Sei sulla pagina 1di 498

















THE present work is the result of an attempt to revise the ,much smaller book which was first published under the 'same title in 1860, and again, somewhat enlarged, in 1865. :When it falls to the lot of a writer to revise, under the greater sense of responsibility which doubled years and more than doubled experience have brought him, a book written in the enthusiasm of youth as an ephemeral production, fue is sure to be his own severest critic; and what he begins as a revision inevitably becomes, as he proceeds, more and more a new and independent work. I cannot forget that this book is adressed in great part to a different ,generation of scholars from that to which the former one ~as offered; and a treatment of the subject which was ~ermissible in 1860 would be far from satisfactory now. I then attempted chiefly to give "a plain and practical statement of the principles which govern the relations of the Greek Moods and Tenses," avoiding theoretical discussions as far as possible. At that time prevailing theories, based chiefly on abstract speculations, had obscured some of the most important facts in the syntax of the moods, and perhaps no better service could be rendered by a new writer than the clearing away of some of the clouds. Few younger scholars are aware how modern are many of the grammatical doctrines which are now taught in all classical schools. It is hard to believe that so elementary a principle as that by which the aorist infinitive is past in <P'l'Jtrtv e"ABe'iv and future in /3ov"Aerat ~"ABE'iv was never distinctly stated until 184 7, when it appeared simultaneously in the Greek



Syntax of Professor Madvig at Copenhagen and in the Greek Grammar of Professor Sophocles at Harvard University. Something more than mere statement of facts has been attell!-pted in the present work, although nothing has been further from my thoughts than a complete theoretical discussion of all the principles which govern the use of the moods. He who ventures far upon that sea is in great danger of being lost in the fog or stranded ; for, while Comparative Philology has thmwn much and most welcome light on the early history of the Greek language, it has also made us more painfully aware of our ignorance, although it is a more enlightened ignorance than that of our predecessors. Since the publication of the :first edition, many most important additions have been made to our resources. Of these I can undertake to notice only a few. Delbriick's elaborate treatise on the Greek Subjunctive and Optative (in his 8yntaktisehe li'orsehungen, vol. i.), with a comparison of Greek and Sanskrit usages, is familiar to all scholars. Whatever may be thought of Delbriick's main thesis, the distinction of the subjunctive as the mood of will from the optative as the mood of wish, none can fail to be impressed and instructed by his attractive and original treatment of the subject, which has made an epoch in grammatical science. Lange's unfinished work on the Particle El in Homer is a model of careful and thorough investigation. When I think of my deep and continued indebtedness to Lange's learned discussions, which include a treatment of all the 200 examples of l with the optative in Homer, I am grieved to dissent so frequently from his most important conclusions. His chief argument is discussed in Appendix I. Schanz, in his Beitrage zur Grieehischen Syntax, has undertaken a work of immense extent, involving an amount of labour which it is hard to over-estimate. His plan is to give full and accurate statistics of the use of every construction bearing on the history of Greek syntax, and thus to make a true historic syntax of the language a possibility. The work of collecting, classifying, and discussing the examples of different constructions has been assigned by him



to a large number of colleagues, and every year testifies to substantial progress. The following treatises bearing on the construction of the moods and tenses have already been published by Schanz : Weber, Entwir:kelungsgesr:hichte der Absichtssiitze; Sturm, Geschichtlir:he Entwir:kelung der Constructionen mit Ilp{v ; Schmitt, Ueber den Ursprung des Substantivsatzes mit Relativpartikeln im Griechischen; Griinewald, IJer freie formelhafte Infinitiv der Limitation im Griechischen; Birklein, Entwickelungsgeschichte des subst'antivirten Infinitivs. The amount of patient labour devoted to these compilations, in which the exact number of examples of each eonstruction in each Greek author beforeAristotle is given, while the most important passages are quoted and nearly all are cited, will be most gratefully appreciated by those who would be least willing to undertake the work themselves. The results of such dry enumerations are often interesting and surprising. No one knows whether statistics will be dry and barren or not, until they are collected and classified; and though it may seem a useless task to count the examples of each of the final particles in all Greek literature before Aristotle, it is interesting to know that in all the Attic prose, except Xenophon, wr; final occurs only five or six times, while va occurs 999 times. Some of the results derived from Weber's statistics of the use of the final particles are given in Appendix III., and an account of Xenophon's peculiar use of wr;, wr; Clv, and 57rW'> &v in Appendix IV., for the benefit of those who have not Weber's book at hand, or want the patience to follow his elaborate historical statements. Monro's Homeric Grammar is one of the best results of recent English scholarship, and for the study of Homeric usages in the moods it is invaluable. I regret that the new edition of this book, soon to be published, has not come in time to benefit the present work. It seems a mere form to acknowledge my obligations to the standard Grammars; but I must repeat my former expression of thanks to Madvig, Kriiger, and Kiihner, not to mention a host of others. To Madvig I am indebted for the first conviction that the syntax of the Greek moods belonged



to the realm of common sense. To Kriiger I have been indebted in the study of every construction; and I have still retained most of the remarks on the tenses of the indicative which were originally borrowed from him. The revised edition of Kiihner's Griechische Grammatik has supplied a large store of examples, to which I have frequently had recourse. I am under especial obligation to him for many of the examples which illustrate the uses of the Supplementary Participle, and the corresponding uses of the infinitive with many of the same verbs. Frequent references are made in the notes to the authorities which I have mentioned, and to many others. It is with pride and pleasure that I acknowledge my deepest indebtedness to an American scholar, whose writings have thrown light upon most of the dark places in Greek syntax. I need not say that I refer to my friend, Professor Gildersleeve of Baltimore. As editor of the American Journal of Philology he has discussed almost every construction of the Greek moods, and he has always left his mark. His two reviews of Weber's work on the Final Sentence in vols. iv. and vi. of his Journal may well save many scholars the trouble of reading the book itself, while they contain much new matter' which is valuable to every one. The acute observation, that the use of liv and ~ee in final constructions depends on the force of cJs-, o1rws-, and 5cf>pa as conditional relative or temporal adverbs, explains much which before seemed inexplicable. His article on 1rptv in vol. ii. stated important principles of classic usage which were confirmed by Sturm's statistics; and this, with the later review of Sturm's volume, has done much to correct current errors and to establish sounder views about 1rptv. His articles on the Articular Infinitive in the Transactions of the American Philological Association for 1878 and in the third volume of his Journal practically anticipated the results of Birklein's statistics. I can mention further only his article in vol. vii. of the Journal on the Consecutive
Sentence, which gives (it seems to me) the clearest state-

ment ever made of the relations of wuTE with the infimtive to wuTE with the finite moods. I have expressed my



indebtedness to these and other writings of Gildersleeve as occasion has required; but I have also often referred to his articles in his Journal by the simple mention of that periodicaL I have sometimes omitted a reference where one might seem proper, lest I should appear to make him responsible for what he might deem some dangerous heresy. I am also under the greatest obligation to my friends and colleagues in Harvard University, Professors Allen, Greenough, Lane, Lanman, White, and Wright, and Dr. Morgan, for valuable suggestions, and to most of them for important help in revising and proof-reading. I regret that I have not had the advantage of their aid in reading the proofs of the last two-thirds of the work. To my friendship of twenty-one years with Dr. Henry Jackson, of Trinity College, Cambridge, I am indebted for some of the most important suggestions which I have received since the publication of the former editions. The Index to the Examples includes all of the more than 4800 examples quoted or cited in the main body of the work, but not those in the Appendix nor those which a,re given in the classified lists in the footnotes on pp. 92, 115, 152, 172, and 290. It may seem useless to index many examples which merely illustrate a common principle, like those of a simple aorist infinitive or present indicative; but it would be difficult to discriminate here, and one seldom knows what may make an example useful to another. The same consideration has induced me to give as great a variety of examples as possible, from authors of different classes, illustrating many constructions which apparently need no such aid. Every teacher will see that many parts of this work, in its present enlarged form, are not adapted to the ordinary uses of a grammatical text-book for the recitation room. On the other hand, it is hoped that the increased fulness and the greater space given to discussions will make the work more useful for private study and for reference. The Dramatists are cited by Dindorf's lines; except the tragic fragments, which follow Nauck's edition, and the


comic fragments, which follow Kock. The lyric fragments follow Bergk's Poetae Lyriei. Plato is cited by the pages and letters of Stephanus, and the Orators by the numbers of the orations and the sections now in universal use. The other citations will be easily understood. In conclusion, I must express my grateful thanks to the University Libraries of Heidelberg and Leyden, and to the Royal Library at The Hague, for the hospitality which was kindly shown me while I was correcting the proofs.

24th September 1889.

In the impression of 1897 many errors have been corrected, some forms of expression have been changed, and some new examples have been added. The most imp?rtant change is that in 572 and 573; this is further explained in the new Appendix VI., page 411. A list of the new examples is given on page 440, omitting those which have been inserted in the regular Index.
RoME, November 1896.


1. The four Moods, the Infinitive, the Participle, and the Verbal in -n!os 2-5. I. Uses of the Indicative 6-11. II. Uses of the Subjunctive 12-17. Ill. Uses of the Optative 18. IV. Uses of the Imperative

1 1, 2

3, 4
4-6 6

THE TENSES. 19, 20. The seven Tenses 21. Primary and Secondary Tenses . 22. Relative and absolute time of the Tenses I. TENSES OF THE INDICATIVE.

7 7
7, 8

23. Meaning of the Present 24-32. Various uses of the Present Indicative 33. Historic Present

8 9-11 11

34. Meaning of the Imperfect 35. Relations of the Imperfect to other Tenses 86-41. Various uses of the Imperfect ,

11 11, 1.2

12, 13

42. Meaning of the Perfect 43. Meaning of the Pluperfect.





44-52. Uses of Perfect and Pluperfect, and compound forms with El}kl and fxw


63-55. Meaning of the Aorist 56, 57. Relation of Aorist to Imperfect 58-62. Various uses of the Aorist

16 16, 17

63. 64-72. 73-75. 76.

Meaning of the Future Various uses of the Future Periphrastic Future with pJ'J,),.w Past Future with tu'Xhw

18, 19 19, 20 20 20, 21

77. Meaning of the Future Perfect 78-84. Various uses and forms of tl1e Future Perfect,

21 21, 22


85. Distinction of Tenses in the Dependent Moods



86. General Principles



87, 88. Distinction of Present and Aorist here 89-93. Present and Aorist Subjunctive and Imperative 94, 95. Present and Aorist Optative , 96-101. Present and Aorist Infinitive .

22, 23 23-27 27, 28 28-31

102. 103, 104. 105-108. 109, 110.

Use of Perfect in the Dependent Moods . Perfect Subjunctive and Optative Perfect Imperative Perfect Infinitive

31, 32 32, 33 33, 34

34, 35

ll1, 112. Future seldom used except in Indirect Discourse

113. Exceptional uses of Future Infinitive elsewhere ll4. Future Perfect Infinitive used only in Indirect Discourse

35, 36 36 37


J.l5. GeneraJI>rinciples of this Construction




116. (1-4.) Four uses of Present Optative 117, 118. Present Infinitive as Prese11t. 119, 120. Present Infinitive as Imperfect 121. (1, 2.) Two uses of Perfect Optative 122. Perfect Infinitive as Perfect 123. Perfect Infinitive as Pluperfect 124 (1-3), 125. Three uses of Aorist Optative 126, 127. Aorist Infinitive 128-134. Fu tme Optative 135, 136. Future Infinitive . 137. Future Perfect Infinitive

37, 38 38 38-40 40 40, 41 41 41, 42 42, 43 43-45 45, 46 47


138. 139-141. 142. 143. 144-147.

Geueral Priuciple . Present Participle as Present and Imperfect Perfect .Participle . Ordinary use of Aorist Participle Aorist Participle (generally not past in time) with "Aa.vlhivw, rvyxavw, and cp!Javw 148-152. Other peculiar uses of Aorist Participle . 153. Future Participle . GNOMIC AND ITERATIVE TENSES.

47 47, 48 48 48, 49 49-51 51-53 53

154-158. 159, 160. 161. 162. 163, 164.

In the Indicative . Gnomic tenses in Optative, Infinitive, and Participle Imperfect not a Gnomic Tense Iterative Imperfect and Aorist with "Av Ionic Iterative Forms in -<rKov and -<rK6p:YJv DEPENDENCE OF MOODS AND TENSES.

