4, although of the same form as the indepen- dent Clauses given in § 278, is practically subordinate, and serves as Object to the Verb. The Verb, it is to be observed, is always in a Present Tense, and in the First Person: z.¢. it is the speaker’s own present fear that is expressed. Such a Clause may also be Object to a Verb of knowing, &c., as— Il. 10. 100 ducpervées 5” dvdpes ox e5dv Hara, ovdé re (Sper ph Tes Kal id viera pevowwnowor paxeOa, The fear expressed by ph ws KtA. is subordinated (or on the way to be subordi- nated) to Spev: we do not know (said apprehensively) whether they will not be eager &e. So Od. 24. 491 efedAOdy tis tor pr) 82) cXEddy Gor KidvTes some one go out and look whether they are not near. And in the Prohibitive use— Il. 5. 411 ppaecdw pn ris of dpetvev ceio paxntar, pay Biv nerd, let him see to it that no one éc., lest de. Od. 22. 367 eimé 58 marpi | wy pe mepraOevéwy SnrAnoerat. (2) Final Clauses: the Verb of the principal Clause being— (a) an Imperative, or equivalent form: as— Il. 3. 414 pap pe épede, oxerAtn, pt) Xooapuery ce pedelw. (8) a Present or Future in the First Person: as— Od. 6. 273 ray ddcelvw pity adevkéa, py tis dticcw papedy. In one place the governing Verb shows that the purpose ex- pressed is not the speaker’s own : Tl. 13. 648 dp 8 Erdpwv els EOvos éxdcero Kip’ ddeclvwv, TaVvTOTE TaTTAalvwH, un TIs xypda XAAKG enravpn. Here (if the reading éxavpy is right) the poet describes the fear as though it were present to himself (see however § 298 fiz.). The two groups of Clauses under discussion agree in using only the pure Subj. (not the Subj. with é or xév). The reason probably is that the Prohibitive form as such expresses a general, unqualified, fear or aversion. 282.| Relative Clauses. These fall into the two groups of Final Clauses and Conditional or limiting Clauses. Final Relative Clauses generally follow a Clause which conveys an expression of will; and the reference to a future occasion is shown in most cases by the use of kév: ¢.g.— 282,] RELATIVE CLAUSES—FINAL. 203 Il. 9. 165 add’ dyere KAntods drptvoper, of Ke TaXLoTa ZOwa’ és kAroinv. 24. 11g dpa 8 ’AXAAHE hepéuev 74 Ke Oupoy ijvy. Od. 13. 399 duct 8& Aaldos Erow, & ke otvyénow day avOpwros éxovTa, With ellipse of the antecedent Pronoun— Il. 7. £71 KAip@ viv remddrayOe diapmepes bs Ke AdxNoL. In other instances the notion of End is less distinctly con- veyed, and the Subj. need only have the emphatic Future mean- ing, as in independent sentences (§ 275, 4): Od. 10. 538 eda rot airixa pdvris eAevoerat, dpyape AaGy, i - y, 8s Kév Tou elmpor «tr. (so 4. 389., 11. 135). 4.756 add’ ere mov tis endacerar bs Kev xno KTA. The prophetic tone prevails in these places: cp. the threat in I. 21.126 pédrauwav opty’ trat&er ixdvs, 8s xe paynow Avkdovos apyéra dnpov. The chief examples of a pure Subj. in a Final Clause are— Il. 3. 286 rity & ’Apyelors dmorivewen iy Tw’ Zorxer, Te Kal €ooopévowon per’ avOpadmovo. méAnTaL, . 18, 7 tis rou Taxe. "Ipos duetvwv dAdos dvacth, Od, 18 i I dA oe -. 3. , 2 / ds Tis o. . O@paTos Exmemypnor. The want of ay or xév is doubtless owing to the vagueness of the future event contemplated, z.¢. the wish to exclude all reference to a particular occasion. Finally, this form is sometimes used after a Negative principal Clause—where there is necessarily no actual purpose :— Od. 2.42 otre tiv ayyedlnv orparod ExAvov épxopevoro - 3 yy 7 p oe / ¥: x ae Ww x’ dpiv cada ein, ote mpdtepds ye TvOotunv. 6. 201 ovk &06? obros dvijp . . ds Kev . . ixnras (v. 0. tkouro). I. 23. 345 odk é08 ds ké 0° Ednot KTA. and without «é, I]. 21.103 viv 8 od« éo6’ ds ris Odvarov pvyn (v. 2. ptyot). The Subj. is doubtless used in order to retain the positiveness of tone which is given by od and od« éy with the Subj. in Simple sentences (odK 06 ds piyn=od tus pdyy). The greater prevalence of these Clauses in the Odyssey is worth notice. Of Delbriick’s examples (Synt. Forsch. I. pp. 130-132)—24 in number, excluding repetitions—17 are from the Odyssey: and of the group which he describes as Subjunctives of Willing with «év, eleven are from the Odyssey, two from the Iliad (9. 166., 24.119). In Attic the idiom only survives in such phrases as yeu & 71 eiwn he has something to say (Goodwin, Moods and Tenses, § 65, 1, N. 3). It is to be noted that the Article (6 4 76) is occasionally used as a Relative to introduce these Clauses. This is in harmony with the quasi-independent character which belongs to them. 204 HOMERIC GRAMMAR. [283. 283.] Conditional Relative Clauses. The numerous Clauses which fall under this heading may be divided again into two classes, distinguished by the presence or absence of dv or Kév. (a) The pure Subj. is used when the speaker wishes to avoid reference to particular cases, especially to any future occasion or state of things. Hence the governing Verb is generally a Pre- sent or Perfect Indicative: examples are— Il. 1. 554 7a dpatea, doo’? é0éAqoda (whatever you choose). 3. 109 ofs 8 6 yépwn perénow, dpa mpdcow kal dnicow Aedoce. (a general reflection or maxim). 14. 81 BéATepov ds hedywv mpoddyn Kakdy He GASN. Od. 8. 546 dvtl xacvyvirov feivds 0 ixérns Te TérvKras avépt 8s 7’ dAlyov mep emupatyn mpatidecor, In Similes this usage is regular ; as— Il. 12. 299 ds te A€wy dpectzpodos, bs 7 emdevijs dnpov én Kpetdv (cp. 5. 5, 138., 10. 185). Od. 13. 31 as & 67° dvyp ddpmov0 AAaleral, O TE Tavijwap velov av’ Edkntov Bde olvowe THKTOV dpoTpov. Where the principal Verb refers to a future event, and &v or xév is not used, the intention is to make the reference quite general and sweeping; so Il. 13. 234, Od. 20. 335: and ep. Il. 5. 33 ommorepoot TaTip Zevs wddos dpéEn to whichever (now or at any time) Zeus shall give the victory. Forms of the 3 Sing. Plqpf. are sometimes given by the MSS. and older editions in Relative Clauses of this kind: as wepdxet (Il. 4. 483), €orqer (Il. 17. 435), ke. These were corrected by Hermann (Opuse. II. 44), reading wedixy, &c.: see La Roche on Jl. 4. 483. (8) The Subj. with «év indicates limitation to particular cir- cumstances: especially limitation or reference to the future. Hence it is used (with the few exceptions just mentioned) when the governing Verb is a Future, or in a Mood that implies futurity, viz. an Imperative, Subjunctive or Optative: as— Il. 1. 139 6 8 key Keyoddoerat ov Kev Tkopat. Od. 2. 25 KéxAure 87) vov pev, TOaxjoror, Orre kev elmo, So after an Infinitive used as an Imperative, or expressing consequence, fitness, obligation, or the like : Od. 1. 316 ddpov & drre K€ por dodvat pidov Hrop dveyn, airis avepxouevp Sdmevar. Tl. 6. 227 moddol pev yap éuot TpGes krertoi 7” énixoupot, kreivew ov Ke Oeds ye TOpn KTA. 19. 228 GdAG xpr) Tov wey KaTabdmrewy ds KE Odvyor. Od. 10. 73 ob ydp jor O€jus eort KoprCeuev ovd dmomeurew x ¢ . 9. L pana tAv Ac ve Acotain AmévAntal unkAnerriy 283.] RELATIVE CLAUSES—CONDITIONAL. 205 And on the same principle after a Verbal in -ros : TL. 1. 527 ob8 drededrynrov 6 tu Kev KrA. 3.65 ob tou andBAnr’ earl. . 6ooa KEV KTA. The reference to the future may be evident from the context : Od. 6. 158 keivos 8 ab Tepl fp paxdpraros e€oxov GAY, Os Ké o €€dvotcr Bpicas otKdve dydynrat (cp. 15. 21). In certain forms of expression the use of xév does not imply that actual particu- lar events or occasions are contemplated. The following cases may be noticed :— 1. In a simile the Verb in the first Clause—the Clause which gives the main characteristic of the object to be described—is usually a pure Subj.; but the sub- sequent Clauses by which the description is carried on may take kév. Thus— I. 14. 414 ds 8 86° bd wAnyiis marpds Ads efepian Spis .. Tov B ob mep Exer Opdcos bs xev USyTaL. 21.22 ds 8 ind SeAdivos . . «. pada yap Te KaTecbie bv Ke AGRO. The reason of this is that when the supposition has once been made, it ceases to be general or typical, and is treated as a particular case about to happen. The principle will be further exemplified in the use of dv and kév with Conjunctions. z. When an opposition is made between two groups, especially between two sub- divisions of a class, the tendency is to regard one or both as definite or particular. The principle is the same as in the use of the Article in the form of pév—oi 8€ meaning some—others. Two indefinite groups, because thought of in contrast, are imagined as definite. For example— IL. 9. 508 bs pev 7° aidécerat xodpas Ads door iovcas, Tov be wey dvncay Kai 7 exdvov edxopevoto* bs 6€ «? dvqvyrat nat Te oTEpemis droeinn, Alcoovra 8 dpa Tai ye KTA. Od. 19. 564 7&v of pév « EOwor bid mproTod erépayTos, of p ehepaipoyrat, ere’ axpaavra pépov res: ot 5 did eoray nepdwy €AOwar KA. So IL. 11. 409., 19. 167., 23. 322, Od. 14.126: cp. the uses with te (§ 289, 2, 8). 3. When a Singular Relative is used after a Plural antecedent, the Subj. may take kév: apparently because with the change of Number we pass from a general description to a particular instance: e.g.— : IL. 3. 278 kat ot brévepOe xapdvras avOpwrous tivuaboy, drs «°° emioprov dpdoon whoever (to take a particular case) shall &c. Od. 15. 421 7a TE ppévas Hrepomever Ondutépnat yuvargt, Kal Hw edepyos enor. So I. 16. 621., 19. 260, Od. 7. 33.5 15- 345+, 20. 295. 4. The kév in the following cases may be due to the contrast implied by the preceding dAXos: Od. 8. 32 0882 yap ob8€ Tis dAdos, Sris «’ End Shpad’ KenTat (so 10. 327). 15. 69 vepeco@pat Se eal dAAwW .. bs Kk’ KTA. It must be remembered that we cannot be sure of the text in all the passages now in question. Such forms (e.g.) as 8s ke and $s te might easily be inter- changed. This may be the case in some of the instances to which the foregoing explanations do not apply: as Il, 1, 218., 9. 313, 615., 24. 335, Od. 4.196. 206 HOMERIC GRAMMAR. [284. (y) The use of ay in the Clauses of this kind is very rare. In the two places Il. 8. 10 and 19. 230 the reference to the future is plain. The remaining instance is Od. 21. 293 6s Te Kat GddAovs Bdanret, ds dv xtd., where either of the last two explanations noticed in the preceding note would be applicable. 284.| The Relative Adverbs. The most important are, the Adverbs of manner, és and étas; iva, originally an Adverb of place (=where); and the Adverbs of time, dpa, éws, eis 6, Ste, and onére, edre, jos. It will be best to take these words separately. 285.| ds, tas: (1) Final Clauses with és or érws and the Subj. generally convey the aim or purpose of something which the speaker him- self does, or desires to be done: as— IL. 1. 32 GAN te an pw’ epéOile, cadrepos as KE vEeNat. 9.293 ayadoyv kat vuxti miecOat, as ob 1 euppiyns Tavtas KTA, (2) With Verbs that by their own meaning imply aim or pur- pose a Clause of this kind becomes an Odject Clause: thus— Il. 4. 66 mewpav 3° bs xe TpGes . . dpEwor xrr. 9. 112 dpatdpecd’ ds Key pv dpecoduevor TeTiOmper. Od. 1. 76 hues © olde Tepidpaépueba TmavTes voorov, Stws €AOyat (how he is to come). 3. 19 AlooesOar 5é pv adrds Emus vyyepréa ely entreat him so that he shall speak (i.e. to speak). Here the subordinate Clause expresses the thing tried, thought about, &c., rather than a distinct consequence (such as we have e.g. in Od. 8. 100 drwy TetpnOdpev Tavtwr, ds x’ 6 Leivos evionn olor pidouot). The instances in which the governing Verb is not an Impera- tive, or akin to one—and consequently the purpose expressed is not the speaker's own—are chiefly Object Clauses: ¢. g.— Od. 1. 205 ppdocerar ds xe vénrar he will devise how he is to return (ep. 2. 368., 14. 329). IL. 1. 558 ry o diw Katavedoat eTiTupoy as "AyiATa Tipnons, dA€ons dé KTA. (hast nodded to the effect Sc.). The text is uncertain in Od. 2. 368 of 5€ ro: adrix’ idvri and ppdocovTa dniccw ds xe BAW HOlys (al. $9elys). The true reading may be $Otys, an Opt. formed like Bvy (§ 83,1). There is a similar doubt as to od@ and ods (Il. 9. 424, 681): see § 83 (note). Regarding dv and xév observe that in Final and Object Clauses after s the Subj. with kév is the commonest, occurring 32 times, while the Subj. with éy and the pure Subj. occur each 8 times, 287.] ‘NS, ‘INA, ’OSPA. 207 After éws, which has a more indefinite meaning (in some such manner that), the pure Subj. occurs 7 times, the Subj. with xév twice (Od. 1. 296., 4. 545). (3) Conditional or limiting Clauses: e. g.— Il. 16. 83 weideo & Bs tor €ya pvOov Tédos ev peor Oelw. Here the Subj. is pure; but we have also the formula as dv éyov elrw weOdyeba, which refers more definitely to a speech about to follow. So dmws e0éAno as he pleases. The use of és and ds te in similes belongs to this head: e.g. Il. 5. 161 as 5 A€wy év Bovol Oopav ef abyéva af KrA. 11.67 of 8 &s 7 dunripes évayrlor GAAHAOLowW Oypov eLadvacw «Kth. In this use, as in the corresponding use of the Relative (§ 283), the Subj. is always pure, the case supposed being not a parti- cular one actually expected, but a typical or recurring one. Delbriick (Synt. Forsch. I. p. 161) makes the curious observation that if the simile begins (as in the second instance quoted) with a Demonstrative denoting the subject of the comparison, then the Adverb used is always 4s te. This rule appears to be without exception. 286.] wa is used in Final Clauses only. With a Subj. it usually expresses the speaker’s own purpose ; even in— Od, 2. 306 ratra d€ tot wdda mdvra TeAevTHTOVELY ’Axatol, vija Kat e€atrovs épétas, va Oacoov tknat Krd. the meaning is ‘I undertake that the Achzans will do this for you.’ Exceptions (out of about 80 instances) are: Il. 1. 203 7j iva bBpw ty is it that you may see Fe.: U. 9. 99., 1%. 435+ 24+ 435 Od. 8. 580., 10. 24., 13. 327. An Object Clause with tva is perhaps to be recognised in— Od. 3. 327 AtacecOan BE pw abrds iva vypepres éviomn if the reading is right. The line may possibly be an incorrect repetition of 3. 19. The pure Subj. only is used with iva, except in Od. 12.156 twa elddres 4} xe Odvaper i} Kev KtA., where the xe—xev may be due to the antithesis of the Clauses (§ 283, . 2). But some MSS. have He Odveysev. 2.87.| 3ppa is sometimes Final, sometimes Conditional. (1) In Final Clauses, 8pa either retains a distinctly ¢emporal force—meaning so long till, to the time when,—or passes into the general meaning ¢o the end that. Thus we have— (2) spa until (as shall be), used with dv or kév, as— Il. 1. 509 réppa 8 emt Tpdecar rider xpdros, bpp’ dv ’Axarot vidv eo ticwou, dpeAAwow TE E Tif. 2.2. 192% avixvetov Oéer Eumedov, Oppa Kev eipy. 208 HOMERIC GRAMMAR. [288. With this meaning the pure Subj. is found in Il. 1. 82 exer kérov Shpa tedrécon he keeps his anger until he accomplishes it—a general reflexion: also in Il. 12. 281 (in a simile). (2) &pa=to the end that, used with the pure Subj., rarely with av or xév. The transition to this meaning may be seen in— Tl. 6. 258 ddd perv’, Spa Ké Tor pedindéa olvoy évelxw stay till I bring (=giving me time to bring). (2) Clauses with &pa may be classed as Conditional when it means so long as; @.g.— Il. 4. 345 évOa td’ daradéa kpéa epevar . . Spp’ eOéAnrov. Od. 2. 123 réppa yap odv Blordy Te Tedv Kal KTHpar’ Govrat, Spa ke Kelvyn ToBrov &xn vdov. The use of av or xév in these Clauses is governed by the same rule as with 8s, viz. it is used when the reference is to the future, and is not expressly meant to be general (as Il. 23. 47 d¢pa Coto. petetw). The combination of dv and xév, in the form é¢p’ dv p& kev, occurs in three places (Il. 11. 202, Od. 5. 361., 6. 259). Where two actions are represented as simultaneous, the general sense of the context may show whether one of them is to be regarded as the condition of the other. In Il. 6.112 dvépes gore, pidror, pyjoacbe Se Oovpidos ddxhs, dpp av eyw Beiw (cp. 8. 375., 17. 186, Od. 13. 412., 19. 17) the Clause seems to mean until I go, i.e. long enough for me to go. We may compare the independent Subj. following an Imperative, as in dye viv émipevov, dphia redxea Suw (§ 275). Delbriick how- ever counts Il. 6, 112, &c. as Conditional (Synt. Forsch. I. p. 170). 288.] éws and eis 8, used with the Subj., always take xév, The meaning wntid, with implied purpose, is the usual one: as— IL. 3. 290 abrap eye kal émeira paxjoouat elveca Kovpns av0. pévav, elds KE TEAOS TOAEUOLO KLXElw. 9. 48 vai 8 eyo Sodverds re paynodped els 6 ke TEKUwp ' TAlov etpapev. The Conditional meaning is only found in the recurring ex- pression els 6 «° ditpa év ornPeror pévy Kal por pidra yotvar’ dpapn (Il. 9. 609., 10. 89)=s0 long as I have life. 2.89.] dre, éwére : (1) The instances in which Clauses with ére and énéte may be counted as Final Clauses are very few. They are— (a) with the pure Subj., Tl. a1. 111 éocerar i Has 7) defn 7 pécov Fuap, Ommdre Tus Kal eueto “Apes ex Ovupov ednrat. So Il. 19. 336 eum woridéypevon alet Avyphy dyyeAlny, Or? aropbi- pevo.o THOnta warting for the message when he shall hear &e., i.e, ‘waiting for the time when the news shall come that &c.’ 289. | ‘OTE, ‘OTIOTE. 209 (8) with & or kév: Tl. 4. 164 éocerat juap 81’ dy Tor’ ASAD KTA. (6. 448., 8. 373). The use of & gives definiteness to the expectation, as though a particular time were contemplated. Cp. also Il. 6. 454 dccov ced een dre Kev Tis... daxpuderoay dynta as I am concerned for you im respect of the time) when Sc. It is obvious that in these places the Clause is not strictly Final, since the Subj. expresses emphatic prediction (§ 275, b) rather than purpose. But they have the essential characteristic of Final Clauses, viz. that the time of the Clause is fixed by that of the governing Verb. (2) Clauses with ére or énére which define the time of the principal Clause may be regarded as Conditional. In regard to the use of dv and kev they follow the rules which hold in the case of Conditional Relative Clauses (§ 283) : viz.— (a) The pure Subj. indicates that the speaker is supposing a case which may occur repeatedly, or at any time: as— Od. 7. 71 of uly pa Gedy ds elaopdwvtes dedexarar pio, dre orelyno ava dorv who look on him as a god, and salute him when he walks 8c. Il. 1. 163 od pev cot more toov éxw yépas, éamdr’ ’Ayatot Tpdwy éxmépowa’ &b vatouevov mroAleOpov whenever the Greeks sack a Trojan town, So in maxims, &c. :— Il. 1. 80 xpeloowy yap Baoireds bre ydoerat dvdph yépni. 15. 207 €oOddv kat 76 réruKras Or’ dyyedos atowa edi. And in similes, as Il. 2. 395 dre xuvijon Néros €hdv. So with the regular ds ére as when, ds Sndte as in any case when. 27s oy, In a few instances as 8’ G7’ dv is found instead of ds 8’ Ste: viz.— Il. 15. 170 ds 8 67° dy é vedéwy mrira KrA, 19. 375 ws 8 br dy éx mévroio céAas vatTyot pavnn I of 9d Od. 5. 394 ds 8 Sr’ av domdowos Biotos maidecot pavnn 23. 233 ws 8 br’ dv dondotos yh vnxopevoior pavyn IL. 11. 269 ds 8 67° dy dBivovoay exn Bédros d¢b yuvaixa 17.520 ds 8 br’ dy dfdby éxwv méAeHuY KTA. Also Il. 10, 5., 24. 480, Od. 22. 468. The resemblance in form that runs through these instances would seem to indicate some common source of the peculiar dv. Perhaps there is ground for the conjecture that the word was originally the Pre- position dva4—not indeed in all the places where this 6t’ dv occurs in the present text, but in a sufficient number of places to form a type, and to lead to confusion between an original 67° av’. . éyn (pavyy, &c.) and the ordinary 87° dv. In the one or two places where the pure Subj. occurs after a Future there is an evident intention to speak quite generally: as Il, a1. 322 otd€ ré puv xpe® eorar TuuBoxdys Gre juv OanTrwow Axatol: so Od. 16. 267., 23.257. But in the similar passage II. 10. 130 éte xév is used. P 210 HOMERIC GRAMMAR. [290. (8) & or xév connects a supposition with a particular event or state of things: hence the following general rules :— 1. The Subj. takes & or xév after a Future, Subjunctive, or Imperative in the principal Clause: e. g.— Il. 4. 53 tas Siamépoat dr’ dy Tor anéxPwvrar Od. 1. 40 ék yap ’Opéorao rlais 2ooerar ’Atpetiao énaér’ dy HBHon TE Kal rs twelperas ains. Tl. 20. 130 deloer’ ere8’, Ste Kev Tug KTA. So after xp4 (Il. 9. 100), and potpa (Od. 4. 475), followed by an Inf. 2. When there is an opposition between two sets of supposed cases, dv or kév may be used: as— Il. 6. 224 76 viv col pev eye Eeivos pidos "Apyet péoow eiul, ov & ev Avkin, bre kev Tay djpov tapas. 20. 166 mpGrov pev. . GAd’ bre Key Tus KA, (ep. Od. 20. 83). Od. 11.17 otf émédr’ dv orelynot. . otf Br’ dv dw KtA. So in Il. 2. 397 navrofwy dvéuav 67’ dv év0’ i &vOa yévovrar: and perhaps in Od. 13. 100 &tocdev b€ 7° dvev decpoto pévover vijes evooehpou br’ dv krA. (in implied contrast to those outside). 3. When there is a change from the Plural to the Singular :— IL. g. 501 Avoodwevor Gre Kev Tus bmEpBiy Kal Guapry. Od. 11. 218 dAd’ abrn bixn eotl Bpordy, dre tis Ke Odvyor. The order dre tis ke does not seem to be Homeric; elsewhere we have Ste kév ats (Il. 9. 501., 10. 130., 20. 130, 167, Od. 13. 180., 19. 567, ep. 20. 83., 24. 7, &e.). ‘We should perhaps read in Od, 11. 218 Ste tis te Oavygor. 290.| eére, jpos. The word ere is only once found with a pure Subj., viz. Od. 7. 202 (in a general assertion) : eé7 dy occurs after a Future (Il. 1. 242., 19. 158), and an Imperative (Il. 2. 34); also in one or two places where the use of év is more difficult to explain, viz. Il, 2. 227, Od. 1. 192., 17. 320, 323., 18.194. The combination eit kev is not found. The pure Subj. with jos occurs in one place— Od. 4. 400 juos 8 Hédvos pécov odrpavdy dpprBeBijxy i.¢. ‘each midday.’ The Subjunctive with ei, Sc. 291.] Clauses with ci. The use of the Particle ei (or at), in the Clauses with which we have now to do, is to make an assump- tion or supposition, In most cases (1) this assumption is made in order to assert a consequence (ei=?/'): in other words, it is a condition. But (2) an assumption may also be made in order to express end (etut..at xe miOnrar I go—suppose he shall listen= 292. | CLAUSES WITH EL BIT I go in the hope that he will listen), and accordingly the Clause may be a Final Clause. Again (3) with certain Verbs the as- sumption may stand as Object (ris of8 ef xev . . dplvw who knows —suppose I shall rowse=who knows whether I shall rouse), We shall take these three groups of Clauses in order. 292.| Conditional Protasis with «i. The chief point of in- terest under this head is the use of xév or dv. The rules will be found to be essentially the same as those already laid down for the corresponding Clauses with the Relative (§ 283, 8) and the Relative Adverbs (see esp. § 289, 2), and to be even more uni- form in their application. (a) The pure Subj. is used in general sayings, and in similes: Il. 1. 80 Kpeloowy yap Bactreds 6 bre xéoerar avdpl xépni* et mEp yap te yoAov ye Kal adriywap Karamewn, GAAG TE Kal peTomaden exer KOTOY. 12. 238 tay 8’ ob Tt petatpemop odd’ ddrcyilo, ei éml deEi” twat mpds 7G 7 HédAtdv TE, el rv em dpiorepa Tol ye KTA, Od. 16. 97 kaotyyyjrous . . olat mep avnp papvapevoios TéeTOLe Kal el pacya vetKos Opyrat. I. 11.116 4 & ef wép re réynor «rd. (in a simile; so 16. 263., 21. 576. 2%. 192), If the principal Verb is a Future (or implies reference to the future), the pure Subj. with ei indicates that the speaker ex- pressly avoids contemplating an actual case arising’: ¢ Il. 12. 223 ds qucts ef wep re wUAas kal refxos "Ayaiay pnfopeba oévei peydro, eEwor 8 "Axacol, ov Kdop@ Tapa vaddiy edevoducO aita KéAevOa. Polydamas is here interpreting an omen which he wishes to remain unfulfilled. Similarly Il. 12.245 ef wep yap Tt’ dAdou ye Tepixtewepeba «tA.: IL 22. 86 ef wep ydp ce Karaxtayy, ov a” er eyoye khadoopat ev Aexéeoor : Od. 1. 204 odd’ ef mép Te ctdijpea dé- opar’ éxnor: Od. 5, 221 ef & av Tis patna OeGy wtA.: Od. 12. 348 ef bé XodAwoduevds Th. ve Oey ddEoat KTA, The object of the speaker, in these examples, is to treat the supposed case as imaginary or unpractical, (8) The Subj. with é or xév indicates that a future occasion is contemplated: hence it is generally used when the govern- ing Verb is a Future, or equivalent to a Future: as— Tl. 4. 353 dear tw €O€AnoOa Kal ef Kev Tow TA weprAr. II. 404 péya pev axon (sc. état) ef Ke PeBwpa, 24. 592 My mo. . oxvdpaiweuer, ef ke TVONaL KTA. P2 212 HOMERIC GRAMMAR. [293.- Od. 2, 218 ef pév kev marpds Blorov Kat vdcTov dkovow, it’ dv rpvxdpuevds ep ere TAalny eviavTdv. 11. 112 ef d€ Ke olvyat, Tore ToL TeKpaipop’ ddreOpov (I prophesy your destruction). 12. 121 iw yap SyOdvycIa Kopvacduevos Tapa TETpN, defo pi} o” eLabris «rd. (ep. Il. 6. 442., 8.478, 482). The chief instances of dv or év after a Present appear to be: Il. 1. 166 dzdp qv more Sacpos txnrat, coi 7d yépas KTA. Here however the Clause gives a particular case of a supposition already made: the context being, I have not an equal prize when the Greeks sack (pure Subj.) a town, but if (in such case) a division is made é&e. Il. 3. 25 wéAa yap Te Kareadier ef wep dv KTA. (SO T2, 302: § 283, %. 1). II. 391 97 GAdws bm epeio, ral ef x” dAlyov wep Exavpy, é¢t Bédos TéAETAL. Od. 11. 158 Tov ov Tws EoTL TEpHTA meCov edvt’, iy ph Tis éxn evepyéa vija. In these cases there is an emphasis due to implied contrast: e.g. ef x’ ddlyov mep éravpy means ‘not only when it reaches him fully, but also if it only touches him slightly’ Cp. § 289, 2, B, 2. 293. Final Clauses with ei.] After a principal Verb expres- sive of the speaker’s w7d/ (an Imperative, or First Person), a Final Clause may be introduced by et kev or jv: as— Tl. 8. 282 Barr’ otires ef xév tt pows Aavavion yevnat. Od. 1. 281 épxeo mevodpevos marpds dy ofxopevoto, iw tis tou eimnoe KTA, IL. 11. 791 tadr’ elrous 7AxATi daidpour ef xe TLOnTaL. 22.418 Nloocwp’ dvépa Todrov .. iv Tews KTA, Od. 4. 34 dS¢Bp" ixdued? at Ké moO Zeds . . maven KA. The effect of using ei (instead of ds or iva) is to express some degree of uncertainty. The end aimed at is represented as a supposition, instead of being a direct purpose. The pure Subj. is not found in any clear example of this use. In Od. 5. 471 ef pe pectin (uednn La R.) pryos wat wdparos, yAuwepds 5€ wor Urvos éréAOn the MSS. have the Optatives peOely and éwéAOor. In Il. 14. 165 the Subj. xevy follows an Opt. (ei mws ineiparro..7g@ 8 .. xevp). 294.| Object Clauses with ci. This term will serve to de- scribe the form of Clause in which the supposition made by ei takes the place of an Acc. of the thing. It may be regarded as a special form of the Final Clause (ep. § 285, 2): thus Il. 18, 600 as Bre Tus Tpoxdv . . medpijoeras et Ke Oégou ‘tries on the suppo- sition that it will run,’ hence tries whether it will run: so— Tl. 4. 249 dopa tyr’ ct « tupuv drepoy7 xeipa Kpovlwy. 15. 32 dppa tdy iv Tor xpatouy KTA. 296. | CLAUSES WITH EI, &c. 213 that you may see whether it will avail, Note that the Subj. here has a distinetly future meaning, as in Final Clauses; the same words taken as a Conditional Protasis would mean if it has availed, So after eimetv, Il. 7. 375 wal 8% 768’ [leg. 70] elmeuevar muiwov émos, al x’ @éAwor say the word supposing that they shall be willing (=ask if they will agree); and ofa, Od. 22. 7 eloopat at xe tUxwp1, and in the phrase ris of8’ ef kev who knows but (I. 15. 403., 16. 860., Od. 2. 332), and od pay oid’ ef (Il. 15. 16). Note the use of an Acc. with ofa in anticipation of the Clause with ‘ei: as— Il. 8. 535 avpiov iv dperiy diaetcerar et x’ euov eyxos welvn erepx duevov meaning ‘he will know as to his prowess whether it will enable him to withstand my spear.’ So Od. 22.6 oxomdv dAdov .. eloo- pat at xe Toxo (ep. § 237, 2, with the note). In one place the Clause with ei serves as explanation of a Neuter Pronoun in the Nominative: TL. 20. 435 GAN’ 7 rot wey Tada OeGv év yovvacr Keirat, ef KE oe XELpdrepds TEP Ewy amd Oupov EXopat. The relation between the Clause with ei and the governing Verb is essentially the same as between a Clause with pq and a Verb of fearing. As the prohibitive py may imply fear, it fits the meaning of the Verb, and comes to serve as ex- planation or Object to it; and similarly the ei implies a more or less uncertain supposition, which may serve as explanation or Object to a Clause expressing doubt, enquiry, or the like. In both cases the essence of the construction is the special harmony of meaning between the governing Verb and the Subordinate Clause. 295.| The Subj. with és ei occurs in a single place only, viz.— Il. 9. 481 kai pe pidno’ as ef te marhp bv Taida pidjoy. Here the assumption ef... @iAjon is made for the purpose of comparison. Thus the meaning is nearly the same as with ds dre (§ 289, 2), and the Clause is essentially Conditional. 296.| éwet with the Subj. The use of éwet implies that the action is prior in time to the action of the principal Clause ; hence Clauses with émet fall under the definition of the Condi- tional Clause. A pure Subj. after émeé is found in two places only, viz. Od. 20, 86 énei dp Brépap’ dudicaddwn (sleep makes men forget every- thing) when it has spread over their eyelids ; and Il. 15. 363 (ina simile). The use of xév or dv (7. ¢, émet kev or émqy, once émel dv) is inva- riable when the principal Verb is future. It is also found after a Present or ‘gnomic’ Aorist, and even in similes; but usually when the Clause is subordinate to the main Clause of the simile: e.g. 214 HOMERIC GRAMMAR. [297- Il. 2. 474 rots 8 ds 7’ ainddta mAaré” alySv ainddot avdpes peta diakpivwow, érel Ke vo@ pLyewow. Here the case has been put by the Clause ds Te. . diaxpivwot, and may then be treated as a particular case (§ 283, B, note I); so Il. 7. 5., 9. 324., 21-575, Od. 10. 411., 19. 206., 24. 7. The explanation here given of the use of émet xe and émfyv in similes does not apply to Il 17. 657 és ris re Adwy .. Os 7 eel dp Ke Kdupor KTA.; but perhaps we should read kexépyot, and similarly in Il. 1. 168 éve? kexdpo (instead of eel xe Kapow). In several instances the use of émet xe, émjy may be due to the emphasis laid on the singleness of the event: viz. in the com- binations émeé xev mpGrov (Il. 13. 285, Od. 11. 221), emqy 7 mpora (Il. 6. 489, Od. 8. 553): so Il. 7. 410 éet xe Odvwor when once they are dead; Tl. 19. 223, Od. 8. 554., 14. 130. The two remaining instances of éwv not referring to a particular occasion (Od. II. 192., 19. 515) are sentences beginning adrdp émijv—,where the use of dv may be suggested by the opposition. 297.] mplv with the Subj. In general, as we have seen (§ 236), mpi is construed with an Infinitive. If, however, the event is insisted upon as a condition,—the principal Verb being an Im- perative or emphatic Future,—the Subj. may be used ; as— Tl. 18. 134 ddAAa od wey pH To Karaddcco podov ”“Apnos mpiv y eye devp’ eOovoay ev d@Oadpotor tdyat do not enter the battle before you see me coming hither. Od. 10.174 & irou, od yap mplv katadvodueh axyvipevol ep eis "Aidao dduous mplv udpoimon juap even. So Il. 18. 190., 24. 551, 781, Od. 13. 336.,17. 9. The Subj. is used in these examples without. év or xév, because it is not meant to refer to a particular occasion when the condition will be ful- filled. When such an occasion is contemplated Homer some- times uses mpiv y 87° dv before the time when (Od. 2. 374., 4.477): ep. IL. 16. 62 od apiv pniOpov Katanavoeuev, GAN’ émdt’ Gv KTA, It is evident that a conditional Clause of this kind can only occur after a negative principal Clause. ‘Do not do this before I come’ makes my coming into a condition, and a condition which may or may not be realised: but ‘do this before I come’ is merely a way of fixing the time of doing. 298.| Subjunctive after a Secondary Tense. The rule in Homer is that the Subj. is not used in a Subordinate Clause to express a past purpose, condition, &c. It may be used however (1) when the governing Verb is a ‘gnomic’ Aorist : Il. 1. 218 és xe Ocots enumelOnrar pada 7’ €kAvov avrod. Od. 20. 85 6 yap v eméAncev amavrav €cOAGy dE KakGv eel Ap BA€Hap’ dupixarvyn, 299. | SUBJUNCTIVE AFTER PAST TENSES. al5 So Il. 17. 98, Od. 14.65., 15. 409: also in general denial :— Od. 10. 327 ot yap ovd¢ Tis GAXos avip rade Pdpyak’ avérdn, bs xe wip «th. (ep. Od. 12. 66-68). And in similes :— Il. 7. 4 as 5& eds vatrnow eedAdopnévoiocw edmxev otpov, ével xe kduwow (Il. 4. 486, Od. 4. 792, &e.). Further (2) if the action expressed by the Subordinate Clause is still future at the time of speaking ; as— IL. 5.127 dyAtv 8 ad rou da’ d6pOadpav €dov f amply erijev, o¢p d yiyvdckns fuev Pedy Oe Kal dvdpa. L have taken away the mist—that you may know &c. 7.394 Kal d¢ rdd judye eimety eos, al x’ COeANTE KTA. 18, 189 paytnp 8 ob pe pirn amply y ela Owphocecbar mpiv ¥ abriy . . Wapa (before I shall see her §c.). Od. 11.434 of re kar’ aloyos éxeve kal écoopevnow dnicow Ondurépyot yuvaréi, kab ij «’ edepyds enor. It will be seen that in such cases the governing Verb is generally to be translated by the English Perfect with have (cp. § 73). The real exceptions are not numerous, and may be due in several cases to alteration of the original text through the influence of the later usage. The reading is uncertain (e. g.) in— Od. 14. 327 Tov & és Awddvnv Paro Bhpeva dppa Geoto (=19. 296) &« Spuds tyixdpoto Ards BovaAy éraxovon, where the Subj. was read by Aristarchus, the Opt. émaxotcat by Aristophanes and Herodian. With the Subj. Byuevar=that he has gone. Again— Od. 10. 65 % yey o° éviueéws dmewéumoper, dp’ dv tena. So La Roche from the best MSS., but other MSS. have dgp’ dv ixoro and d¢p’ dpixowo. The Subj. gives the meaning ‘we have but now sent you away, intend- ing that you shall arrive &. In IL. 5. 567 (fe) wh Te wdO01, péya 5é opas dwoopnAce, and 15. 598 (€BovAeTo . a) éuBddro . . OéTiBos F efaloroy dpivy wacay émxphvee, most MSS. have 740 and épBdrdy. But the Opt. in the Clause following supports wa@or and ép- Pddor. Other places where the Subj. is supported by external authority, but is contrary to the rule now laid down, are Il. 13. 649., 14. 165, 522., 15. 23., 16. 650 (see La R.)., Od. 9. 102., 10. 24., 16. 369.,17.60. In all the Opt. may be substituted without affecting the metre: and when we consider the number of places where the MSS. vary between Subj. and Opt. forms, we can hardly doubt that it would often be right to make the change. The Homeric rule is observed by Plato (see Riddell, Dig. §§ 90, 91), but not by Attic writers in general. The Optative in Simple Sentences. 299.| The uses of the Optative in Simple Sentences range 216 HOMERIC GRAMMAR. [299- fyom the expression of a wish on the part of the speaker to the expression of mere supposition, or admission of possibility. Without dy or xév the Optative may express— (2) Simple wish or prayer: as— Il. 1. 42 ticecav Aavaol éua ddxpva cotot BéArcoot. Od. 1. 403 ph yap 8 y' @XOou krA. never may he come §e. Regarding the Opt. of wish with ei or ai, ele, aie, &c. see § 311. (2) A gentle or deferential Imperative, conveying advice, sug- gestion, or the like: as— Il. 4.17 i 8 ad ws rdde Tact Pirov Kal Hdd yevo.ro, 9} Tou pev oikéowro wéALS [I pudovo KTA. (=I presume the city is to remain inhabited). Od. 4. 735 GAAd tus dtpnpds Aodiov xahécere yepovta (as we say, would some one call Sc.) 18. 141 7@ pH tis ToTe maumay dviip aOeulotios etn, GAN 6 ye oryh dGpa Bev Exor L would have a man not be lawless, but §e. Note especially this use of the Second Person: e. g.— Od. 4. 193 miOord pow pray listen to me: so in the phrase 7 pd vd pol te mido.o (Il. 4. 93, &e.). Tl. 11.791 ratr’ etmous "Axiaqi suppose you say this to Achilles. Od. 15. 24 GAAG od y’ €AOdv adros emuTpeWeras ExaoTa. Il. 3. 406 fjoo map’ adrév iotca, cGy 8 amdetke KeAevOov, pnd ere cotot médecow trootpeweras "OAvpTov. Hence in Il. 1. 20 we should read (with the best MSS.) zaiéa & éuol Moatte (not Adoai re, Wolf’s conjecture). (c) Rhetorical wish, implying willingness, or at least indiffer- ence to the happening of some evil: as in imprecations— Il. 2: 340 év rupt 1) Bovdal re yevotaro pajdea 8 avdpav. 6.164 reOvains, & Mpotr’, 7} naxtave BeAAepodovtny (=L care not if you were dead, unless you §c.). Od. 7. 224 iddvra je kal Altos aley KTiow euHr KTA, (=I am content to die when I have seen Se.) (¢) Concession or acquiescence : IL. 21. 359 Any’ Epidos, Tpdas 8% Kal adrixa dios "AyiAAcds doreos é£ehdoeve (cease strife, and I consent that &¢.). Od. 1. 402 xtijuara 8 adrés exols Kal Space cotow avdcoos. 2.232 adN’ aiel xaderds 7’ ein kat aiovaa peor (i.e. he may as well be unjust as just). The following are instances of the First Person used in this way : Il. 15.45 adrdp ro Kal welvw eye mapapvOnoaipny Lam willing to advise him (a concession), 300. | THE OPTATIVE. aI7 Il. 23. 150 viv & énel od véopai ye pidny és narpida yaiav, Tlarpéxry tpwi nopny dadacayun pépecbau since Iam not to return, I may give the lock to Patroclus. Od. 16. 383 GAAd POéwpev EAdvTES em” Aypod vdaqu TEANOS i} év 688, Biorov 8 avtot nai erqpar’ Exwper daccdpevor Kata potpay ép’ Hpéas, ola 8 adre xelvou pnrép. Soluev Exew 715° bs Tes dmviot. Here what the Suitors are to do for themselves is put in the Subj., what they do or allow to be done for Penelope in the Opt. Compare Hat. 7. 5. 4 70 pev viv radra mphocos 74 wep ev xepat exes, Puepwoas 8& Alyunroy Thy evBpicacay orparnddree én rds ’AOhvas, i.e. I consent to your doing what you have in hand, but when it is done, march against Athens.’ (e) Strong denial is sometimes implied, under the form of de- precation, by the Opt. with py: as— Od. 7. 316 uh todro idov Aut aarp! yevorro let us not admit that this is the will of father Zeus, 22. 462 pi) wey 01) KabapS Oavatw ard Ovjdv Edolunv. (7) From acquiescence, or willingness that something shall happen, the Opt. passes to admission of possibility, i.e. will- ingness to suppose or believe that the thing will happen. This use is rarely found without dv or xév: an instance 1s— Od. 3. 231 peta Oeds y’ COeAwv Kal TnAdOev avdpa Tawra. This is said as a concession : ‘we men must allow that a god can save even from afar.’ So perhaps Il. 10. 247 (if the text is right). Other instances are negative, viz. Il. 19. 321 od prev ydp te Kaxorepov dAAo maéOouut. Od. 14.122 ® yépov, ob tus Keivov drip dhadijpevos eAdav ayyéd\Awy tmeloee yuvaikd Te Kal pldov vidv. The absence of dv or xév marks the negation as sweeping and unconditional. We should compare the corresponding Homeric use of o8 with the pure Subj., which differs in the degree of con- fidence expressed: otd¢ fdwpar I am sure I shall never see, ob mado. I suppose I shall never suffer. 300.| With a or xév the Optative does not express wish (which is essentially unconditional), or even direct zw2/dimguess on the part of the speaker, but only willingness to admit a consequence : hence expectation in view of particular circumstances: €.g.— Tl, 1. 100 rére Kev py iAaoodpevoe weT(Ooev then we may enpect to appease him and gain grace. The character of a Clause of this kind depends chiefly on the manner in which the condition is indicated. The following are the main points to be observed :— (a) An Opt. with é& or «év often follows an independent Clause with a Future, Imperative, &c. 218 HOMERIC GRAMMAR. [ 300. Tl. 22. 108 ds épéovary, enol 8& tr’ dv word Képd.ov etn KTA. Od. 10. 269 dedyoper* rt ydp kev Gddgayev KaKOV tuap. Il. 3. 410 keive & eydy ode elut, veweoontov 8é kev ely. (2) Or the preceding Clause may contain a wish : Il. 7.157 €i0 ds HBdoun, Bin 5€ pou eumedos Et T@ KE TAY’ GVTHOELE KTA. Cp. Il. 4. 93 (where the Opt. is a gentle Imper.). (c) The case supposed may be in past time, so that the Opta- tive expresses what would have followed on an event which did not occur: @.g.— Tl. 5. 311 Kal wd kev &v6 anddouto ava dvdpSv “Ayapeuvor, ei pn dp’ d€v vonoe Krr, Od. 5. 73 &0a x’ érera kal dOdvards mep éneAOav Onjoaro dev. So IL 2. 81., 3. 220., 4. 223, 429, 539., 5- 85, 388., 12. 58., 13. GAS, U5. O67, LF 70, 306., OU, 1, 296,, 7, 29%., 6. 241., 13, 86. This use of the Optative is confined to Homer. (@) The case supposed may be vague or imaginary : Il. 8. 143 dump 3€ Kev ob re Ards vdov eiptacaito, where the emphatic dvjp suggests a condition: if a man, he cannot &e.; ep. Od. 4. 78., 23. 125. Od. 12. 102 mAnoiov dAdjAwy' Kal Kev d.0icTEVoeas one may (on occasion arising) shoot an arrow across. 9. 131 ov pey yap Tt Kak ye, pépor S€ Kev Spa Twavta. It is natural that an admission that something may happen should generally be made more or less in view of circumstances, given or supposed. Hence the use of é or xév with an Opt. of this force became the prevailing use, and exceptions are rare, even in Homer. The Principal Clause or Apodosis of an ordinary Complex Con- ditional Sentence belongs to this head. It is erroneous, how- ever, to regard the varieties now explained as complex sentences with the Protasis understood. In this, as in some other cases, the complex is to be explained from the simple, not vice versd. In some instances the Opt. with dv or Kév appears to be concessive (expressing willingness). Delbriick (Synt. Forsch. I. p. 200) gives as examples— Il. 22. 252 viv avré pe Oupds aviKe oTnpevat dvria celo* €douut Kev H Kev adoiny. Od. 8.570 Ta 5é rev Beds 7} TEAC ELEY iw arédAea7’ ein, ds of pidov émdeTo upd. To which may be added Od. 14. 183 # Kev GAoin @ KE puyor KA. (but IL. 13. 486 is different). Possibly the use of «év in these places is due to the opposition made between the two alternatives. 301.] THE OPTATIVE. 219 Il. 24.618 Gar’ dye 87 ral vBi peddpeba, Ste -yepaé, airov’ émeta Key avre pidov maida xdalowba. Also Od. 16. 391., 21.161. But these instances need not be separated from others in which expectation rather than concession is recognised. We may notice espe- cially, as on the border between the two meanings— (a) Uses of the First Person (esp. in the Odyssey): e. g.— Od. 15. 506 7G0ev 5€ xev bupw ddoumdproy wapadeipnv. 22. 262 @ pido, 78n pév Kev eydy elrotpe Kal Gppuy KTA, 16. 304 GAN’ ofa ot 7° éyw Te -yuvaiKay yydopev iduy, Kal xé Teo Suwwv dvdpav ert weipyOet per. 14.155 mply 5€ Ke, nal pada wep Kexpnpévos, ov Ti Sexotpny. So Od. 2. 219., 4. 347, 12. 387., 15. 313, 449., 18. 166., 19. 579., 20. 326., 21. 113, Il. 9. 417., 24. 664. (8) Negative Clauses, with the Second Person : IL, 14. 126 7G obm dy pe. . pavres | wiOoy dripnoare I do not think you will (I expect you not to) &c. Od. 20. 135 ov« dy pu viv, réxvov, dvaitioy aitidwo. So probably I. 2. 250 7@ ode dv Baotdjas dvd ordy’ Exwv ayopevors KA. is ironical courtesy (you will not if you are advised by me). This, again, when turned into a question yields another form of polite Imperative: as Il. 3. 52 od« dy 51 pelveras will you not await? So Il. 5. 32, 456., 10. 204., Od. 6. 57., 7. 22. But we have to remember that it is one thing to convey or imply wish or willingness, another to use a form which expresses it. The fact that od is the negative Particle in all these instances shows that the Optative is grammatically more akin to a Future than to an Imperative. So far as an Imperative meaning is intended, the use is a rhetorical one, like the similar use of the Future Indica- tive in Attic. It will be seen that, except in one or two rare Homeric uses of the pure Opt., the usage of the Opt. in independent Sen- tences is nearly the same in Homer as in later Greek. Optative in Subordinate Clauses. 301.| The classification which has been followed in discussing the Subordinate Clauses with the Subjunctive will also be the most convenient in the case of the Optative. Indeed there is so close a parallelism between the uses of these two Moods that little is now left to do except to take Clauses of the several types already analysed, and show in each case the nature of the differ- ence which determines the use of one Mood rather than the other. The reason for using an Optative will generally be found in the circumstance that the governing Verb is incompatible with a subordinate Clause expressing either the widd or the assured expectation of the speaker. If the occasion to which the whole sentence refers is past, or is a mere possibility, or an imaginary case, these two meanings of the Subjunctive are generally out of place—and we can only have the Mood which expresses a wish, 220 HOMERIC GRAMMAR, [302. or an admission of possibility. Hence the general rule—to which however we have seen that there are important excep- tions (§ 298)—that the Optative must be used when the principal Verb is an Optative, or one of the Secondary Tenses. 302.] Clauses with j¢—fe. The Optative in the Homeric examples is generally to be explained as the translation of the Subjunctive into oratio obliqua; that is to say, it expresses a doubt or deliberation thrown back into the past. Thus we have past deliberation in Tl. 16. 713 di€e yap je pdyouro Kara KAdvov adris éAdocas, H Aaovds és Telxos dpuoKAjoeey aAfjvat he debated—should he fight §c., or should he call to the people &e.: so Il. 1. 189., 5.671, Od. 4. 117., 6. 141., 10. 50, &e. Past doubt is less common: the examples are— Od. 4. 789 éppyatvovo’ 7 of Odvaroy iyo. vids autor 4 Oy tnd prnotipow bneppidro.or Sapyein. 15. 304 ovBdtew Tetpnticav q pu er’ evdukews dtdcot pelvai Te KeAevou abtod evt crabue, 7 dtptvere TOALVOE Ulysses tried the swineherd—whether would he stilt be hospitable and bid him stay, or Sc. In this use we once find kev, viz. Od. 15. 300 dpyatvwv i} Kev Odvarov piyor 7} Kev GAoin (where La Roche reads cq). 303.] Clauses with py. These are of two kinds, answering to the similar Clauses with the Subj. (§ 281): (1) Object Clauses, with Verbs of thinking, &e.:— Il. 21. 516 péuBarcro yap ol teixos evdprto.o TéAnos, pa) Aavaol mépoceayv (his care being that) the Tro- jans should not §¢.: so Od. 19. 390. Od. 21. 394 Teip@mevos EvOa Kal évda fa) Képa. times dove AtoLyouevoro dvakTos The double past—e tried to see that worms should not have eaten it—has no distinct form of expression: cp. the use of the Aorist as a Pluperfect ($ 76). So in the common use with Verbs of fearing: as Il. 18. 34 delive yap jun) Aaywov enrapjoee he feared lest Sc. But in— IL. 9. 244 radr’ aivds deldorxa xara dpéva py of amerhas €xTed€owot Geol, Huty dé by aloimov ein KTA, the Opt. stands for the more remote event: see § 304,a. The true reading however may be ely, a Subj. like per-etw (Il. 23. 47). These Object Clauses may be regarded as the negative forms answering to the Clauses expressing past deliberation. As in the corresponding uses of pq with 304. | RELATIVE CLAUSES. 221 the Subj. and Opt. in principal Clauses (§ 278), the Mood is never qualified by dv or kév. (2) Final Clauses: an example will suffice :-— IL. 5. 845 div’ “Aidos kuvény yh pv tSou 6Bpiyos ”Apns (so that) Ares should not see her. 304.]| Relative Clauses—Final. In many instances the Opt. in a Relative Clause is used precisely as in an independent sentence: the wish or supposition being expressed from the speaker’s present point of view, not subordinated to the point of view fixed by the governing Verb. Thus in— Od. 4. 698 GAA OAD peitdv Te Kal dpyadedrepov GAO punotipes ppacovrat, 0 ph TeAcoete Kpoviwy we have an independent parenthetical wish: and in— Tl. 3. 234 viv 8 Gddovs pev Tavtas 6pG .. obs Kev eb yvoinv KTr. 5. 303 péeya epyov, d ov bo y’ avdpe péporev (§ 299, 7) a parenthetical expectation. In other places, however, the Rela- tive Clause is connected, by implication at least, with the action of the principal Clause, and expresses an intended or expected consequence. We may distinguish the following cases : (a) The choice of the Opt. shows want of confident expectation of the result intended : IL, 1. 62 GAd’ dye 84 Tia pdtv epeloper 7) tepya, . . 8s «° elrot rr. (with the view that he may tell: cp. YBa. Bi, BBO, Ou.§. 206). 7.231 typels 8 eiuev rotor ot dy cer avtidoaypev kat Todées (= many of us are ready to meet thee). Od, 10. 431 tt kax@p tyelpere TovTor, Klpxns és uéyapov xaraBijpevat, } Kev &ravras obs He AvKovS ToujoeTat HE A€ovras, of xév of péya bGpa pvddocomen Kal dvayKny. Here zoujoerat (which is either a Fut. or a Subj.) expresses the certain immediate result, dvddooomev the further and therefore (in the nature of things) dess certain consequence. In this group of Clauses the Opt. always takes ay or xév (cp. the corresponding Subj., § 282). (2) The Opt. with xév is especially common after a Principal Clause of negative meaning (in which case the consequence is necessarily matter of mere supposition): as— Tl. 5. 192 tao 8 ob mapéace Kat Gppara ray « émBalnv. Od. 1. 253 } 8) TOAACY arroLyopuévov ’Odvajjos dedn, 8 Ke pnoThpow dvadéor xeipas épely. The pure Opt. occurs in Il. 22. 348 ov« E06" ds . . dwaddAxou, 22% HOMERIC GRAMMAR. [305- (c) The Opt. is used if the governing Verb is an Optative, or a Secondary Tense: e.g.— Il. 14. 107 viv 8 ein ds Thode y dpelvova pir evlonoL. Od. 6. 113 &s "Odvceds eyporro, tdou 7 ebdmida Kovpny, fH of Daujxwv dvdpdv wodw iyjoatro. 9. 331 abrap rods GAdovs KAjpo TeTaddoOar dvwyov ds Tis ToApHoerev KTr. (for the man) who should §e. In the last instance the Clause approaches the character of a Dependent Interrogative, and therefore of an Object Clause. Similarly, after Verbs that imply asking— Tl. 3. 316 «Arjpous .. tédAov . . dwadrepos kth. (cast lots for which of the two §c.). 14. 507 ndaryvev 5 Exaotos San piyor KTA. And with mevderdat, épecivw, &c. Od. g. 88., 10. 100, 109., 19. 463. 305.] Relative Clauses—Conditional. When the event to which the condition attaches is matter of wish or mere expecta- tion, or is in the past time, the condition is generally expressed by the Optative. Hence we find the Optative— (a) After an Optative of wish in the principal Clause :— Il. 3. 299 éamérepor mpdrepor brép Epxia mHunverav, Gdé of” eyxeparos xauddrs peor as bbe otros. Od. 1.47 &s amddouto Kal GAXOos GTis ToLadTa ye peor. (2) After an Optative of expectation : Il. 9. 125 ob kev dArjios ein dvip @ té00a yévo.to he will not be poor to whom such things come. 12.228 &8€ x’ dmoxpivaito Oeompdros bs cada Oupd elde(n Tepdwy kai of TeLolaro Aaol so will a diviner answer, who knows §e. Od. 4. 222 ds Td karaBpdferev .. ov Key epnuepids ye Bddot kata Saxpy mapeav. (c) After a Present or Future, in one or two places where the time is purposely vague : Od. 6. 286 Kal 8 GAAn veueos, i Tis Tovadrd ye péCor =I am ready to be angry with any other who §c. 1g. 510 kat yap 8y Kolrowo Tay’ €ocerat Hddos dpn, bv twa y Umvos Edor KTA, The Opt. avoids assuming that the case will ever occur. The reading is very doubtful in Tl, 5. 407 S77 par’ ob Snvaids ds dbavdrowws pd- xorro, the Ambrosian and some others having paxynrar. (2) When the principal Verb is in a past Tense; the Relative Clause generally expressing indefinite frequency, iteration, &ec. : as— 306. ] 5 ‘NS, ‘ONS, ‘INA. 2.2.3 Tl. 2. 188 dy twa pev Bacidfja cal @oxov dvdpa Kixein, tov & dyavots éméecow épnticacke. 15. 22 dy 5é AdBouut | pintackoy terayav kth. Od. 22. 315 maveckov prynoriipas ris tovadra ye peor. In these uses, and generally, the Opt. is pure. Exceptions are— Od. 4. 600 8Gpov 8 Srre xé poe Soins KeynAtov EoTw (where the Opt. may be substituted for the Subj. for the sake of courtesy, to avoid assuming the certainty of the gift), and— Od. 21. 161 4 O€ énerra yyual’ bs Ke TAcioTa Tépot Kal pdpotyos EOL. Clauses formed by a Relative and the pure Optative are strictly parallel to the Conditional Clauses formed by a Relative and the pure Subjunctive, such as Xaiper dé pry bs Tis eOeipp, or BeArepov ds HPevywy mpopdyn (§ 283, a). In both groups of Clauses the reference is indefinite; but with the Subj. the instances must be thought of as real future instances, and consequently the governing Verb must not imply that they are past or imaginary. For the exceptions to this rule, see § 298. It may happen that the condition is expressed by the Subj. (because regarded as certain to be fulfilled), while the main action is uncertain, and therefore put in the Opt.: as— Il. 14. 126 7@ odm dy pe yévos ye nakdv Kal dvadnida pdyres poor atiphoare mepacpévoy, bv ev eimw, 20. 250 érmoidy x’ elmnada énos, Toidy x énaxovoats., But the general rule is to let the subordinate Clause follow the Mood of the go- verning Verb: hence the so-called ‘ Attraction’ of the Optative. 306.] Clauses with ds, éras, iva and the Opt. are either Final or Object Clauses (not Conditional, as with the Subj., § 285, 3). (1) In Final Clauses the Opt. may be used either (a) to indi- cate that the consequence is not immediate or certain (the go- verning Verb having a present or future meaning), or (4) because the governing Verb is an Opt., or (c) a Secondary Tense. Thus we have— (a) With a Present, &c. in the principal Clause; especially when the Clause bears a negative meaning (so that the occasion is necessarily imaginary) : Il. 1. 343 od8€ 71 ofde voroa dua mpdoow kat drlocw, dames of mapa vnval cdot paxéowro ‘Axatol, Od. 2. 52 ot marpos yey és otkoy dmepplyact véco Oat *Ixaplov, ds « abros eedvacaito Ovyarpa. But also after an affirmative Clause : Od. 23. 134 hyelodw piromalypyovos épxnOuoto, ds xév tis hain yapov eupevat éxtds dxovwv =s0 that any one who happens to hear may think §c. 294 HOMERIC GRAMMAR. [307- Od. 17. 249 rév Tor’ éyav ént vyds eiooédpoto pedalyys do rH’ 1Od«ns, va por Bloroy TodAdv GAqou (moré indicates a distant occasion). 13. 401 Kvucdow 5é Tor dave Tapos TEpLKAdre’ edvTE, Os dv detxéAros tact pynotijpat paveins (perhaps however we should read paviys). 24. 532 loxerbe .. ds kev. . SiaxpwOeire (qu. diaxpwvOyre). (2) After an Optative, either of wish or of expectation : Od. 14. 407 TdxtoTa pot évdov ETatpor elev, ty’ ev xAucin Aapov TervKoiyeba ddpTovy. 15. 537 TO ke Taxa yvolns .. os dv rls ce. . paxapicor. So a fortiori after an Optative that implies prohibitcon— Od. 3. 346 Zevs 16 y’ ddebjoere . . OS tuels . . Kloure Zeus avert that you should go Se. (c) After a Past Tense—a use of which it is needless to give examples. Regarding the use of dv and kev, it is to be observed that— 1. The Opt. with tva and émws is always pure. 2. The Opt. with és takes dv or xév in a few places, where there is clear reference to a single occasion, as in Od, 2. 52 (quoted above), Il. 19. 331, Od. 17. 362; and in the combinations @s Gy Tus (Od. 15. 538); @s Kév Tis (Od. 23. 135). (2) The corresponding Object Clause with és is seen in— Il. 2.3 Gar’ 6 ye peppajpice kata ppeva as "AxtAja Tiunoer dAgoat O€ KTA. 9. 181 mepay as menlOorev (bade them try how to persuade). And similarly with ézws: Il. 21.137 Spynvev & dvd Ovpdy drs tadcere (so 24. 680), Od. 8. 344 Alovero 8 aiet “Hoatorov kdutoepyov dmws Adoeev “Apna. In Od. 14. 329 Ommws voornon xTA., we should read vooryeer’, as is shown by the parallel passage— Od. 19. 298 damws voornoee pidny és marpisa yatay where the Opt. is secured by the metre. In Il. 2. 4 the reading tiwqce’ is sup- ported by the Ven. A., which has myLhoNL, and the Scholia (rine: edxrixdy Schol. A. B.): but all other authorities have typqoy, and all have éAéoy. 307.] Clauses with és and dpa, These also are Final in character: 7.e. the Conjunction has the meaning ¢l/ the time that, hence (commonly) ix order that,—not while, so long as. The notion of ¢ime is distinct in— Od. 12. 437 vadreuews exdunv opp’ ebeuéoeev dalacw until it should vomit forth again(so 17.298., 20, 80), 308.] ‘ENS, ‘OSPA, ‘OTE. 225 Od. 23. 151 eipvoda péya ddua biaymepes elos tkouro tell he should come (so 5. 386., 9. 376). It is indistinct, or lost, in the ordinary use of dpa, as— Il. 6. 170 deigar 8 jvdyer 6 tevdepG dp’ dxddouro and with gs in Od. 4. 799 méume 5é ww. . elos [InveAdmeay . . Tavoere KKavOoto. The corresponding form of Object Clause with these Conjunc- tions may be traced in one instance of each, viz. II. 4. 46 5 Aeduy= HEvos Opa Taxiota Tevxea cvAjoee, and Od. 19. 367 dpodyevos elos ixovo. Here, after a Verb of wishing, the meaning until passes into the simple that. With éws and dpa the Opt. is nearly always pure: but we have é$p’ dv in Od. 17. 398 (until), 24. 334: and gus key in— Od. 2. 977 toppa yap dv Kara dorty Totintvcoofpeda pvdw Xenpwar damairiCovres, Ews «’ amd TdvTa d00ety where there is a stress on the particular time contemplated. There is only one place where éppa with an Opt. follows a Fut. or Subj., viz. Il. 7. 339 mUAas mojoouev .. dppa.. 68ds ein. But the example is open to doubt, partly because there may be a Subj. ety (see § 80), partly because the line also occurs (7. 349) where the governing Verb is an Imperfect, and therefore it may have been wrongly inserted in v. 339. It is true that the Opt. is found after the Future with other Conjunctions, to express remoteness or uncertainty; but a word which literally means ¢iiJ the time that is not quite so likely to have been used to express a remote end or consequence. 308.] Clauses with ére, émére, &c. Most Clauses of this kind are essentially — (1) Conditional. The Verb of the principal Clause may be— (2) An Optative of wish: as— Il. 21. 428 rovotros viv mavres, 8c01 Tpdecow apwyoi, vo) 4% elev 81’ Apyelovor paxotaro (cp. Il. 18. 465, &c.). (6) An Optative of expectation: e.9. Od. 13. 390 Kal Ke Tpinxoclotow eyav dvdpecor payotunv atv col, ToTva Ged, STE por TPdPpaca’ enapyyors. I. 14. 247 Znyvos 8’ odk av éywye Kpoviovos Gocov tkotuny, ovde KaTevrysaip Ore pH avtds ye KeAevou, (c) A Future: in one place, viz. Il. 13. 317 ainv of éooeira . . vipas évimpioat bre ph adtds ye Kpoviwy éuBddo. xrA., where the speaker does not wish to imply the fulfilment of the condition. In Od. 24. 343 év0a 3° dvd orapvdal navrota éaow, émmdre 5x Ards Gpar émBpi- ceav the Present éacw is open to suspicion, because all the rest of the description is in the past tense; with which the Opt. is in harmony. In IL. 4. 263 €o7nx’ dowep éuol, miéev bre Ovpds dvwyor the Opt. is read by most MSS. It may be regarded as an Opt. of the remoter event (§ 305, c), depending on méev, which is an Inf. of purpose, La Roche reads avwyp. Q 226 HOMERIC GRAMMAR. [ 309. (d@) A Past Tense, generally of an event which happens re- peatedly or habitually : as— Il. 1. 610 é00a mdpos oad’ bre puv yrvkds Orvos ixdvor. 21. 265 dood. 8 dpyrorere kth. as often as he started &e. Od. 8. 87 7 rou bre Aij€evev . . EAeoxev (iterative). So with ére after mpiv, in Il. 9. 486 odx eOddAeckes .. amply 7’ bre 8). . doaipe=you would only .. when dc.: ep. § 297. In this group of uses the Opt. is pure, except in— Il. 9. 524 oft kal Tdv Tpdabev emevddueba Kréa avdpav Hpdwv, Ste Kév Tuy? emiCaedos xddos tkou where the «év may be accounted for by the change from the Plural to the Singular; ep. § 283, note 3. (2) After a Past Tense of a Verb of waiting éxéte with the Aorist Opt. forms a kind of Object Clause: as I. 7. 415 moriéy- pevor Ommdr’ dp’ Oo. waiting for (the time) when he should come ; so Il. 9. 191., 18. 524, and (after pevovres) 4. 334. 309.] Clauses with énei. The few examples of this use show the same varieties as with ére, except that after another Opt. éwet takes dv: viz, in— Il. 9. 304 viv ydp x’ “Exrop’ édous, émel dv pada Tor oxedov edOor. So émiy in Il. 24. 227 (Opt. of wish), Od. 4.222. It is also found without dv or xév, in the statement of a supposition, Od. 24. 254 Toiovtm b& Eouxas, émet Aovoaito Payor Te, evSeevat padraxGs: and in the iterative use, Il. 24. 14. 310.| wet. The peculiar way of expressing a condition by a Negative followed by apiv (§ 297) is transferred to the past, the Subj. becoming an Opt., in— Tl. 21. 580 obk Gerken hevyew mplv werpijoatr’ ’AxiAfpos. The Optative with ei, S'c. _311.] Optative with c-—Wish. It will be convenient to be- gin with examples of this form of expression, although it is a question (§ 318) whether they are true Subordinate Clauses. Dl. 15. 