53-55 55, 56 56 56 56

165-169. 170-173. 174, 175. 176-186. 187-191.

General Principles of Dependence . Tenses of the Indicative as Primary or Secondary . Tenses of the Subjunctive and Imperative Tenses of the Optative in various Constructions Tenses of the Infinitive and Participle

57 57, 58 58 59-62 62, 63

192, 193. Two uses and meaning of liv 194. Distinctions of liv and Ke

av. 65

64 65

195. Av not used with the Present and Perfect Indicative



196. "Av with Future Indicative in early Poets 197. "Av with Future Indicative in Attic Greek (rare) 198, 199. "Av with Secondary Tenses of the Indicative .

65, 66 66 66

66 66, 67 67

200. "Av with the Subjunctive in Conditional and Final

201. "Av with the Subjunctive in Potential sense and in

other uses (Epic) 202, 203. "Av with the Optative


67 67, 68
68. 68 68, 69 69, 70

204. General Principle of t1v with the Infinitive 205. Present Infinitive' with t1v 206. Perfect Infinitive with t1v 207. Aorist Infinitive with t1v 208. Future Infinitive with t1v 209-212. Various uses of the Infinitive with t1v

flv, 70
70, 71

213. General Principle of /lv with the Participle 214. Present Participle with t1v 215. Aorist Participle with t1v 216. Future Participle with t1v 217. Participle with IJ,v never forms a Protasis 218-222. Position of IJ,v 223-226. Repetition of IJ,v 227-229. Elliptical uses of IJ,v


71-73 73, 74





230, 231. Classification of Constructions (I.-X.)


The Potential Optative and Indicative with &v and

232. Meaning of Potential Forms



233-235. 236. 237. 238.


Origin and Relations of the Potential Optative Examples of pure Potential Optative with IJ,v
Potential Optative in Commands and Exhortations

77, 78
78, 79 79

Potential Optative expressing what may prove to be





239. Potential Optative with definite condition implied or expressed (as Apodosis) 240-242. Potential Optative without liv or Kt!

80, 81



243. Meaning and Relation of Potential Indicative 244. Pure Potential Indicative, with no unfulfilled condition implied 245. Potential Indicative with indefinite unfulfilled condition implied 246. Time of Potential Indicative . 247, 248. Potential Indicative with definite unfulfilled condition implied or expressed. 249. Iterative Indicative with liv

81, 82

82, 83


84, 85



Imperative and Subjunctive in Commands, Exhortations, and Prohibitions.-Subjunctive and Indicative with p,~ and p,~ ov in Cautious Assertions.--''01rws and o1rws p,~ with the Independent Future Indicative or Subjunctive.
250. 251. 252. 253. 254. 255-258. 259, 260, 261, 262. 263. 264. 265-268.

Imperative in Commands etc.

"A-y<, <f>ip<, etc. with Imperative

orrre' oetc. with Imperative

IIiis with second person of Imperative

86 86, 87 87

87-89 89

270. 271-282. 283.

Imperative in Assumptions First person of Subjunctive as Imperative Present Imperative or Aorist Subjunctive with p.-1} in Prohibitions Independent Subjunctive with p.-1} (Homeric) expressing object of Fea.r to be averted MT] ou with the Subjunctive Independent p.-1} and p.T) ou after Homer . M-1} and p.T) oil with Subjunctive in cautious assertions and negations (chiefly in Plato) M-I} and p.T) oil with Indicative (as above) MT] ou in dependent clauses . "071'ws and li'll'ws p.-1} generally with Future Indicative (independent) . Subjunctive rarely used with o'JI'ws p.-1}

90, 91 91 91 92, 93

93 94-96 96, 97


Subjunctive, like Future Indicative, in Independent Sentences in Homer.-Interrogative Subjunctive.

284-286. Independent Subjunctive in Homer
287-290. Interrogative Subjunctive (with or without {3ov)\EL or Qt!Xm)

97, 98 98-100



291. Negative p.~ of Interrogative Subjunctive 292. Future Indicative and Potential Optative with sense of Interrogative Subjunctive 293. M?) with Subjunctive in Affirmative Questions



lOO, 101 101


101, 102 295. Ov p,?) with Subjunctive or Future Indicative in Denials. 102, 103 296. Same construction in dependent sentences 103 297. Ov p,?) with Future Indicative (sometimes with Subjunctive) in Prohibitions , 103, 104 298. Ov p,1j followed by p,rJol, b/aAM or Ot!, or by Kal 104 299. Ou and p,f} in separate questions 104 300. Interrogative theory of Elmsley 105 301. Subjunctive in Prohibitions with ov p,f} 105 SECTION V.

Ov fL1 with Subjunctive and Future Indicative. 294. Double negative ov p,?j in Denials and Prohibitions.

Final and Object Clauses after i'va,

302. The Final Particles 303, 304. Classification . 305, 306. Negative p,f}, and p,i}



Ocppa, and /h~

10& 106, 107 107

tva, 6!!1,




307-310. 311-314. 315. 316.

History of Final Constructions Use oftva, ws, li?rws, and o<f>pa Negative Final Clauses . Attic final use of Past Tenses of Indicative

107-109 109-112 112 112



317. Subjunctive and Optative after tva, ws, o1rws, o<f>pa, and 318-321. 322, 323. 324. 325-328. 329, 3:JO. 331, 332. 333-337. 338. Subjuuctive after Secondary Tenses Optative after Primary Tenses Future Indicative in Final Clauses .A.v or d iu Final Clauses with Subjunctive .A.v or d in Final Clauses with Optative "Iva Tl ;-Omission of leading Verb Secondary Tenses of Indicative with va etc. Other expressions of Purpose 113, 114 114, 115 115 115, 116 116, 117 117-119 120 120-122 122




339, 340. Future Indicative etc. after o1rws and o1rws p,f}




341-347. Homeric construction after, and other early 124, 125 uses. 348. "01rws lLv with Subjunctive in Attic Greek 125 126 349 350. "O,.ws lLv and oKws lLv with Optative (rare) ' 351. 'Os, ws lLv, and il7rws lLv in Xenophon . 126, 127 352 353. Negative Object Clauses 127 ' 354. M'>I for lhrws p.?j in Object Clauses 127 355-360. "O,.ws after Verbs of Asking, Cornrnanding, etc. 128, 129 361. Object Infinitive for Clause with IJ1rws 129 362. Indirect Questions 129 363. "O,.ws with sigmatic Aorist Subjunctive 129 364. Dawes's Canon 129, 130

365, 366. 367. 368. 369. 370. 371. 372-375. 376. 377.

Subjunctive and Optative after p.?j Future Indicative after p.~ M?j with Optative and lLv M?j with Present and Past Tenses of Indicative "01rws p.?j for p.?j with Verbs of Fearing Indirect Discourse with ws or 81rws Infinitive after Verbs of Fearing etc. Indirect Questions Causal on SECTION VI.

131, 132 132 133 133, 134 134 134, 135 135, 136 136 136, 137

Conditional Sentences.
378-380. 381, 382. 383-387. 388-397. 398. 399-401.

Protasis and Apodosis.-Conditional Particles

"Av or Kl in Protasis and Apodosis

Negative Particles. Classification of Conditional Sentences Origin of the Conditional Sentence Early combinations of <I with Ke or lLv

137 137, 138 138, 139 139-142 142, 143 143-14.5


402-404. 405, 406. 407, 408. 409.


1. Sirnple Suppositions (chiefly Particular).

Simple Present and Past Suppositions . General Suppositions sometimes with Indicative Future Indicative expressing Present Intention Potential Optative and Indicative in these conditions 145, 146 146 146, 147 147

2. With Supposition contrary to Fact.

HO, 411. 412. 413. 414.

Secondary Tenses of Indicative with lLv in Apodosis Apodosis not always denied here . Relations of Tenses in unreal Conditions Aorist Indicative of Present Time .

147-149 149, 150

151 161



415-422. 423. 424-432. 433. 434-442. 443.

"Eoe<, xp'i]v, etc. with Infinitive (without l!v)

"Eo<L liv . "O<j>e'Aov, lfJovMJL'Y)V, l!jLe'A'Aov, etc. with Infinitive

Pl'Otasis for Infinitive in construction of 419 Homeric Peculiarities Homeric usages in Herodotus and Attic Greek

151-156 156, 157 157-160 160 16G163 168



1. Subjunctive or Future Indicative in Protasis with a Future Apodosis.

444446. Subjunctive with 1jv, Uv, or U.v in Protasis 447-449. Future Indicative in hotasis. 450-454. Homeric Peculiarities 163-165 165, 166 166, 167

2. Optative in Protasis and Apodosis.

455-459. Optative with El in Protasis, with liv in Apodosis 460, 461. Homeric Peculiarities 168, 169 169, 170


462-466. Subjunctive and Optative in Protasis 467. Indicative for Subjunctive or Optative 468471. Homeric and other Poetic Peculiarities

170, 171 171, 172 172, 173

Substitution and Ellipsis in Protasis-Protasis without a Verb.

472. 473. 4 74. 475. 476, 477.

Protasis contained in Participle, Adverb, or other word , 173, 174 Future Participle representing Future Indicative (407) 175 Homeric Ei o' li-ye 175 'Os Ei or WS er Tf in Comparisons 175, 176 El fLY, without a Ve1b.-IIMw El 176, 177 478. El o /LT,, otherwise . 177

Substitution and Ellipsis in Apodosis.

479-481. 482. 483, 484. 485.

Apodosis contained in Infinitive, Participle, Noun, etc. Apodosis omitted for effect Apodosis represented by liv 'Os el and tfJfnrep el in Si)lliles .

177-179 179 179 179, 180

Apodosis contained in Protasis.

486. General statement of Principle 487. Er K< or 1jv with Subjunctive in Homer, expressing Hope 180 180-182 182 182-184

or Desire .
488. El (or d Ke) with Optative in Homer, in same sense 489, 490. Similar constructions in Attic Greek and Herodotus



491. Subjunctive and Optative with r K, 1, etc. after o!oa,, 185, 186 elOov, etc .. 492. Comparison of the Protases of 491 with Clauses with 186 p.i} (366) 186 493. Relation of the Protases of 491 to Indirect Questions 186, 187 494A97. Ei after expressions of Wonder etc.

498. Protasis and Apodosis differing in Form


I. Optatime in Protasis, with


or Present Indicative etc. in

499. Optative in Protasis with Future Indicative etc. in

Apodosis (chiefly in Homer) 188 500, 501. Optative in Protasis with Present Indicative in Apodosis 188, 189 502. Optative in Protasis depending on Present like Ot or lv 189, 190

II. Indicative or Subjunctive in Protasis, with Potential Optative or

Indicative in Apodosis.
603, 504. Present or Past Indicative with Potential Optative or

606. Subjunctive or Future Indicative with Potential Opta-

190, 191 191, 192

tive .

III. Potential Optative or Indicative (with av) in Protasis.

506, 607. Potential Optative or Indicative with ilv in present or past Conditions with el


IV. l1-regular Combinations.-Present or Past with Future Conditions

in one Protasis.
508. Cases of Anacoluthon 609. Protases of different times combined in one 192, 193 193

V. Several Protases in one Sentence.

510. Several Protases (not co-ordinate) with one Apodosis 511. Relation of leading and subordinate Conditions 193, 194

194, 195

fle, d,\,\6.,




512. Apodosis introduced by word meaning but 513. 'AXM vOv or ciXM (elliptical)

195 195


Relative and Temporal Sentences.

ln4. Relative and Temporal Words
' 195, 196



515-517. Definite and Indefinite Antecedent 518. Negative Particles in Relative Sentences

196 196



519. Indicative and other constructions after Relative with Definite Antecedent. 196, 197



520. Conditional Relative explained 197, 198 521. Conditional Relative Clauses and forms of Protasis compared 198, 199 522. "Av or d with Conditional Relatives and Subjunctive 199 523. Classification (same as in Conditional Sentences) 199

524. Two forms of Present and Past, and two of Future, Conditions


525, 526. With Simple Indicative (like Protasis of 402) 527. With Future Indicative to express PTesent Intention 528. 'With Secondary Tenses of Indicative (like 410)

199, 200 200 200-202



529, 530. With Subjunctive (like 444) 531. With Optative (like 455)

102, 203 203, 204

532, 533. Subjunctive and Optative in general Conditions 534-537. Indicative in general Conditions

204-206 206, 207

Homeric and other Poetic Peculiarities.

538-541. 542. 543-549. 550, 551.

Subjunctive without KE pr llv Relative with KE or llv and the Optative . Homeric Similes with J,, etc.. "0 n p.1} and orrov p.1} without a Verb (Homeric)

207-209 209 209-211 211

Special Forms of Antecedent Clause.

552. Infinitive, Participle, Noun, etc. in Antecedent Clause 553. '01r6TE after past Verbs of Expecting (Homeric) 212 212

Mixed Conditional Constructions.

554. Optative depending on Present or Future 555. Optative depending on o<l, xp1}, etc. 212 212, 213



556. Indicative or Subjunctive depending on Potential Op-

557. Potential Optative or Indicative in Conditional Relative

213 213, 214


Assimilation in Conditional Relative Clauses.

558. 559. 560-562. 563.

Assimilation by Subjunctive or Optative Assimilation by Secondary Tenses of Indicative Princi pies of Assimilation Assimilation after General Conditions (variable)

214 214 214, 215 215

6. in the Antecedent Clause.

564. Antecedent Clause introduced by M

215, 216

565, 566. 567. 568-571. 572. 573. 57 4.

Future Indicative in Attic Greek . Past Purpose expressed by Imperfect of p.\\w S'ubjunctive and Optative in Homer Subjunctive not used in Attic Greek Optative rare in Attic Greek . Future Optative occasionally used .

216 216 216, 217 217 218 218


575. 576, 577. 578. 579.

Iudicative (with negative ou) Future Indicative (with p.-1)) "01fw< as Relative thus used . Occasional use of the Optative

218 218, 219 219 219

580. Causal

Rel~tive with Indicative 581. Causal and Conditional Forces united (with JL7J)

220 220







{> OR e<f>' {>Te,

220, 221 221, 222

582, 583. General Distinction of Infinitive and Indicative with

584. Meaniug of Cxne and principles of its use 585. "Ocrre in Homer and the early Poets 586. 'Os for ws n




587. 1. Expressing a Result to wl1ich an act tends

223, 224

2. Expressing a Condition or Limitation 3. Expressing a Purpose (like Final Clause) 588. After Verbs of Wishing, Gommaruling, 'etc. 589. Two examples only in Homer 690. Tenses of Infinitive after wcrre

224 224
224, 225 225, 226 226



591. Future Infinitive rare (except in Indirect Discourse) 226, 227 592. Infinitive with llv 227 593. Oih-w (fxrT< in Herodotus 227 594-599. Infinitive in Indirect Discourse and other constructions (w(jT oo) 227-229 600. Omission of W(J"Te 229


with Indicative and other constructions 604. Optative by Assimilation 605. Optative of Indirect Diseourse 606. "D(jTE p:f} with Finite Moods (rare)

229, 230 280, 231 231 231


607. Participle (by Assimilatioi1) after


231, 232



608. Chiefly in Aeschylus, Sophocles, Herodotus, Xenophon 609. Rarely in other Authors

232, 233 233





610. With Iufiniti ve and Future Indicative .


Temporal Particles signifying Until and Before.

611, 612. Meaning of a clause with Until 613. 1-5. Five constructions with i!ws (as Relative) 614. Final use of Mws (especially in Odyssey) . 615. "Oq,pa in Epic poetry 616. Eh I! KE in Homer; is I! and is oii in Herodctus 617. "E(jTE (after Homer) 618. "Axp< and p.xp< (chiefly in Prose) . 619. "Axp< oiJ and p.fxp< o1i 620. Omission of llv with i!ws etc. and Subjunctive
B. ITp{v, BEFORE, UNTIL. 621, 622. Meaning and general use of 1rplv 623-625. Development of constructions with 1rplv.
240, 241 241243

234, 235 235-237 237, 238 238

238, 239

239 239 239, 240



626. llplv used regularly with Infinitive in Homer. 627. Later than Homer: chiefly after Affirmatives 628-630. Infinitive with 1rplv after Negatives 631. H 1rpv with Infinitive


243, 244
244, 245 245





632. 633. 634. 635. 636. 637.

IIplv with Indicative in early Poets

In Attic Poets In Prose : chiefly after Negatives . In Prose : exceptional use after Aflirmatives IIpiv y' ore with Indicative in Homer Indicative with 1rplv in unreal Conditions

245 245, 246 246 246, 247 247 247

638. 639, 640. 641. 642. 643. 644. 645, 646. 647.


IIplv with Subjunctive only after Negatives 248 Without liv or d in Homer and Hesiod . 248 IIplv 'Y' or' liv with Subjunctive in Odyssey 248 IIp~v av with Subjunctive after Homer and Hesiod 248, 249 IIplv with Optative only after Negatives 249 Optative with 1rplv in Indirect Discourse 249, 250 IIplv with Subjunctive in General Conditions. 250 Apparent Aflirmatives followed by 1rplv and Subjunc250, 251 tive. 251 648. IIplv without liv followed by Subjunctive 251 649. Doubtful cases of 1rplv IJ.v with Optative 251 650. IIplv with Subjunctive depending on Optative with liv
IIp~v ~' 1rp6upov ~' AND m5.pos.

651, 652. IIplv 11 in sense of IIplv 653, 654. IIp6repov 11 655. "Trnepov 11 with Infinitive, once in Thucydides 656. IIapos with Infinitive in Homer

251, 252 252, 253 253 253

IJpv, mxpos,


657. IIplv (as adverb) etc. in leading Clause in Homer 658, 659. IIpbrepov, 7rp6(]'1Jev, etc. after Homer 660, 661. <PIJavw as correlative to following 1rplv or 11

253 253, 254 254


Indirect Discourse.
662. Direct and Indirect Quotations distinguished. 663. Manner of introducing Indirect Quotations 664. Relation of Indirect Discourse to other substantive Clauses 665. Indirect Questions. 666. Extent of term Indirect Discourse 667, 668. General Principles of Indirect Discourse 256, 254
255 255

256 256




669. (1, 2.) Indicative and Optative with Sn and ws, and in Indirect Questions 670. Indicative and Optative in same Sentence 671. Indirect Questions and Quotations in Homer . 672. Imperfect and Pluperfect retained . 673. Present Optative as Imperfect 674. Imperfect and Pluperfect for Present and Imperfect 675. Independent Optative, generally with "'(rip 676. Optative with on or after Present Tense implying Past. 677-680. Subjunctive or Optative representing Interrogative Subjunctive 681. Indicative or Optative with 11 682. Secondary Tense of Indicative without liP (Potential) 683. Infinitive in Indirect Discourse 684. When Infinitive stands in Indirect Discourse . 685, 686. M?) with Infinitive in Indirect Discourse 687. Participle in Indirect Discourse 688. Negative p.?j with Participle


258-260 261 261, 262 262 263 263, 264 264 264 265, 266 266 266, 267 267, 268 269 269, 270 270-272 272


689. General Principles and Examples . 272-276 690. Mixture of 1\loods in Quotations 276, 277 691. Imperfect and Pluperfect representing dependent Present and Perfect Indicative 277 '692. A inegularly retaiJH;d with Optative from the direct form 277 693. Aorist Indicative in dependent Clause rarely changed to Optative 277, 278


694. General Principles of these Clauses. 695-700. Six classes of these Clauses:1. After Infinitive following Verbs of TVishing, Com'manding, etc. . II. Protases with Apodosis implied in leading Verb III. Protases after past Verbs of E-motion IV. Temporal Sentences expressing Past Intention, Purpose, etc. V. Past Causal Sentences with assigned cause VI. Relative Clauses containing another's thought 701. Imperfect and Pluperfect for Present and Imperfect 702. "Av irregularly retained with Optative 703, 704. Same principle applied to Final Clauses etc. 705, OUi' liTt without a Verb


278, 279 279, 280 280, 281

281 281 281, 282 282 282 282, 283 283



o'lrws, 8, oliveKa,



706. "01rws like ws in Indirect Quotations 707' 708. Ovx O'lrWS, OVx on, etc. . 709. "0 or a7' (for a7) in Homer for lin 710. Ovv<Ka, oOoUP<Ka, and o,dJn 711. "On before Direct Quotations .

283 283, 284 284, 285 285 285, 286


Causal Sentences.
28u 712. Causal Sentences aiHl Causal Particles . 286, 287 713. Indicative in Causal Sentences 714716. Optative, to express cause assigned by another, after 287 Past Tenses 717. Cause expressed by Potential Optative or Indicative 287 718. Interrogative Causal Sentences etc. 288 719. 1. Cause implied (not expressed) in leading Sentence 288 2. 'E,.<l, although, referring to something implied 288


Expression of a Wish.
720. Two classes of Wishes 288

721. 722. 723. 724. 725. 726. 727. 728. 729. 730.

Two forms in Future Wishes. I. Pure Optative II. Optative with ,re., Ei 'Yap, or Ei Present Optative in Homer in Present Wishes Optative in Commands aml Exhortations 'Os with Optative in Wisbes OV7ws with Optative in Protestations Wish expressed by Potential Optative Infinitive in Wishes (see 785 and 786) Wish in Homer followed by Apodosis PRESENT OR PAST WISHES (NOT ATTAINED).

289 289 289, 290 290, 291 291 291 291 291 291 291, 292

731. 732, 733. 734. 735. 736. 737.



292, 293 Two forms in Present or Past Wishes 293 I. Past Tenses of Indicative with <lll< or Ei 'Yap II. "Dtpi'Aov and (Hom.) (i;rpi'A'Aov with Infinitive 293, 294 Form with CJrp<'Aov or the Optative in Present Wishes in 294 Homer (739) 294 Etll<, El 'Yap, and wfJ before (/;tpe'Aov 294, 295 'Os before t:Jrpe'Aov (poetic) 2!!5 Simple El not used witl1 w<{>e'Aov or Indicative in Wisl1es 295 Present Optative in Preiient Wishes in Homer



740. Greek and Latin expressions of Wish compared.-Optative and Indicative in Wishes distinguished by Time 295, 296


741. Infinitive as a Verbal Noun .

742-744. Origin of Infinitive and development of its use

297 297-299

A. INFINITIVE WITHOUT THE ARTICLE. 745. Infinitive as Subject, Predicate or Appositive

299, 300

Infinitive as Object.
746. Two classes of Object Infinitive

Object I nfirnitive not in Indirect Discourse.

. 748. Peculiar Object Infinitives in Homer 749. Infinitive after Nouns with Verbs (as Object) 750. Infinitive in Laws, Treaties, etc.
747. Ordinary Object Infinitive after Verbs

300, 301 301

301, 302


Infinitive in lndi1ect Discourse.

751. Infinitive after Verbs of Saying, Thinking, etc. 302, 303 752. Infinitive after Verbs of Hoping, Promising, etc. (See
753. Constructions after </17J~t1, Ei'll"ov, and A.-yw

303 303

754. Personal and Impersonal Constructions with A.l-yerat etc .. 303 755, 756. Infinitive with Relatives etc. (by Assimilation) 303, 304 7o7. Infinitive in Narration (like Indicative). 304, 305

Infinitive after Adjectives, Adverbs, and Noum. 758. Infinitive with Adjectives denoting Ability, Fitness, etc. . 305 '/59, 760. Infinitive with rowvros olos etc., and (in Homer) with roZos etc . . 305, 306 761. Infinitive with v6vra, 7rpou~Kovra, etc. used personally 306 762. Infinitive with lilK<uos etc. used personally 306 306, 307 763. Adjectives with limiting Infiuitive 764. Infinitive with Comparative and ij. 307 765-768. Similar use with Adverbs, Verbs, and Nouns. 307, 308 769. Infinitive with op.oZos in Homer 308



Infonitive of Purpose

770. 771. 772. 773, 77 4.

Chiefly used with Verbs of Choosing, Giving, or Taking. 308, 309 Infinitive Active or Middle (rather than Passive) . 309 In poetry with Verbs of J,fotion, and with Elp.! etc. 309 Eivcu denoting Purpose (chiefly Ionic) 309, 310 775. Infinitive expressing Result (in Homer). 310

Absolute Infinitive.
776. 777. 778. 779. .780, 781. 782. 783. Infinitive expressing Limitation (parenthetical)
'Os g,.os d?r<w, ws El?r<'i:v, or <i?r<'i:v, etc. tOs OoKe'iv, Ws ElKd.<rat, Ws l0'iv, &.KoVa-at, etc. '01\!-yov O<tv, jJ-LKpoO O<tv, or o/\i-yov, JLLKpoO Absolute <ivcu (as in hwv iva.<)

310 310, 311

311 311, 312

Absolnte Infinitive in Herodotus . Absolute Irrfinitive as Accusative of Limitation

312 312, 313 313

Infinitive in Oornmands, Pmhibitions, Wishes, and Exclamations.