569 ’Avridoy’, ob tis ceio vewrepos GAXos “Ayadn, ovTe TooW Odoowy obT GAKiwos Hs ov pdyeoOaL" el tid Tov Tpdwy e€ddrpevos dvdpa Bddrowba. More frequently a wish is introduced by ei ydp or at yép, as in— at ydp, Zed re matep Kai ?’AOnvatn Kal” AmodAov, KTA. Such a wish is sometimes used as a form of asseveration, as— 319.| CLAUSES WITH EL 229 Tl. 18. 464 at yap puv Oavdrovo Bvonyéos dde duvatunv voor amoKkpt at, OTe pty pdpos aives tkdvot, &s ot Tevyea KaAa Tapéooerat 2.é. ‘as surely as I wish I could save him from death’: so Il. 8. 538 (et yép), Od. 9. 523, &e.: and ironically— Od. 21. 402 al yap 5) roccobror dvijo.os dytidoeer, os otTds ToTE TOTO SurHceTas evTavicacbat. Here also we must place the wishes expressed by cide or aide, which have generally the character of hopeless regret: as ei’ os HBadouys Kerr. Note that in the Odyssey wish is not expressed by ei except in the combinations ei yap and ei@e. A wish is often followed by a Clause expressing an expected consequence of its fulfilment; ¢.g.— Il. 2. 371 at yap, Zet re mdrep.. TO ke TAY’ Hudoee TALS TIpedpovo dvakros. Od. 7. 331 Zeb wdrep, ail? 80a cine TeAevTHOELEY GnavTA *AAkivoos’ Tob pév Kev emt Celdwpov dpovpay do Beortov KAéos ein. For other examples see § 318, 4. So we should probably punc- tuate— Tl. 13. 485 ef yap éundtkln ye yevotueba TOS emt Ovyo" aiwd kev He pépoiro péya Kpdros He pepolunv. It is important to observe here that the combination of 4 Clause of wish with a Clause in the form of a Conditional Apodosis is not confined to examples with ei: see § 318, 2. 312.] Optative with ei—Conditional Protasis. Tlie distinc- tion between wish and supposition in the case of Clauses with et is mainly a question of punctuation; which must be decided by the general sense of the context. If the ‘Apodosis’ follows closely, the Clause with ei is necessarily subordinate, and instead of an independent wish we have a mere supposition, made in order to lead up to the Clause of consequence: as— Od. 1. 163 ef Keivdv y’ "Iaxnvde ido(aro vootjcavta, mavtes K dpnoaiar’ edadpdrepor médas eivat KTA. Here, although the first line expresses what the speaker would gladly see, it is better to take it as in form a supposition, or Con- ditional Protasis. This interpretation is required by the context in some places: as— Il. 7. 129 tovs vbv ei ardocovras ip’ "Exrops mdvtas axovoat, TOoAAG Kev GOavdrowot Pldras ava xeipas delpas. Od. 16.105 €f 8 ad pe TANOvi Sapacaiaro KrA. Q 2 2.28 HOMERIC GRAMMAR. [aaa When the Clause with «i follows the other, the Opt. expresses supposition, or at most concession : as— Tl. 22.20 9 o dv ticalunv, ef pou dbvapls ye wapely. Od. 8.138 ob yap éywyé ti pnue Kaxérepov dAdo Oadrdoons dvdpa ye cvyxedat, ef kal pada Kaprepos etn no matter if he is very strong (= even if he should be). The combination ds ei (or ds ei te) expresses comparison ; the principal Clause being in a past Tense, as— Il. 2. 780 08 8 dp’ icav ws ef re mupl xOwv Taca véepoLTO (cp. Il. 11. 467., 22. 410, Od. 9. 314., 10. 416, 420., 17. 366). Or negative— Il. 11. 389 odk drdyo as ef pe yuri) Bardon 7) mais appov. The use of et with the Opt. in the iterative sense (/f ever, whenever), which is common in later Greek, is not Homeric: the only passage which might be quoted as an example is— Il. 24.768 Gdn ef ris pe Kal dAAos évl peydporow evintor . GAAG od Tév y’ exéerot Tapapdpevos KarépuKes. 313.| Optative with «i xev—Conditional Protasis. This is a comparatively rare form; it can generally be explained in accordance with the other uses of xév: IL. 5. 273 ef rovrw xe AdBowmer Gpolpedd Ke Kr€os eoOAdv if (as I propose) we take them, we should Sc. Q. 141 ef b¢ kev “Apyos tkolued” "Ayarixdy KrA. if (as a further step) we reach Argos §e. 23. 591 intov b€ Tou adtos ddoe, Thy apounv’ ef kal vd Kev olkoOev GAXO pido enaitioeas, Apap Ké Tot adrixa dudvat Bovdotuny if (after that) you demand more §e. Od. 2. 76 «t y’ tyets ye payoure, Tay’ dv Tote Kal tics ef x ye pay x 7 ff (as I say is better, see v.74) you devour, then Sc. See also Il. 2. 123., 8.196, 205, Od. 2. 246., 12. 345., 13. 389., 19.590. And with the Clause with ei following the other— Il. 6. 49 rév Kév ror xapioaito Tarhp dnepeiow’ anowwa, ei Kev eve Cwov meTVOowr’ emt vyvoly ’Ayadv. So I. 1. 60,, 10. 381; ep. Od. 7. 314, and the use of od8 et kev not even in case, Il. 9. 445., 19. 322., 22. 220. 314.| Opt. with ci—Final and Object Clauses. These are generally found after a past Tense in the Principal Clause ; e.g.— Tl. 2.97 kijpuxes Bodwrres epirvov, ef mor’ dirhs oxolar’, dxovoeray d& KrA (in view that they should Sc.) 316.] SUMMARY OF USES. 2.29 Od. 4. 317 7AvOov, ef twd jou KAnnddva marpds evlorols I have come in the hope that you may tell §c. With Verbs of seeking, trying, desiring, &c. the Clause with ei has the character of an Object Clause : as— Il. 4. 88 [dvdapov dvriOeov duCyuevy ef tov epedpor seeking in the hope of finding (= seeking to find). With Verbs of telling, knowing, seeing, thinking, &c. the Clause with «i serves as Object. Examples of this idiom are almost con- fined to the Odyssey ; e.g.— Od. 1.115 dccdpevos narép eoOddv er) fpecty, et roev edOdv PynoTHpwr tov pev oxédaow Kata dduara ely z.e. ‘with the thought in his heart, whether his father would come and scatter the suitors:’ ep. 2. 351., 5.4395 9. 317, 421. Od. 12.112 €1 8 dye 81 roe Todro, Oed, vnweprés eviones el Tas Ti ddoiy yey drexTpopdyouys KadpuBdw tell me as to the hope that I may escape Sc. In three places the Clause follows a Primary Tense : Od. 14. 119 Zebs .. ode . . ef dé pov dyyelraupe lddv. 20.224 GAN rt tov dvoTnvo diopat et moOev . . Oly. Tl. 11. 792 ris & of’ ef Kev of ody daluovt Oupodv dpivass ; The pure Optative is used in all the places quoted, except the two in which et kev follows oi8e (Il. 11. 792, Od. 14.119). In these the structure is the same as in the corresponding indepen- dent Clauses. History of the Subjunctive and Optative. 315] Uses in Independent Clauses. The uses of the Subj. and Opt. in independent Clauses have been shown to fall in each case into two main groups. In one set of meanings the Mood expresses desire on the part of the speaker ; to this belong the Subj. of command and prohibition, and the Opt. of wish. In the other the Mood is a kind of Future; the Subj. being an emphatic or confident Future (like our Future with shall), the Opt. a softened Future, expressing ex- pectation, or mere admission of possibility (the English may or should). These two sets of meanings may be called the ‘quasi-Imperative,’ and the ‘quasi-Future.? We must remember however that they are not always clearly separable, but are connected by transitional or intermediate uses: such as (e.g.) the Subj. which expresses necessity (§ 277), and the Opt. of concession (§ 299, @). 316.] Uses in Subordinate Clauses. Passing over for the present the question whether the quasi-Imperative or the quasi-Future use is to be regarded in each case as representing the original meaning of the Mood, we proceed to con- sider the uses in Subordinate Clauses. Here the main distinction is that between « Final’ and ‘ Conditional,’ if these terms are used with some latitude: especially if we rank with the Final Clauses not only those which distinctly express the end or purpose of an action, but also all Clauses which are referred to the time of the 2.30 HOMERIC GRAMMAR. [316. governing Verb. It is true that this distinction does not always apply; e.g. to the Subj. in— Aavady dropupdped’ aixunraev, ot xev 52) xaxov ofroy dvanAnaaytes BAwvrar™ or to the Opt. in— GAAA TOAD pelCov . . punoripes ppacovras, d pr) TeA€oee Kpoviwv. ¥or there the Relative Clause is not really subordinate to the Verb of the Principal Clause : it is in sense a parenthesis, and is construed accordingly as an independent Sentence. Again, in— éooerau jyuap br dv wor’ bAwAD KTA. gppacadped” Ae vedped” ep’ Huerep’ He pévwpev. Belde yop pr) Acupov dmotpHoee KTX. and generally in Object Clauses, the Subordinate Clause does not express end ; but the time from which it is regarded as spoken is fixed by the governing Verb, in the same way that the time of a true Final Clause is fixed by the action of which it gives the end. For the present purpose, accordingly, there are two kinds of Clause to be considered, (1) Final and Object Clauses, and (2) Conditional Clauses. Regarding the meaning of the Subjunctive and Optative in Final Clauses there can be little doubt. The Subj. in most instances follows either a First Person (Present or Future), or an Imperative; that is to say, it expresses the immediate purpose with which the speaker announces his own action, or commands the action of others. Hence, by a natural transference, it comes to express the purpose of another person (viz. the Subject of the Principal Clause). Similarly the Opt., whether as the Mood of wish or of expectation, comes to express a wish or expec- tation not felt, but spoken of; and moreover, by virtue of its character as a softened or less confident Future, it naturally expresses a purpose that does not lie within the speaker’s own sphere of action or direct influence. It should be observed, however, that the shifting of the point of view from which a purpose is regarded is comparatively rare in Homer: see §§ 280, 281, 285,.286. This shifting is of course a criterion of a fully formed Subordinate Clause. In Conditional Clauses, on the other hand, the condition or supposition is not subordinated to the time of the governing Verb, but is made from the present point of view of the speaker. The question arises: What is the original force of the Subj. and Opt. in this use ? The answer seems to be that the Subj. is used in its quasi-Imperative sense, the Opt. as an Opt. of wish or concession. Both these ways of expressing a supposi- tion are natural enough, as may be shown by parallels from other languages, We say (¢.g.) ‘let it be so,’ meaning ‘if it is so’: and ‘were he here’ both in express- ing wish, and with the meaning ‘if he were here.’ Moreover, as is plain in these English uses, the choice of the Mood does not ultimately depend, as is sometimes said, on the greater or less probability of the supposition being true, but on the tone in which it is made—on the degree of vividness, as Mr. Goodwin says, with which it is expressed (Goodwin’s Moods and Tenses, § 50, 2). That is to say, the difference between et épxerat, édv €AOp, ei EAPOL, is not a question between he comes,—he shall come,—he may come, but rather between I suppose him to come, —L will have you suppose him to come,—I pray you to suppose him to come. This account of the matter is confirmed by the fact that negative Conditional Clauses take pq, not od: in other words they are forms of prohibition, not denial, Thus (¢.9.) ds py &AOD literally means, not ‘who will not come (od# dv €A6n), but, a7, ORIGINAL MEANING OF THE MOODS. 231 who musé not come, 7. e. ‘whom we must not suppose coming’; and similarly és #7} €A80. ‘ whom we do not choose to suppose coming.’ It may be objected that, if the ground of the use of py were the quasi-Impera- tive sense of the Mood, we should have et od with the Indicative. But there is no difficulty in supposing that 4 was extended to the Indicative on the analogy of the (probably more familiar) Clauses with the Subj. and Opt.; just as p17) dpedrov is an extension from the common use of p4 in wishes, and as in Final Clauses pq is used with the past Tenses of the Indicative. And this view is strongly sup- ported by the circumstance that in fact ei od with the Indicative occurs several times in Homer, in places to which the ordinary explanation (that od goes closely with some one word) does not apply: e.g.— Tl. 15. 162 ef 5€ pou ove énéeao” émmetoera: xTA. (so 178). 20. 129 ef & ‘Axideds od Tada Oedy éx mevoerar dupys 24. 296 ed 5€ Tor ob Biaer Edy dyyeAov KTA, Od. 2. 274 ef 8 ob Keivou y éaol yovos TA, See also Il. 4. 160, Od. 12. 382., 13.143. On the other hand, in the very few ex- amples of et od with a Subj., the od goes closely with the Verb, viz. Il. 3. 289 (ov« 20éAwoiv), 20. 139 (ove €idor), On the whole, therefore, it is probable that the Subj. and Opt. in Conditional Clauses represent the tone of requirement or desire (respectively) in which the speaker asks us to suppose the condition to be true. 317 | Original meaning. The question whether the use of the Subj. as an emphatic Future was derived from its use to express Will, or vice versa, and the distinct though singularly parallel question whether the Optative originally ex- pressed wish or supposition, take us back to a very early period in the history of Indo-European speech. The two Moods are found in the same uses (generally speaking) in Homer and in the Veda: the formation of these uses therefore belongs in the main to the period before the separation of the different languages, —to the period, indeed, when the original parent language was itself in course of formation. The problem therefore is one on which comparison of the earliest forms of the known Indo-European languages can hardly throw any light. It is as though we were asked to divine whether the use of shall in commands (thou shalt not—) or in predictions (ye shall see me) is the older, without being allowed to compare the uses of the word in Old English, German, &c. Perhaps, however, the fact that the Endings of the Subj. are generally the same as those of the Present and Future Indicative creates a slight probability against regarding it as an Imperative. The use of the Subj. for the Imperative of the First Person proves nothing on the other side, because a true First Person Imperative is in the nature of things impossible. Similarly the use of the Secondary Endings in the Optative points to the con- clusion that it was originally a kind of Past Tense. The loss of the Augment, being common to Greek and Sanscrit, would have to be referred to a very early period, but is not therefore incredible. The tendency to use a past Tense in wishes, and in some kinds of suppositions, may be amply illustrated. On the other hand it may be said that the simplest use of the Optative is that in which it expresses wish : yévouro means may if be/ but an additional word is necessary to turn this into yévour’ dv it may be. This argument must be allowed to have considerable weight, since it is likely @ priori that the simplest use of a form will approach most nearly to the primitive use; but it is not quite conclusive. We may suppose that yévouro originally meant it might be, and was often used to 232 HOMERIC GRAMMAR. [318. express a wish: whereas yévour’ dv (it might be in that case) could not express wish, but only supposition. Under such circumstances the tendency would be to avoid ambiguity by generally confining the ‘pure’ yévorro to wishes. An argument against Delbriick’s theory, that the Opt. originally expressed wish, may perhaps be found in the use of the pure Opt. with ot (§ 299, f). If the meaning passed gradually, as he supposes, from wish to expectation, we should expect to find pq retained (as in the Conditional Clauses). It may be observed, too, that in the case of the Subj. the tendency is to drop the quasi-Future uses, and retain those which are quasi-Imperative ; see § 322. 31 8.] Conditional Protasis with ei. The derivations that have been pro- posed for the Particle et or ai are too uncertain to furnish any ground for deter- mining its original meaning. It shows no trace of a temporal sense; and the form eita, which might be thought to be the corresponding Demonstrative (cp. érre(—érrevta) is not found in Homer. Curtius is disposed to identify ei with Lat. si, Oscan svai, from the Stem sva (ofe-). This derivation, however, does not suggest any theory as to the manner in which the Conditional Protasis may be supposed to have been formed. The question arises for us on the passages in which ei with the Opt. is used to express a wish. Thus in et tis kaAéoae J pray some one to call it is usual to take the Clause as Conditional, with a suppressed Apodosis («ad@s ay €xo: or the like). Or we may follow L. Lange in holding that the Clause is not Subordinate at all, the Particle ei being originally a kind of affirmative Interjection, used to introduce expressions of wish and supposition ; and we can thus explain the ordinary Complex Conditional Sentence as made up of two originally independent Clauses, viz. (1) a wish or supposition, introduced by ei, and (2) an assertion of the consequence to be expected from its being realised. The latter of these views has a priori the advantage of deriving the complex from the simple: and it is supported by Homeric usage. We find in Homer— (1) Wish, standing alone: ds amréAato Kal dAAos Otis ToLadra ye féCor. (2) Wish, followed by an independent Clause expressing expectation of a con- sequence : Od, 15. 180 obrw viv Leds Gein, épiySourTos méars “Hpns* TO kéy Tot Kal KE Oe@ ws edyerowpny. Il. 13.55 opaiv 8 Ge Gedy tis évl ppect Tomoeer, airh @ éordpevar kparepds nat dvwyépev dAAous* 7G xe kat éoovpevov wep epwnoat’ and vyay. 4.93 % 6a vd pot te miBo10, Avcdovos vit Saippor- Trains Kev KTA, So Il. 14. 107., 21. 428., 22, 285, Od. 8. 465., 14.193, &c. With these we may com- pare isolated cases in which the expression of wish takes a different form : Il. 3. 52 ob dy 37 peiveras dpyipirov Mevédaoy ; ywotns x’ otov pwrds éxes Oarepiy mapdxorriv. Od. 17. 242 7685€ por Kpnnvar’ éédAdwp, ds €AOot prey Keivos dvnp, aydyo dé € Saipwr TO Ké ror dydalas ye SiaoKxedaceey dndcas. (3) Wish, with ei, et yap, ei@e, &c., but without ‘ Apodosis’ I. 4. 189 ai yap 57 obrws ein, pidos & Mevérae. 320.] CLAUSES WITH EI. 430 11.670 «#0? ds HBdorpi, Bin BE por EumeBos ely, ws 6167 KA, (4) Wish, with ei, et ydp, ele, &., followed by a Clause of Consequence : Il. 7.157 €i0" ds ABdwoums, Bin dé por umedos ely TO Ké Thy’ dvrhoee KTA. Od. 15. 536 at yap Toiiro, feive, nos TeAgoere Kpoviwy" yoins x’ oin éur Sdvapus wal yeipes Exovra. (5) Supposition, with ¢i, followed by a Clause of expectation : I, 7. 129 robs viv «i mrwaooovras bp’ “Exrop: ndvtas dxovoat, MOAAG Kev GBavdro.r piras dvd yelpas delpar. The similarity in these examples is manifest. The type in the first four sets consists of a Clause of Wish, either standing alone or followed by a Clause of Consequence. As the Clauses are independent in (2), so they must he in (4): and if the Clause with the Opt. and «év is not the grammatical Apodosis to the Clause with ei, then there is no grammatical Apodosis. If so, there is no reason to supply an Apodosis in (3). Again, (5) only differs from (4) in punctuation, so to speak: the two Clauses are taken together, instead of being separated by a dis- tinct pause. On this theory, then, the Clause of Wish introduced by et is not an incomplete Sentence, derived from a Complex Sentence by omission of the Apodosis, but is one of the elements from which the Complex Sentence was itself developed. And regarding ei we have to suppose that it was originally used to introduce a Clause of Wish, whether followed by an ‘ Apodosis’ or not, and was confined by degrees to the Conditional « Protasis’: the use of «{@ and et yép being a survival of the earlier and simpler syntax. arg.] Final Clauses with el. This view of the original force of et makes it easier to explain the use of ei in Final Clauses, such as ef: ef xe mlOnrat, lit. I go —(suppose) he shall listen. The meaning here is essentially different from that of the Conditional sentence I go if he listens; and on the ordinary hypothesis, that the Clause with et originally expressed a condition, it is difficult to account for the two uses. If however ei merely indicates a hope or supposition, it is intel- ligible that the Clause should admit of the Conditional or the Final sense, as the context may determine. 320. | Order of the Clauses. A further argument on the same side may be founded on the observation that a Conditional Clause introduced by e in most cases precedes the Principal Clause. The reverse order is mostly found where the condition is a kind of afterthought, as in Clauses introduced by et wep dv even if, 008’ el kev not even if, and the like. Most Relative Clauses, on the other hand, follow the Principal Clause, and this may be regarded as the normal order—the Relative ‘ referring’ to an ‘antecedent’ Noun or Pronoun. In this point a slight difference is observed between the Iliad and the Odyssey. Lange has shown that the number of instances in which a Conditional et-Clause with the Opt. precedes is, in the Iliad 22, in the Odyssey 15. As the Iliad is longer than the Odyssey in the ratio 5: 4, this shows a slight difference in favour of the Iliad. Again, the instances in which the Clause follows are, in the Iliad 15, in the Odyssey 26; giving the Odyssey an advantage (when the length of the poems has been taken into account) represented by 13:6. It is also worth notice that (as Lange has pointed out) the form et mwep dv is confined to the Iliad, occur- 234 HOMERIC GRAMMAR. [42t. ring 4 times; also that ot8’ ev kev occurs 7 times in the Iliad, but only once in the Odyssey. That is to say, in the Iliad the inversion of the order is oftener marked by the use of a special form. gt, | The formula ei 8’ dye, with the varieties ei 8’ dyer’ (Il. 22. 381) and ei 8é (IL. 9. 46, 262), is often used in Homer to introduce an Imperative or Sub- junctive (§ 275). It has generally been supposed to be elliptical, standing for « 8° 20€Acis dye, or the like. And «i 8’ éAets is actually found with an Imperative in a few places: Il. 19. 142 ef 3 @édes émipewov, Od. 16. 82., 17. 277 (cp. 3- 324): It has been pointed out, however, by Prof. L. Lange, in his dissertation on this question,* that «i 8’ é0éAes is only found where it introduces a distinct second alternative. Thus in Od. 16. 82 the context is: ‘I will send the stranger where- ever he desires; or if you choose (ei 8’ é6éAcis) take him into your house.’ So Od. 3. 323 GAA’ 10 viv oby vnt.. ei 8 U€Aas weCds HTA. But with ei 8’ dye this is not the case. We find it at the beginning of a speech; as— Il. 6. 376 ef & dye por, Suwal, vnpéprea pvOnoacde, Od. 2.178 & yépov, ei & aye viv pavreveo xrd.: also Il. 16. 667., 17. 685, Od. 12. 112., 22. 391., 23. 35. Or in the Apodosis of a Conditional sentence, as— Od. 4.831 ef pev 84 beds doar, Oeoid re ExdAves aidijs, «i 8’ dye por eTA.: so IL. 22. 379-381. Or to express an appeal which is consequent upon something just said: as— Il. 1. 301 ray ote dv tT pépors dveAdy GéxovTos epeto ei 8 dye phy meipnoa (ay, come now and try): ep. Il. 8.18. 1.523 éuol 5é xe TadTa peAnoeTar Oppa TEACTTW" ei 8 dye To Kepadry Katravevoopa: (so come, I will nod my head). 23.579 «8 dy eyav adtods dixdow, Kai p ov Tivd pnt dddov émmdrngev Aavady eta yap éoTar" *Aytiroy’, ei 5° dye Setpo . . Suvuds KrA. come I will be judge myself . . so come, Antilochus, take this oath &e.: see also Od. 1. 271., 9.37. 21. 217.5 24. 336. Hence, as Lange argues, it is much more probable that ei does not express condi- tion, but has an interjectional character (cp. Latin eia age): and if so it offers a perfect parallel to the use with the Opt. of Wish (§ 311). We obtain a ‘survival’ —a use of ei with the Imper. and the Subj. in Simple Clauses—bearing the same relation to the Conditional Protasis with the Subj. (ei €d@y let him come) which we have supposed ei with the Opt. of Wish («i éX@0: may he come) to bear to the Conditional Protasis with the Opt. 322, | Homeric and Attic uses. The main difference between Homer and later writers in regard to the Moods may be said to be that the later uses are much more restricted. Thus the Subj. is used by Homer in Principal Clauses of every kind—Affirmative and Negative, as well as Prohibitive, Interrogative, &e. In Attic it is confined to the Prohibitive use with «7, and the idiomatic ‘ Horta- tory’ and ‘ Deliberative ’ uses. Again, in Subordinate Clauses the important Homeric distinction between the ‘pure’ Subj. and the Subj. with dv or «év is almost wholly lost in Attic. In * De formula Homerica ei 8’ dye commentatio, Lipsiae 1873. 323.] THE INDICATIVE. 235 Clauses of Conditional meaning, whether Relatival, Temporal, or introduced by ei, the Subj. with dv has become the only generally allowable construction: the pure Subj. being confined to a few instances in poetry. With the Optative, on the other hand, an equal uniformity has been attained by the loss of the use with dv or kév. In short, of the four distinct Homeric constructions— I. 0s EAD (57 ZAOn, i ZdAGy, Ke.) 2. 6s dv (or ds Kev) EXO (S7’ dv ZAOn, edv ZAOy, Kc.) 3. Os €XOor (Gre €AOor, ei ZADOr, Ke.) 4. 8s v (or ds Kev) EAOou (87° dy EAOou, Edy ZdADor, Ke.) the language dropped the first and last: with the result that as dv always accom- panied the Subj. and was absent from the Opt., it ceased to convey a distinct meaning, independent of the meaning given by the Mood. In other words, the use became a mere idiom. The change, though apparently slight, is very signi- ficant as an evidence of linguistic progress: for it is in such minutiae that real familiarity with a language is most certainly distinguished from a merely imitative or conventional use of it. In regard to Final Clauses the most noticeable point is the use of the Relative with a Subjunctive. In this respect Homeric Greek agrees with Latin: while in later Greek the Subj. was replaced, generally speaking, by the Future Indicative. It is also worth observing here that in Homer, as has been said (§ 316), the Final Clause in the great majority of instances expresses the speaker’s own purpose, not a purpose which he attributes to » person spoken of: see §§ 280, 281, 285, 286. In other words, the subordination of the Clause to the governing Verb does not often go so far as to put the Third Person for the First (e.g. ppaccera: ts ke vé- nrat = he will consider— how am I to return’). The further license by which a past purpose is thought of as if still present—so that the Subj. is used instead of the Opt.—is not Homeric (§ 298). Modal Uses of the Indicative. 323.| The Indicative is primarily the Mood of assertion: from which it is an easy step to the use in Negative and In- terrogative sentences. It is also used in Greek (as in other languages) to express mere supposition: thus we have «i in a Conditional Protasis with all Tenses (ei jv, éori, ef ora), where there need be no implication either for or against the truth of the supposition thus made. Further, the Indicative may be used in certain cases in a Conditional Apodosis, expressing an Imaginary consequence. Again, it may be used in Final and Object Clauses referring to the past or to future. All such uses, in which the Indicative does not assert, may be called Modal Uses. The tendency of language appears to be to extend the Modal Uses of the Indica- tive, and consequently to diminish the range of the other Moods. It is found possible, and more convenient, to show the modal character of a Clause by means of Particles, or from the drift of the context, without a distinct Verbal form. It will be seen, on comparing the Homeric and Attic usage, that the Indicative has encroached in several points upon the other Moods. 236 HOMERIC GRAMMAR. [324. 324.] Conditional Clauses (Apodosis). The Secondary Tenses, or Tenses of past time (Aor. Impf. and Plupf.), are used with dv or xév to express a supposed consequence: ¢. g.— Tl. 4. 420 dewdv 8 EBpaxe xadkds ent aryBecow dvakros épvvpevov' td Kev Tadacippova wep S€os cidev fear would have seized even the stout-hearted. This way of speaking of a conditional event ordinarily implies that the condition on which it depended was not fulfilled. For if (e.g.) the assertion #\ev he came is true, we can hardly ever have occasion to limit it by saying 7AOev dv, lit. he came in that case. Hence a Past Tense with dv or xév naturally came to be used where the event in question had not happened, owing to the non-fulfilment of the condition. Sometimes, of course, the event is not sufficiently definite to suggest anything as to its happening or not: e.g. in— Tl. 16. 638 08 dy ért ppdduwv rep dvip Sapnnbdva Siov éyvw (not even a shrewd man would have known Sarpedon) there is no implication that some one did know him. Again, the rule does not apply to events that occur repeatedly, or on no parti- cular occasion ; for there is no contradiction in saying of such an event that it happened when a condition was fulfilled. Hence the use in the iterative sense (as Hat. 3. 119 #daleoxe dv nal ddupécxero, Thuc. 7. 71 ef tues iSorev . . dveBdponodv te dv xrA.). This use, however, can hardly be shown to be Homeric. In Od. 2. 104 Oa kev juarin pev bpaiveoxer has slender authority, most MSS. reading évéa kat, Another doubtful instance is— Od. 18. 263 inmwy 7° dxuTddwv émBHTopas, of Ke TAXLOTA éxpway péya veikos KTA. where the commentators (Fisi, Ameis, Merry) take éxptvav as a ‘gnomic’ Aorist. On this view the use of «év has no parallel in Homer. An exceptional use of a different kind is— ; Od. 4. 546 7) yap puv Gwdv ye mexnoent, 7 Kev ’Opeatns areivey bTopOdpevos. Here «év marks the alternative (§ 283, . 2): either you will find him alive or (in the other case) Orestes has killed him (i.e. must have killed him). Thrown into « Conditional form the sentence would be: ‘if you do not find him alive, then Orestes has killed him,’ The Imperfect in this use may express either a continuous action which would have occurred at some past time, or an action (continuous or momentary) which would have been occurring at the moment of speaking. This, at least, is the later usage. Mr. Goodwin holds that in Homer the Im- perfect Indicative never refers to the time of speaking (Greek Moods and Tenses, § 49, 2, N.1): but the following is perhaps an instance :— Od. 4. 