784. 785. 786. 787. Infinitive in sense of Imperative Infinitive like Optative in Wishes . Infinitive with a.t -yap in Wishes (twice in Odyssey) Infinitive Subject Accusative in Exclamations 313 313, 314


B. INFINITIVE WITH THE ARTICLE. 788. General use of Articular Infinitive 314, 315

Articular Infinitive as Subject or Object.

789. 790. 791, 792. 793. 794. Article makes the Infinitive more distinctly a Noun Infinitive with r6 as Subject Infinitive with r6 as Object Infinitive with roG as Object . Infinitive with ro in Indirect Discourse (rare) 316

316, 317 317 317

Infinitive with

T6 after Adjectives and Nouns.


'795-797. Infinitive as Accusative after Adjectives and Nouns.

Infinitive with
798. 799. .S00-803. 804. 805. 806.

Tov, Tljl, and T6 in various Oonstructiow.


Constructions of Infinitive as Genitive with roO Constructions of Infinitive as Dative with r</i. Infinitive with roO, rei>, and TO with Prepositions Artiaula.r l!lfi11itivs IJ.g Appositive . Infinitive with ro in Exclamations. Infinitive with dependent Clauses with ro as Noun

319, 320
320, 321

321 821, 322



Simple Infinitive and Infinitive with


after Verbs of Hindrance etc.


807-810. Four expressions after Verbs implying Hindrance: (a) ?rotdv, (b) rou 1rOt<tv, (c) 11-iJ 1roteiv, (d) rou JLr) 1rot<Zv.-Mij ov when leading Verb has Negative 322, 323

Infinitive with T6 fl~ or T6

fJ-0 oll.
324, 325 325, 326

811-813. After expressions implying Hindrance or Denial 814. Infinitive with ro 11-iJ ov (or ro p:q) iu negative sense



815-817. M?j and 1-'iJ ov with Infinitive. 818, 819. Mij ov with Participles and 'Nouns . 820. M1] otl forming one Syllable

326, 327 327, 328 328

THE PARTICIPLE. 821. Participle as Verbal Adjective 822, 823. Three uses of Participle distinguished A. ATTRIBUTIVE PAP..TICIPLE. '824. 825. 826. 827. 828. 829. 329 329

830. 831.

Participle as Adjective . 329, 330 Participle with Article used snbstantively 330 Future Participle in these nses 330 Participles (generally plural) used substantively without Article 330, 331 Participles (as Substantives) with adnominal Genitive 331 (a) N enter singular of Participle with A1ticle in sense of Articular Infinitive 331 (b) Similar constructions without Article 332 Participle as Predicate Adjective . 332 Participle with elJLl or txw as periphrastic Perfect, Pluperfect, or Future Perfect 332

832. Participle defining circumstances of acticn, expressing various relations 333 833, 834. I. Time 333 835. II. Means . 333 836, 837. Ill. Manner, including manner of Employment 334 838, 839. IV. Cause or ground of Action 384, 335 840. V. Purpose, Object, or Intention 335



841 VI. Condition (Participle in Protasis) 335, 336 842. VII. Opposition, Limitation, or Concession 336 843, 844. VIII. Any attendant circumstance 336, 337 845. IX. That in which an action consists . 337 846. No exact distinction of all circumstantial Participles possible . 337

Genitive Absolute.
847. 848. 849. 850. Genitive Absolute independent of main construction Participle alone used absolutely Passive Participle in Genitive Absolute with Clause Genitive Absolute rare with suLject already belonging to the sentence

337 338 338


Accusative Absolute.
851. Impersonal Participles in Accusative Absolute 338, 339 852. Rarely with Infinitive and r6 339 853, 854. Personal Participles sometimes in Accusative Absolute (generally with ws or &o-7rEp) 339, 340

Adverbs with Circumstantial Participle.

T6re, 1fon, ivraDIJa, tra, l7reLra, oi!Tws, etc. "A11-a, Jl-ETa.~u, euiJus, aUr[Ka1 etc. Kal7rp (Kal trep), ovoe, JJ.?]De,, etc. "Are, ota or otov "Oo-re in Herodotus, like liT< . 'Os, when thought of leading suLject is expressed 867, 868. "0J'7r<p and &J'7r<p <l 869-874. Remarks on &J'7r<p and ws with Participle
855-857. 858. 859-861. 862. 863. 864-866. 340
340, 341

341, 342 34.2 342 342, 343 343, 344 344-346

Omission of <:lv.
875. Cases of omission of &v : 1. After liu, ola, ws, or Kal1rep
2. Rarely without these Particles (poetic) 3. With hwv and 11Kwv

4. When another Participle precedes

346 346 346 346, 347

ComMnations of Circumstantial Participles.

876. Participles belonging to main construction combined with those in Genitive or Accusative Absolute in one sentence .


877. Nature of Supplementary Participle 847, 848 878. Two uses, corresponding to those of the Object Infinitive ( 746) 348





879, 880. 881, 882. 883. 884-886. gs7-894. 895. 896. 897. 898. 899. 900. 901.

I. With Verbs signifying to begin, endure, cease, stop, permit, etc. 348, 349 II. With Verbs denoting states of feeling (to repent etc.). 349 Ill. With Verbs denoting to find, detect, represent 350 IV. With Verbs of Perceptidn (hearing, seeing, etc.) 350-352 V. With Xav&rlvw, ru-yxrlvw, tp8rlvw, etc. 352-354 VI. With, 1jKw,, etc. 354 VII. With, 1roXMs dp.t, etc. in Herodotus 354 With etc. in Attic Greek 354, 355 VIII. With chrooElKvup.t, Ka8l5w, and 355 IX. With apKew, iKav6s elp.t, etc.. 355 X. Dative of f3ouMwvos, rjlioJ.'EVOS, etc. with Dative after dp.l etc. 355, 356 XI. Dative of Participle with Impersonal Expressions (it is fitting, pleasant, etc.) 356

Omission of t5v.

occasionally omitted in constructions of 879-901


Infinitive with Vmbs of 879-901.

903. Infinitive sometimes used with Verbs which take Supplementary Participle:1. With and 357 2. With avexoJ.'aL) V'TrOJ.'EVW, rMw, TOAJ.'W 357 3. With a'TrOKap.vw 357 4. With 357, 358 5. With 1ravw 358 6. With 7rEpwpw, overlook, pern~it, etc. 358 7. With the Impersonal Expressions of 901. 358 8. Probably never with Xav8rlvw, rv-yxrlvw, and <jJ8rlvw 358, 359



904. Participle with Verbs of seeing, hearing, knowing, showing, etc., and with 6.-y-ye>.Xw, like Iufinitive of Indirect Discourse. 359, 360 905. Participle agreeing with Accusative of Reflexive 360 906. Participle of Impersonals in Accusative . 360 907. Participle with ofi;\os and <f>avp6s elp.t 360 908. With o-vvotoa and o-u-y-yt-yvwo-Kw and Dative of Reflexive 361 909. With Infinitive depending on Verb with Dative 361 910. Occasionally with Verbs like vop.l5"w 361 911. "Ov sometimes omitted in Indirect Discourse 361



912. Clause with on or ws for Participle 913. Mlp,vw-<a< oTE for Participial Construction

361 362

Infinitive with the Vmbs



914. Infinitive of In,'<irect Discourse with some of the Verbs of 904:1. With aKovw, 7rvv1Javof1a<, and alrr1Javof1a< 36'2 2. With opw 362 3. With U'YJ'fAAW 362 4. With OflOAo-ylw 362 5. With <f>alvof1a< 362, 363 915. Infinitive in various uses and senses with other Verbs of 904:1. With f1av1Javw, fllflV'fJflU<, and E7r<Aav1Javof1a< 363 2. With oToa and E7rirrraf1a< (two uses with Infinitive) 363 3. With -y<-yvwrrKw (three uses with Infinitive) 364 4. With OeiKVVflL 364 5. With O'fJAw 364 6. With d;pirrKw (three uses with Infinitive)

'Q,- with PaTticiple in IndiTect Discoune,

916. 'Os showing that Participle expresses thought of leading subject 3G.! 917, 918. 'Os with Circumstantial Participle, equivalent to Indirect Discourse . 365, 366 919. Participle with ws (peculiarly used) after certain Verbs 366, 36i of saying and thinking


920. Two Constructions of the Verbal in -ros 368 921. Personal Construction . 368 922. Agent expressed by Dative in Personal Construction. -Omission of elfll 368 923. Impersonal Construction with lrrrl. 368, 369 924. Comparison with Latin Participle in -dus 369 925. Verbal in -rlov and Infinitive (se. M') in same Construc369 tion . !l26. Agent expressed by Dative or Accusative in ~m personal Construction 369

I. The Relation of the Optative to the Subjunctive and other Moods




11. The Origin of the Construction of ov 1"-IJ with the Subjunctive and the Future Indicative III. Statistics of the use of the Final Particles IV. Xenophon's peculiar use of ws, ws liv, and B.,.ws liv in Final and Object Clauses . V. On some disputed points in the Construction of i!oeL etc. with the Infinitive (Supplement to 415-423)

389 398 400 403 413 441 452


1. THE Mood of a verb sh~ws the rnanne1 in which the assertion of the verb is made. The Greek verb has four moods, properly so called,-the indicative, the subjunctive, the optative, and the imperative. The infinitive, which is a verbal noun, and the participle and the verbal in -T~o<;, which are verbal adjectives, are so closely connected with the moods in many constructions, that they .are discussed with them in Syntax.
The four proper moods, as opposed to the infinitive, are sometimes called the finite moods. The subjunctive, optative, imperative, and infinitive, as opposed to the indicative, are sometimes called the de/pendent moods.



2. The indicative, in its most primitive use, makes a simple, absolute assertion, or asks a question which includes or concerns such an assertion. E.g.

r pacpH, he is writing j ypacpev, he Was writing j ypatfev, he WrOte; ypatj;H, he will Write. r paq)H; is he W1'iting? Jypatj;aT; did you write? ypatfen; will you WTite? T 'Eypatfev; what did he write?
3. The indicative may also express (a) A dependent statement (or quotation) of such an absolute assertion or question. E.g.
AeyH on ypacpH, he says that he is writing (he says ypacpw); A.eyet on;Et, he says that he will write (he says ypatfw); ~pwri- ..-{ Eypatfap.ev, he asks what we wrote; f.pwT(j. El eypafa, he asks whether I



(b) A distinct statement of an object aimed at or feared. E.g.

'E1Tt,UEAEtrat 1 OUTO YEII'IJO'ETat, he takes cure tha~ this shall be done (339); <j>o(3o-6,ue8a .U'l d,u<j>orpwv ~,uapr~Ka,uev, we fear that we have
missed both (369, 2).

(c) A distinct supposition of an absolute statement, that is, a supposition that such a statement is, was, or will be true. E.g.
El ypa<j>et, if he is 1vriting; el f.ypafev, if he wrote; el yypa<j>e, if he has wTitten , el ypafet, if he shall write or if he is to wTite. What is supposed in each case could be expressed by ypaq>et, f.ypu-fev, yypa<j>e, or ypafet.

if I

4. The past tenses of the indicative may, further, express a supposition that some statement either had been or wen now true, while it is implied that really it was not or is not true. E.g. El f.ypafa, if I had wTitten; el f.ypa<f>ov, if I were now w1'iting or

hnd been wTiting; the context indicating that really I did not vnite or am not wn'ting (410). These expressions originally always referred io the past, as they do in Homer.

5. Out of the form of unreal supposition (4) were developed aih~r Homer the use of the past tenses of the indicative with dBE or El ryap in wishes (732); and also the Attic construction of the past tenses of the indicative to express an unaccon!plished purpose (333), where there is an assimilation of the final clause to a preceding indicative.

El yd.p rovro J1rolrwa, 0 if I had only done this! EWe rovro el'xes, 0 if you only had this ! EWe ror' d7T'eeavov, (va .U'J TOVTO E7T'<L8ov, would that I had then perished, that I might not have sujj'wed this. For the indicative with dv or KE, the potential indicative, see 243.



6. (a) The subjunctive, in its simplest and appai:ently most primitive use, seen in Homer (284), expresses futurity, like the future indicative, and has for its negative. E.g. Ov ydp 1rw ro[ovc; i:oov avf.pas ovo i:ow,uo.t, for never did I see such


rnen nor shall I ever see them, Il. i. 262; Ka[ 1ror ns el1r'{/O'tv, and sorne one 1cill some time say, Il. vi. 459.

(b) Though this primitive use disappears in the later language,



the subjunctive still remains closely related in sense to the future indicative, and in most of its constructions can be interchanged with it. 7. The subjunctive in questions of appeal as to the future (28 7) has, even in Homer, developed the idea of propriety or expediency. E.g.
NMi p.vw ~e ()f.w; shtill I remain here 01' run? Il. x. 62. So 'TI'll But the future indicative can be used in the same sense; as -r IS~-ra /Spwp.ev; Jl-YJTEP 1) <j>ovdJ<rop.ev; what aTe we to do? shall we slay our mother? EuR. El. 967. (See 68.)

Zw; whithm shall I go ? Od. xv. 509.

8. (a) In exhortations and in prohibitions with f-L?J (250259) the subjunctive has an imperative force, and is always future; as in rwuEV, let us go; f-L~ Oavf-LaCT'TJTE, do not wonder.
The future indicative occasionally occurs in prohibitions with p.~ (70).

(b) The subjunctive with f-LtJ, especially in Homer, may express a future object of fear with a desire to avert it; as in f-L~ vijar; f!t..,wcr~, may they not seize the ships (as I fear they will). (See 2 61.) From such expressions combined with verbs of fearing arose the dependent use of f-Ltl with the subjunctive expressing a future object of fear; as cpo(3ouf-La~ f-L~ a:JTot..,nmt, I fear that he may perish.
9. In the constructions with ov f-Ltl (294) the subjunctive and the future indicative are used, without apparent distinction, in a future sense; as ov f-L?J ryevnm~ and ov f-L~ ryEv+ crETa~, it will not happen. 10. The subjunctive may express a future purpose or a future object of care or exertion. E.g.
"Epxe-rai 07TWS TOVTO he comes that he may see this (317); l'TI'ijl-EAel!rui 07TWS -rovro "jEvYJTa i (or yev~O'e-rai), he takes care that this shall be done (339). In clauses of purpose the future indicative is sometimes used (324), and in the construction of 339 it became the regular Attic form.


11. In conditional clauses the subjuncti Ye expresses either a future supposition (444), or a general supposition which is indefinite (never strictly present) in its time (462).
(a) In the former it supposes 8Uch a future case as the Homeric subjunctive (6) states ; as Uv -rt> Ef7T)), if one shall say (the thing supposed being d7TYI ns, one will say) ; here the future indicative may be used



in essentially the same sense (44 7). In the general condition it supposes an event to occur at any time, as we say if any one ever goes or whoever goes, with an apodosis expressing repetition or a general truth ; as M.v n<; KA~tfTJ (or il<; aJ' KA.efv), KoM.(ETat, if any one steals (or whoever steals), he is always punished.

(b) The subjunctive in general suppositions is the only sulJjunctive which does not 1efer to future time, and here the future indicative can never be used. In most other languages (as in English and generally in Latin), and sometimes in Greek, such a condition is expressed by the present indicative, like an ordinary present supposition; but the Grr.ek, in its desire to avoid a form denoting present time, generally fell into one which it uses elsewhere only for future time. The construction, however, appears in Homer imperfectly established, except in relative clauses (468): this indicates that it does not belong to the primitive uses of the subjunctive. (See 17.) For the Homeric subjunctive with KE or d1' in independent sentenees, which does not differ perceptibly in meaning from the future with KE or dv, see 201, 1.



12, The optative is commonly a less distinct and direct form of expression than the subjunctive, imperative, or indicative, in constructions of the same general character as those in which these moods are used. 13. This is seen. especially in independent sentences, where the optative either expresses a wish or exhortation, or is used (regularly with or lcf) in a potential sense.


Thus l'o,fl-EV, may we go, corresponds as a weaker form to l'w;ull, let 1iS go. Corresponding to ~A.8wl' n'> lof.Tw, let some one yo out and see, we have E~EABcfw n<; l'oo t, may some one go out and see, Od. xxiv. 491. EA.o,To dv, he would lake or he might take, corresponds to the Homeric EAYJTO.t or EAY)Ta. KE, he will take or he may take (201, 1). We find in Homer a few Ol)tatives expressing concession or permission, which have a neutral sense and can hardly be classed as either potential or wishing. See Il. iv. l 7, El o' 1rWS TOOE 7rUCH cpA.ov Kat ~ov 1rEAOtTo, -i} TOt fl-~V 0 l KED tTO 7rOAiS ITpi~Jp.ow avaKTos, a.Vns o' 'ApyE[1)V 'EA.evryv llfEJIEAO.OS dyo tTO, where we may translate the apodosis either let the city still be a habitation and let M. cany away Helen, or the city rnay still be a habitation and JJI[. may carry away Helen. In iii. 72 we have yvva.ZKa TE Otl<a.O' ayf.c:rew, and in.iii. 255 T0 KE VtK~fTO.VTi yvvry Ka.i KT~fl-a()' E'lrOiTO, where J.yc:r8w and E71'0tT6 KE refer to essentially the same thing with dyo,To in iv. 19. Following I1. iii. 255 (above) we have of o' aAAot va.fol.fl-EV Tpo[Y)v, Toi o~ vovTa.t, i.e. the 1est of us may Ternain dwelleTs in Troy, while





they will return to Greece. From such neutral future expressions were probably developed the two distinct uses of the optative. In its hortatory sense as a form of wishing, the optative was distinguished by the use of JkYJ as a negative ; while in its potential sense it had ov as its negative (as in ov P'l'' yap TL KaKrhupov aAAO 7ra80LJkL, for really I can s1~ffer nothing wnne, Il. xix. 321), and it was soon further marked by the addition of KE or &.v. (See .Appendix I.)

14. In dependent clauses expressing purpose or the object of exertion or of fear, the optative is never an original form; but it always represents a dependent subjunctive or future indicative (8, b; 1 0) in the changed relation m which either of them is placed when its leading verb is changed from present or future to past time.
We represent this change in English by a change from may to might, or from shall or will to shmdrl or would; as epxErat Zva Uiv, he comes tlutt he may see, i)>.8Ev Zva Zoo t, he came that he might see; E7rLfLEAEtrat o1rws rovro yEvljfJETat, he takes caTe that this shall be done, E7rEfLEAeiro o1rws rovro Y'VYJfJO tro, he took mre that thiH should be done; <f>of3EZrat fk'l rovro 1rafJv, he feaTs that he may S?<jfer this; <f>of31JfJYJ fL0 TOVTO 7r a() 0 L, he feared that he might sujfe? this. Here the original subjunctive or future indicative (especially the latter) is very often used in place of the optative.

15. In all forms (14) holds, that the a changed relation) direct form, which present and future tenses (667, 1).

of indirect discourse the same principle optative after past tenses represents (in an indicative or a subjunctive of the original mood is always used after tenses, and may be retained after past

Here again we see what the change is, for we represent it by onr change from is to was, hr<ve to had, shall and will to should and 'lmnld, etc. ; as AEYEL on dA1)8s E(J"TL v, he says that it is tnw; EAE~EJI OTL aAYJBf.s [,7 (or EfJr[v), he said that it was true; A.!:ya on ypafEL, he oays that he will write; EAE~EV on ypafot (or ypafH), he said that he would write. So ovK oioa r[ d1rw, I lcnow not whrd I shall say; oi~< ifoEtv r[ ,;:7rOLfLL (or d1rw), I lcnew not what I should say.

16. In future conditions the optative expresses the supposition in a weakened future form, as compared with the stronger future of the subjunctive and the future indicative.
Compare a.v f!A.fJw, if I (shalT:) go (444), with El A.8otfLt, if I should go (455). Often the form of the leading sentence (the apodosis) decides
whether a given supposition shall be expressed by a subjunctive or by

an optative; thus in


iv. 11 we have av oi5r6s


7ra8v, 1j any-



thing happens (shall happen) to liim (Philip), depending on TrOL~aue; and in the next sentence, referring to precisely the same contingency, we have eZ n 1raOot, depending on two optatives with d.v.

17. The only remaining form of dependent optative is that found in past general suppositions, as t Tl'> tc"A."/rtev (or o\ tcA-e'ft'ELEv), tcoA.cil;ero, if ever any one stole (or whoever stole), he was (always) punished (462; 531).
Here the optative after a past tense represents an original subjunctive after a present tense (11), differing in this from the optative in future conditions (16), which is in an original construction. The late development of this optative appears from its almost total absence in protasis with e1 in Homer ( 468), where the corresponding subjunctive in protasis is also infrequent. It may therefore be disregarded in considering the primitive uses of the optative. (See 11, b.) For a more full discussion of the relations of the optative to the other moods, see Appendix I.



18. The imperative expresses et command, exhortation, entreaty, or prohibition (250 and 259). E.g.
<I>evye, begone!
thi.~. M~ 7r0LEL aOLKa,

'E>..O.irw, let him come. b.os p.ot rovro, give me do not do what is unjust.

19. THERE are seven Tenses,-the present, imperfect, perfect, pluperfect, aorist, future, and future perfect. The imperfect and pluperfect occur only in the indicative; the futures are wanting in the subjunctive and imperative. 20. These tenses may express two relations. They may designate the time of an action as present, past, or jut~tre _- and also its character as going on, finished, or simply taking place. The latter relation is expressed by the tenses in all the moods and in the infinitive and the participle; the former is always expressed in the indicative, and to a certain extent (to be explained below) in the dependent moods and the participle. 21. The tenses are divided into p1irnary tenses, which denote present or future time, and seconda1y or histmical tenses, which denote past time. This distinction applies properly only to the tenses of the indicative; but it may be extended to any forms of the dependent moods which have the same distinction of time as the tenses of the indicative. The primary tenses of the indicative are the present (in its ordinary uses), perfect, future, and future perfect. The secondary tenses are the imperfect, pluperfect, and aorist (in its ordinary uses).
This distinction will be more fully explained at the end of this chapter (165-191). It must be noted that the historic present (33) is a secondary tense, and the gnomic aorist (154) is a primary tense. 22. In speaking of the time denoted by any verb, we must

distinguish between time which is present, past, or future with


reference to the time of speaking or writing (that is, time

absolutely present, etc.), and time which is present, past, or

future with reference to the time of some verb with which the verbal form in question is connected (that is, time relatively present, etc.) Thus, when we say -rain-a &i\.7]8&c; i.crnv, this is true,!v is present with reference to the time of speaking; but when _we say ~</>7J -rov-ro 0.A:ry8~c; elv<u or el1rev on -roln-o &A.r]fJ, (or el7J), he said that this was true, (i.e. he said "this is true"), the present tense which we use denotes time present to the time of the leading verb, i.e. time absolutely past and only felatively present. The same distinction is seen between the future in -rov-ro yv~cre-rat, this will happen, and that in ~<f>YJ -rov-ro yev~crecr8at or el-;rev on yev~crerat (yev~crot-ro), he said that this would happen/ where the future in the first case is absolutely future, but in the other cases is only relatively future and ma.y be even absolutely past. Again, in -rov-ro i.y&ve-ro, this happened, the aorist is absolutely past j but in ~cf>YJ -rov-ro yeva-8at, or <l?rev on -rovro i.y&v-ro (or yvot-ro), he said t/w,t this had happened, it denotes time past to the time of the past leading verb, and so is doubly past. But in connection with a future expression an aorist, though relatively past, may be absolutely future j as in PLAT. Rep. 478 D, r6 <f>avv as subject of rrerr8o.t means that which will hereafter have appeaTed. So &a?rpataJLevoc; in 406 E. (See 143.) It is a special distinction between the Greek and the English idioms, that the Greek uses its verbal forms much more freely to denote merely relative time. Thus, we translate the Greek presents el'vat and f.crr after <f)YJ or e'l-;rev (ahove) by our was/ the futures '}'EV~crErrBat and ywqcrETat by 1vould happen/ and the aorists yEvcr8at and i.yv<ro by had happened. This distinction appears especially in the indicative, optative, and infinitive of indirect discourse ; in future forms after past tenses in final and object clauses with i:va> o-;rwc;, etc. ; and usually in the participle; but not in protasis.