174 wal «é of “Apyct vaooa méAw .. * * * * 178 Kai xe Ody? évOG5’ Edy tes Emrorysped’, ovSE xev Fucas dAdo Siéxpivey pidéovTé TE TEpTopevw TE ply 7° bre 8) Oavdroo wéday végos dupendruper, 326.] THE INDICATIVE. 237 Here the Impf. éyro-yéyeGa takes in the present time: we should (from that time till now) have been meeting. The Impf. without dv or «év may express what ought to have been, if the meaning of fitness, obligation, &c. is given by the Verb or Predicate. Thus we have Od. 20. 331 wépduoy fev it would have been better. So in Attic with éxpiy, Set, and similar words. The Opt. with dv or xév, as we have seen (§ 300, c), is not un- frequently used in Homer with the same meaning as the Aor. or Impf. Indicative. The latter is the only construction in later Greek ; so that this is one of the points in which the use of the Indicative gained on that of the Optative. 325.| Past Tense by ‘Assimilation.’ When a Past Tense relating to an event which has not happened is followed by a Subordinate Clause, the Verb of the Subordinate Clause may also be in a Past Tense (the event which it expresses being equally imaginary): as— Il. 6. 345 ds we Sper’ quate TO... olxeoOat mpodepovea kaxh avewoto OvedAa, évOa pe kip. améepoe KTA. and so v. 350 dvdpos emeir’ Gheddov .. 85 4Sy KrA., and Od, 1. 318: also the use with mpiv, Od. 4.178 odd€ Kev tuéas UAdo Suexpivev . . mpiv y’ bre 8) Oavaroio pédav védos dudexddrvwev nothing would have parted us before the dark cloud of death had wrapped us round. See Kiihner, § 399, 6,a. This idiom is the same in principle as the use of Past Tenses in Final Clauses (Kiihner, § 553, 7), which is common in Attic with ta and &s: as Soph. O. T. 1393 Ti p’ od AaBdy Exrevas ebOUs, ws ede~a ph woTE KTA. that so I might never have shown &e. When the context has once shown that we are dealing with a purely imaginary event, the Indicative serves to carry on the train of suppositions. The Opt. would suggest a real purpose, and thus would be at least ambiguous. 326.] Future Indicative. The following points have to be noticed : 1. Homer not unfrequently uses xév with the Future, the effect. being (as with the Subj.) to indicate a limitation or con- dition: as— Il. 1. 139 6 8 kev Kexoddoerar and he (if I do so) will be angry. 52% Grd ob pev viv adris dmdoTiXE pH Te vonTY “Hpn’ euol b€ xe radra pedjoerar (to me, as my part). 4.196 kat Ké Tis OO epee in such case men will say. This use of xév is chiefly found after 8é, as I]. 1. 139., 6. 260., 8. 419, &c.: and in Relative Clauses, as Il. 12. 226., 17. 241., 22. 70, Od. 5. 36., 8. 318., 16.438. Cp. the use of kév with the Subj., § 275, 8. . . The Future with dy is very rare: see Il. 9. 167., 22. 66. 2.38 HOMERIC GRAMMAR. [326. 2. The use of the Future with the force of a gentle Imperative has been ascribed to Homer, but without sufficient ground. Where it appears to take the place of an Imperative it will be found in reality to express the zndifference of the speaker ; as— Tl. 6. 70 GAN avdpas kretvwper? erecta be Kal Ta Exndou vexpous dp mediov cvdnoete TeOvNSTas then you can (if you like) strip the dead of their arms. The forms olcete and dere, which are sometimes given as instances of this use, do not belong to the Future, but are Imperatives of an Aorist (§ 41). 3. The Future is occasionally found in Final Clauses with nearly the force of the Subj.: viz. with the Conjunctions mus (in Od. 1. 57 Oédyer Stas “lOdkns emudjoera charms so that he may JSorget Ithaca), and 8>pa, as— IL. 8. 110 Tpwoty 颒 inmoddpors lObvopev, dppa Kal “Exrwp eloerat «rd. (so Il. 16. 242, Od. 4. 163., 6. 218). So with pH, Il. 20. 301 py tws Kat Kpovidns cexoddoerat. The Future with xév in Relative Clauses sometimes appears to express end, as in I. 1.174 map eyorye wal GAAot ot KE pe Tysoouct: cp. 2. 229., 10. 44, 282., 23.675., 24.154, Od. 14. 333. In such places, however, as in the corresponding uses of the Subj. (§ 282), and Opt. (§ 304), it is difficult to say how far the notion of end is distinctly ex- pressed: ia other words, how far the future action is subordinated to that of the main Verb. The use of the Future in Object Clauses (common in Attic after Verbs of striving, &c.) may perhaps be seen in I]. 12. 59 pevoiveov ei teA€ovar, also Od. 5. 24., 13. 376. It is sometimes impossible to decide whether a form is a Future or an Aorist Subj.: e.g.in Od. 1. 269 o2 58 ppdfecba dvarya Ormws Ke pynoTipas amdoear, where the Verb may be either a Future, as in the places now quoted, or a Subj., according to the commoner Homeric construction. Mr. Goodwin (Moods and Tenses, § 44,1, N.1) holds that in these construc- tions the Future differs from the Subjunctive only by being a more vivid form of statement. The instances are hardly numerous enough for a good induction ; but a comparison of other uses of the Subj. does not bear out Mr. Goodwin’s view. In general, as we have. seen, the Subj. is akin to the Imperative, and therefore ex- presses the speaker’s purpose directly, by its own force; whereas the Fut. Ind. properly expresses sequence, and so comes indirectly to imply purpose. Thus OéAyer Ws AdOnTar literally means ‘charms so that he shall forget’: Oéryer Srws Anoera ‘charms so that he will forget.’ Hence the Subj. is naturally the more direct and vivid Mood. The same conclusion seems to follow from the rule that érws and éppa may be used with w Future, but not és or tva (Goodwin, zbid.). For és in the manner that fits a direct purpose better than 6Tws in some such manner that, or dpa till the time that. It would seem probable, then, that in Final Clauses the Future is a less emphatic and positive form of expression. And the Homeric instances do not prove the contrary. Thus when Achilles prays (Il. 16. 242), ‘embolden him so that Hector may know,’ the Future may well convey a shade of indifference, as though Hector’s knowledge were the natural conse- 328.] THE IMPERATIVE. 239 quence rather than the direct object. And so in Il. 1.175 of #é we Tipqoovar who will (I presume) honour me. 4. In Clauses with ei the Future is chiefly used (1) to make suppositions about definite events or issues: as— Il. 12. 248 ef 8& ob dniorhros apéLear, Hé Tw’ UAdov mappaevos enéecow amotpéwets KTA. Do Il, 1. 294, 2. 8874.5. 950y 19. 875., 15. 762, Od. 2. 115, Bees And with cév— Il. 15. 213 al xev dvev euédev . . wepidyoerat KrA. It is also used (2) to express a present intention or necessity : Il. 14. 61 tysets 8% hpalsyeO Gras ~orat rade Epya, ei te vdos pe€e (if wit is to be of any avait). 17.418 ef rodrov Tpdecat peOjooper (if we are going to §c.). ‘We may compare the Conditional Relative Clause— Il. 23. 753 dpvvcd? ot kal rodrov déOdov Teipioecde rise, ye that will make trial of this contest. Other instances of this use of the Future to express present conditions are given by Mr. Goodwin (§ 49,1, N. 3). Regarding the use in future conditions, he holds that the Future is a more vivid form than the Subj. Thus there would be three degrees of vividness, (1) «i with the Future, (2) éav with the Subj., (3) e with the Opt. The view taken above (§ 316) of the nature of the Moods in Conditional Clauses points rather to the conclusion that while the Subj. gives vividness or emphasis, and the Opt. has the opposite character (expressly avoiding a positive tone), the Indicative makes a supposition in a neutral manner. Accordingly it is generally found in Homer when the supposition is of an obvious or familiar kind. The Imperative. 327.| The Homeric uses of the Imperative present little or no difficulty. We may notice the use in concession, ironical or real : Il. 4. 29 &p8’, drap od rou waves ermaweopev Oeol GAAou. The forms dye and dyere are often combined with other Imperatives for the sake of emphasis: and sometimes dye is treated as indeclinable, and used where the context requires a Plural; as— Il. 2. 331 GAN’ dye pipvere mayTes KTA. (so I, 62., 6. 376, &e.). Similarly t@. is a kind of Interjection in Il. 4 362 dAX it, rata 8 dmiobev dpec- adpe? xrd.; and so we have Bacn’ i& (like ein’ dye). And Sette hither / is evi- dently an Imperative: cp. Il. 14. 128 deb7’ Youev wédeudvde. The corresponding 2 Sing. doubtless enters into the formation of Sefpo; but it is not clear how that word is to be analysed. 328.| Prohibition. The Aorist Imperative is very rarely used with pi: examples are— Il. 4. 410 7 pr} wor warépas 70” Guoin evOeo Tish (so Od. 24.248 ob dé put) xdAov evOeo Buys). 18, 134 od pey pa] TH Kataddceo pOdov “Apros. 240 HOMERIC GRAMMAR. [329- For the rule which is the complement of this one, forbidding the use of the Present Subj. with py, see § 278 jin. Regarding the origin of this curious idiom a very probable conjecture has recently been made by Prof. Delbriick (Synt. Forsch, IV. p. 120). In the Veda, it has been shown by Grassmann, the prohibitive Particle md is never found with the forms of the Imperative proper, but only with the so-called ‘spurious Con- junctive.’ Hence it may be inferred that the Imperative was only used originally in positive commands, not in prohibitions, Again, it appears that in Sanscrit the Imperative is nearly confined to the Present Tense: and in Greek the forms of the First Aor. Imper. (#Aé~or, Mid. «Aépar) are certainly of late origin. The fine distinction which is made, in the Imperative as well as in other Moods, between the continuous action expressed by the Present Stem and the momentary action expressed by the Aorist belongs to the specific development of Greek. Accordingly Delbriick suggests that the extension of the Imperative to express prohibition took place at a time in the growth of Greek syntax when the Aorist Imperative had not come into general use: and hence that it was only carried into the Present Tense. In other words, the form pa KAémte came into use in pre-historic Greek as an extension of the positive «Aémre: while px) KA€bns, which represents an older form of prohibition, kept its ground (side by side with the new pz) xAémre) because the form KAdpov did not then exist. This account of the idiom seems much more probable than any attempt to explain it on psychological grounds. CHAPTER XIII. Tur ParrticLEes. 329.] Under the term Particles it is convenient to group together a number of words that are mainly used to show the relations between other words, and between Clauses. In respect of this office they are akin to the various syllables or letters used as Endings: and with them go to constitute what are called the ‘formal elements’ of the language, in contradistinction to the roots or stems which compose its ‘ matter.’ The Particles which connect successive Clauses in any way form the Conjunctions. As such they may be distinguished, according to the nature of the connexion which they indicate, as Copulative (wat, ré, 48é, &e.), Adversative (8€, ddd, adtdép), Dis- junctive (j—h), Conditional (ct, av, xév), Illative (dpa, 8, obv), Causal (ydp), &e. Those Particles, again, which affect single Clauses may either serve to show the character of the whole Clause (as Affirmative, Interrogative, Conditional, &c.), or to influence particular words in it. We cannot, however, make a satisfactory classification of 331.| KAI, TE. 241 the Particles on the basis of these uses, because some of them are employed in several distinct ways: and moreover they enter into various combinations in which they often acquire new meanings. It will be best therefore to take them separately, beginning with the most familiar. , Kal, 330.] The uses of kat are in the main the same in all periods of Greek. It is (1) a Copulative Conjunction, conveying the idea of addition to what has preceded: Znvi ddws epéovaa kal dddows to Zeus and the others besides: ds ap eon kab xtdr. thus he spoke and thereupon Sc.: and (2) a strengthening or emphasising Particle meaning also, even, just: as— Il. 1. 63 7) cal dveipomddov or even a dream-prophet. 3.176 76 Kal khalovoa rérynka, which is the very reason that I am wasted with weeping. It is especially used with words that imply comparison, increase or diminution, extension of time or the reverse, &c.; as kal dAAos another (not this only), cal airds himself (as well as others): kat nddat long ago (not merely now), cal addis another time (if not now), kat pdda, cal Atqy (in a high degree, not merely in an ordinary degree): so with Comparatives, cal pet(ov, Kal piyo, &c. Both terms of a comparison may be strengthened in this way ; as— IL. 1. 81 ef wep ydp Te xodov ye Kai adthwap Karanéwy, GNAG Te Kal peTomiobev KTA. Notice, too, the use at the beginning of an Apodosis, esp. with Adverbs of tzme, as— Il. 1. 477 jjyos 8 apryévera pdvy pododaxrvdros Ads, kal Tor mer’ KTA. kat precedes the word which it emphasises, but is sometimes separated from it by other Particles, enclitic Pronouns, &c. ; as Il. 1.213 xal mor€ tot tpis téooa (not merely compensation but) three times as much: 2. 292 Kal yap tls @ éva piva pévav a man who stays even one month. So 7. 281 kal tOyev dravres (=topev kal mavTes). kal et and ei kal, The combination kai ei indicates that the whole condition is an extreme one: even on the supposition that—. But with the order ei kal the kat emphasises particular words: ei cal pada xaprepds éore even of he is (I will go so far as to say) very strong. Hence ei kat usually implies that the supposition is more or less true. é, 331.] The enclitic ré has two main uses which it is essential to distinguish ; besides one or two special uses of less importance. R 242 HOMERIC GRAMMAR. [332 (a) As a Conjunction ré connects clauses and single words. Tt is especially used when a new fact or new object is to take its place part passu with what has been already said: kéveoouw oiwvoicl re naot to dogs and birds as well: at act xaxov Tpdecor yévovto of 1” aire which were a bane to all the Trojans, and to himself (equally). This meaning is given still more distinctly by the Correlative 7é—1é: thus we have the pairs avdpGv Te Oedv te, dijuds Te TéALS Te, KAayyH T evomy Te, dAlyov Te pidov te, &c. and the pairs of Clauses expressing s¢multaneous action, such as— Avodpevds Te Oyatpa, pépwv 7’ amepelo’ dnowwa. dy 7’ dvexdpynoev, Bxpds Té puv etre maperas. The combinations té—xai and ré—a8é (or i8é) are also common in Homer, and not sensibly different in meaning from ré—ré: as— gyoker 7 dp’ erera kal & mendyero pps. As to the place of ré the general rule is that it follows the first word in the Clause. Hence ré (esp. when standing first in the pair té—ré) does not always follow the word which it couples: e.g. Il. 6. 317 éyyi@. re Mpidyoto kal “Exropos near both Priam and Hector ; Il. 5. 878 cot ? énimeiGovtar cat Sedunueocba Exaotos (cp. 2.198., 7. 294-5). The use of ré as a Particle of transition (to begin a fresh sentence after a pause) is not Homeric, though common in later Greek. This may indicate that the use as a connecting Particle was originally confined to the Correlative ré—ré. (Delbr. iv. p. 145.) 332-] (4) In its other use—which is distinctively Homeric— 7é serves to mark an assertion as general or indefinite. Hence it is found in guomic passages: ¢.g.— Il. 1. 218 ds xe Oeots éxumetOnrat, pada 7 éxdvov adrod. 9. 509 Tov d& yey Svncav kal 7’ xdvoy edEauevoro. Od. 6.185 udduora 5€ 7’ éxAvov adrol. Tl. 16. 688 Gan’ aiel re Atos kpeloowy védos é wep dvdpav. 19.221 ald re puddmibdos méAera Kdpos (ep. Od. 1. 392). In similes it is very common, and is often repeated in the suc- cessive Clauses: ¢,g.— Il. 4. 482 68 éy xovinot xayal mévev, aiyerpos ds, H pat ev elapevf EXeos peyddrowo Tepvuy Aeln, ardp té of Oot ex’ akpordrn meddace THY wey OF dpuaromnyds dvijp alOwvr oidipo e€érap’, Spa «tA, (cp. 5. 136-9., 15. 271-5, &c.). So where the meaning is frequentative :— ‘ Od. 4. 102 dddore pév re yop ppéva réptopar (ep. 5. 65., 12. 64). I. 19. 86 kal ré jue verxelerxov (20. 28, Od. 5. 331, &e.). 332.] TE. 243 So Il. 1. 521 verkel, kai ré we yor wth. and says (habitually) that I §e.: ep. 9. 410, Od. 1. 215., 4. 387., 17. 25. Hence it is used of names, as Il. 1. 403 avdpes b€ Te mdvres (kaddovor), 2. 814., 5. 306, &c. And of characteristic attributes, as— Tl. 2. 753 088 6 ye Wnverd ouppicyerar. GAA TE piv KAOUTEpHED emippéer HT’ EAatov. It may be laid down as a general rule that ré in the combina- tions péy te, 5€ Te, kai Te, ydp Te, GAAG Te, and the like, is not a Conjunction, and does not affect the meaning of the Conjunction which it follows. In a Complex: Conditional sentence of gnomic character the ré is often used in both members, as— Tl. 1. 81 ef wep ydp re xddov ye kal adriwap Karanéyn, GAAd Te Kal peTOmicben exer KdTOv. Under this head comes the use with the Article and the different forms of the Relative, a use which has been already discussed in the chapter on the Pronouns (see §§ 263, 266, 272 2.). It was there pointed out that ré is used when the Relative Clause serves to describe a class, as— dypia mdvra, Ta Te Tpéhet ovpeow VAN. peta & aplyvaros yovos avépos w te Kpovtwy xr. or to express a permanent characteristic, as— ynpas kat Odvaros, rad rT em avOpadmo.ot médovTaL. xoros, bs 7 epenxe ToAVHpoVA TEP YarETHvat. Awrtopdyer, of 7° dvOwvov efdap over. So ds te, Ste te, iva te, 2vOa te, Soo Te, ods te, &e. Of these ds te (or Sore) and oids te, with the adverbial dre and 颒 G te, are the only forms in which this use of ré has remained in Attic Greek. émet te, which is regular in Herodotus, is rare in Homer. It has been supposed that the ré in these uses originally had a connecting force, so that $s re was=and he; just as Lat. qui is often=et is. But this hypothesis is refuted by the circumstance that although other forms of the Relative may be explained as=and he, that explanation never applies to és te. Distinguishing between the Relative Clauses which add a new particular to our knowledge of a definite antecedent (KaéAxas bs 5n Calchas—he knew &c.), and those which define an indefinite antecedent (the man who—), we see that the meaning and he only suits the former, and that ré is chiefly used in the latter. The Indefinite tis is not unfrequently strengthened in its meaning (any one) by té:— Il. 3. 12 réccov tis 7° emdevooe: doov 7’ én) AGay iow. 14. 90 olya, uy tis 7 dAdos . . dkovon (so Od. 19. 486). So cal ydp ris re, kat yey tis re, and in Relative Clauses, és ris te, bre rls Te, ds ths re &Ke.: also iy ris re (Od. 5.120). Note however that the ré does not unite closely with ris (as in Latin guisque, &e.), R 2 244 HOMERIC GRAMMAR. [332- The uses of ré may sometimes be distinguished by its place. Thus the Conjunction ré comes before dpa, as Il. 2. 522 of 1’ apa and who Sc. (cp. & 7 apa, obr apa); and before ms, as Il. 8.7 aire tus kTA. Both uses may even occur in the same Clause, as Il. 24. 337 ds pjr’ dp tus ld pair’ ap Te vojon. Of the places in which ré appears to be used of single or definite facts, some at. least may be corrected without difficulty. Thus Il. 6. 367 od yap 7° of5a (so Od. Io. 190, &c.) was originally ob yép Foida. For Od. 20. 252 év 8€ re ofvoy there was an ancient variant év 8’ dpa olvoy; and so perhaps in Od. ro. 317. In Il. 16. 96 for rovs 5€ 7’ éay we should perhaps read rots 8 édav. In Il. 23. 277 dOdvarot Te yap clot, Moceddev 8 nop’ abrovs the té is a Conjunction (ré—Bé being occasionally found): otherwise we should have the order d0dvaro. yap 7° cici.* Some isolated Epic uses remain to be noticed :— (1) After an Interrogative in the combination 1’ &pa, 1° dp : as— IL. 1. 8 ris 7’ Gp owe OeGy pid. Evvenxe paxerOar ; 18. 188 més 7° dp’ tw pera prov; (so mH 7 Gp Il. 13. 307). Od. 1. 346 piirep euy, thr’ dpa pOovées KTA. The ancient grammarians regarded rdép as a single enclitic Particle (Herodian ap. Schol. Il. 1.65). As the force of the ré seems to have merged in the com- pound, this is probably right: just as y’ dp having become a single Particle is written yap. If so, we must also recognise the form Tapa, (2) With 4 in strong Affirmation: as 7 7° épdyny I did indeed think, This may originally belong to the same head as the in- definite use: 4 te= surely anyhow. But a distinct force of the té is no longer perceptible. (3) With the Disjunctive 4, after a Comparative: as Od. 16. 216 ddworepov 7 7’ olwvol. So in IL. 4. 277 peddyrepor ire niooa (blacker like pitch) the true reading may be qé te micca. This was suggested by Bekker (Hom. BI. I. p. 312), though not adopted by him. On the other hand Buttmann (Lexil. s. v.) takes Hite as used = than, and would read nor’ in Od. 16. 216. On 4 Te—7 TE either—or, see § 340 fin. (4) After od8é and pSé: as— Tl. 1. 406 rév kat brédercav pdxapes Geol, obdé 7” ebnoav. 2.199 GAN’ tO vov kara Aady ’Ayatdy, unde 7” epder. There is generally some marked parallelism between the words of the two Clauses: Il. 15. yo9 réfew ., ov5é 7° axdvTwv, 17.42 ameipnros .. ovde 7 GdNpiTOS, 22, 300 éyyU . . ovd5€ 7” dvevOev, 23. 622 madaices obb€ 7° dxovriaTiv eadvccat, 23. 730 Kivnoev .. obE 7” deipev, 24.52 HddAALOV OvdE 7° dpevov, Od. 12.198 pboyyijs . . ovd€ 7’ dodfjs. The remaining instances are, ovdé 7” €ace (Il. 11. 437.5 21. 596), ob5é 7’ éAnye (Il, 21, 248). In most, if not all, instances it is at least possible to read o¥8? é1’, py €r’ (for rz): ep. Il. 12, 106, Od. 12. 198., 24. 401. * The account now given of the uses of té was suggested (in substance) by Dr. Wentzel, whose dissertation (Ueber den Gebrauch der Partikel ré bei Homer, Glogau 1847) appears to have been overlooked by subsequent writers. 334-] AE. 24.5 The Latin que, which is originally identical with ré, shows the same separation into two main uses. In the use as a Conjunction the agreement is close. It is less so in the other use, chiefly because ré in Homer is still a distinct word, whereas gue in Latin is confined to certain combinations, viz. a¢-que, nam-que (ep. nal Te, GAAG Te, yap Te, &c.), ita-que, the Indefinite quisque (with the correspond- ing forms uwbique, quandoque, uterque, &c.) and the Relative quicunque. The two uses are also united in the Sanserit ca, which as a connecting Particle agrees closely with té, and is also found after the Indefinite kas, especially in the com- bination ydh kdg ca (és tls te). See Delbriick, Synt. Forsch. iv. p. 144. 8é. 333-| The Adversative 8€ properly indicates that the new Clause stands in some contrast to what has preceded. Ordinarily, however, it is used in the continuation of a narrative (2. e. to show that the new fact is not simultaneous, as té would imply). It often serves also to introduce a parenthesis or subordinate state- ment: ¢.9.— vodooy ava otpardv apoe kaxny, GAéxovto dé Aaol, otvena KTA, Here a prose writer would say dAcOplav, or dote Aatdd\Avabat tovs otparidras, or tf’ is of orpaTiGtat dm@drdvyTO, &e, So— Aptidoxos 5& Midwva Bad’, jvioyov Oepanovra, écbdov ’Atupreddyy, 6 8 tréoTpepe pavvyxas inmovs, Xepuadlo dykGva TuxXe@y jécon, T.e. ‘struck him as he was turning the horses.’ 8é is nearly always the second word in the Clause. It is ocea- sionally put after (1) a Preposition and Case-form, as éa’ airay & dpodérncay, or (2) an Article and Adjective, as 77 dexdtyn 8 KTA.; but not after other combinations. Hence «ai 5é, as Il. 7.113 kal 8 "Axideds and even Achilles (never xai—é, as in later Greek). 334.] 3€ of the Apodosis. While 8é serves in general to mark the beginning of a new independent Sentence, there are certain uses, especially in Homer, in which it appears to connect a principal Clause with a Relatival or Conditional Protasis. This is occasionally found where there is an opposition of some kind between the two members of the Sentence: e¢. g.— Il. 4. 261 ef wep ydp 7’ Gddou ye Kdpy Koudwvres ’AXasol daurpov aivacw, sdv d& mAciov démas KTA. (SO 12. 245). 5. 260 at xév por ToAdBovdos ’AOjvn Kidos dpefy dypotépw Kretvat, oy Ot. . €pvKkakéery KTA. Od. 7.108 dccov Painxes mepi mavtwv idpres avdpav via Sony évi mévrm ehavvewer, Ss b& yuvatkes ioroy rexviooat (ep. IL. 592., 14. 178, 405., 18, 62). 2,46 HOMERIC GRAMMAR. [335 And with pév in the Protasis— Il. 2. 188 8vriva pev Bacidfja cab eoxov dvdpa Ktxein, Tov 8 dyavots émeecow KTA, Much more commonly, the Clause which 8¢ ‘of the Apodosis’ introduces is opposed to a preceding Principal Clause (from which of course it is separated by the Protasis) : as— Il. 5. 436 tpls pev eneur’ erdpovce karaxtéuevat pevealvav, Tpis d€ of eorupedice paewiy donid’ ’AndAdwv" GAN’ bre 8) 7d Téraprov éemécovto daipovr icos, dewva. 8 duokAjoas mpoocépy Exdepyos "ATO\AwY. Here the last line contains an opposition not so much to the Protasis add’ dre «rd. as to the first Apodosis rpls 5¢ of éorupércte xt\. So in many places the 8¢ of an Apodosis appears merely to repeat the 8€ (or equivalent Particle) with which the whole sentence was introduced: ¢.g.— D. 1.57 of & eet ody tyepdev . . toict 8 dviordpevos petépy mddas axis "AythAeds. 137 ef 8€ Ke py ddwow, eyw dE Kev adTos Elapan. 16. 198 adrap éreidy mavtas dw yeusverow "Axsddreds orhoev eb xpivas, kpatepov & emt pov éredre. Od. 3. 473 adrap énel mdotos Kal ednrvos && Epov évto, rotor dé pUOwy Tpxe KTAr, So Il. 2. 718., 4. 213., 7. 549, 314., 9. 301, 511, &e., Od. 7. 47, 185., 9: 311+, 10.112, 365, &e. When the Protasis is not introduced by any Particle, there may nevertheless be an implied opposition such as to account for the appearance of 8€ in the Apodosis * Thus we have in Il. 1. 188-194 &s Paro, TnAcinn 8 dxos yéver’, év 5é of Hrop KTA. . . efos 6 Tad6’ Wppawe . . HAGE & ’AOHy KTA., 7.e. Achilles deliberated whether he should kill Agamemnon, but while he was deliberating Athene came. The words eios 6 kA, are a kind of Asyndeton, equivalent to adrdp éws rav@’ xrA.: ep. Il. 10. 507., II. 41I., 15. §39., 17. 106, Od. 4. 120., 5. 365, 424. The idiom is also found with dpa (IL. 4. 220, Od. 10. 125), and edre (IL. 6. 392., 12. 373, Od. 3.9., 17. 359): It has been observed that when the Protasis is a Relative Clause, S€ of the Apodosis is generally found after a Demonstrative. The only exceptions to this rule are, Il. 9. 510 ds 5€ «’ dvqvyta . . Atooovta 8 dpa rai ye eTd., and Il. 23. 319 GAN bs pév O tnmocot . . twmor 88 mAGvdavTat «TA, (Schémann, Opuse. Acad. ii. p. 97.) 335.| Enclitic 5¢. There are two uses which may be noticed under this heading :— (1) The 8¢ of 8-8¢, téc00-8e, toidc-8e is properly an Enclitic (as the accent shows). * This was pointed out by Nigelsbach in his Anmerkungen zur Ilias (p. 261 and p. 271 ff. ed. 1834). The Excursus which contains his discussion of the point is oitted in the later editions. 338.] *AAAA, ’AYTAP, ’AY, ’H. 247 The form toto-Seo or toic-Seaa may be » trace of an inflected Pronoun akin to 8€ (related to it perhaps, as tis to ré); or it may be merely a form created by the analogy of other Datives in -eoo., -eov. (2) The 8¢ which is suffixed to Accusatives expressing motion to is generally treated as an Enclitic in respect of accent: as oixdvde, méAeuovde. The ancient grammarians, however, wrote 8é as a distinct orthotone word, hence ofkoy 5€, méAquov dé &e. (but olkade, pvyade were made exceptions). It seems likely that the -8¢ of these two uses is originally the same. The force in both cases is that of a local Adverb. Whether it is to be identified with the Conjunction 6é is a further question. GNA, avtdp, dtdp, at, atte. 336.| The remaining Adversative Particles do not need much explanation. édhd and adtép are used (like 8€) i apodosi, especially after a Clause with ef mep: as— Tl. 1. 81 ef wep yap re. . GAAG Te (ep. 8. 153., 19. 164). 22. 390 ef b& Oavdvray TEp .. abrap ey KTA. atrép and drdp express a slighter opposition than dddd, and accordingly are often used as Particles of transition, e.g. in such formulae as ds of pév.. adrap xrA. A similar use of é\\d may be seen with Imperatives; as GAA’ 101, GAN’ dye wou Tdde eine, and the like. It is evident that the stronger Adversative is chosen where greater /iveliness of tone is to be conveyed. 337.| a8 and aire (again, on the contrary) have nearly the same force as attép, but do not begin the sentence: hence viv ad, tis 8 ad, rinr’ atre, &c. Originally, doubtless, aé meant backwards, but in Homer this sense is only found in the form aéms: though perhaps it survives in the sacrificial word avdépucay. The form 8pus is later, the Homeric word being guns. Spos is usually read in Il. 12. 393 Spuws 8 ob AnOeTo xdpyns, and Od. 11. 565 évba x’ 8pws mpocépyv. In both places however the Scholia indicate that the word was anciently circumflexed by some authorities. m He 338.] The Particle 7 at the beginning of a sentence gives it the character of a strong affirmation: Il. 1. 240 9 wor’ ’AytAAjos 10O% tera be sure that one day Sc. So, with an ironical tone, Il. 1. 229 7} ToAd Adidy eos KaTa oTpaTdv edpdy ’AxaLdv dGp’ AmoaipetaOae KTA, 248 HOMERIC GRAMMAR. [339. It is often used interrogatively, esp. in questions of surprise, indignation, irony, &c.: as— I. 2. 229 3 ére cal ypuood emideveat xrr. 15. 504 4} €AtreoO” iv vias €An KopvOaiodros “Extwp euBadoy t€ecOar kth (do you really hope §c.). Od. 2. 312 i odx Aus ds Krad, (is it not— ?= surely it is): ep. § 358, ¢. Occasionally, in short parenthetical sentences, 4 has a concessive force, it is true that, hence and yet, although: as— Il. 3. 214 radpa per, adda pddra Aryéws, evel od ToAvpvO0s, ovd ddapaproemis’ 7} cal yeéves torepos rev. 11. 362 e& ad viv epvyes Odvatov, Kdov" 7} TE ToL ayxe 7AGe kaxdv (so 18. 13, and ep. F phy in Il. 7. 393). 2.2. 280 9 ror &pns ye (=though you said so), ep. § 346. The question whether 4 (or 4) can be used to introduce a Dependent Interroga- tive depends upon a few passages. Bekker favours 4 in this use, and reads accordingly, e. g. Il. 1. 