23. The present indicative represents an action as going

on at the time of speaking or writing ; as or I am writing.

rypacpw, I w1ite,

An important exception occurs when the present indicative in indirect discourse denotes time wbich is present relatively to the leading verb. See above, 22; 669, 2 ; G7 4, l.



24. As the limits of such an action on either side of the present moment are n?t defined, the present may express a c-ustornary or 1epeated actwn or a geneTal truth. E.g. I ~ , I " , J(')IlOV 'Ae'JVO.LO 7rfh7rOV(TV, the I 'H 7rpVfhVa. TOV ,1/'llOtoV 0 ELS, I..A~,
ste?"n of the shtp whtch the Athemans seud to l.Jelos (eve?"y yea1). PLAT. Phaed. 58 A. T KT TO KOpos iif3pw, OTav KaK<{) oA.f3os f7r")Tat, satiety begets insolence, wheneve1 p1'ospe1ity follows the wicked. THEOG. 153. 'Ev xpovrp a7rOq)evE TO Tapf:Jor; &v8pw7rotrrw, in time timidity dies out in men. AESCH. Ag. 857.

25. The present denotes merel! the continuance or progress of an action, without reference to 1ts completion. It may, however, be implied by the context that the action is not to be completed, so that the present denotes an attempted or intended action. Especially 38wfht, in the sense of offeT, and 71'E8w, try to persuade, are thus used. E.g. l'\vv 8' fifha ! a-8TKa 7roAAa 3 too~ he oife1's many things. Il. ix.
519. ITd8ovrrt VfhUS f.vavT[a Kat Tol:r; VOfhOtr; Kat T<j) OtKa[rp f"lcprrarr8at, they aTe t1'ying to peTSuade you to vote contrary both to the laws and to justice. I sA E. i. 26. This conative signification is much more common in the imperfect. See 3 6 and the examples.

26. The present is often used with expressions denoting past time, especially 7raAat, in the sense of a pel'fect and a present combined. E.g.
KEI:'Jiov lxJIEVW 1l'aAat, I have been tracking him a long time (and . . \ I " , I sttll conttnue t 't). sOPH. A' 20 . 0, 7rallat rrot AE')'W on TaVTOV J. v I\ cpr}fht Etvat; i.e. have I not lm1~1 "'!In told you (and do I not still repeat) tlwt I call it the same thing? l'LAT. Gorg. 489 C. 8wvs alTw . q)povp&.s f7'Elas fh'JKoS. AE~CH. A g. l. So 1/'0AVV xpovov TOVTO 71'0 tW. So in Latin, iaru dudum loquor.
27. The presents i]Kw, I am come, and oi:xofhat, I am gone, are used in the sense of the perfect, An approaeh to the perfect sense is sometimes founu in such presents as 1)<1!yw, in the sense I am banished, 6..\[U'Kafhat, I arn captwed, vu<w and KpaTw, I am victorious, ?)TTWfha, I am COJM[1W1'ed, aOtKW, I ha-ve been unjust (I am liotKos). So the Epic ZKw and iKavw, with oAAvfhat and sometimes TiKTW in tragedy. E.g. 0 LX cra t Eis aAa oi:av, he is gone to the divine sea, Il. XV, 223.

8EfhtrrToKA0s ~Kw napa rr, I, Themistocles, am conM to you. THUC. i. 137. Tovs aOLKWS cpEvyovTaS OtKa{ws KaT~yayov, they justly re8tored those who were nnjustly banished. PLAT. l\lenex. 242 B. 'IAtov aAtrrKOfhEVOV, after the captnn of Ilium. Tm:c. vi. 2, So aA(TKOfhEJIOV Tov TExws. HDT. i. 85. "07rtu8E T?/s dvotyofhEV")S (){.p")>, beh,:nd the open door. HDT. i. g, Et n<fVTa TQVTa EA11fha{vETO Tol:s oAots, ws aTIETPElfE, T IJ."]f1.0rr8v")s aO!KELj how is Demosthenes to




blame? DEM. xviii. 303. ITvpywv oA.A.vp.hwv Jv vaV<Ttv lf3av, 1 embarked after the towers had been destroyed. EuR. I. T. 1108. ''Hoe -rlKTEL <TE, this woman is thy mother. Id. Ion. 1560.

Present participles are given in some examples here where they illustrate the meaning of the tense.

28. The Greek, like other languages, often uses such presents as I hear, I lemn, I suy, even when their action is finished before the time to which they strictly refer. E.g.
El <J"Ta<J"ut&ovr:nv, IJJ<J"'TrP 7rvv8av6p.e8a, if they (the Sicilians) are in discord, as we leam. THUC. vi. 16. 'E7ri 7!"0AHS, WS eyw aKoii al<J", p.aA.op.iiV levaL JLii'}'aAas. Id. vi. 20.

Nub. 125. El o' OVTOL a7r[a(]"LJJ, ~;ui;s fJ-OVOL f.LEVOVfJ-IiV, but if they (shall) rlepa1t, we alone shall remain. XE!<. Cyr. iv. 5, 24. In Homer /{p.t is usd also as a present ; as oio> o' a<rT~P ei<J"t fJ-ET a<J"Tpa<rL, I!. xxii. 317. So ii. 87, xi. 415; OJ. iv. 401; and often in similes. This is doubtful in Attic; as in 7rpo<rELJLL owl'-a Ko.i (3phas TU <J"ov, AEsCH. Eum. 242, where 7rpo<rHf.U may be 1rp6s + iilp.. See Kriiger and Cla~sen on e7!"[a<J"LV, THUC. iv. 61.

(E ip. t as Future.) . 29. The present Elp.t, I am going, and its compounds, have a future sense. Eip.t thus became a future of, the future EAEV<J"OJLO.L not being in good use in Attic prose. E.g. 2:ev V<rTEpos dtf v1r6 yaZav, I shall go. Il. xviii. 333. E lp.t 1raALV e1r' KeZva, I shall recur to that. PLAT. Phaed. 100 B. "',Q cp[A.', yw p.'Ev ll7rELJLL, <J"vas Kat Ketva cpvA.r.f~wv. Od. xvii. 593. 'AA.A.' d<J"etp.t, <J"ov 8' ov <f>povnw, but I'll go in and not mi'(l.d you. AR.

30. 'The future sense of Elp.t and its compounds extends to the optative, infinitive, and participle in indirect discourse, and often to the participle in other uses (especially when it expresses pur])Ose with ~s). E.g. IIpoEC71"0JJ on, el !'-~ 7rapE<J"OjLE8a <J"V<J"TpaTEVlfOfJ-EVOL, EKiivoL 1} ~/ha> !o LE v, i.e. that they would er-me against us. XEN. Hell. v. 2, 13. See also V. 1, 34, where Elfh'l a7r{OLEV corresponds to Ell'-~ EK7rEfhlfOLV. As tOLfJ-L in this use is equivalent to a future optative, it is naturally rare (128). 'A7rtEvat eVOfJ-L(w OTav (3ovA'JTat, he believed he could depart (a7rELfJ-t) whenever he pleased. THUC. v. 7. So oi>K <f>a<J"av (E<f>YJ) lE vat, XEN. An. i. 3, 1 and 8 ; i. 4, 12: cf. ii. 1, 3, ii. 6, 10. Kai TU 1rvp

ye aD 7rp0<J"LOVTO') TOV fvxpov

a~Tcfl ~ V71"~LEVO.L ~ a7rOAE;;<r8at.

PLAT. Phaed. 103 D. (Ilpo<J"tovTo> is an ordinary present participle : see 31.) Ov yd.p ifonv ~twv, for he did not know that he was to go. AR. Pac. 1182. '0 ll' els Ilep<J"as l w 1rapYjv <rvve<J"KEVa<J"p.evos. XEN. v Cyr. iv. 5, 26. TavT el7rwv avL<rTaJL']V ~. U7rLWV. PLAT. Prot. 335 c (this might come under 31). So avE<J"T<JKYJ ws e~twv, ib. 335 D. ITapE<J"KEva(e.,-o WS ar.LOV<J"a. XEN. Cyr. i. 3, 13. So THUC. vi. 63. 31. In the optative and infinitive not in indirect discourse, and




often in the participle, the same forms of dfLt are used as ordinary presents. E.g. 00ov &v otdcpopov 71'0 w ~ dAA' ~7/'t TUVTOV Eo LV. PLAT. Rep. 360 C. El 71'0A~f1W'i t'o t. Ib. 415 E. See lot in Rep. 490 B, in a peculiar indirect quotation. "On Eot. Id. Tim. 78 C. In XEN. An. i. 3, I, after lEvat as future (30), we have ~(3td~To lE vat and 1Jp~aTo 11'podvat. 'E~uv a-&T<iJ cl(rt6 vn El> Tt1s olK[as o-vyyyver8at DT'{' (3ovAotTO. PLAT. Rep. 360 C. 'A71'oTp11'Df1Vos d~p Ka.i Ota Tov a-wfLaTos li~w lwv. Id. Tim. 79 C. So ANT. v. 78, vi. 45. Jn the subjunctive and imperative there can of course be no special future sense in these verbs.

32. In animated language the present often refers to the future, to express likelihood, intention, or danger. E.g. El aVT7J 1] 71'6Ats- A'I]</:>B!Jo-erat, <!X ETa t 1) 1rao-a LtKcA[a, if this city
shall be captund, all Sicily is (at once) in thei1 JJOssession. THuc. vi. 91. .M vo fLEV i!ws &v i!KatTTOt KaTa 71'6AHs AYjcp8wfLEV; shall we wait until we are each captu?ed, city by city? Id. vi. 77. El o </:>'I]CTLV ovTos, OH~aTw Kat 11'apatTXECT8w, Kdyw KaTa(3avw, and I will talce 'fi1I!J seat. DE11. xix. 32. So d71'6AAvf1at, I am to JJcrish, LYS. xii. 14. For a similar use of the perfect, see 51. (See also 61.)

33. (Historic b'esent.) The present is often used in narration for the aorist, sometimes for the imperfect, to give a more animated statement of past events. This is called the historic present. E.g.
BovA1)v J71'tTexvaTat o1rwc; fL?J d.AttT8eiev 'A8?)vafot, he contrives a plan to prevent the Athenians from assembling. HD1'. i. 63. KeA.evet 7/'EfLfat dvopas d1rOO"TEAAovO"tl! ovv, KO.t 7rpt aDTWV 0 efLtO"TOKMjs Kp!!<f:>a 7TEfL7r<t. TRue. i. 91. !J.apdov Kat ITapVO"d.noos y[yvovTat 1rai:8es o1!o. XEN. AN. i. 1, 1. TowvTa Tov 11'apovTos ~vK' 1)A'P odKVVO"t Tovvap iKAvov ~~YJ'}'OVfLEJ!OV. SoPH. El. 424. The historic present is not found in Homer.


34. The imperfect represents an action as going on m past time ; as i!'Ypa~ov, I was WTiting. 35. The imperfect is thus a present transferred to the past, retaining all the peculiarities of the present which are consistent with the change. Thus it may denote a customary or repeated action, or a series of actions ; or, if it refers to a single action (as it very frequently does), it represents it in its progress rather than as a simple past occurrence (like the




aorist), In narration it dwells on the course of an event instead of merely stating its occurrence. E.g. 'E1n Kp07rOS 'fj TTLK'fj KO.Ta 7TOA!S <fKE!TO, KO.! OV c;;VV'(Ia"UV 'K' ''A ' ' '' ' ,. , ' ' c '
(3ovAn>u6JLVO!, d..\A.' av-rot EKO.a"TO! E1TOA!T1JOVTO Kai f3ovAE-60VTO. 'E1r!8~ 8~ 8ryrnvs (3au..\<va-EV, s -r~v vvv 1r6..\w oiia-av ~vv0K!a"E 1TavTas. Tnoo. ii. 15. (Here the imperfects refer to the state of the country or to customs; the aorists state events, i(3ao-[A.wo-E, became king, ~vv<{!K !a-, collected into one state.) Kat 1rapao-Td8 6 JLEV eJ!thv o 8' eJI8<JI, (36wJI, ~eKpov6v JL<, -r<Aw-rwJITES Jx..\da(ov {>JL<'is o' lycAaT, Kat OVT' dKOVf!V ;)8EA7< OW 1T!a"TEVHJI (3ov..\w8c, they kept on shouting, etc., and you laughed, etc. DEAL xix. 23, '1;' / ,, R ,, ' / \ "'1TE<pWfL'fjJI T! ACf!V TOVTWV WV E!S n)V 1-'0VA'fjV a1T'fJ'Y'YHAa. Ib'd , 1 IT6-rEpov Tavra 7T<fv-ra 1Totwv lj8KH Kat 7rap<a-1T6vfht Kat f!..\vE T1JV tlp;)v'fJV 'IJ ov; in doing all these things was he acting unjustly and breaking the peace, etc.? Id. xviii. 71 ; see also ib. 69. (Compare r1}v dp-IJV'fjV n VU Ta 1TAo'ia A.a(Jc!Jv, of the event, ib. 73.) ITapEABwv J1ri 8ptf.K'fJS Bv(avrovs 'l~[ov UVJL7TOAEJLELJ'. lb. 87. 'YJLE'is yap rav/ E7rpUTTT, Kat TU.VTU. 7TUO"!JI DJL'iv .Y)p<UKEV (of a course of action), Id. xix. 189. 'E1r<tO~ yap <lAEv ''OA.vv8ov <l>[A!1T1Tos, '0..\vJLma E7TOH, ds 8 T~V eva-av 7TUVTUS TOVS T<xvras a-vv~yayEv. lb. 192. Eim r6r' OVK EAEYES 1Tapaxp1)JLa TUVTa ov8' 8t8no-KES ?JJLUSj did you then not tell this at once on the spot, o1. instruct us? lb. 25. The same action (as in the last two examples) couH easily have been mentioned, without reference to its continuance, as a mere event. For the relations of the imperfect to the aorist, see 56.
I "' ' '

. 36. The imperfect, like the present (25), sometimes denotes rtitempted action, being here strictly an im:pe1ji;ct tense. So especially 88ovv and E1TH(}ov. E.g.
(<l>..\t7T7ros) 'A..\6vvryo-ov 88ov, Philip offered Halonne:n/.8 (lit. tried to give it). AESORIN. iii. 83. "EKaaros E1TH8Ev avrov D1roaT~Vat T~v dpx~v, each one tried to pe1suade him to undertake the command. XEN, An. vi. 1, 19. KvJLa ru-rar' d!p6JLVOV, KOTU 8' 1JPEE IT'fjAE[wva, and was about to overpower the son of Pele?ls. IJ. xxi. 327. 'EJLw8ovro 1rap ovK EK8t86vro<; T~V avA~v, he tried to hire the yard of one who 14used to let it. HDT. i. 68. ITJLfavns ES 'Lapot<; xpvo-ov C:,vovro, they sent to Sardis and wanted to buy gold. HDT. i. 69. 'E7TE(j1JJL1)a"E T~S xA.av8os, Kat avr~v 7rpoo-EA8wv ~VETO, he took a fancy (aor.) to the cloak, and tried to buy it. HDT. iii. 139. "A i1Tpaa-a-Ero ovK yevEro, what was attempted did not happen. THUC. vi. 7 4. So 7rpoo-Er(:)H, she wanted to add, AR. Nub. 63.

37. When the present has the force of the perfect (27), the imperfect has regularly .the force of a pluperfect. E.g.

'0 oxAo<; KaTa Bf:av i]KEV, the crowd had come to loolc on. THUG. vi. 'E1rEt <tJxEo V'fJL ITvA.ovo<, ajte1 thou wast gone by ship to Pylas.

Od. xvi. 24.




38. The imperfect sometimes denotes likelilwod, intention, or danger in past time (see 32). E.g. 'E7TEtch) -rti} ifm)8,<T8at d1rwAAv-ro, when he was on the point of ntin through his deceit. ANT. v. 37. Kai -rajl Eevr;<TKE -rEKV', d?TwAAVJ-LYJV
8' f:yw, and my children were about tn die, and I was about to perish. EuR. H. F. 538. 'EKatVOJ-LYJV gtcpEt, I was to be slain. Id. I. T. 27. 39. The imperfect ~v (generally with llpa) may express a fact
which is just recognised as such by the s1Jeaker or writer, having previously been denied, over-looked, or not understood. E.g. "'Q 7T<57Tot, oiJK apa. 7!'UVTa VO~fLOVf<; OVb~ O[KaWt ~<Ta V <i>atqKW1' ~Y+ -rop> 'JO~ J-LE8ovns, i.e. they an not, as I once irnagined. Od. xiii. 209. OvK apa fLOVVOV EYJV lp[owv yvo<;, &AA' hi -yai:av Ei<TL ovw, there is not after all merely one 1ace of discards, but there are two on ea1th. HEs. Op. 11. ''Oo' 1J V apa 0 tvAA.af36Jv J-L, this is then the one who seized ?ne. SoPH. Ph. 978. Ov <TV J-LOVO<; ap' 1J<T8' E7rotf j a?e you not then the only epops (as I thought)? AR. Av. 280. "'Ap ov -roo i) v -r6 8vopov, </) o7rEp '1-ys ~/La>; is not this then the tree to which you wme bringing us? PLAT. Phaedr. 230 A. Other imperfects are rare ; as 1J7rt<T-rw, XEN. Hell. iii. 4, 9.

40. In like manner the imperfect may express something which is the result of a previous discussion, with reference to which the past form is used. This is sometirnes called the philosophic imperfect. E.g. "'I{ V 1J fLDV<TtKlJ av-rt<T-rpo<f>o<; rijs 'YVfLVa<YTLK'/j<;, EL fLEfLVYJ<Tat, music then (as we p1oved) cmTesponds, if you rernembm, to gymnastics. Pr,AT. Hep. 522 A. Kai OLKCUOV 01J </>~<TOJ-LEV avopa Elvat -rrp av-rrp -rp07r<{', 07Tp Kai 1roAts ~ v OtKa[a, and now we shall say that a man is just in the sarne way in which also a state was (shown to be) just. Ib. 441 D. !1vacpBEpovJ-Lv EKEtvo, & T~ fL~V OtKa['{' f3EA-rwv -yt-yv-ro, Tti) 8 dotK'{' &1rwAA.vTo, we shall destToy that 1vhich (as we proved) becomes better by justice and is ru~:ned by injustice. PLA T. Crit. 4 7 D. 41. The Greek sometimes uses an idiom like the English he was the one who did it for he is th" one who did it; as .ryv T~v -yvWfL'Y)V -ra<JTYJV l1ri1v ITEI<Tavopo>, THuc. viii. 68; -r[s ~v a f3ol)e~m.Js Tots Bv(av-rfots Kat <Tw<YaS av-rovs; DEM. xviii. 88.


42. The perfect represents an action as already finished at the present time; as ryrypa~a, I have w1itten (that is, my writing is now finished). 43. The pluperfect represents an action as already finished at a given past time; as E"frypa~Hv, I had written (that is, my writing was fin.ished at some specified past time).
44. The perfect, although it implies the performance of the action




in past time, yet states only that jt stands completed at the present time. This explains why the perfect is classed with the present as a primary tense, that is, as a tense of present time.

45. The perfect and the pluperfect may be expressed by the perfect participle with the present and imperfect of ipl. Here, however, each part of the compound generally retains its own signification, so that this form expresses more fully the continuance of the result of the action of the perfect to the present time, and of that of the plupel'fect to the past time referred to.

IIE7ro~1JKW> ECTnv (or 1jv), he il (or was) in the condition of having done,-he has done (or. had done). 'Ep.ov oi v6p.o~ oD p.6vov d?TyvwKOTS dCTt /1-~ &o~Kll', &_,\,\a Kat f(KA1!KOTES TaVTYJV T~V o[KYJV Aap.f3avw, it is the laws which not only have acqnitted me of injustice, but have commanded me to inflict this punishment. LYs. i. 34. 'Er6A.,ua AeyHv w<; yw ru 7rpayfl d,u2 rovro oopaKws, he daTed to say that I was the one who had done this deed. DEAf. xxi. 104. In DEM. xviii. 23, OVT yap 1J V 7rpCTf3da 7rpOr; ovova Q7THTTaAJ1-EVYJ TO'Tf TWV 'EA.A~vwv means for there was no entbassy then out on a mission to any of the GTeelcs; whereas a7rECTTaAro would have given the meaning no embassy had eveT been sent out (see 8 31 ). This of course does not apply to cases where the compound form is the only one in use, as in the third per:;on plnral of the perfect and pluperfect passive and midtlle of ruute.amlliq uitl verbs. 46. On the other hand, although tl1e simple form very often implies . the continuance of the result of the action down to the present time or to a specified past time, it does so less distinctly than the compound form, and not necessarily (see the last two examples below). E.g. 'Emp.Aws o[ 8wc iSv o1 av8pw1ro~ oovra~ l<aTCTI<E1> aKacnv, the Gods have caTefully pTovided what men need. XEN. Mem. iv. :3, 3. Twv 7rOH)TWV nv> il1ro8~Kas ws XP1J Nv Kara AE Ao I. ?TaCT~ v, some of tlte puets have left ns suggestions how to li'Ve. Isoc. ii. 3. 'AK~Koa p.ev rovvop.a, Jl-VYJ,UOVElJW o' ov, I have heaTd the name, but I clo not ?'emembeT it. PLAT. Theaet. 144 B. "A (TO~ TVXYJ KEXPYJI<, ra.vr' a<f>EATO, Fortune has taken back what ohe has lent you. MEN. Fr. 598.

47. ''Exw with the aorist and sometimes the perfect participle may form a ])eriphrastic j)Crfect (831). In tragedy and in Herodotus this is often fully equivalent to our perfect with have; elsewhere, especially in Attic prose, the participle aml :Jxw are more or less distinct in their force. Still, this is the beginning of the modern perfect. E.g. Ilo[cp (Tl!J' EflY(P rovr' d7rHA1)o-a<; EXH>; have yo1~ made this threat? SoPH. 0. c. 817. Tuv ,u~v 7rporCTa<;, TUV o' d.nJl-aCTa<; EX~; Id. Ant. 22; see ib. 32. ' 7rpayos aCTKo7rov <!xo 7r<pavas. Id. Aj. 21. H15Saro ya.p ravr', ovo 1rw ,\~~avr' ffx, i.e. the story has not yet ceased to be told. Id. 0. T. 731; see Tr. 37, Tapf3~o-ac:l i!xw. "Os o-<f> vvv d.T<.,uaCTa> EXH Eun. Med. 33; see ib. 90. "Apws T




fW'ipav JLera.\af3wv iixE nvd. Id. Bacch. 302. 2:ov 8avJLd.rras 768E. SoPH. Ph. 1362; so PLAT. Phaedr. 257 C (in poetic language). Oiri JLOL {3Ej:3ovAEvKws iixo. SorH. 0. T. 701 (after rrTfJrras i!xas " I o ~I in 699). "0 rrnr; y> XE JLOV 'C .,ap7rarrar; TO 1rawwv, whoever 7 w,~


snatched away (though here EX! may mean lceeps). AR. Th. 706. 'E y K.\1) rrarr' lixa .,a, rrtT[a. Id. Eccl. 355. 'Y1r'Ep TWV 'E,\.\1)vwv Tov> rrv oovAu)rras i!xn>, Le. whom you hold in slavery or whom you 'A -1 > \ l d have ensa.v~. H DT. : "'' , fLp~rEpo;v fLE rovrwv a,7r~KA"J~O'~<;
nM I I I

i!xa>. Id. 1. 37; so J. 41. AAaCovt E7rtrpEfavn> 17fLWS avrov<; exofLEV, we have entrusted ourselves, etc. Id. vi. 12. IloAAa XP~0ara exofLEV dv"Jp7raK6rl!<;. XE~. An. i. 3, 14 (here f.xofLEV expresses possession). See THUC. i. GS; DEM. ix. 12, xxvii. 17. The beginning of this usage appears in l:!Es. Op. 42 : 1 Kpv'faVTE<; yap EXOVO' Wi {3 tov av pw1rotrrt. I ' '' ' "


48. Eixov or EO'xov with the participle may form a periphrastic

pluperfect in tl~e same way (4 7). E.g. "Ov y' ElXOJI 1}8,/ xpovwv JK{3E{3A'I)I<OTES See HDT. i. 28, 73, and 75; XEN. An. iv. 7, 1. SorH. Ph. 600.