83 ob 58 ¢pdou 7 we cawoes. The majority of the editors recognise it in three or four places: Tl. 8, 111 efcerar 4 wat nov Sdpu paiverar xd. Od. 13. 415 @xeTo wevadpevos pera adv KAé€os, i Tov er’ ens. 16.137 GAd’ bye por 7é5¢€ cine nal drpexéws KaTadrctov, f} nal Aaéprn abtiy b8dv ayyedos EAOw. 19. 325 THs ydp eyed ad, feive, Banoea, 7 Te -yuvarkay GhAdow reEpicrpme ; In all these places, however, there is manuscript support for ei, and so La Roche reads in the two last. For the use of ei with the Subj. see § 294, with the Opt. § 314. It is difficult to derive the general use of 4 which Bekker supposes either from the emphatic 4, or from the disjunctive 7é or 7 (Hom. BI. p. 59). In any case there is no sufficient ground for deserting the MSS. 4 is often combined more or less closely with other Particles : as y te (§ 332, 2), f pdv, &e. (§§ 343-5), F tor (or Frox), 43 (for 7 64), and the correlative jpev—adé. In these combinations 7 strengthens the other Particle. Note that— i jpev—ibé are used of slightly opposed things, especially when alternation is implied: as— Od. 2. 68 Aiccopat juev Zyvds ’Orvurlov 7de O€uaros, HT avdpGv ayopas huey Aver HSE KadéCer. a.e, ‘assembles and dissolves again in turn’ (Lat. ¢wm—tum). Cp. Il. 8. 395 juév dvaxdivar . . 4d’ émOeivar: and so probably Il. 6. 149 nuey vee 7d’ doAnyet 48 is also used (=and) without a preceding jpév: but not to begin a fresh sentence. Cp. § 331 fiz. for the similar use of ré. 339.] # after ti, émel. It is a question whether we should write rin, émem, or ré i, eet 7}, as is done by the most recent 340.] "HE, ’H. 249 editors, The form ézet 7 is supported by the fact that it is chiefly found in the combination émet 4 modd «rd. (Il. 1. 169., 4. 56, 307, &e.); also with pada (Il. 1. 156 énel 7 pdda ToAAG perags xrh., Od. 10. 465), and kat (Il. 20. 437, Od. 16. 442). The case is different with ri; we must surely write rin 8€ od (Il. 6. 55.) 14. 264, &c.), not ri # 8. The ancient grammarians recognise a difference, their rule being that y after émet is circumflexed, after ti acute (La Roche H.T. p. 267). Trypho, however, wrote tiy as one word; while in Attic we are told that it was accented tq (cbid.). Whether the 1 in either case is identical with the Affirmative 7 seems very doubt- ful. It is at least a distinct use, marked by a different place in the sentence. With rim is to be placed the emphatic Nom. tév-y thou, a form which occurs in the Iliad only (cp. the Doric éyév-n). Ne, 1. 340.| jé and 4 are used in Homer as equivalent forms of the same Particle: which is (1) Disjunctive (=or) and (2) used after Comparatives (= than). The use of the Correlative 4é (4)—aé (4) =cither—or is also common in Homer: as Il. 1. 504 7 emer i} epyw: 3. 239 7) odx éorréoOny . . 7) Sevpw pev Emovto KTA, When a question is asked in a disjunctive form, the second member of the sentence is introduced by #e or 4 (2.¢. the accent of the Particle 4é, 4 is thrown back) : Il. 13. 251 Hé re BEBAnat, Béreos 5€ ce Telper dkwxi, He Tev dyyedins per ew’ HArvoes 5 Od. 4. 632 ‘Avrivo’, 7 6a tt per evi dpecty, je Kal odkl; So when the first part of the question is not introduced by a Particle; Il. 10, 534 wevoouar 7 ervpov epéw; shall I speak Jalsehood or the truth ? Od. 1.226 eidanivy rie yauos; Indeed the first half of the sentence need not be interrogative; as Od. 21. 193 émos Ti xe mvOnoaluny, 7 adiros KevOw; I would say a word ; or shall I keep it to myself ? (so Il. 14. 190., 23. 465). One of the members of a disjunctive question may be itself Disjunctive: ¢.g.— Il. 6. 377 aij 2Bn ’Avdpopdxn AevKdAevos ek peydpoto ; He 7 2s yadrdwor 7 elvarépov einénAwr, 9 es AOnvains e€olyerat xrA. Here 7) elvarépwy offers an alternative for yaddwy, but the main question is between these two alternatives on one side and és ’AOnvatns xrA. on the other. Most editors of Homer recognise an interrogative use of the form We, but erroneously.* The questions in which je is found * This has been well shown by Dr. Praetorius, in a dissertation to which I am largely indebted (Der homerische Gebrauch von 4 (#¢) in Fragesitzen, Cassel 1873). 250 HOMERIC GRAMMAR. [341. are all disjunctive, so that we must write #é—je (Il. 6. 378., 13. 251, 15. 735+, 16.12, 13-17, Od. 1. 408., 2. 30-, 11. 399). In— Od. 13. 233 ris yj3 tls djpos; rlves dvépes eyyeydacuy ; q mod Tis Vicon eddeledros, HE Tis AKTH | Ket KTA, 4 wou means surely methinks: the sense being, ‘ what land is this? It must be some island or else promontory.’ Hence we should read 4é, not we (as Ameis, &c.). né or = than is found after Comparatives ; also after Verbs implying comparison, as BovAoua: I prefer, bOdve I come sooner. The correlative 4 te—# Te appears in three places, viz. Il. 9. 276 4 7” dvipav Te yuvaix@v (where it seems to be=juév—75é), 11. 410 @ 7’ EBANT’ H 7 EBan’ GAdov, and 17. 42 #7’ dAxis 4 re pdB00 (where however Aristarchus read 45— 7182). The single 4 te occurs with the meaning or in Il. 19. 148 4 7 éxémev mapa cot: with the meaning than in Od. 16. 216 (§ 332, 3). Considering the general difficulty of deciding between «i and 4 in the text of Homer, we cannot regard the form # Te as resting on good evidence: see the next section. 341.] Dependent Interrogative Clauses. A Disjunctive question after a Verb of asking, saying, knowing, &c. is generally expressed by the Correlatives i (j)—ye (): as— Od. 1.174 kat pos tot’ dydpevaon erntupov, opp’ ev €id6, He véov peOerets, 7} Kal TaTpaids eoot KTA. IL. 2. 299 rAfre pidor kal pelvar’ emt xpdvov, dpa Sayer, h ereov Kddyas pavreverat, je kal ovxi. Other examples have been given in the account of the Subjunc- tive (§ 280) and the Optative ($ 302). In general it will be seen that these Dependent Clauses are the same in form as the cor- responding direct questions. In a very few instances the first member of a sentence of this kind is without #é (4): viz.— Od. 4. 109 ovd€ rr Hyer | Gerd y’ 7 TEOvnKE (cp. 837., 11. 464). So I. 10. 544 elm’ dye... Oates todos innovs AdBerov, karaddvres Surrov Todwv, 7 tls cpwe mépev KTA. The combination ei—ye (4) is often found in the MSS. of Homer; see Il. 2. 367., 8. 532, Od. 4. 28, 712, 789., 16. 238, 260., 17. 308., 18. 265., 24. 217. La Roche (following Bekker) reads i—1e (7) in all these places. The common texts have in one place et te—He, Il. 2. 349 yuepevar ef Te WedSos Erdoxears He eat odxi. In this instance, if the reading is right, there is a slight irregularity: the speaker beginning as if he meant to use et re—ei re, and changing to the familiar je kat ovxi. But the best MSS. have et re—et Te. A change of construction may also be seen in Od. 24. 235-8 pepunpige . . evooa nal wepupivac .. 7 mpOr’ eepéorto he debated about embracing &c., or should he Jirst ask &e. 345:| MAN, MHN, MEN. 251 pad, piv, per. 342.] The three words pd, pry, pév agree so nearly in meaning and usage that they are probably to be regarded as etymologi- cally connected, if not merely varieties of the same original form. They express strong affirmation (= surely, indeed, &c.), and thus acquire two main uses: (1) the concessive use, preparing us for a Clause with an Adversative 8¢, airdp, adddd, &c.: and (2) the use in the second of two Clauses with the meaning yet, never- theless. 443,] pdv is only found in the Iliad. It has an affirmative and generally a hortatory or interjectional force: as in dype pav nay come! (Il. 5. 765., 7- 459), and 7 pdv, od pdv, used when a speech begins in a tone of Saris, triumph, or the like ; as— Il. 2. 370 7 paw adr’ dyopn vikas, yépov, vias *Ayadv. 12. 318 ov pay dxAnets Auxiny Kata Kowpavéovaw neétepor Bactdijes (cp. 4. 512). The force of an emphatic yet appears in— Il. 8. 373 éora pay br dv atre pidnv yAavedmda elny’ and in dAX’ od pay (II. 5. 895., 17. 448, &e.), wry pay (Il. 8. 512, &c.). 344. py with a hortatory force occurs (¢.g.) in Il. 1. 302 8 aye piv melpnoa do but try. In the combination piv it is pointedly concessive, admitting an objection or reply: as Il. 7. 393 7) pay Tpdés ye Kédovtas that notwithstanding that the Trojans bid him: Tl. 9. 57 7 pay kat véos éoot true that you are young. In kat pv it emphasises the new fact introduced by nat: Il. 19. 45 kal pny ot rére y’ els dyopiy toav observe that even these went &c. 345.] pév is very common in Homer. An instance of a simply affirmative use may be found in Il, 1. 234 val pa tdde oxjrTpov, TO pev od more HUAAA Kal dCovs pice. which will surely never put JSorth leaves or shoots, Generally it marks an opposition, either (1) to a following Clause—a use which is common to all periods of Greek,— or (2) to what precedes, with the meaning yet, how- ever, for which Attic writers use py, as— Il. 1. 603 od pev pdpysyyos nor yet of the phormina. 2.703 ovde pev odd’ ot dvapxo. Ecav, TéOedv ye pev apxov. Od. 15. 405 ob T TepitAnOijs Alyy Técov, AAN ayabh per. Similarly, 7 pév is a strong affirmation = Attic 7 yyy), as— 11. 7.97 7 mey 89 AGBy rade y’ Eooerar (9. 252, 348., 16. 362, Od. 14. 160). Similarly we have pi pév (in negative oaths, as Il. 10. 330, &c.), 25% HOMERIC GRAMMAR. [346. and kat pév (= yes and, yes too), answering to the later kat pujy: also érap pév (= GAG pv). . From this second meaning is derived the use of pév with a resumptive Article or Relative, which is common in Homer: as in Od. 6. 2-13 abrap ’AGjun Bij p és Paujxwy «rd, (then follows a digression about the Phaeacians, ending) *AAkivoos 8& rdr PXE, Ocv amo pidea €ldds* Tod pev &Bn mpds dana now it was to his house that she went Sc. Here pév marks the slight contrast between the digression and the main thread of the story. ’ TOL, 346.] The enclitic toi seems properly to express a restricted affirmation, generally qualifying a preceding statement : at least, yet surely, &e. It is especially used where a speaker wishes to imply that he is saying as little as possible: as Il. 4. 405 jets To. Tatépwov wey’ dmelvoves ebydued’ iva: 8. 294 od pév tow bon ddvapls ye Tdpeott Tavopar: cp. 5. 873., 6. 211., 10. 250, Od. 2. 280, &. So again in maxims, Od. 2. 276 madpor yap tor maides xtr. few children, it must be said, &e.: Tl. 23. 315 pajre roe SpuTdpos KTA, it is by understanding, after all, that the woodman Se.: Od. 9. 2.7 ot Tot éywye is yains Sdvauat «tr. L cannot, when all is said, &e.: Od. 8. 329, &e. rol is combined in Homer with Adversative Particles, as autép rot, addd tor (11.15. 45, Od. 18. 230); and with the Affirmative 4 in ¥ tot (or Hrov), which expresses a restricted concession (Il. 1. 140, 211., 5. 724, &c.). But the combinations kairo and yet, pévtot but yet, toivuv so then, are post-Homeric: and so is the Disjunctive qrow either, or. toi has the first place in the sentence in the compound rovydp, which is used to begin speeches; as Il. 1. 76 rovyap éyav épéw so then I will speak. It is generally used with the First Person, and has a kind of apologetic force (=J will say, since I must speak), In Attic it survives in the compounds rovydpro, tol- yapodv: and the same meaning is commonly expressed by roivuv. It has sometimes been thought that tof is originally the same as the Dat. of ov, meaning ‘I tell you’ or the like. The orthotone tovydp (or rot yap, as some MSS. read) is difficult to explain on this view. It has also been explained as the Locative of +6: cp. the Dat. 7G =in that case, therefore. Or it may be from the same stem as tis and ré (as Kiihner holds, § 507). Cp. rou (84 mov) = somehow, thence surely. But these conjectures do not reach a high degree of probability. dpa, ydp. 347-| The Adverb apa properly means fittingly, accordingly oe dp- to fit). The forms ap and $¢ are merely varieties pro- uced by apocope, in hasty pronunciation. Of these dp retains its accent, but 64, the shortest form, is enclitic. 348.] TOI, APA, TAP. 253 The ordinary place of dpa is at the beginning of a Clause which expresses what is consequent upon something already said. But occasionally it follows a Participle in the same Clause, as in ioe ‘alias 4 tor & y ds elnav Kar’ dp’ Cero (cp. IL. 2. 310., 5. 740). It is to be observed, however, that apa may indicate a reason (as well as a consequence): that is to say, we may go back from a fact to the antecedent which falls in with and so explains it. Fig. V1. 429 yodpevov xara Oupov eidvo10 yuvaixds, tiv pa. . amnvpwv whom (and this was the reason of his anger) they had taken away. So in the combinations ds fa, émet fa, dtu fa, over’ dpa because (and this is the explanation): also in yép fa, as Il, 1.113 kat yap pa KAvratprynorpys mpo8éBovda. dpa is also found in the first of two correlative Clauses; as— er dp’ by’ ebywAgs émpéeuderat ef 0 ExatduPns. Os dyay’ ws pyr dp tis 1dn par’ ap re vonon. The parallel form of the sentence enables us to regard the first Clause, by anticipation, as falling in with and completing the second. The Attic dpa is unknown to Homer. Whether it is identical with dpa seems doubtful. It is worth while noticing that (as Hartung points out, I. p. 455) dpa answers in usage to the Homeric combination 7 fa (is it then—?). 348.] The Causal Particle yép is originally a compound of yé and dpa, but the two elements have so completely united into a new whole that the fresh combination ydép fa is found in Homer. yép serves to indicate that the Clause in which it is used is a reason or explanation, usually of something just mentioned or suggested : as T@ yap emt pect Ofjxe Oed AevKwAEvos “Hpn’ KydETO yap Aavaéy, xrA. Thus it follows the sequence of thought—by which we go back from a consequent to an antecedent—whereas dpa more commonly indicates the sequence of the facts them- selves. Compare the double use of 8, dt, 5 Te (1) to express a cause, (2) to express & consequent used as an argument (cp. Tolov yap kal matpés, d xal memvupéva Bates, and other examples in § 269). To understand the ordinary use of yap, we have only to suppose that when a speaker was going back upon an antecedent fact, he generally used the combination yé dpa (7 dp, yap), rather than the simple dpa. The principle of this usage is that a causal relation may be indicated by a distinction of emphasis, such as yé would express (as indeed yé alone sometimes has a distinctly causal force). As subordinate’ or exceptional uses, we have to note the fol- lowing :— 1. The use of yép to introduce a mere explanation, which became very common in Attic (e.g. Thue. 1. 8 papripioy dé AjAov yep krA.) and may be traced back to Homer. Thus— 254 HOMERIC GRAMMAR. [348. Tl. 8.147 GAA& 768" aivov dxos Kpadinu kal Ovpov ixdver "Extap yap wore pices KTA. This idiom—by which the Clause with yép becomes a kind of Object-Clause, in apposition to a Pronoun—may be compared with the use of Sr and otvexa with the meaning ¢hat, instead of because: see §§ 268, 269. In both cases the lan- guage does not clearly distinguish between the ground of a fact (being properly a separate and prior fact), and a mere analysis, or statement of circumstances in which a fact consists. 2. The inversion (as it may be regarded) by which the Clause with ydép precedes the fact explained; as— Il. 2. 802 “Exrop, col 8& pddior’ emiréd\Aopar GE ye pear" mo\dol yap Kata Gotu péya Tpiduov ésixoupos, dAAn 8 &d\wv yAdooa TodvoTepéwy avOpdmeov' totow exactos avyp onuawere xrA, (II. 7. 328-331., 17. 221-97, Od. 1. 337). Here the speaker begins by stating something that leads up to his main point. This use of ydp is also found in combination with adversative Conjunctions, as— Il. 12. 326 viv & éumns yap Kijpes épeotaoww Oavdrouo . . touev xrh. (cp. 7. 735 17. 338., 24. 223). Od. 14. 355 GAN ob ydp ow epalveto Képdioy civar patecOa: mporépw’ Tot wey madwy adris {Bavov. é&\\a—ydép also occurs without a subsequent Clause : Od. 10. 201 kAatov 8& Atyéws, Oadrepdv xara ddaxpu x€ovTes* GAN’ od ydp tis mpHis eylyveto pupopevorct (so Il. 7. 237-242). Here it has the force of ‘but be that as it may,’ ‘but the truth is’ (Riddell, Dig. § 147). That is, d\Aa—ydp meets what has preceded not by a simple opposition, but by going back to a reason for the opposite: which may be enough to convey the speaker’s meaning. In these uses of ydp the peculiarity is more logical than grammatical. The yap (or rather the dpa contained in it) indicates that the Clause gives a reason or explanation, but the consequent or thing to be explained has not been already mentioned. With 8’—ydp and dAAd—yép it is incorrect (as Riddell shows, Zc.) to treat the Clause with yap as a parenthesis (writing e.g. viv 8—éumns ydp xra.). The Clause so introduced is always in opposition to the preceding context, so that the Sé or GAAG has its full force. This opposition naturally prevents yép in such cases from giving an explanation of the preceding context (unless the explanation is ironical): hence yap after dAA& or S€ generally refers to something that follows, or that is not expressed at all. 3. After the Relative és, 4, 8: as— Dl. 12. 344 dyporépw pev paddrov' d ydp x dy? dpiocrov amdvrev ely (so Il. 23. 9, Od. 24. 190). 349.] TAP, ‘OTN. 255 Od. 1. 286 (MevéAaos) ds yap dedraros Aden (ep. 17. 172). So with és yép=/or thus, and tva ydp (Il. 10. 127). These are generally regarded as instances of the original use of 8s as a Demon- strative (§ 265). But it is at least as probable that they exhibit the original use of yap (=e dpa). It will be seen that ds yap may always be replaced by és dpa or és fa without changing the sense. 4. In abrupt questions, and expressions of surprise: as— Il. 1.123 ws ydp tor ddcover yépas peydduuor ’Axatol ; why, how are the Greeks to giwe you a prize ? 18. 182 "Ipe ded, tis ydp oe OeGy euol ayyedov Ke ; 1. 293 9 ydp Kev detdds Te Kal odridavds Kadeoluny KTA. why, I should be a coward Sc. So in the formulae of wish, et ydp, at ydp, &c. In all such cases the yép seems to be mainly interjectional. Properly it implies that the speaker is taking up the thread of a previous speech, and as it were continuing the construction: the new Clause being one that gives a reason, or affects to do so ironically. Particles so used easily acquire an irrational character. We may compare the use of 5€ and 7’ dpa in questions, 4s in expressions of wish, GAA before an Imperative (§ 336): also the English use of why, well, and similar pleonasms. ody, 8H, vd, On. 349.| ody is not an Illative Particle in Homer; it does not express inference, or even consequence (like dpa). Its use is to affirm something with reference to other facts, already men- tioned or known; hence it may generally be represented by a phrase such as after all, be this as it may, &e. HE. g.— Il. 2. 350 pnpl yap oby for I do dectare that Se. Od. 11. 350 eivos 38 TART, pddra TEP véoTOLO XariCwy, umns ody émipeivas és adprov (nevertheless to wait). Like dpa, it is used to emphasise correlative Clauses, but only with the negative odre—oire and piyre—prjre: as— Od. 6. 192 ot ody éoOfros Sevjoeat ode rev dAXov. IL. 16.97 at yap .. pire tus obv Tpéwy ., pyre Tis "Apyetav, KA. (so Il. 8. 7., 17. 20., 20. 7, Od. 1. 414., 2. 200., II. 200., 16, 302., 17. 401), The combination y’ ody (not to be written yodv in Homer) occurs only twice, with the meaning in any case :— IL. 5.258 ef y’ oty Exepds ye piynoe if one of the two does (after all) escape. 16, 30 pi) Que y’ obv obrds ye AdBou xddos (cp. 19. 94 xara 8 oby Erepdv ye wédncer). As an emphatic Particle of ¢ransition ofv is found in pev of (LI. g. 550, and several times in the Odyssey), much more frequently 2.56 HOMERIC GRAMMAR. [3.50 in the combinations émet ofv, és ofv. In these an approach to the illative force may perhaps be observed. 350.] 84 is properly a demporal Particle, meaning now, at length that jam): hence it implies arriving at a result, as ef ob 3) Ta mpGra diacrirny from the time that the point was reached when they quarrelled: «i di if it has come to this that, and so if finally, of really. With Superlatives it expresses that the highest stage has been reached, as Il. 1. 266 kdpruorou 51) Keivor xr. these were quite (finally) the mightiest. So in questions, 74s 64 how has it come to be that— ; and prohibitions, yi 87 do not go so far as to—. 84 may begin a sentence in Homer, as Il. 15. 437 Tedxpe méTov, 3} vdiv améxtato motos éraipos: and often in the combinations 8) réte (dum vero), and 8 yép. The original meaning is best seen in these forms (where 67 is emphatic), and in 43y (for 7 87), and éme.dy. As 8% is one of the words which unite with a following vowel, so as to form one syllable, it is sometimes written 8’, and so is liable to be confused with 8é. This occurs especially in the com- binations 8) ad, 8} adtés, 8) oftws: as Il. 1. 131 pt) 87) otras, I. 340 el more di) adre, 10. 385 7H bi obras, 20. 220 bs 89 apvedtaros xtA. So in et § aye the sense generally requires 84: see § 321. Note that 8#ra, 89@ev (cognate or derivative forms) are post- Homeric ; as also are the combinations 8y7ov, kat 8. 351.] vs is obviously a shortened form of vév now. It is used as an affirmative Particle (like 84, but somewhat less emphatic), especially in combinations such as 9 fd vu, kai vd ke, of vu, pH vu, énet vy, and after Interrogatives, as tis vu who now, tt vu why now (see Od. 1. 59-62). The form w is exclusively Epic: vév (v), which is used by Attic poets (Ellendt, Lex. Soph. II. p. 183) appears in Il. 10. 105 dca mov vey eeAverar, and Il. 23. 485 debpd vuv, 7) tptmodos KTA.: but it is probably not Homeric. In IL. 10. 105 the sense is distinctly temporal, and accordingly we should pro- bably read viv ¢Awerat. The temporal sense also suits Il. 23. 485, where moreover there is a variant 5edpé ye viv rpimodos, found in the Scholia on Aristophanes (Ach. 771, Eq. 788). 352.] Ov is an affirmative enclitic, giving a mocking or ironical force, like the later 84mou and 80ev (which is perhaps originally 8 Onv): as Il. 2. 276 ob Onv wiv médAw adris dujcer Ovpos ayjvwp his bold spirit will not I imagine impel him again: Tl. 13. 620 Aciperé Onv otrw ye methinks in this fashion you will leave 8c. 353°] MEP. 257 Tép, 353-| The Particle wép is evidently a shorter form of the Pre- position zépi, which in its adverbial use has the meaning Jeyond, exceedingly (§ 185). Accordingly wép is intensive, denoting that the word to which it is subjoined is true in a high degree, in its fullest sense, &c.: ¢.g.— Il. 1. 353 marep, emel pw? erexés ye pivyvOddudv mep edvra (cp. plvuvdd wep, ot Te pada, div). 23.79 dNdxe yewdsuevoy Tep was my fate quite from my birth. Od. 1. 315 py pe ere viv Karépuce AAatduevdy TEp ddoio. 8. 187 oriBapsrepov odk dAlyov TeEp. Or that the word is used with full confidence and emphasis :— Il. 2.236 ofkadé wep obv vnvol vedpcba (= let us have nothing short of return home). 8.452 opGiv dé mpiv wep Tpdmos EAdaBe haldima yvia even beforehand trembling seized your knees. 13.72 dplyvewror dé Geol ep gods, surely, are easily known. Od. 4. 34 af xé 708. Zebs | Lorlicw rep navon diGvos. So with Relatives, 8s wep the very one who, ws éverat mep (Attic dotep kal €orat) just as it will be. Also ei wep even if, and we wep or 4 wep even than. Usually, however, mép implies a sense of opposition; z.¢. it emphasises something as true in spite of a preceding assertion : as ov Tt duvjceat dxvipevds wep thou wilt not be able, however much vened, Tohees ep edvtes many as they are, mlvovTa mep Eumns even though drinking, &c.; and with Substantives, I. 20. 65 ra re atvyéovor Oeol wep which even the gods (gods though they are) dread. uess commonly wép implies compensation for the absence of something else: Il. 1. 508 GAA ov mép yuv ticov do thou honour him (since Agamemnon will not); 17. 121 al ke véxuy mep ’AXtrAgi mpopepopey yupvdv’ atdp td ye Tevxe KTA, In Homer the intensive kai and mép are often used with the same word or phrase: as kal dé mep even though late, wal mpos datuovd wep even though it were against a higher power, kat me(os mep edv though only on foot: dpedos (where ob5év may be adverbial). A trace of the Gen. Neut. appears in the Compound odSevdo-wpos worth nothing (Il. 8.178). The Masc. occurs only in the phrase 70 dv pévos ovdert etewy (Il. 22. 459, Od. 11. 515). . The form pyfdets is entirely post-Homeric. 357-] Double negation. This characteristic feature of Greek is caused by the tendency to repeat the negative Particle with any word or phrase to which the negation especially applies: as Tl. 1. 114 érel ob €0& eats xepelwv, od déuas KA. since she is not $2 2,60 HOMERIC GRAMMAR. [358. inferior—not in form &e. The emphatic ovSé and pn8é are chiefly used in this way: as od pay odd’ ’Axideds Krd. no, not even Achilles &e.; Od. 8.176 od8€ kev dAAws odde Oeds Tedfere nor could even a god make one otherwise: Od. 8. 280 Td y' ot Ké Tis 0d8E Torro, ovdE Gedy paxdpov: Il. 2. 703 088 pev odd’ ot dvapxor oar: Il. 6. 58 pnd Sv twa yaorépe pajtnp Kodpoy edvta pépot pnd’ ds iyo. 358.] pi is commonly used (as we should expect) with the Moods expressive of command or wish, viz. the Imperative Sub- junctive and Optative. These uses having been already dis- cussed (&§ 278, 281, 299, 303, &e.), it only remains to notice some idiomatic uses in which py is found with the Mood of simple assertion or denial. With the Indicative p is used in Homer— (2) In the phrase pi) Speddov (or Spedrov) would that I had not §c. Logically the pj in this idiom belongs to the following Infinitive (cp. § 355)- (2) In oaths, to express solemn or impassioned denial : Il. 10. 329 torw viv Zeds airds, éplySoumos méovs “Hpns, 4) wey Tots Imnovow avyp émoxjoerat GAAos KTA, (I swear that no one else shall ride Sc.) 15. 36 lorw viv rdde yaia... pay dC euny lornta Tocewddwv évoolydwv aypatver KTA. In this use py denies by disclaiming (as it were) or protesting against a fact supposed to be within the speaker’s power (=/far be it from me that §.). (c) After 4, to express incredulity, &e.: Od. 6. 200 7} py mod Tiva dvopevéwr pacd Euperar avdpav (surely you do not suppose it is any enemy !) 9. 405 7} py tis oev pda Bpotdy déxovtos ehadver ; H pn tis o adrov Kreiver d0A@ HE Binds; (surely no one is driving off your sheep ? &c.) This is the common type of ‘question expecting a negative answer,’ viz. an impassioned denial uttered in a hesitating or interrogative tone. Compare the quasi-interrogative use of qf (§ 338) to indicate surprise or indignation. (d) After Verbs of fearing which relate to a past event : Od. 5. 300 defS pr) 5% mdvra Ged vnuepréa Etmev. Here, as with the Subj. (§ 281, 1), the Clause with pj passes into an Object-Clause. The difference is that the Indicative shows the event to be past; so that the Clause expresses literally a 359-] MH. 261 refusal to believe what the speaker nevertheless ‘fears’ to be true. So perhaps Od. 13. 216 pa ri wor ofxovra I fear they are gone: but the better reading is otxwvrat, the Subj. being understood, as in Il. 1.555 pn oe wapeinn lest she have persuaded thee (i.e. prove to have persuaded); cp. the Opt. in Od. 21. 395 wh Képa Tres Sore lest worms should (be found to) have eaten (§ 303,1). Cp. also Matth. xvi. 5 éweAdOovro dprous AaBeiv they found that they had forgotten (Field’s Otiwm Norvicense, Pt. 3, p. 7). The essence of these idioms is the combination of the impera- tive tone—which shows itself in the Particle—with the Mood proper to simple assertion. The disposition to resort to the form of prohibition in order to express strong or passionate denial may be seen in the occasional use of the Optative with py in depre- cating a supposition (§ 299, e), and of the Subj. with pA in oaths, as Od. 12. 300., 18. 56. The use of the Indicative with p4 may have been derived from forms of prohi- bition which originally used the Subj. or Opt., the change being due to the sense of the speaker that the fact in question was really certain: ep. the Modal uses noticed in §§ 324-326, and the remark in § 323 as to the tendency in favour of the Indicative. The use of the Past Indicative after Verbs of fearing is closely parallel to the use in Final Clauses, noticed in § 325. While the Clause, as an expression of the speaker’s mind about an event—his fear or his purpose—should have a Subj. or Opt., the sense that the happening of the event is matter of past fact causes the Indicative to be preferred. 359.] Conditional Clauses. The rule which prescribes pi as the negative Particle to be used in every Clause of Conditional (or quasi-Conditional) meaning does not hold universally for Homer. In Homeric Conditional Clauses— (2) When the Verb is a Subjunctive or Optative p4 is used: the very few exceptions being confined to od« é6¢dq (II. 3. 289., 15. 492) and ovk édw (II. 20. 139), which are treated almost as Compounds. See § 355, and cp. the use of od« é0é\w in Final Clauses, as IL. 5. 233 #). . Hae Thaerov ovd eO€AnTov KTA, (2) With the Relatives 8s, 800s, &c. and an Indicative od is generally used ; as— Il. 2. 143 mor peta TANOdy, Boor od Bovdrjs endxovoav. Od. 3. 348 dote rev 7} mapa mduTav dvelpovos H& Terixpod, @ ov tt xAaivat kth. (a general description). Tl. 2. 338 (matoly eorxdres) vntidxots, ofs of Te pert KTA, (sO 7, 236., 18. 363). The only instance of py is Il. 2. 301 éoré 38 wavres pdprupor, ods py Kijpes @Bav Oavdrowo pépovoa, where the speaker wishes to make an exception to what he has just said. 262 HOMERIC GRAMMAR. [360. (c) With et and the Indicative o¥ is used when the Clause with ei precedes the Principal Clause: as— Il. 4. 160 ef wep ydp re kat adrix’ ’Odtpmos odk érédcoce KTA., and similarly in Il. 9. 435., 15. 213, Od. 19. 85, and the (eight) other places quoted in § 317. But when the Clause with ei fol- lows the other, pj is used, as in the numerous sentences of the form— Il. 2.155 évOa xev . , vdaros ervxOn | ef iy KTA. The only instance in which the rule fails seems to be— Od, 9. 410 ef pev 8} pp tls oe udCera ofov édvta, n , . yx yo \ t 2 , vododv y ob Tas eat Atos peyddou ahéeacGat. Here py ts may be used rather than od ts in order to bring out more clearly the misunderstanding of the Odrs of Polyphemus. This curious law was pointed out by A. R. Vierke, in a valuable dissertation De py particulae cum indicativo conjunctae usu antiquiore (Lipsiae, 1876), With regard to the ground of it, we have seen (§ 320) that a Clause with ei in most cases precedes the apodosis; and that this is probably the original order. When it is inverted it may be that the use of py instead of od has a prohibitive character, as though the condition were added as an afterthought, in bar of what has been already said. In any case the inversion throws an emphasis on the Clause, which would account for the preference for py; see § 358. 360.] Infinitive and Participle. It appears from comparison wih the forms of negation in the oldest Sanscrit that the nega- tive Particles were originally used only with jinite Verbs. The negation of a Noun was expressed by forming it into a Com- pound with the prefix an- or a- (Greek dv-, d-): and the Infini- tives and Participles were treated in this respect as Nouns. The first exception to this rule in Greek was probably the use of 08 with the Participle—a use which is well established in Homer. od with the Infinitive is used in Homer (as in Attic) after Verbs of saying, thinking, knowing, &c. (§ 237); as Il. 16. 61 # rou Epyy ye ob amply pyviduov Katanavoguev xTA.: Od. 5. 342 doxéers d€ ol ovK amuvvocety. This use however is to be compared with that noticed above (§ 355), in which an ot which belongs in sense to the Infinitive is placed before the governing Verb; as od pnow buoev he says he will not give. Sometimes the Homeric language seems to hesitate between the two forms, or to use them indifferently: compare (e.g.) Il. 12. 106 ob8" é7’ épavro oxnoec# rd, and (a few lines further) v. 125 épavto yap obuér’ Axatods oxjoec6 xrrA. Occasionally the negative is used with the Verb and repeated with the Infinitive : Tl. 17. 641 éet ob pay dtopar ov5e mendaGa (cp. 12. 73). Od. 3.27 ob yap dtw | ot ce Oedv déxnte yevéoOat KTA, It may be conjectured that the use of od with the governing Verb is the more ancient ; while the use with the Infinitive is certainly the more logical. 362.] °AN, KEN, 263 361.] w} with the Infinitive and Participle. The Homeric uses of this kind are few and simple in comparison with those of later Greek. The Infinitive when used for the Imperative (§ 241) naturally takes py instead of ov: as Il. 4. 42 pr} te duatpiBew tov euov xddov, GAG p eaoat. An Infinitive which stands as Object of a Verb of saying, &c. takes pj when it expresses command or wish: as Il. 3. 434 mad- erOat KeAopat nde «tA. I bid you stop and not Se. (so 9. 12): Od. I, 37 émel mpd of eltopey Huels pojr’ KtA. we told him before not to $e. So Od. g. 530 d0s pH "Odvecfa.. ixéoOa grant that Ulysses may not come. Again, a dependent Infinitive takes pj in oaths, as Il. 19. 176 GuvveT@ . . ua} ToTE Ths ebvhs emBruevar «tr. let him swear that he never Sc. cp. V. 258 torw viv Zeds.. uy pev eyo xotpn Bpionide xeip’ éevetcat. So generally after Verbs of asseveration, as Il. 14. 45 ds mor éemnmeiAncev .. un amply xtd, threatened that he would not &e.; Il. 18. 500 6 & dvatvero undév édécOau denied that he had recewed anything. This use of pis evidently parallel to the use with the Indicative, § 358. Compare also Il. 19. 22 of” émetes épy euev AOavdrov pndé Bpordv dvdpa teA€coa, where the pi may be emphatic (such as we must not suppose any mortal to have made).* Or this may be an instance of the use of p4 in Relative Clauses containing a general description (§ 359, 0). The use of 4 with the Participle appears in one Homeric instance :— Od. 4. 684 um pvnotedcavres pnd’ GAO GudAnoavres borata kal Tipara viv évOdde Sermvjceray. Here py belongs to éj:Ajoavres, and expresses a wish: ‘ may they (after their wooing) have no other meeting, but sup now for the last time.’ For the parenthetical pvyoredoavres and the repeti- tion of the negative with dAAore, cp. the parallel place Od. 11. 613 pt Texvnodpevos pnd GANo TL TexvIoaLTO. dv, Kev, 362.] So much has been said in the last chapter about the uses of é and xév in combination with the different Moods that it is hardly necessary to add anything here. It may be con- venient howevcr to recapitulate shortly the results arrived at, 1. The primary use of dy and xév, from which the special uses may be more or less easily derived, is to show that the speaker * This would be akin to the later use with Verbs of belief. As to the Verbs which take 4 see Prof. Gildersleeve in the American Journal of Philology, vol. I. P. 49. 264 HOMERIC GRAMMAR. [362. is thinking of particular instances or occasions. The Homeric use of ré, as we have seen (§ 332), is precisely the opposite, viz. to mark a sentence as applicable generally or in an indefinite number of instances. It is to be observed accordingly that in Homer dv and xév are hardly ever used in the same Clause with ré, except in the combinations 4 7’ dv, 4 Té kev. Similarly the use of dv and kév with the Relatives which contain an Indefinite Pronoun, as Straws and drére, is comparatively rare: see §§ 285, 306. 2. In Simple Sentences, and in the apodosis of Complex Sen- tences, é& and xév express limitation by circumstances or condi- tions. This applies to the common uses with the Opt., and with Past Tenses; also to the Homeric uses with the Subj. (§ 275), and Fut. Ind. (§ 326). But the Subj. and Opt. are ‘pure’ when used in their quasi-Imperative sense, and in waconditional nega- tion (otde yévnrar and never will be), and rhetorical questions (7s T dp’ tw how am I to go’) which are equivalent to negation. 3. In Final Clauses which refer to the Future, the use of av or xév prevails ($$ 282, 285, 288, 293, 304). But with certain Con- junctions (especially ds, &tws, iva, dpa) it is liable to considerable exceptions; see $$ 285-9, 306-7. When the purpose spoken of is not an actual one (but either past or imaginary), the Verb is generally ‘ pure.’ 4. In Conditional Clauses the Subj. and Opt. generally take dv or xév when the governing Verb is a Future, or in a Mood which implies a future occasion (Imperative, Subjunctive, Opta- tive with dy or xév). On the other hand in similes, maxims, and all references to frequent or indefinite occasions, av and xév are not used. 5. In several places the pure Subj. is used after a Future, to show that the speaker avoids contemplating particular occasions: see § 283, a, § 289, I. a, § 292, a. 6. Exceptions of the opposite kind—in which & or xév is used although the reference is zudefinite—ave chiefly found (1) in Clauses which restrict or qualify a general supposition already stated, and (2) where a distinction or contrast is implied. As examples of the former, compare— Tl. 3. 26 para ydp re xarec Ofer ef wep dy adrov gevwvtat xTr. (even in the case when Sc.). 21.24 pdda ydp re karecOier Sv Ke AdByor. Od. 21. 293 olvds oe Tpdet pedindis, Os Te Kal GAdovs BaAdaret, ds dv piv yavddy Edy KTA. Il. 3. 65 ob ror andBanr’ orl Oedv eprxvdéa Sdpa, doa Kev avrol ddow «rd. (i.¢. the real gifts of the gods). 363.] *AN, KEN. 265 Tl. 9. 524 otrw kal rdv mpdabev emevOdueda Kr€a avopar Hpowv, Ore Kev tiv” emiCadedos xddos tkot. 19. 259 7Epivies, al 0 tnd yatay avOpdrovs rivuvrat, Stis x” enlopKoy dudoon. Od. 19. 566 of 6° éruya Kpatvovor, Bpotéy bre Kév tis WSnTa. There is an evident tendency in the language to extend the use of dv or kev beyond its original limits,—that is, to state indefinite cases as if they were de- Jinite. The change is analogous ta that which is seen in the use of the Indicative in a general Conditional Protasis: in which use, as Goodwin says, ‘the speaker states the supposition as if it were particular and not general’ (Moods and Tenses, § 51, N. 3). The use of the Indicative in similes is of the same kind (cp. dv and xév in the two first examples above): and it is especially to be observed that, as Delbriick has pointed out (Synt. Forsch. I. p. 135), the most important point of comparison is usually expressed by the Subjunctive, while details and subordinate incidents are given in the Indicative: as— Il. 4. 482 6 8 év xovinat xapal récev, aiyerpos ds, h pd 7’ év eiapera Arcos peydroro mepueyn Aein, drdp Té of Bou em” axpotary wepiact. The loss of the Homeric use of ré is another example of the tendency in the language to abandon the fine but cumbrous distinction between general and par- ticular statement. The use of xév to mark contrast may be seen in— Il. 11. 408 ofda yap 8rre xaxol pev amolxovrat ToA€po10, ds O¢ « dpiotednat KTA, So we find és 8€ xev in Il. 19. 167, Od. 14. 126, and even és pév te —s 8¢ kev, Il. 9. 508 ff., 23. 319 ff.; whereas és 8é te does not occur, 7. The anomalous use of év and xév may sometimes be due to the influence of familiar combinations: e.g. 61 av (cp. especially the use in similes, § 289, 2, a), edr’ dv (§ 290), émei kev ($ 296). 8. The use of dv with the Infinitive occurs once in Homer, in a piece of oratio obliqua (Il. 9. 684, ep. v. 417). The use with the Participle is wholly post-Homeric. 363.] Difference of dv and xév. Up to this point it has been assumed that the two Particles are practically equivalent. It will be proper, however, before leaving the subject to bring together the differences of usage which have been pointed out between them, and to consider whether any difference of meaning can be detected or inferred. 1. xv ig much commoner than é& in Homer; the respective numbers in the Iliad being 621 and 155, or 4: 1.* * Statistics as to kév are given by Hinrichs, De Homericae elocutionis vestigiis Aeolicis (Jena, 1875), as to dv, by Leo Meyer (see the next note). 266 HOMERIC GRAMMAR. [363. 2. In Negative Clauses there is a marked preference for dv. In the Iliad & is found with a negative 53 times (being one- third of the whole number of instances), xév is similarly used 33 times (about one-twentieth). The difference is especially to be noticed in the Homeric use of the Subj. as a kind of Future ($§ 275, 276). In Affirmative instances of this use xév is fre- quent, dv very rare: in Negative instances dv only is found. 3. &y is very rarely used with the Relative. Thus in the large class of Conditional Clauses with 8s (érr, &e.) and the Subj. there are more than 100 instances with xév, against four or five with é, Again, in Final Relative Clauses xév is common, both with the Subj. (§ 282) and the Opt. (§ 304); dy is never found with the Subj., hardly ever with the Opt. (the only clear instance being I]. 7. 231 roto: ot av odOev dvtidoaer). On the other hand ay is frequently used with certain Temporal and Final Conjunctions, viz. éte, éméte, edre, 8ppa, ds, also with ei and énei (in the forms iy, émjy). : 4. xév is often used in two or more successive Clauses of a Sentence: e.g. in both protasis and apodosis, as— Tl. 1. 324 €f b€ Ke put) SoNowW, eyo 5€ Kev adds EXwpat KTA. 5.273 el rovrw xe AdBowuer, dpoiueba Ke kréos ec OAdv. In Disjunctive Sentences, as— I]. 18. 308 orjooua, 7 Ke pépnor péya patos 7 Ke hepolunu. Od. 4.692 GdAov x’ éxOaipnor Bpordv, dddov Ke frdoén. And in parallel and Correlative Clauses of all kinds :— Il. 3. 41 kat xe rd Bovdoluny xal Kev word Képdiov etn. 23.855 Os wév Ke Badn . . ds b€ Ke pnplyOo10 Tdyn, KTD. Od. 11. 110 ras ef pév «’ dowdas edas vdotou Te uédqaL, kat kev ér’ eis 1Odxny Kaxd wep Tacyovtes txowoOe* ei 8€ Ke oivnat KTr. av, on the other hand, is especially used in the second of two parallel or connected Clauses, as— Tl. 19. 228 GAAG xpi Toy pev Karabdarew bs Ke Odvyor.. doco 8’ by TOCuoLo TEpl OTVyEpoto AitwyTaL KTA. Od. 19. 329 ds wen amnris adros én Kal amnvéa cldq.. ds © ay dutpwv aitos én rr. So Il. 21. 553 «f pev xev.. ei 8 dy xrd., and in a negative apo- dosis, N. 2. 488 mAnObv 8 ove dv eye xrd. (§ 276, 8), Il. 14. 247 Znvos 8 od« dv xrh.: also Il. 3. 288 ff. ef pv kev—el 8 xe—el 8 éy (the last being an alternative to the second), The only instance of av in two parallel Clauses is— Od. 11.17 086 émér’ dy areiynot apis obpavdy dorepdevra ov0? br dv dw én yaiay Kerr, 363.] ’AN, KEN. 267 and there we may read éadére oreiynot, obtaining the regular Homeric use of the Subj. in general statements (§ 289, 2, a). Instances of év in the first of two such Clauses are very rare. In three places we have the combination épp’ av yey «ey (Il, 11.187, Od. 5. 361., 6. 259), cp. also Od. 15. 513 dAAws pev o° dv éywye xTA., where dAAws means under other circum- stances (not now), The use of jv and éwfy is worth notice here. jv is nearly always used either in Final Clauses, or when the condition is put after the Principal Clause: once or twice in Adversative Clauses (iv 8&—, drap v—), and in the form jv wep ydp, cp. et 8’ av and et mep av. So émjy is used either (1) after the Principal Clause or (2) in the form aérdp émjy. 5. There are several indications of the use of av as a more emphatic Particle than xév. Thus the combination 4 7’ éy surely in that case occurs 7 times in the Iliad, 4 té kev only twice. Compare the force of kai av in— Il. 5. 362 ds viv ye cal dv Act rarpt paxorro Od. 6. 300 feta 8 dpiyver’ éori, kal dv mais Hyjoatro. So Il. 14. 244 dddov pév xev .. pela xarevijoaut, kat dv totapoto peeOpa Oxeavot I would put any other to sleep, even Oceanus Se. Cp. also 161’ dv (then indeed, then at length), i— Il. 18. 397 rér dv madov dAyea Oupa. 22. 108 éyol 6€ tdr” dv TOAD Képdiov ely KTA. 24.213 Tor’ dp Tira Epya yévorro. Od. 9. 211 tér’ dv ot Tot dmocyxéobar iron ijev. The general effect of these differences of usage between the two Particles seems to be that é is used either in an adversative sense—with a second or opposed alternative—or when greater emphasis has to be expressed. Thus while the force of xév might be given approximately by such words as then, im that case (and that of xév—xév by in one case—in another case), év might be translated then indeed, then rather, even in that case, in that other case, &e. This account of the matter is in harmony with the pre- dominance of éy in negative sentences. When we speak of an event as xot happening in certain circumstances, we almost neces- sarily think of the opposite circumstances, those in which it will happen ; as od« dy tow xpalopun KiBapis the lyre will not avail you (viz. in battle—whatever it may do elsewhere). The accent of the Particles must not be overlooked as a con- firmation of the view now taken. Evidently dv is more likely to convey emphasis than the enclitic xév. We may find an analogy in the orthotone and adversative 8¢, which stands to ré and the correlated té—ré somewhat as we have supposed dy to stand to xév and Kéy—kev, 268 HOMERIC GRAMMAR. [364. 364.] Original meaning of dv. The original identity of the Greek dv with the Latin and Gothic an has been recently maintained with much force and ingenuity by Prof. Leo Meyer of Dorpat. The following are some of the chief points of his dissertation.* The Latin an is used by the older poets in the second member of a disjunctive question, either direct, as, egone an ille injurie facimus? or indirect, as, wérum seapulae plus an collus calli habeat nescio (both from Naevius). The use in single questions is a derivative one, and properly implies that the question is put as an alternative: as— Plaut. Asin. 5,1, 10 credam istuc, si te esse hilarum videro. AR. An tu me tristem putas ? do you then think me (the opposite, viz.) sad? Amph. 3, 3, 8 derides qui scis haec dudum me dixisse per jocum. SO. an illut joculo dixisti? equidem serio ac vero ratus. In these places + we see how ax comes to mean then on the contrary, then an the other case, &c. So in Naevius, eho an vicimus? what then, have we conquered ? In Gothic, again, an is used in questions of an adversative character: as in Luke x. 29 an hvas ist mis néhvundja (‘he willing to justify himself, said): and who is my neighbour?’ John xviii. 37 an nuh thiudans is thu ‘art thou a king then ?’ These instances exhibit a close similarity between the Latin and the Gothic an, and suggest the possibility of a Disjunctive Particle (or, or else) coming to ex- press recourse to a second alternative (if not, then—), and so acquiring the uses of the Greek dv. This supposition, as Leo Meyer goes on to show, is confirmed by the Gothic aiththau and thau, which are employed (1) as Disjunctive Particles, or, or else, and (2) to render the Greek dv, chiefly in the use with the Past Indicative. Thus we have :— Matth. v. 36 ni magt ain tagl hveit aiththau svart gataujan thou canst not make one hair white or black. Matth. ix. 17 aiththau distaurnand balgeis (neither do men put new wine into old bottles) else the bottles break. Jobn xiv. 2 niba véseina, aiththau qvéthjau if it were not so, I would have told you [==it is not so, else I would have told you]. John xiv. 7 ith kunthédeith mik, aiththau kunthédeith &e. if ye had known me, ye should have known ke. The Gothic thau, again, is used to translate 4 in double questions, as Matth. xxvii. 47 whom will ye that I release unto you, Barabbas or (thau) Jesus? and after a Comparative (=than): frequently also in a Conditional Apodosis, esp. to translate dv with Past Tenses, as— Luke vii. 39 sa ith vési praufétus ufkunthedi thau this man, if he were a prophet, would have known. Sometimes also with the Present (where there is no dv in the Greek),—the meaning being that of a solemn or emphatic Future: Mark xi. 26 ith jabai jus ni aflétith, ni thau .. aflétith if ye do not forgive neither will . . forgive (ov8i . . dace). Matth. v. 20 ni thau qvimith (except your righteousness shall exceed &e.) ye shall in no case enter &c. (ob pi eicédAOnrTe). * "AN im Griechischen, Lateinischen und Gothischen, Berlin 1880, + Taken from Draeger’s Historische Syntax, I. p. 321, where many other examples will be found. . 365.] "AN, KEN. 269 This use evidently answers to the Homeric dv or ké with the Subj. and Fut. Ind. : ni thau qvimith =o dy éA@yre, ni thau aflétith = ovd€ nev dgioe. If now we suppose that dv, like aiththaw and thau, had originally two main uses, (1) in the second member of a Disjunctive sentence (=else, or else), and (2) in the Conditional Apodosis (=in that case rather), we can explain the Gothic and Latin an from the former, the Greek dv from the latter. The idiomatic ‘ellipse’ in # ydp dy . . Sorara AwBjoao else you would outrage for the last time will represent an intermediate or transitional use. We can then understand why dv should often accompany Negatives, and why it should be used in the latter Clause of a sentence. The main difference of the two uses evidently is that in the first the Clauses are co-ordinate, in the second the Clause with dv is the apodosis or principal Clause. Thus the two uses may be said to be related to each other as the two uses of 5€ (1) as an adversative Conjunction, (2) in apodosi. The use of &v in Final Clauses may be illustrated by that of thaw in Mark vi. 56 bédun ina ei thau . . attaitokeina mapexddouy adrdv iva xdy . . dwvra that they might touch if it were but de. With tva, &s, &c. dv may have had originally the same kind of emphasis as kdv in this passage: ‘that in any case,’ ‘ that if no more at least &.” The use in a Conditional Protasis following the Principal Clause may be compared with Luke ix. 13 niba thau . . bugjaima (we have no more) except we should buy (=unless indeed we should buy). 365.] The Particle kév is generally identified with the Sanscrit kam; the exact meaning of which, however, appears to be unknown (Delbr. IV. p. 144). It is found in Holic (in the same form as in Homer), and in Doric (in the form «4). Some writers even hold that it is properly Molic, and passed into the Homeric poems, with other archaic forms, from an early Aolic literature.* Without entering upon the question of the existence of Holic forms in Homer, it will be enough to point out here that the use of a Particle such as kév stands upon very different ground from the other peculiarities which would now be counted as Holic. The ancient grammarians ascribed nearly everything un-Attic in Homer to admixture of dialect; but this was due to their inability to conceive of language as subject to gradual change. It may be going too far to deny altogether the existence of a foreign or non-Jonic element in Homer, but in all probability it is to be found, if at all, in isolated words and phrases. That a Particle which is essential to the syntax in hundreds of passages should be a mere piece of literary tradition or imitation seems impossible. It has been maintained indeed that dv and «év are exactly synonymous, and therefore cannot have been used together in the same spoken dialect. But the abundance of Particles in Greek is such that we cannot be surprised at the difficulty of making a distinction of meaning in this case. Compare the analogous Gothic aiththau and thaw; also the redundance of negatives in some languages, as Latin non and haud, French pas and point. * So G. Hinrichs, De Homericae elocutionis vestigiis Aeolicis, pp. 148, 166. See also Bergk’s Griech. Literaturgeschichte, p. 130, n. 214. 270 HOMERIC GRAMMAR. [366. CHAPTER XIV. Merre anp Quantity. The Hexameter. 366.] The verse in which the Homeric poems are composed— the heroic hexameter—consists of six feet, of equal length, each of which again is divided into two equal parts, viz. an accented part or arsis (on which the rhythmical beat or zctus falls), and an unaccented part or ¢éhesis. In each foot the arsis consists of one long syllable, the thesis of one long or two short syllables ; except the last thesis, which consists of one syllable, either long or short. The fifth thesis nearly always consists of two short syllables, thus producing the characteristic — Vv - * which marks the end of each hexameter. The last foot is probably to be regarded as a little shorter than the others, the time being filled up by the pause at the end of the verse. The effect of this shortening is heightened by the dactyl in the fifth place, since the two short syllables take the full time of half a foot. 367.| Diaeresis and Caesura. Besides the recognised stops or pauses which mark the separation of sentences and clauses there is in general a slight pause or break of the voice between successive words in the same clause, sufficient to affect the rhythm of the verse. Hence the rules regarding Diaeresis and Caesura. By Diaeresis is meant the coincidence of the division between words with the division into feet. The commonest place of diaeresis in the hexameter is after the fourth foot: as— « a ’ A. € , Li 4 jodwv adtovs b& Edrdipra | redxe Kiverow. This is called the Bucolie Diaeresis. Caesura, (roy:)) occurs when the pause between two words falls within a foot, so as to ‘cut’ it into two parts. The caesura which separates the arsis from the thesis (so as to divide the foot equally) is called the strong or masculine caesura: that which falls between the two short syllables of the thesis is called the weak or feminine or trochaic caesura. The chief points to be observed regarding caesura in the Homeric hexameter are as follows :— 367.] CAESURA. 271 1, There is nearly always a caesura in the third foot. Of the two caesuras the more frequent in this place is the trochaic (rouy kara. Tpirov tpoxatov), as— dvdpa jou €vvee Modoa | todvrpomov ds pada TOAAG. The strong caesura, or ‘caesura after the fifth half-foot’ (roi mevOnutuepys), is rather less common: as— Miu Gede, Oed, | Tnaniddew ’AyiAfjos. In the first book of the Iliad, which contains 611 lines, the trochaic caesura of the third foot occurs in 356, and the cor- responding strong caesura in 247.* On the other hand, there must be no diaeresis after the third foot; and in the few cases in which the third foot lies wholly in one word there is always a strong caesura in the fourth foot (ropa EpPOnurprepijs), as— &s Ke Oeots ememelOnrat | wada 7’ ExAvoy avrod "Hpn 7 75 Mocedawy | kat Maddds ’AOnvn. The division between an enclitic and the preceding word is not sufficient for the caesura in the third foot: hence in Od. 10. 58 we should read— + _% 5 ‘ 7. Pt 2 , 9 o avrap énel cirod 7 | énarodued’ de ToriTos not cirow re tacodued’ (as La Roche). The remaining exceptions to these rules are— IL. 1.179 ot¢a8 iy oy vnvoi re ofs Kai cots Erdpowt, which is an adaptation of the (probably conventional) form abv ynt 7’ éuf xat épots érdpo.ot (v. 183). We may help the rhythm by taking vyvoi re ofs closely together, so as to avoid the break in the middle of the line. IL. 15.18 7 od pépyy Ore 7” expéuw tober, &e TE modoiiv. We may read Gre re xpéuw: but possibly the peculiar rhythm is intentional, as being adapted to the sense. 2. Trochaic caesura of the fourth foot is very rare, and is only found under certain conditions, viz.— (1) when there is a strong caesura in the fifth foot; as— TOAAG 8’ dp’ evOa kal évO? vce | wdxn TredioL0, (2) when the caesura is preceded by an enclitic or short mono- syllable (such as pév, 8¢, &c.) ; as— kat kev Todr’ éO€Aoumt Ards ye diddvT0s aperOat, (3) when the line ends with a word of the form v—-——¥*; as— adrap 6 podvos env peta TevTE KaTLyYTT|CL. The commonest form of this kind of caesura (especially in the * In this calculation no lines are reckoned twice, short monosyllables being taken either with the preceding or the following word, according to the sense, 27% HOMERIC GRAMMAR. [368. Tliad) is that in which the two former alleviations are both pre- sent ; as— Oepatr’ axpirdyvbe, Aiyds mep Cov ayopyTis. The first fifteen books of the Iliad contain eleven instances of trochaic caesura in the fourth foot, of which seven are of this form. In IL. 9. 394 the MSS. give— TIqacts fv pot Emerra. yuvatna | yapeooerat adrés. But we should doubtless read, with Aristarchus,— yuvaind ye paooerat abrés, Similarly we should probably read ra 8é y’ ob dpa pédAAov évacer (Il. 5. 205, &.), instead of gueAAoy: and conversely Oadepi) 3° émiaivero xairy (Il. 17. 439), and pagal 3° édéavvro indvtwy (Od. 22. 186), instead of pucivero, A€AvvTo. In Od. 5. 272 we may perhaps treat éyé Svevra as one word in rhythm. But it is not easy to account for the rhythm in Od. 12. 47 él & ovar’ ddeiiba: Eraipow. The result of these rules evidently is that there are two chief breaks or pauses in the verse—the caesura in the third foot, and the diaeresis between the fourth and fifth—and that the /or- bidden divisions are the diaeresis and caesura which lie nearest to these pauses. Thus— Best caesura —-W -w Yu | Vr ws —- ww Worst diaeresis — GWU — Gu — Gu | -— Ww -—- ww Again— Best diaeresis — UU —Uu — Gu — vu | -w -- Worst caesura — GU — Gu —- GW -— vu | Ve 368.| Spondaic verses. The use of a spondee in the fifth place occurs most commonly in verses which end with a word of four or more syllables, as— oréupar’ &xwy ev xepoly éxnBdrov ’AmdANovos “Apet 5€ Céunv, orépvov 5& Tloceddwvt. It is also found with words of three long syllables, as— T@ 9° 7dn do ev yeveal pepdtwy avOpsTwv. And once or twice when the last word is a monosyllable: vw- pijoat Bdv (Il. 7. 238) éorixes pels (Il, 19. 117). A spondee in the fifth place ought not to end with a word. Hence we should correct the endings 76 diav &c. by reading 4a, and djpou Pijyws (Od. 14. 239) by restoring the archaic diju00, In Od. 12. 64 the words Als wérpy at the end of the line are scanned together, Words of three long syllables are very seldom found before the Bucolic diaeresis. Examples are :— Il. 13. 713 ob yap ogu radi | doptvn | wluve pidrov Kp: Od. 10. 492 Wuxi xpnoopevous | ©nBalov | Tetpectao. 370. | POSITION. 273 The rarity of verses with this rhythm may be judged from the fact that it is never found with the oblique cases of dvOpwzos (av0pdrov &c.), although these occur about 150 times, and in every other part of the verse: or with dAAjAwy &e., which occur about Ico times. Syllabic Quantity—Position. 369.] The quantity of a syllable—that is to say, the time which it takes in pronunciation—may be determined either by the length of the vowel (or vowels) which it contains, or by the character of the consonants which separate it from the next vowel sound. In ancient technical language, the vowel may be long by its own nature (picet), or by its position (Bécet). The assumptions that all long syllables are equal, and that a long syllable is equal in quantity to two short syllables, are not strictly true of the natural quantity in ordinary pronunciation. Since every consonant takes some time to pronounce, it is evident that the first syllables of the words équs, d@pis, dud), SuBpos are different in length: and so again are the first syllables of *Qros, Srpvvov. Again, the diphthongs 4, nu, &c. are longer than the single vowels n, w, &c., and also longer than the diphthongs et, eu, ot, ov. In short, the poetical ‘ quantities’ must not be sup- posed to answer exactly to the natural or inherent length of the syllables. The poetical or metrical value is founded upon the natural length, but is the result of a sort of compromise, by which minor varieties of quantity are neglected, and the syl- lables thereby adapted to the demands of a simple rhythm. It has been shown, however, that the general rule of Position rests upon a sound physiological basis. ‘The insertion of a consonant may be regarded as equivalent in respect of time to the change of w short vowel into a long one.’ (Briicke, Die physiologischen Grundlagen der neuhochdeutschen Verskunst, p. 70; quoted by Hartel). 370.| Position. The general rule is that when a short vowel is followed by two consonants the syllable is long. Regarding this rule it is to be observed that— (1) Exceptions are almost wholly confined to combinations of a Mute (esp. a ¢enuis) with a following Liquid. But even with these combinations the general rule is observed in the great majority of the instances. (2) Most of the exceptions are found with words which could not otherwise be brought into the hexameter: such as ’Agpodirn, *"Apditpiwy, Bpotav, tpdmea, mpoontda, &e. (3) The remaining exceptions are nearly all instances in which the vowel is separated by Diaeresis from the following con- T 274 HOMERIC GRAMMAR. [370. sonants: as Il. 18. 122 xal rwd Tpwiddwy, 24. 795 Kal ta ye xpuceinu. The chief exceptions are as follows * :— ap: in ’Apgirptwr, exrpdgny (Il. 23. 84), rerpdavedov (Il. 24. 324), paperpys i. 8. 323), Orpuvreds (Il. 20. 383-4) ; and in dAAdrpros (unless we scan -ios, -tov, &e.). Before rpdme(a, rpiawa, rpirn (rpijxovra, &c.), Tpameioper (rpdmovTo, mpo-T pamreé- aba, &c.), tpdyous, tpomois, rpéper (Od. 5. 422., 13. 410), Tpopod (Od. 19. 489), Tpépov (Od. 11. 527). Before a diaeresis, cai T:va Tpwiadwy (Il. 18. 122). mp: in ddAompéaaados (Il. 5. 831); before mpoonvda, mpdawnov, mpolkTns, mpbow, and other Compounds of mpé and mpds (mpoxeiyeva, mpocaitas, &c.); also before mpos ddAdAous, Tpd doreos, and one or two similar phrases (cp. Il. 13. 799., 17. 726). Before Upraptdys (I1.), mpiv (Ll. 1.97 008 & ye mpiv erA., cp. 19. 313, Od. 14. 334. 17.597); mparos (Od. 3. 320., 17. 275), mpoopdaba (Od. 23. 106). kp: in daxptous (Od. 18.173), SaxpuTAbew (Od. 19.122), evéxpupe (Od. 5. 488), xexpuppeva (Od, 23. 110). Before Kpoviwy, Kpévov, mais, paraids, Kparauis, xpdros péya (Il. 20. 121), xpaveca, xpupnddv, padaivw, kparevtdov, kpeav, Add IL. 11, 697 eidero xpivdpevos; Od. 8. 92 xara xpGra, 12. 99 5é Te Kpari. Bp: in Bpords and its derivatives, as dBpéry, dupiBporos : also before Bpaxiwv. Bp: in dudu-Spupys (Il. 2. 700), and before Spdxwv, Apvas, Spéyous. Also in Il. II, 69 7a S& dpdypyara (unless we read Sépypara, as Hartel suggests). Op: in dAAd-Opoos (Od. 1. 183, &c.), and before Opdvewv, &c, and Opacedav, Also in Ll. 5. 462 Hynrope Opnxav, op: in ’Appodiry: and Od. 15. 444 jpiv 8 ém-ppdacer’ ddrcOpov. xp: before xpéos or xpéws (Od. 8. 353): and in IL. 23. 186 foddevrs 52 xptev, Il. 24. 795 Kal Ta ye xpuceiny. vA: in oxerAtn (IL. 3. 414), which however may be scanned — -. kA: in Tdrpowde (IL, 19. 287), edly (Od. 19. 470—should perhaps be read éré- puce KAiOn), mpooéxrAuve (Od, 21, 138, 165—read perhaps mpdoxdue or éxAuve): and before KAuraipyqjorpyn, KAewval, KAvowv, erendwy, KALAvat (Od, 1. 366). Also, in Od. 12. 215 TUrrere KAnidecow, 20, 92 Ths 8 dpa xAatovons. mA: in the Compounds rexeot-mAfra (Il. 5. 31, 455), mpurd-tA00s, mpooénAace (Od. 11. §83—read perhaps mpda-mAaCe) : before MAdtara, mA€ww sailing, théov more (IL, 10. 252), mAdoy full (Od. 20. 355). Add Aiyunrias, 6 mAciora (Il. 9. 382, Od. 4.127), and IL. 4. 329 adrap 6 mAnoiov. XA: in Od. 10. 234 Kal perc xAwpédy, 14. 429 Gypl 58 yAaivav. To these have to be added the very few examples of a vowel remaining short before ox and £: viz. ok: before Sxduavbpos, cxémapvov (Od. 5. 237., 9. 391)- £: before ZdxuvOos (Il. 2. 634, Od. 1. 246, &c.), ZéAea (Il. 2. 824, &e.). A comparison of these exceptions will show that in a sense we are right in attributing them to metrical necessity. There are comparatively few instances in which the two consonants * They are enumerated by La Roche, Homerische Untersuchungen, pp. 1-41, with his usual care and completeness, 371.] “POSITION. 275 are not followed by syllables of the quantity u —, necessitating a short vowel before them. The metre, therefore, must be an important element in the case. On the other hand, the extent to which neglect of position is allowed for metrical convenience is limited, and depends on the natural quantity—i.e. the time occupied by the pronunciation—of the consonants in question. Sonant mutes (mediae) and gutturals are longer than the others: and of the liquids \ is longer than p. Thus shortening is tolerably frequent before mp and tp, less so before xp, mA, kA, ®p, xp. With other combinations of mute and liquid, as $p, Bp, 8p, and with ox and ¢, it seems to be only admitted for the sake of words which the poet was absolutely compelled to bring in: such as ’Adpodizn, Apias, Zkduavdpos, ZaxvvOos, Bpotds with its compounds, &c. No exceptions are found before yp, yA, gad, xv, ku, or any combination other than those mentioned. In short, the harshness tolerated in a violation of the rule usually bears a direct relation to its necessity. It was impossible to have an Iliad without the names Aphrodite and Scamander, but these are felt and treated as exceptions. The word avSpérns, which appears in the fixed ending Armoto’ dvipérynra Kai %Bnv, should probably be written dSporys. As the original pp of Bpords becomes either pBp (as d-uBporos, pérci-uBporos), or Bp (as vi¢ d-Bpdrn, dupt-Bpotos), so vp might become v8p (as dvdpés), or 8p. So perhaps "Evuadip dvdpeipévry should be ‘Evvadiy ddpipdvrp (uu -——-): ep. avdpe-pévos (Hdn. ap. Eustath. 183, 6). Neglect of Position is perceptibly commoner in the Odyssey than in the Dliad. Apart from cases in which the necessities of metre can be pleaded, viz. proper names and words beginning with uv -, it will be found that the proportion of examples is about 3:1. It will be seen, too, that some marked instances occur in Books 23 and 24 of the Iliad. In Hesiod and the Homeric Hymns the rule is still more lax. Thus in Hesiod a vowel remains short before xv (Op. 567, Fr. 95), and mv (Theog. 319). In the scanty fragments of the Cyclic poets we find némpwrat (Cypria), warpi (Little Iliad), “Ayxioao «dvTdv KrA, (id.), axpiBéa (Hiupersis). 371.] Lengthening before p, \, p, v, ¢,8. There are various words beginning with one of these letters (the liquids p, X, p, v, the spirant o, and the media 8), before which a short final vowel is often allowed to have the metrical value of a long syllable. Initial p appears always to have this power of lengthening a preceding vowel; but in the case of the other letters mentioned it is generally confined to certain words. Thus we have ex- amples before— A, in Alocopar, Ajjyw, efBw, Avyds, Mrapds, Autapds, Als, Aamdpy, Ad= gos, and occasionally in a few others: but not (¢..) in such frequently occurring words as Avkvos, A€xos, Acimw. p, in péyas, wéyapov, poipa, padaxds, wédos, medin, pdoreE, wdOos : but not (e.g.) maxopat, mévos, péAas, paxap, p50os, T2 276 HOMERIC GRAMMAR. (371. v, in vevph, vépos, vids, viudn, vdros, vytds, viooa: once only before a Case of vnis (Il. 13. 472): not before véxus, vos, véweots, &e. o, in cevo, dp£; once before ov (Il. 20. 434), and once before avpeds (Od. 10. 238). 8, in 80s, dewds, def-cas &c. (Stem Se-), dyv, Snpdv (§ 393). This lengthening, it is to be observed, is almost wholly con- fined to the syllables which have the metrical ictus: the excep- tions are, ToAAd Awoopevy (Il. 5. 358, so Il. 21. 368., 22. 91), muKva poyaréqy (Od. 13. 438, a TOAAG puoTaceckev (0. 24. 755). Further, it is chiefly found where the sense requires the two words to be closely joined in pronunciation: in particular— (1) In the final vowel of Prepositions followed by a Case-form: as énl pnyytn, tort Adpov, dnd Armapoiat, Kata potpav, evi peydpo, kata pd0ov, dia vedéwv, amd veupipi, kara cvpeoioww, Kata dewwods, én dnpdv, and similar combinations. (2) In fixed phrases: ds te Ads (Il. 11. 239., 17. 109., 18. 318), khatov 8& Aryéws (Od. 10. 201, &e.), dajpova te Acapdy re (II. 14. 164, &e.), Kadx te peyddn Te, €ldds Te péyeOds Te, Tpdes d€ peyd~ Ovpot, tplroda péyav, TIyAudda pedinu, ds te vipddes, iv dé ve- péecou kddve, Gre cevaito, ov TL pada Syv, and the like. These facts lead us to connect the lengthening now in ques- tion with the peculiar doubling of the initial consonant which we see in Compounds, as dmo-ppitrw, ev-ppoos, &-ppnxtos, Tpi-AAtoTos, ei-puperins, ayd-vvios, émi-coetw, ed-coeAwos, a-ddeqs: and after the Augment (§ 67), as é-ppupa, é-ppnfa, é-ppeov, €-AAlocero, é-puope, &prveov, &-coeva, @-dderca. The words and stems in which this doubling occurs are in the main the same as those which lengthen a preceding final vowel: and the explanation, whatever it be, must be one that will apply to both groups of phenomena. With most of these words the lengthening of a preceding vowel (or doubling of the consonant, as the case may be) is optional. But there is no clear instance in Homer of a short vowel remaining short before the root 5fi- or 5fet- (e.g. in the 2 Aor. Siov, the 1 Aor. é5erca, the Nouns déos, devds, SerAds, even the proper names Acohvep, &c.), or the Adverb Syv. The same may be said of fdxos, pryvuys, ftopa, pntds, pirtw, plov, also padaxds, pedln, vipds. Lengthening is also the rule, subject to few exceptions, with Alccopuar, Adpos, védos, veuph, fivds, fbos, pdBbos, si¢a, and some others (La Roche, Hom. Unters. pp. 47 ff.). 37% al Origin of the lengthening*. The most probable account of the matter is that most of the rocts or stems affected originally began with two consonants, * On this subject the chief sources of information are, La Roche, Homerische Untersuchungen (pp. 49-65); Hartel, Homerische Studien (Pt. I. pp. 1-55); and Knis, De Digammo Homerico Quaestiones (Pt. III. pp. 225 ff.). 372.] LENGTHENING BEFORE LIQUIDS. 257 one of which was lost by phonetic decay. Thus initial p may stand for Fp (as in Fevy-vupr), or op (as *opéw, Sanscrit sravdmi): Ms is probably for AFls (with a shorter Stem than the form seen in AéF-wv): vuds is for ovuds (Sanscr. suushd) : wp-ds goes back to a root svibh (Goth. snaivs, snow): potpa is probably from a a root smar: oéApo, is for oFéApa (Curt. s.v.): and Be- in deidw, &. is for &Fet- (ep. dei-doxa for 5€-5F ona). It is not indeed necessary to maintain that in all these cases (or indeed in any of them) the lost consonant was pronounced at the time when the Homeric poems were composed. We have only to suppose, in most of the cases, that the particular combination in question had established itself in the usage of the language before the two consonants were reduced by phonetic decay to one. Thus we may either suppose (e.g.) that kard féov in the time of Homer was still pronounced kard opéov, or that (while opéw, opdos in most cir- cumstances became féw, 660s) certain combinations—Kata-cpéw, év-cpoos, kata opéov, &e.—passed into kata-ppew, é-ppoos, card ppdov (or kata pédov). Simi- larly, to take an instance of poetical rather than colloquial usage, when the phrase ot 7 dda BF qv was formed the final a of pada was long by position before 5F. But it is not necessary (though allowable) to suppose 5fyv to be the Homeric form; it is enough to suppose that the phrase was handed down as a whole, from a time when the word was still pronounced 5f7v. Whether, in case of the loss of F in the initial group §/, the reciter of the poems lengthened the preceding vowel, or doubled the consonant, or left the metre in its defective condition, is a further question (La Roche, H. 7. pp. 389 ff.). If any difficulty is felt in supposing that the author (or authors) of the Homeric poems sometimes used two forms of the same word, differing only in their initial consonants, it will be worth while to point out that there are several undoubted examples of this. Moreover, in some instances a second form only appears in combinations of a fixed type. Thus we have the form wrdAts, in woz! mrdAu0s, "AXIAARa TroAlmopOov, &c.: wréAepos, in peya mrorguowo peundws, dvd mrodéporo yepupas. Similarly a primitive ySotmos survives in épi-y5oumos (also épi-Soumos), é-y5ovmjoe: and yvéos in d-yvoew. Cp. also the pairs opixpés and pixpés, oid- vapat and kl8vapat, ods and ts, Evv and ovv. It is at least conceivable that in the same way the poet of the Iliad said kard potpav and also katd cpoipdy, pe- Sidwv but dro-cpedis, div jv at the beginning of a line, but pada SFHv at the end: and so in other cases. It is true that the proportion of the words (or stems) now in question which can be proved to have originally had an initial double consonant is not very great. Of the liquids, the method is most successful with initial p, which can nearly always be traced back to vr or sr, And among the words with initial v a fair proportion can be shown to have begun originally with ov (veupy, vuds, vidas, véw, vupgy). The chief exception is vépos (vepéAn). But with p and A the etymological explanation more often fails. The two words which furnish the greatest number of instances of lengthening are péyas (with its derivatives) and péyapov: and in these words there is no independent ground for supposing a primitive opey- or the like. This difficulty is partly met by the further supposition that the habit of length- ening before initial liquids was extended by analogy, from the stems in which it was originally due to a double consonant to others in which it had no such etymo- logical ground, This supposition is certainly well founded in the case of p, before which lengthening became the rule. It will also serve to account for many isolated cases of lengthening: as iva vypeprés (Od. 3. 327), Wa py (Il. 7. 253), TO 8 dp’ dao paryp (IL. 14. 492), Stt ov (Il. 20, 434): as well as occasional doubling, 278 HOMERIC GRAMMAR. [372- as tppae (Od. 17. 226., 18. 362), &c. But, as Prof. Hartel points out, it can hardly be applied to péyas and péyapov, before which short vowels are lengthened altogether about 200 times. Had this been the result of extension of usage by analogy, it is difficult to see why it should have been confined to these words. We should have found the same phenomenon with every initial p—which is by no means the case. In such cases, then, we must fall back upon the consideration that the etymological record is necessarily incomplete. All that we can expect, from the nature of the case, is that in a certain number of instances the older forms should survive in dialects, cognate languages, &c. To argue from the instances in which evidence of this kind fails us is in effect to argue from the silence of witnesses whom we have been unable to summon. 373.] Finals of the Dat. Sing. In some cases of apparently irregular lengthening of a final vowel the ground is to be sought, not in the following word, but in the history of a par- ticular grammatical form. Under this head falls the exceptional quantity of the Dat. Sing. in Homer. It has long been observed that the final « of the Dat. Sing. is so frequently long that it may be regarded as a ‘ doubtful vowel.’ The examples are especially found in lines and phrases of a fixed or archaic type ; 7 pa, kat év devo odxel dao’ 6Bpisov &yxos. otrw mov Auk péddee breppevdi pidov eivar (thrice in the II.) 76 tplrov ab@ ari (Od. 10. 520., 11. 28). airod map unt re weve (Od. 9. 194., 10. 444). HAvOov eixooT@ ret és xtA, (6 times in the Od.). So in Afavtt 8% pddwora, "Odvacji de padtora, &c. and the fixed epithet Avi piAos. Considering also that this vowel is rarely elided (§ 376), and that the corresponding Latin ending is 2, it becomes highly probable that the Greek -: was originally long, or at least that t as well as ¢ was originally in use. That this -2 should be found in Homer chiefly in combinations which may have had a traditional sanction and value, is quite in harmony with the general character of the Epic language.* 374.| Final o of the Neut. Plur. The considerations, me- trical and etymological, which lead us to recognise -i in the Dat. Sing. also tell, though not so decisively, in favour of an original -a@ as the ending of the Neut. Plur. in all Declensions. We have— * The priority in this as in so many fine inferences from Homeric usage belongs (as Hartel notices) to H. L. Ahrens (Philologus, IV. pp. 593 ff.). If the -i is the representative of the Sanscrit -é, original ai, then the Tt and t of the Greek Dative answer to the original Dative and Locative Cases, But if the -a. of the Greek Infinitive is the Sanscrit Datival -é (§ 231), can the same account be given of the Greek -? 375.] SHORT VOWELS IN ARSIS. 279 Il. 5. 745 (=8. 389) és 8 dyea prdyea toot Bycero. 8. 556 gatver’ dpumpenéa, bre xrA. 20.255 MOAN’ ered re kal ovké. 21.352 Ta mepl Kad pecOpa. 23.240 apippadéa dé réruxrat. 24.49 bn0ca ToAUTEUCE. Od. 9. 109 damapra kal dvjpora. 10. 353 moppdpea xabdmepé’, 1%. 396 daradea Te Kal dud. 14. 343 pwyadéa, 7a Kal ards. 23.225 dpiuppadéa xarédeLas. It is worth observing, however, that in the majority of these instances the final a is preceded by the vowel ¢, from which it was originally separated by a spirant (xE-0-a, moppupe-y-a). Cp. Il. 1. 45 dupnpepéa re papérpny, 5. 576 TvAaipevea éAerqy, 5. 827 "Apna 76 ye, 14. 329 Meponad mavrwv, Od. 1. 40 é« yap ’Opéorao ricts. As two successive vowels are often found to interchange their quantity (Bac:Aja, Baordéa), so perhaps, even when the first vowel retains its metrical value, there may be a slight transference of quantity, sufficient to allow the final vowel, when reinforced by the icéus, to count as a long syllable. Cp. § 375, 3. The scanning ¢@ (in Il. 4. 321 ei rére xodpos éa viv xrh., cp. 5. 887, Od. 14. 352) may be explained by transference of quantity, from ja. 375-| Short syllables ending in a consonant are also occa- sionally lengthened in arsis, although the next word begins with a vowel: as— ore mor’ és moAEuon Ga Aag PwpnxOjvat. al’ dsedres dyovds T” Euevar KTA, xepoly bm ’Apyetwv POipevos ev marpid. yaln. The circumstances under which this metrical lengthening is generally found differ remarkably, as has been recently shown,* from those which prevail where short final vowels are lengthened before an initial consonant. In those cases, as we saw (§ 371), the rule is that the two words are closely connected, usually in a set phrase or piece of epic commonplace. In the examples now in question the words are often separated by the punctuation: and where this is not the case it will usually be found that there is a slight pause. In half of the instances the words are separated by the penthemimeral caesura, which always marks a pause in the rhythm. Further, this lengthening is only found in the syllable with the ictus. The explanation, therefore, must be sought either in the force of the zctws, or in * By Prof. Hartel, in the Homeric Studies already quoted, I. p.10, 280 HOMERIC GRAMMAR. [375 the pause (which necessarily adds something to the time of a preceding syllable), or in the combination of these two causes. In some instances, however, a different account of the matter has to be given: in particular— (x) In phrases with és following the word to which it refers: as Il. 2. 190 xaxov &s (U ——), and so Oeds ds, Kdves ds, Spyies &s, d0dvaros &s, &e. In these instances the lengthening is to be referred to the original palatal y of the Pronoun (Sanser. yas, yd, yo =ss, ff, 8). It is not likely, indeed, that the actual form yds existed in Homeric times: but the habit of treating a preceding syllable as long by Position survived in the particular group of phrases, (2) In a number of words ending with the syllables -ts, -w, -us, -uv, where there is some evidence to show that the vowel was long, or at least ‘ doubtful,’ in Homer. It has been already pointed out (§ 116, 3) that the -. of Feminine Stems was originally long, and that this fact appears clearly in the Homeric scansion of BAoovpims and jus, the -ts being long before a vowel even in thesis. So thet may have been long in Oodpis (the only evidence being the phrase Oodpy émetpévos ddxnv): and traces of the same scansion may be seen in the phrases épis dyorov pepavia, Aut pari ardédayros, although épis, pris are more common. The short te of the Gen. -vos or -.50s proves nothing, since t is naturally shortened before a vowel: -tos becoming -tyos, then -ios or -t5os. Final -vs (Gen. -vos) is long in Feminine Substantives (§ 116, 4), as ius aim (@ in thesis, Il. 6. 79., 21. 303), mAnOus (Il. 11. 305), axAvs (Il. 20. 421), iAvs (Gen. -os), Bpwrds (Od. 18. 407) and other Nouns in -rvs: also in the Mase. ix@vs, véxus, Bérpus (Borpvddv), and perhaps méAexus (Il. 17. 520). Probably, therefore, the inflexion -vs, -vos points always to original tj: and similarly we may conjecture that -ts, -tos, and -ts, -\80s point to t. (3) Where the vowel of the final syllable is preceded by another, especially by a long vowel; as olxijas dAoyév re (Il. 6. 366), "AxtdAjjos ddodp xijp (Il. 14. 139), ds Aadv iyeipa (Od. 2. 41), dudes evi olkw (Od. 11. 190), mAciov eA€Aeumro (Od. 8. 475), xpetos tradvéa (with v.1. xpelws, Od. 8. 355): and so in vias (4, Il. 2. 165., 18. 260), vnds (Od. 12. 329), Tpwes (Il. 17.730), Bods (IL. 11. 776), also “Apna, Ilepojja, and the other examples given in § 374. In such cases there is a tendency to lengthen the second vowel, as in the Attic forms BacAéa, "AxiAAéws, &c. In Homer we may suppose that the second of the two vowels borrows some of the quantity of the other, so that with the help of the ictus it can form the arsis of a foot. Actual lengthening of the second vowel may be seen in Homer in the form dn-jwpos hanging loose (cp. met-fjopos and the later wer-éwpos), and the Perfects pepadre, reOvnara, &e. (4) In the Ending -ouiy of the Dual, as Gyotiy (Il. 13. 511., 16. 560, Od. 6. 219), tnmouiv, craduotiv: also in véiv, opdiv. We may compare the doubtful « of jyiv, juiv, and the two forms of the Dat. Plur. in Latin (-d%s, -67s), Similarly there are traces of t in piv (Il. 5. 385., 6. 501., 10. 347., 11. 376, &e.). In the 377.] ELISION, CRASIS, SYNIZESIS. 281 case of -ovtv and -wiv the account given under the last head would apply. In a few places it appears as though the 3 Plur. of Secondary Tenses in -v (for -vt) were allowed to be long: as épav dmdyres (Od. 9. 413), nal Kiveov d-yara(é- pevor (Od. 17. 35, &c.), &c. This is confined (curiously enough) to the Odyssey and the Catalogue of the Ships. In the latter it occurs seven times: in the Odyssey eleven times, in the rest of the Iliad once (7. 206). Elision, Crasis, §c. 376.| A final vowel cut off before a word beginning with a vowel is said to suffer Hlision (kOAujis): as pupl’ Axatois adye €Onke. Whether an elided vowel was entirely silent, or merely slurred over in such w way that it did not form a distinct syllable, is a question which can hardly be determined. The vowels that are generally liable to elision are a, ¢, 0, u But— (1) The o of 6, 76, mpd, and the Genitives in -oto, -e10, -ao, is not elided. (2) The ¢ of ti, dvrt, wept is not elided: regarding é7 see § 269. (3) The -. of the Dat. Sing. is rarely elided, ep. § 373. Ex- ceptions are to be seen in Il. 4. 259 70 ev dai’ Gre KTA.3 5. 5 dorép’ démwpiwe xrA.; and ten or twelve other places. The ¢ of the Dat. Plur. is often elided in the First and Second Declen- sions, and in the forms in -oo of the Third Declension. On the other hand, elision is very rare in the forms in -eot, -dot, -vou, &c, The diphthong -o: of the Person-Endingss -pat, -cor, -row, -vrac, -Oa. is frequently elided: as BotvAop’ éyo, keloovr’ év mpobdpocr, mplv Aboac6’ érdpovs. But not the -a of the 1 Aor. Inf. Act. ; hence in Il. 21. 323 read rupBoxdns, not tyuBoxojo’. The diphthong -o. of the Pronouns pot and oot or to. appears to be elided in a few places: Il. 6. 165 ds ps.’ Geren piddrnte peyy- pevat ovx eedovon; 13. 481 Kal p’ ol dydvere (cp. Od. 4. 367); 17. 100 To p’ ob Tis... veuconoerar: Od. 1. 347 ob vd 7 doidol airiou: also Il. 1. 170., 9. 673., 10. 544., 23. 310, 579, Od. 1. 60., 23. 21 (Cobet, Mise. Crit. p. 345). 377.] Crasis. When a final vowel, instead of being elided, coalesces with the initial vowel of the next word, the process is termed Crasis. The use of Crasis in Homer is limited. It is seen in ofdvexa and totvexa, also in tad\Aa for 7a GAAa (II. 1. 465, &c.), kavrés for kal adrds (in Il, 6. 260., 13. 734, Od, 3. 255., 6. 282—the three last being passages where x adrés for xe aitds is inadmissible), 282 HOMERIC GRAMMAR. [377- and yipets for kal jets (Il. 2. 238). In these cases either Crasis or Elision is required by the metre. Most texts also have Spiros, obpds (Il. 8. 360), witds for 6 adrds (Il. 5. 396), Kayo, Tayud, thy: also mpod- for mpo-e- (in mpotpave, mpodxovoas, &ec.). But since the full forms 6 dpuctos, &e. are equally allowed by the metre we cannot but suspect that the spelling with Crasis may be due to later usage. The forms kdxeivos, kdxeioe, &e. (for kal «eivos, &e.) are certainly wrong, as éxeivos is not the Ho- meric form. 378.] Synizesis is the term used when the two coalescing vowels are written in full, but ‘sink together’ (cvv{avw) into one syllable in pronunciation. The Particle 84 unites with the initial vowel of a following vowel, especially with ad, airdés and ottws (§ 350); also with *Avtimdxoto (Il. 11. 138), ddvedraros (Il. 20, 220), aypnv (Od. 12. 330. Synizesis is also found with #, in the combination 7 ody (II. 5. 439, &c.), 7 eis & wey (Il. 5. 466), j etméuevar (Od. 4. 682): and with émei ov (Od. 4. 352, &c.), and ph ardor (Od. 4. 165). The remaining instances are— IL. 2.651 "Evuadiw dvdpeipévrn: see § 370, note. 17.89 doBéor@ od vidv AdGev ’Arpéos: where we may perhaps read aoBéore od8 via Ad6’ ’Arpéos. 18, 458 viet én dxupdpy, for which one or two MSS. give vi éud. Od. 1. 226 eiAamivn 78 yapos KTA. Tl. 1. 277 IlyAcl5n eed’, and Od. 17. 375 © dpiyywre are somewhat different. In Od. 7. 261., 14. 287 GAN’ dre 57 dy5odv wor we may either scan dySoov or read éyddarov, with Synizesis of 54 (as Dindorf proposes). Other examples of Synizesis are to be found in the mono- syllabic pronunciation of ea, eo, ew, both in Verbs (§ 57) and Nouns (§ 105, 3). It will be seen that in the cases now in question (apart from some doubtful forms) an E-sound (q or et) merges in a following « or o. The term Synizesis may also be applied to the monosyllabic pronunciation of the vowels in wéAcos (scanned uv — in Il. 2. 811., 21. 567), méAtas (v — in Od. 8. 560, 574), Atyunrin (Od. 4. 229), &c., exeTALy (Il. 3. 414), ‘Ioriaa (Il. 2. 537). It has been thought that in these cases the t was pronounced like our y: but this is not a necessary inference from the scansion. In Italian verse, for instance, such words as mio, mia count as monosyllables, but are not pronounced myo, mya. The corresponding Synizesis of v is generally recognised in the word ’Evvadiy (com- monly scanned va in the phrase Evvadiy dvipeipévrn) : but see § 370 ad fin. Matus, 379.| Hiatus is a term which is used by writers on metre in more than one sense. It will be convenient here to apply it to 380.] HIATUS. 283 every case in which a word ending with a vowel or diphthong is followed by a word beginning with a vowel, and the two vowel- sounds are not merged together (as by elision, crasis, &c.) so as to form one syllable for the metre. It would be more scientific, perhaps, to understand the word Hiatus as implying that the two vowels are separated by a break or stoppage of vocal sound, so that the second begins with either the rough or the smooth ‘breathing. Thus it would be opposed to every form of diphthong (including synizesis), the charac- teristic of which is that the two vowels are slurred together, by shifting the position of the organs, without any perceptible interruption of the current of breath. This definition, however, might exclude the case of a long vowel or diphthong shortened before an initial vowel (as 7A 8 éya ov, where the final w seems to be partly merged in the following ov). Again when a final t or v comes before a vowel without suffering elision, it is probable that the corresponding ‘semi-vowel’ (more correctly ‘spirant,’=our y or w) is developed from the vowel- sound, and prevents complete hiatus, 380.| Long vowels before Hiatus. The general rule is that a long final vowel or diphthong coming before a vowel forms a short syllable in the metre. This shortening is very common in Homer: ep. Il. 1. 299 otre cot otre tw GAdAq, éwel KTA., where it occurs in three successive feet. But the natural quantity may be retained before Hiatus when the vowel is in the arsis of the foot, as ’Arpeldn ’Ayaueuvort, 8s «° elmot Ort xtA, And in a few instances a long vowel or diphthong is allowed to remain long in thesis, as I]. 1. 39 DpuvOed" ef aoré TOL KTA, The readiness with which long syllables are allowed before Hiatus varies with the several long vowels and diphthongs ; partly also it depends on the pauses of the sense. The long diphthongs (as they may be called), viz. y and o, are the most capable of resisting the shortening influence of Hiatus; next to them are ev and ov, and the long vowels y and w: while «, o. and a are at the other end of the scale. A measure of this may be gained by observing how often each of these terminations is long before a vowel, and comparing the number with the total number of times that the same termina- tion occurs. Thus it appears that out of every 100 instances of final », it is long before Hiatus about 23 times. Similarly final -y is long 19 times, -ev 6°7 times, -ov 6 times, -y 5°7 times, -w 4 times, -e 1°8 times, -o: 1°6 times, and -a only 1°3 times. Thus Hiatus after » and y is scarcely avoided, while after e, o. and a it is very rare. In a large proportion of the instances in which a long vowel retains its quantity before Hiatus it will be found that the Hiatus coincides with a division either in the sense or the rhythm, Of the examples in the arsis of the foot, more than half occur 284 HOMERIC GRAMMAR. [ 380. before the penthemimeral caesura, where there is almost always a pause: while in thesis the same thing is chiefly found to occur either after the first foot, as Il. 2. 209 7xf, ws Ore xrA., Od. II. 188 dyp, odd xrA.; or after the fourth foot (in the Bucolic diaeresis). 381.] Shortening of diphthongs before Hiatus. Regarding the nature of the process by which a diphthong before Hiatus was reduced to the time or metrical value of a short syllable two probable views have been maintained. 1. Curtius holds that whenever long syllables are shortened by the effect of Hiatus something of the nature of Elision takes place. Thus y and w lose the second half of the vowel sound, while at, et, ot lose the. In support of this he points to the facts of Crasis: thus sai éyw in becoming xdyw may be supposed to pass through the stage xa éyu. 2, According to an older view, which has been revived and defended with great ingenuity by Prof. Hartel,* the second vowel in a diphthong, being t or v, is turned into the corresponding spirant; so that ai éyw becomes xa-y-eyw, and é« TlvAov 2Adcuv becomes éx TlvA0-F-eAOuv. It is certainly in favour of this latter supposition that (as Hartel urges) it does not oblige us to suppose the frequent elision of the two vowels which in general are the least liable to be elided. The same scholar has also considerable ground for his contention that the two sounds which we have denoted by y and F—the palatal and labial spirant—may have existed in the spoken Greek of Homeric times. The explanation however is not a complete one. It does not account for the shortening of a, 9, @, which on the principle assumed by Hartel would become -ay-, -ny-, -wy-. Nor does it account very well for the shortening which is found with Synizesis, as -eat in yvwoeat emer’, and even -ep in xpuaéy dvd oxnnTpy, Bevdpew epeCSpevos. On the whole it seems most probable that the shortening in question was effected, for diphthongs as well as for simple long vowels, by a process in which ancient grammarians would have recognised rather ‘Synizesis’—viz. the slurring of vowels together without complete loss of any sound—than either Elision or Con- traction («paois). And this conclusion is supported by the general tendencies of the Ionic dialect, which was especially tolerant of hiatus, and allowed numerous combinations, such as ea, €0, ew, eot, to have the value either of one syllable or two.t 382.] Hiatus after short syllables. The vowels which are not liable to elision may generally stand before Hiatus: thus we find (warhpt dpnpdrt (§ 376, 3), mpd 6500, mpd ’"Axaidv, adrap 6 €upeuads, érdpoto évnéos, and the like. Hiatus is also tolerated occasionally in the pauses of the verse ; * Homerische Studien, III. pp. 7 ff. + The use of €o for ev in Ionic inscriptions shows, not indeed that ev and eo were identical in pronunciation, or that eo was a true diphthong, but certainly that €o was very like ev, and might be monosyllabic in scansion. Probably monosyllabic eo (when it was not a mere error for ev) stood to ev as the Synizesis €a, ew, €or, dc, to the contracted y, w, or. See Erman in Curt Stud. V. pp. 292 ff. 383.] DOUBTFUL VOWELS. 285 (1) In the trochaic caesura of the third foot ; as— Tl. 1. 569 «al p° dxéovoa xabfjoro, émyvdurpaca Krad. Od. 3.175 réuvew, dhpa tdxiora baek KA. (2) In the Bucolic diaeresis: as— Il. 8. 66 dopa yey Has iv Kat défero fepdv apap. Od. 2. 57 eiAamvdtovow mivovol te aldota otvov. The vowel of the Person-endings -ro, -vro seems to be especially capable of standing before Hiatus in these places. It appears in more than a fourth of the whole number of instances given by Knis (pp. 42-45). The original -& of the Neut. Plur. (especially after another vowel) may account for some instances, such as xdAxea dxrdxvnua (Il. 5. 723), cp. Il. 2. 87., 5. 90, 568., 11. 678., 16, 404., 20. 170., 22. 266., 23. 465, Od. 10. 458., 14. 432., 24. 273, 466. Hiatus in the Bucolic dieresis is commoner in the Odyssey than in the Iliad, in the proportion 2:1. Hiatus ufter the vowel ¢ is also comparatively rare in the Tliad: Knés reckons 22 instances (many of them doubtful), against 40 in the Odyssey. It is worth notice that in both these points books 23 and 24 of the Iliad agree with the Odyssey. Doubtful Syllables. 383.] Besides the cases in which the metrical value of a syllable may be made uncertain by its place in a particular verse—i.e. by the circumstances of Position, Hiatus, Ictus, &c. —there are many instances in which the ‘natural’ quantity of the vowel appears to be indeterminate. Two probable examples of this have been already noticed, viz. the -. of the Dat, Sing. (§ 373), and the -a of the Neut. Plur. (§ 374). Under the heading of ‘ doubtful vowels’ should be classed, not only the words in which the same letter may stand either for a long or a short vowel, as “Apys, dvyp, but also those in which the change is shown by the spelling, ¢.¢. in which a short vowel interchanges with a long vowel or diphthong: as veds and iis, évona and otvoua, &e. And with these variations, again, we may place, as at least kindred phenomena, the doubtful syllables which arise from the interchange of single and double con- sonants: ’Odvcce’ds and ’Odvceds, "AxsAdeds and ’AyiAeds. As we speak of doubtful vowels, these might similarly be called ‘doubtful consonants.’ In all such words the variation of quantity may either mean that there were two distinct forms in actual use, between which accordingly the poet had a choice, or that the ‘natural’ quantity —the quantity as it existed in the spoken language—was in fact intermediate. In the latter case the poet could often give the syllable either metrical value ; or (as in so many instances) he might treat the syllable as ordinarily short, but capable of being lengthened by the ictus, or by the pauses of the verse. 286 HOMERIC GRAMMAR. [384. 384.] Doubtful vowels appear to arise chiefly in two ways: (1) By the shortening of a long vowel or diphthong before a vowel : viz.— n, in the oblique cases of vnis (except the Dat. vni): apén for apn (§ 80); js and éds, Aniorot and Actor (Il. 9. 408). z, in tepds, xovin, Av: Comparatives in -.wv: Patronymics, as Kpoviev : tower, inus (ater, &c.), laivw, and Verbs in -1w, as tlw, otw (§ 51, 1). 0, in Verbs in -vw (§ 51, 4). w, 12 fpwos (- v v in Od. 6. 303): ep. § 55, y- at, in det for aiel, umaos (-- v vu in Od. 20. 379), and the Com- pound xapyaedvat, xaparevyddes. er, in tAéos, oméos, xpéos, xped (for mAcios, &c.), véos (Super. velatos): @kéa, Babéns (for oxeta, Babetns): Adjectives in -elos, aS xdAKetos and xdAkeos: peta and fea: Aelwy and Aéwy: Betouat and Béowar, and many Verbs in -ew (§ 51, 3). ot, in ddods and ddouds; also ofos (vv), as in Il. 13. 275 off dpetny olds éoot, ep. Il. 18. 105, Od. 7. 312., 20, 89. ev, in devouar and déopuat, éxeva and éxea, ov, in Aodw, Aor. Aogcoaro. ut, in vids (Il. 4. 473., 5. 612, &e.), Interchange of quantity is occasionally found: éws, réws for elos (or fjos) and retos (or rijos): oréwper, Teper, POémpen (for atyoper, &e.), § 80. (2) By compensatory lengthening, of— -e to ev, in Geivos (¢vFos) but Levin, xewwds and xevds, meipap and mépas, elvatos, elvexa. o to ov, podvos (but povwOels Il. 11.470); otpos (a watcher) but ép-dw: perhaps otvoya for é-yvopa. a in mapéxy (map-c€x@), and J in cvvexés (Od. 19.113., Il. 12.26). Under this head we should place double forms arising by Epen- thesis, as érapos and éraipos (for érap-yos): évi, év and «ivi, civ: Gmepetovos (for amepéctos). Other variations, of which no general account can be given, are seen in “Apns, dup (a in arsis); plAos (t in ire kaclyvyte) ; drtros and ritds; tdwp, davtixpd (d in arsis); dvo0 and btw, deipo and (once) dedpw, Ardvucos and Arsdyvaos, rods and movaAds, dpos and otpea, "OAvymos and OvAvymouo. The chief cases of a doubtful vowel being long in ¢esis as well as in arsis are, dp7, GAGvat (GAdvre with a in Il. 5. 487), tuds, mupadoxe. 385.] Double consonants, causing doubtful syllables: chiefly— oo, in the First Aorist (§ 40, 1), and Dat. Plur. (§ 102); also dc008, wéooos, veuecoaw (where co=Ty), Odvaceds. 388.] METRICAL LICENCE—VOCATIVES. 287 AA, in "AyudAeds. wm and tr, in the Pronominal Stem mo-, m-, as Oamws, 87, &e. kk, In Tedéxx@ (xk= «KF ?), ep. wéAexus, 386.] Metrical licence. In a few cases the use of a vowel as long appears to be merely due to the necessities of the metre. Such are: a in d0dvaros, axdpatos, dmovéecOat, dnodlapat, dyopdacbe. € in énlrovos (Od. 12. 423), Cepuptn (Od. 7. 119). vin TIpsaptdns, did (in ba wey domidos xr. Il. 3. 357, &e.). u in Ovyarepes (Il. 2. 492, &c.), dvvayevoro (Od. 1. 276, &c.). In these cases there is every reason to believe that the vowel was naturally short, and the lengthening must therefore be regarded as a “cence, to be compared with the neglect of Position before Exduavdpos, &e. (§ 370), or the exceptional Synizesis of Alyunrin and ‘Ioriaa (§ 378 fin.). Vocatives. 387.| The short final syllable of the Vocative appears in several places as a metrically long syllable: as— Il. 4.155 pide xaolyvyte, Odvarov «td. and so 5. 359: also Tl. 19. 400 RdvOe re wal Badse, 21, 474 vyriree, ; Od. 3. 230 Tyréuaxe. 4. 338 @ vie [leredo xrd. 18. 385 dpco Ger ravitenrAe: so Od, 24. 192 Aa€prao mai. 14. 357 Moceidaoy érduve: so Il. 24. 569., Od. 8. 408, &e. 23.493 Aiav [domeved re: cp. 17. 142 “Exrop, eldos dpiore. The reason may be found (as Hartel thinks*) in the nature of the Vocative as an interruption of the natural flow of a sentence. It is very possible, however, that the Nominative ought to be read in these places: see §164. In Il. 4. 338 & vie Ieredo the Plur, ufes may be read. The Digamma. 388.] In seeking to arrive at general conclusions as to the rules and structure of the Homeric hexameter, it was necessary to leave out of sight all the words whose metrical form is uncertain on account of the possible or probable loss of an initial consonant. It is time to return to this disturbing element of the enquiry. The scholars who first wrote on this subject had few materials for their investigations outside of the Homeric poems. To them, * Homerische Studien, I. p. 64. 288 HOMERIC GRAMMAR. [388. therefore, the ‘Digamma’ was little more than a symbol—the unknown cause of a series of metrical anomalies. In the pre- sent state of etymological knowledge the order of the enquiry has been to a great extent reversed. It is known in most cases which of the original sounds of the Indo-European languages have been lost in Greek, and where in each word the loss has taken place. Hence we now come to Homer with this know- ledge already in our possession. Instead of asking what sounds are wanting, we have only to ask whether certain sounds, of whose former existence we have no doubt, were still living at the time when the poems were composed, and how far they can be traced in their effect on the versification. 389.] Nature of the evidence from metre. The questions which are suggested by the discovery in Homer of traces of a lost ‘Digamma’ cannot be answered without some reference to the very exceptional circumstances of the text. Whatever may be the date at_which writing was first used in Greece for literary purposes, there can be no doubt that the Homeric poems were chiefly known for some centuries through the medium of oral recitation, and that it was not till the time of the Alexandrian grammarians that adequate materials were brought together for the study and correction of the text. Accordingly when these scholars began to collect and compare the manuscripts of Homer, they found themselves engaged in a problem of great complexity. The various readings, to judge from the brief notices of them preserved in the Scholia, were very numerous; and they are often of a kind which must be attributed to failure of memory, or the licence of oral recita- tion, rather than to errors of transcription. And the amount of interpolation must have been considerable, if there was any ground for the suspicions so often expressed by the ancient critics. It follows from these circumstances that an attempt to restore the lost F throughout the text of Homer cannot be expected to succeed. Such an attempt necessarily proceeds on the assump- tion that the text which we have is sound as far as it goes, or that it is so nearly right that we can recover the original by conjecture. With an imperfect text the process can only be approximate. We may be satisfied if the proportion of failure is not greater than the (probable) condition of the text would lead us to expect. The loss of the F-sound, moreover, must have been itself a cause of textual corruption. It led to irregularities of metre, especially to frequent hiatus, and there would be a constant tendency to cure these defects by some slight change. The insertion of the v épeAxvorixéy was almost a matter of course 390.] INITIAL DIGAMMA, 2,89 (see however § 391). The numerous alternative forms used in the poetical language, and the abundance of short Particles such as yé, té, 64, &c. made it easy to disguise the loss of F in many places. We cannot be surprised, therefore, if we have often to make the reverse changes. Indeed it is part of the case in favour of the restoration of the Digamma in Homer that there are so many places in which emendations of an easy and obvious kind are sufficient for the purpose. A few instances will serve to show the existence in pre-Alexandrian times of corruption such as would naturally arise from a tendency to repair defects of metre In Il. 9. 73 the common text has wodéeoot 8 dvaooes, but Aristarchus read mrodéow yap avacoes. These are evidently both derived from the original woAé- ow 8é dvaocets (i.e. Favaooers), corrected in two different ways. In ll. 13.107 the MSS. have viv 8 écaQev, the reading of Aristarchus: but Zenodotus and Aristophanes had viv 8€ kas (i.e. Féxas). In IL. 13. 609 most MSS. have kat éAero vinnv, Zenodotus read péya 8 Amero. In Il. 9. 88 the reading tifevro 5 Sépma exacros was that of Aristarchus, other ancient sources had 8éprov (the reading of most MSS.): Zenodotus had TidevTo 6& Saira Oadeav, In IL. 14. 235 weiBev, éy@ 5€ Ké Tor idéw yap para mayra, the ancient sources were divided between ciSéo xdpw and ydpw eidéw, the latter being preferred by Aristarchus. Two very similar instances are— Il. 5. 787 wax’ édéyxea, 250s d-ynrot (Ar. éXeyxées). g. 128 yuvaixas dudpova épya idvias (Ar. dptpovas). In O4. 5. 34 Hyati ’ eixoor@..ixorro the ‘common’ texts of Alexandrian times (ai xowvdrepat) omitted the x’, which is not necessary, and may have been inserted in imitation of jyari xe TpiTatw KTA, (Il. 9. 363). It should be observed that the argument from these instances is equally good, whether the readings ascribed to Zenodotus, Aristarchus, &c. are conjectures made by them, or were derived (as is more probable) from older sources. They equally serve to illustrate the process by which traces of an original F were liable to be gradually effaced. And it is not necessary to suppose any deliberate attempt to emend Homer on metrical grounds. It is enough to assume that the metre helped to determine the preference given (consciously or unconsciously) to one or other of the existing variants. 390.] Words with initial F. The former existence of the F in a given Homeric word may be inferred either from its appear- ance in sume other dialect of Greek, or (where this kind of evidence fails) from the corresponding forms in the cognate languages. Thus an original Fetxoot is supported by the forms Fixatt and Feéxare on Dorie and Boeotian inscriptions, by the Laconian Beixars (given by Hesychius), and again by Latin viginti, Sanserit vimngati, &e.: an original Féowepos by the form Feonapiwy on a Locrian inscription, as well as by Latin vesper : original Fi8etv, Fotda, &e. by Fioropes on inscriptions, yoiéa and yoiéne in Hesychius (erroneously so written, as Ahrens showed, U 290 HOMERIC GRAMMAR. [390. for Foida and Foidnu), Bide. and Prdvor (officials at Sparta, lit. ‘witnesses ’), and also by Latin video, Sanserit vedmi, veda, Eng]. wit, &e. We do not, however, propose to discuss the external evidence, as it may be called, by which the loss of an initial F is proved, but only to consider the degree and manner in which the former existence of such a letter can be shown to have affected the versification of Homer. For this purpose it will be enough to give a list of the chief words in question, and in a few cases a statement, by way of specimen, of some of the attempts made to restore the F to the text. dyvupe, The initial F is to be traced by the hiatus in Il. 5. 161 e& avxyéva af, Il. 8. 403 ward 0 appara aéw (similar phrases in 8. 417., 23. 341, 467), and Il. 16. 769 amdrayos b¢ Te dyvupevdwn (so Od. 10. 123): and less decisively by the lengthening of the final -w of the preceding word in Il. 4. 214 mdAw adyev d&€es byxot. The evidence against an initial consonant is very slight. In Od. 19. 539 maou kar’ adyévas 7£e we should read adyéva a&e (or adyev’ éagée), understanding the Singular distributively (§170). In IL. 23. 392 for tmmeov b€ of 7& may be read tnmedy ot age (as Hoff- mann proposed, Quaest. Hom. II. 93). dvag (dvacoa, dvdocety). The words of this group occur in Homer about 300 times, and in about 80 instances they are preceded by a final short vowel which would ordinarily be elided. This calculation does not include the numerous examples of hiatus after the Dat. Sing. in -, after the Genitives in ~owo, -e1o, -ao, and in the phrase ig. avdooew.* The cases in which a slight correction of the text is needed to make room for the F are as follows: Il. 1. 288 mdvreca & dvdocew (read waovw 8&é€). 9. 73 wodéeoa 5 dvdoces (read moAéow 5é, § 389). 2.672 xapdémo.d 7’ dvaxros (read xapomou Te). 7. 162 (= 23. 288) mp@ros pev dvag (read perhaps mpetiora), 15. 453 xporéovres: dvag (read kporéovre, the Dual). 16, 371 (=507) Atwov pyar’ dvaxrwy (read dppa, § 170). 523 ov mép por dvag Téd€ KapTepoy EAKos dkecoat (read pe). 23. 49 Orpuvov, dvag (read étpuve, the Pres. Imper.). 517 os pa 7’ dvaxra (read bs te or Ss pa). Od. 9. 452 9 ov 7 dvaxros (omit y’). 17. 189 yaAenal 5é 7’ dvaxtwy (omit 7’). 21.56 (=83) régov dvaxros (read réta). * For a complete analysis of the examples in the Iliad see Dawes, Miscellanea Critica, Sect. 1V. 390.] INITIAL DIGAMMA. 291 The Imperfect fvacce, which occurs five times, can always be changed into éivacce. The remaining passages are :— Il. 19. 124 ody yévos od of deikes dvaccépev “Apyeloiow (a verse which is possibly interpolated). 20. 67 évayra Moceddwvos dvaxros (in the probably spurious Peopaxia). 24. 449, 452 molnoay avanti, Od. 14. 40 dv7.éou yap dvaxtos KTA, 305 ei wey Kev vootnon dvat. 438 xddauve 82 Oupdy dvaxros, 24. 30 fs wep dvaooes (perhaps js édvacces). dpva (dpves, &c.). The F is supported by three instances of hiatus, viz. Il. 4. 158 aid te dpvdv, 4. 435 dna dpvdv, 8.131 nite dpves: and by the metrical length given to the preceding syllable in Il. 3. 103 és dippov dpvas, 16. 352 AvKot dpveoar, The passages which need correction are— Il. 3. 103 ofcere & dpv’ (the 8€ is better omitted). 11g 98 dpv’ éxédever (read i8e apv’). 22, 263 ob5e Adxoe Te Kal dpves (omit TE). Od. 4. 86 iva 7 dpves dap xepaol reAEPovar (omit 7’). 9. 226 épipous Te nal dpvas (at the end of a line). Note, however, that the evidence for F is confined to the Iliad, and that the derivative apverds shows no trace of it. dotu, The presence of an initial consonant is shown by hiatus in nearly 80 places. In two places the text is uncertain: Il. 24. 320 bmép doreos (but da doreos in the Bankes papyrus, and several MSS.), Od. 3. 260 éxas doreos (Exas “Apyeos in most MSS.). Two passages admit of the easiest correction : J. 3.140 dvipdés re mporéporo kal doreos (read mpotépov). 15. 455 Tous pev & y' ’AoTuvdpw (omit y’, or read Tovs 5 ye). Two remain, viz. Il. 11. 733 dupioravro 51 doru (aphéorav Bekk.). 18, 274 viera per civ d-yoph oOévos Efoper, dotu 58 mupyo (bere Bekk,). The changes made by Bekker in these places are not improbable, but are hardly so obvious as to exclude other hypotheses. €ap, elapuvds, Hiatus is found in IL. 8. 307 vorinoi re elapwijor, and a short final syllable is lengthened in Od. 19. 519 deldnow €apos. In the phrase &pn év eiapwh we should probably omit the év, as in dpy xeseply (Od. 5. 485). elkoo, The F appears in ava efkoo. (Od. 9. 209), and the combination v2 292 HOMERIC GRAMMAR. [390 cai. elkoot (which occurs 9 times, including the compounds with dvaKareckoct-). In Il. 11. 25 ypvcoto wai efor read xpvood: and so in the combination re xal eizoo: (in three places) omit re. In the recurring #Av@oy elxoor@ éret nerd. Bekker reads }AQov éexootd (Cobet well compares Od. 23. 102 EAOot éetkogT@ wTA.). On Od. 5. 34 juati x’ eixoor@ K.7.A0. see § 388. eka, Two instances of hiatus indicate F, in Il. 24. 100, 718, besides many places in which the word is preceded by a Dat. Sing., as ovdéve elkwv, kapted elkwv. Two places may be easily corrected: IL. 4. 509 und etxere (prj etxere), and 12. 48 Th 7’ eixovor (omit te). In Od. 12.117 for Geotow imeigea read Oeois imocigeat. There remains Il. 1. 294 €i 5) cot nay épyov imeitopa. Zouka, étokw, etkedos. The F of gouxe appears from hiatus in 46 instances (not counting the numerous places in which it follows a Dative in -.). The adverse instances are 11 in number, besides the form éz-éouxe (which occurs 11 times). The corresponding Present eikw is generally recognised in I]. 18. 520 801 ogiow eke Noxjoa where at suited them to be in ambush. The form étcxw has hiatus before it in 3 places, but rejects F in 2 (Od. g. 321., 11. 363). The adjective eixehos or ikehos usually needs an initial consonant (ex- cept Il. 19. 282, Od. 11. 207). It seems probable that this is the same word as eikw to yield. The notion of giving way easily passes into that of suiting or fitting, hence conforming to, resembling. éxav, Exntt, Exndos. Hiatus indicating F is found in 22 places (not reckoning o¢ 7 éxoy, Il. 8. 81, &e.). In Od. 4. 649 for aivés éxdy we may read aids éywy (cp. Od. 2. 133, where these forms are both found in good MSS.). In Od. 17. 478 é060 enAos two MSS. have éo@’ (i.e. éo@e). The remaining exceptions are; with écav, Il. 23. 434, 585, Od. 5. 100 (where we may perhaps read ris 8 éxdv.. diadpdyon; the Opt. without dv being used as in negative Clauses, § 299 f): with &enAos, IL. 8. 512, Od. 2. 311. éxds, Ekatos, &e, Traces of F are to be seen in the hiatus viv 8% éxds (Il. 5. 791., 13.107), GAAG éxds (Od. 15. 33), ode ExnBodlar (Il. 5. 54): and in the lengthening in ’AmdAdA@vos éxdrouo (LL. 97. 83.; 20. 295), eUTAdKapos ‘Exayndn, &e. The exceptions are, Il. 1. 21, 438., 17. 333.) 20. 422., 22.15, 302, Od. 7. 321—mostly admitting of easy correction. €xaoTos. The original F of this word (recently found on a Locrian inscription, see Curt. Stud. II. 441 ff.) is traced by means of 390.] INITIAL DIGAMMA. 293 hiatus in 115 places, The adverse instances, however, are about 50 in number, and the proportion that can be removed by emendation is not so large as in most cases. About a fourth of the exceptions appear in the recurring phrase pévos kal Ovpdv * , €KaOTOV. The form éxdtepGe shows slight traces of initial F in Od. 6.19 cradpotiv éxdrepoe, 11. 578 yire 5é€ muy Exarepbe, 22.181 rw 5’ goray ExdrepOe. It is preceded by elision in Jl. 20. 153 (omit 6’), and in Il. 24. 273, Od. 7. 91 (omit 8’). eihw (hoa, edAqv), Gddvar, &Ars. The F is shown by hiatus in I]. 1. 409 dud’ GAa éAcat: 16.403 joto adels (and five other examples of this Tense, viz. Il. 5. 823., 21. 571, 607., 22. 308, Od. 24. 538): Il. 18. 287 kexdpnode eed- pevor: Il. 12.172 ne GAGvat (so 14. $1), Il, 21. 281 eluapro GA@var (so Od. 5. 342., 24. 34), Il. 21. 495 tH ye GAdueva. Before édts hiatus occurs in about 12 places: cp. also Il. 23. 420 elvd- Tepes GAs Hoar, In II. 21. 236 nar’ abrdv dus écay some MSS, read éovav GAts, and at v. 344 the same transposition may be made. The only other instance against F is 0.17. 54 ag 80 Gus dvaBéBpuxev (dvaBéBpoxey Zenod.), where Bentley read 6 &Ats dvaBéBpoxev. EXloow, eidw, Before é\iccw hiatus is found in four places, and the recur- ring phrases kal €Atxas Bods and etAimodas EAtkas Bods point in the same direction. The only exceptions are Od. 12. 355 Booké- oxovd’ €dtxes xrA., and Il, 18. 401 yvaymras & Edixas (spirals ?). It is probable that in many places the forms éAéAuero, édeAtxOn, &e. are old errors for éf éAucro, éFeAixOn, &c.; see Cobet’s Misc. Crit. pp. 275 ff. Traces of F in etAdw should perhaps be recognised in Od. 5. 403 (épevydpevoy, elAvro) and 14. 479 odxeow eidupevor: cp. Il. 20. 492 prdya eiAvpdter, In Tl. 18. 522 iovr’ eiAvpévor it is easy to read ifov (as Bekker does). The Aor. Part. éXvoGeis has no F: but it may be from a different Verb-stem (see Buttm. Lexil. s. v. elAdw). ékrw (ZodTa). The initial F of this word is proved by 10 instances of hiatus (including kai éAmidos, Od. 16. 101., 19. 84). The Perfect codna also shows traces of F in the reduplicated syllable, viz. in Od. 2. 275+ 3+ 375+ 5+ 379: In I. 8. 526 edxopar éAndpevos should perhaps be (as Zenodotus read) éA7ropat edxépevos. In four places FéATrw can be restored by very slight corrections : IL. 15. 701 Tpwotv 8 éAnero (Tpwot 5€ Hoffm.). 18. 194 GAAA xal ards 83’, €Awop’ (adrds 2éAwop’ Hoffm.). Od. 2. 91 (=13. 380) tdvras pev p EX (omit p’). Two others are less easy; Il. 15. 539 é7e 8 Amero, and Il. 24. 491 émi 7 éAmeTat. Hoffmann reads «al éAmero, kal EXreTat. The passages which tell against FéFoAma are, Il. 20. 186 xadends 5é 0° €oATa 71d Aros pete (vead ot ohmra), 21. 583 wad’ Eodmas (uaa game’ Hoffm.), 22. 216 vai 7 Eodma (omit y’), Od. 8. 315., 24. 313- 294 HOMERIC GRAMMAR. [399. tos, €imeiv. ; The F of émos is supported by about 26 instances of hiatus, and a much larger number in which preceding syllables are length- ened (as in the common line kai piv dperBdouevos éxea KrA.). Of the apparent exceptions, about 35 are removed by reading émeoor for éréeoor (as in IL. 5. 40 yetpds édrovo’ éréeoor mpoonvia, read EModoa émeoor): and 11 more by scanning émea as a dissyllable (u -) in the formula gwvjoaca érea mrepdevTa mpoonvsa. Another small group of exceptions is formed by phrases such as Od. 4. 706 dpe 58 84 wv émecow wrd., where perhaps é may be put for wv. There remain two instances in the Iliad (5. 689., 7. 108), and seven in the Odyssey (11. 146, 561., 14. 509., 15. 375+) 16. 469., 17. 374.) 24. 161). In cineiy the F is proved by about 80 instances of hiatus, be- sides lengthening such as we have in the forms dde d€ Tis elmecke, Os dpa of éimév7t, &. The exceptions number about 35. Of these exceptions 10 are found in the recurring line d¢p’ eimw Td pe Oupds évt oTnbeca. kehever. It has been suggested as possible that elmw has here taken the place of an older émw (Fémw), or €omw (cp. éomere), This supposition would of course explain other instances of neglected F. épdw, épyov, &e. The Verb ép8w is preceded by hiatus in two clear instances, II. 14. 261, Od. 15. 360. In Il. 9. 540 adAX epdecxev there is an ancient v.1. éppetev. In I]. 10. 503 ére xtvrarov épdo. we may read kévrata. But there are several instances on the other side in the Odyssey (viz. 1. 293., 5+ 342, 360., 6. 258., 7. 202., 8. 490., 11. 80). ihe seduplicated form éopya (for FéFopya) is preceded by hiatus in 7 places. Instances on the other side are, Il. 3. 351 6 pe mpd- Tepos Kak’ Zopye (where the Aor. épefev is more Homeric, ep. § 28), 21. 399 doo pw’ Eopyas, 22. 347 ofa yw’ gopyas (in both places pe may be omitted), Od. 22. 318 ovdev eopyds (read ot m, cp. § 356). The Noun épyov, with its derivation épya¢ouat, occurs in Homer about 250 times, and the F is required to prevent hiatus in about 165 places. There are about 18 instances against F, cipw, épéw. The F of eipw is required by hiatus in the three places where it oceurs, viz. Od. 2. 162., 11. 137., 13. 73 that of épéw by about 50 instances of lengthening (such as dA’ ék ro. épéw, ds moré tis épéer, and the like), against which are to be set three instances of elision (Il. 4. 176., 23. 787, Od. 12. 156). Evvupe, etpa, eoOys. The F is shown by hiatus in more than 80 places, including the instances of the Perfect Mid. (lua, gooa, &c., see § 23, 5). The contrary instances are of no weight. The superfluous p may be omitted in émeé 6” €ocavro (three places), and 7° similarly in Od. 14. 510., 24.67. This leaves Il. 3. 57, Od. 6. 83., 7.259. 390.] INITIAL DIGAMMA. 295 €omepos, Hiatus occurs in six places, after the Prepositions rori (Od. 17. 191) and éné, There are no instances against F. €tos, The F is supported by the lengthening of the preceding syllable in five places, such as Il. 24. 765 éevxoordv ros éari. In the only adverse instance, Il. 2. 328 toccair’ érea, we may read and scan toccadra érea, as in the case of éxea (supra). idxo, taxh, AXh- The F in idxw and lax7 is chiefly indicated by 23 instances of a peculiar hiatus, viz. after a naturally short final vowel in arsis ; as 7) 6@ wéya idxovoa, Huts b& idyorres, yévero tax7, and the like. There are also 3 instances of lengthening by Position. The F is also proved by avlaxos (=4-FlFaxos) without a ery. The excep- tions, 11 in number, are confined to the Impf. taxov (%), which never admits F in Homer. Hence it is probable that the true form of the Impf. was eiaxov, standing for éFiFaxov. The derivative yx7ets follows hiatus in two places (Il. 1. 157, Od. 4. 72): elsewhere in Homer 7}x7 only occurs at the beginning of the line. iSetv, ofda, Eidos. In the different forms of the Second Aor. i8etv the F is shown by upwards of 180 instances of hiatus, and about 12 instances of lengthening of a short syllable. The Indicative (cidov in Attic) is nearly always a trisyllable (7.¢. éFidov) in Homer. On the other side we have to set nearly 50 instances of neglected F, about half of which are susceptible of easy emendation (such as putting id