49. (a) The perfect of many verbs has the signification of a present, w hi eh may usually he explain eel by the peculiar meaning of the verbs. Thus Bvr/O'Kav, to die, TE8V?, to be dead; KaA Etv, to call, t<EKA-Ij0'8at, to be called or named; yyvE0'8at, to become,, to be; fLLfLVllO'Kav, to 1ernind, JLEfLV~0'8at, to remember/ EloJiat, to know; tO'n1Fa.t, to place, eO'ravat, to stand. So {3Ef3"JKf.vat, to stand; JyvwKf.vat, to bww , ?JfL<Pt0'8at, to wear/ KEKr~rr8a,, to possess,- 7rE7rot8vaL, to trust, 7rE<j;vKf.Fat, to be (by nature) ; etc. (b) The pluperfect of such verbs has the signification of the imperfect; as oioa, I know, 1JOELv, I knew.
50. In epistles, the perfect and aorist are sometimes used where we might expect the present, the writer transferrillg himself to the time of the reader. E.g. 'A7rEO'TaAKri (J'Oi rovOE TOV ,\6yov, I send yon this speech. Isoc. i. 2. MEr' 'Apraf30.Cov, oF <TOt E7rEfLfa, 1rpaO'O'E. THl;c. i. 129. (Here 6v E7rEfLfa refers to the man who was to carry the letter.) So scTipsi and rnisi in Latin.

51. The j)erfect sometimes refers to the future, to denote certainty or likelihood that an action will immediately take place, in a sense similar to that of the present (32), but with more emphasis, as the , change in time is greater. E.g. "!J.O'r' d fLE ,-6~wv JyKpar?JS a1rr8~rrErat, &AwAa, I shall JJIYrish at once. SOPH. Ph. 7 5. K&JI rovro ViKWJLEV, 7rd.v8' 'JJL'i:V 71'71'0 ["JTCH. XEN. An. i. S, 12. So peTii in Latin.

52. In a somewhat similar sense (51), the pluperfect may express the immediate or sudden occurrence of a past action. This occurs especially in Homer and Herodotus. E.g.




OilS' &:TrterJcm' fl-ue~) 'ABryvaf,r; ~ 3' 00Avf1-7rfw3< (3<{3-.]KHv, and she was gone to Ulympus. 11. i. 221. 'H p..~v 8ap..(31jawra 1nfA.w oiK6v3e {3ef3lJKHV. Od. i. 360. T2,v o' EAt71'E lfvxiJ, KaTa 3' 6,P8afi.p..wv KExvr' dxAvs. Il. v. 696. "AA.A.ot 3~ ~yep..6vas iixovres ~PJ-t~aTo E7rt TO 1p6v, i.e. they were on their way (at once). HDT. viii. 35 ; see ix. 61. For the gnomic perfect, see 154 and 155.


53. The aorist indicative expresses the simple occurrence of an action in past time; as ~rypa"fra, I wrote. 54. This fundamental idea of simple occurrence remains the essential
characteristic of the aorist through all the dependent moods, however indefinite they may be in regard to time. The aorist takes its name (a6pt<TTos, unlimited\ unqualified) from its thus denoting merely the occurrence of an action, without any of the limitations (opot) as to completion, continuance, repetition, etc., which belong to other tenses. It corresponds to the ordinary preterite (e.g. did, went, said) in English, whereas the Greek imperfect corresponds generally to the forms I was doing, etc. Thus, 71' o [o TovTo is he was doing this or he did this habitually ; 71' e 1r o ["7 Ke rovro is he has alnady done this ; J1r E 1r o t-.] KEt rovro is he had already (at some past time) done this; but E71'o ['I] erE -rovro is simply he did th'is, without qualification of any kind.

55. The aorist of verbs which denote a stctte or condition generally expresses the entrance into that state or condition. E.g.
Ba<TtAEVw, I am lcing, J{3r:tcrA.evr.ra, I became lcing; flpxw, I hold office, took office; 1rAovrw, J71'fi.ol'!rrJfra, I became rich. Tz7 aA'f}BEfrc<Tvv{lKH Kat ovbE71'W Kat r1jp..Epov d7roAeAot7rev aAAa 1rapa (wvros 1\p..oKpdTovs lKEiv<p r.rvv<{>Krycr.:, she was his wife in good faith, and has not yet even to this day been div01ced ; but she went to live with him front Timocmtes while 1'. was stillliving. DllM. xxx. 33.
ijp~a, I

56. The aorist is distinguished from the imperfect by expressing only the occurrent:e of an action or the entrance into a state or condition, while the imperfect properly represents an action or state as going on or as repeated. t)ee the examples of the imperfect and aorist in 35, and compare crvv<~J<Et and crvv~K'f}rTE in DEM. xxx. 33 (in 55). The aorist is therefore more common in rapid narration, the imperfect in detailed description. It must be remembered that the same event may be looked upon from different points of view by the same person; thus in DEM. xviii. 71 and 73 (quoted in 35) EAVE T~V elp'iJV'IJI' and T1JV EipfJv'IJV eAvrrE refer to the same thing, once as an act in progress, and once as a fact accomplished. No amount of duration in an act, therefore,




can make the aorist an improper form to express it, provided it is stated as a single past event viewed as a whole. Thus ~f3arr A.wcH oeKa lrry (see HDT. ii. 15 7) means he had a reign of ten years, (which is viewed as a single past event), while Jf3aaD..wE oeKa ~T'l} might refer to the same reign in the sense he was reigning during ten years. The aorist may refer even to a series of repetitions ; but it takes them collectively as a whole, while the imperfect would take them separately as individuals. See DEM. xviii 80, Jl-ETa rai!ra o~ roi!s d1roar6Aovs a1ravras d7TearELAa, and afterwa1ds I sent out all the naval armaments/ and xviii. 60, a Jl'~V 1rp'O roi! 7TOA,TEVEa8a, Kat Sryj1-'l}yopE'iv f-Jl'~ 7TpoVAa(3E Kat KaraxE ip[A,7T7Tos, the (succession of) advantages which Philip secured during the peTiod before I enteTed public life, emphatically opposed (as a whole) to Philip's many failures after that time, which are mentioned in a 0~ Kat 0UKWAV8ry. If the Orator had wished to dwell on the number of the advantages or failures, or on their duration, he could have used the imperfect. See the last example under 35. 57. Since the same event may thus be stated by the aorist or the imperfect according to the writer's point of view, it is natural that it should occasionally be a matter of indifference which form is used, especially when the action is of such a nature that it is not important to distinguish its duration from its occurrence. For example, this distinction can seldom be important in such expressions as he said, he commanded; and we find eAEyov and ~K!Awov in the historians where no idea of duration can have been in mind. See oi o' JK AEV61' TE E7T,eva,, KO-~ 7TOpEA86vTE<; OL 'A8ryva'i:ot eAEyov TOUtOE, THUC. i. 72, followed, at the end of the speech in 79, by rowi!ra OE oi 'A8ryva/;o~ d1rov and 'Apx8aJ1-0> lAE~E ro,aSE. In such cases as the following (eited with others by Kriiger) it was not important to the narrative whether the idea of duration was included in the expression or not: f3aAAEro and (3aAEro, Il. ii. 43 and 45; 80KEV and r8a, xxiii. 65:3 and 656; SwKE and 0/lov, vii. 303 and 305; >u7TEV and AEZ7TE, ii. 106 and 107 ; compare also Jl'[arvAAov with e1mpav, CJ"Trrryaav, and epvcravro, i. 46 5 and 4 6 G. In all these cases the fundamental distinction of the tenses, which was inherent in 1he form, remained; only it happeJ}(:'d that either of the two distinct forms , expressed the meaning which was here needed equally well. It must not be thought, from these occasional examples, that the Greeks of any period were not fully alive to the distinction of the two tenses and could not use it with skill and nicety. But the Greeks, like other workmen, did not care to use their finest tools on every occasion ; and it is often 11ecessary to remember this if we would avoid hair-splitting,




58. The aorist, expressing simply a past occurrence, is sometimes used where we slwuld expect a llerfect or pluperfect, the action being merely referred to the past without the more exact specification which these tenses would give. E.g. Twv olKETWV o-&8va KaTAt7rev, dAA' d7ral'Ta 7rE7rpaKev, he (has) left none of the servants, but has sold everything. AESOHIN. i. 99. 'ETpct1TOVTO J, TdV ITdvopfLOV, oBev7rEp dv?)'}'ct'}'OVTO, they turned towaTds PanoTmus, whence they (had) set sail. Tauc. ii. 92. Kvpov 8 fLETa1TEfliTI.ETat d1ro Try> dpxfis i)s aBTov a-a-rpct7r?)l' E7ro [1} ere v, j1om the dominion of which he (had once) made him satmp. XEN. An. i. 1, 2. 59. The aorist is generally used with J1rd or E1Tet8,], aftm that, the aorist with the particle being equivalent to our pluperfect. So after i!ws and 1rpv, until. . E.g. 'E1rnory JTeAn'IT?JtTE 6.ape'ios Kai KaTEtTTYJ 'Apra~p~?)'>, ajte1 DaTius (had) died and ATtaxeTxes had become established. XEN. An. i. 1, 3. Ov 1rpoa-Bev J~eve'}'Kei:v ETDAJL?Ja-av 1rpils ?)JLfi> 7ToAefLOl' 1rp!v Tovs a-TpaT?)yovs TJJLWV a-vv Aaf3 ov, they did not dare to bring warlt)?On us until they (had) seized our generals. Ib. iii. 2, 29. But the plnpedcct may still be used after E1re or J7Tet8,), to give additional emphasis to the doubl:r past action; as in DEM. xviii. 42, E7rHO?J E~>J7rUT?)uBe JLEV {JLE'i~, f.~')'lr<LT?)vTo 8~ a' <PwKEts Kat dvvp>JvTo a' 7rOAHs, T f.yveTo; So in Latin we have generally postquam venit, but occasioually postquam venemt. 60. The aorist is sometimes used colloquially by the poets (especially the dramatists), when a sudden action, which is just taking pla.ce, is spoken of as if it had already hap1Jened. E.g. 'E1r7l vea-' (p'}'OV Kaf 1rpovowv 1)1' Bov, I must apJ.nove you.r act, etc. SoPH. Aj. 536. ''Ha-B?)l' d1rHAaZs, E'}'EAarra lf;oAo~<DfL7r[ats, I am amused uy you1 thTeats, I cannot help laughing, etc. AR. Eq. 696.
61. The aorist sometimes refers vividly to the future, like the present (32) or perfect (51); as d7rwADJL1JV d' fL Adrfw;, I pmish if you, leave rne. Eua.Alc. 386: so Med. 78. See also wAeTo, Il.ix.413 ancl415. 62. In questions with ~r oB, expressing surprise that sometl1ing is uot already done, and implying an exhortation to do it, the aorist is sometimes used strangely like a future. E.g. T o~v oB 8 t?)y~crw 1JJLZv n)v ~vvova-av; why then don't ymt tell us about the meeting? PLAT. Prot. 310 A. T oiiv oB Kat IIp6otKOV Kat 'I?T?dav f.KaAa-aJLEV; why then don't we call Prodicus and Hippias too ? lb. 31 7 D. So T o~v oB . . . Ea-KElfW; Id. So ph. 251 E. See also SoPH. 0. T. 1003. For the gnomic aorist see 154.


63. The future denotes that an action is to take place




in time to come; as rypatw, I shall write or I shall be writ1>ng, sometimes I will write; 7Tcucrat, he will suffer, sometimes he shall snffer.
64. In indirect discourse and in all final constructions the future expresses time future relatively to the leading verb. See 22.

65. The future may represent an action in its duration its mere occurrence, or its inception ; as ~w, I shall have, or I ;hall obtain/ Tol',.,-o Mo-w, I shall give this; lip~w, I shall rule, or I shall obtain power (cf. 55). E.g.
ITpay}haTdiovTaL o1rws lip~ova-Lv, they take trouble to gain power XEN. Rep. Lac. xiv. 5. "'Ap' ov (ow,LpETEov) oZnvEs lip~ova-{v 'TE Kai ({ p ~ o v 'Tat ; must we not distinguish between those who a1e to rule and those who a1e to be ruled? PLAT. Rep. 412 B. IIi) o-Taa-Lao-ov<nv ol 7r{Kovpot Kat ol flpxovns; how will they fall into faction? Ib 545 D (see below, o1rws o~ 1rpWTOI1 a-Taa-Ls Eft7rECTE).

66. The future may be used in a gqwmic sense, denoting that something will always happen when an occasion offers. E.g.
'Av~p o <j>Evywv Kat 1raALv JkUX~a-eTaL. l'.fEN. Mon. 45. fights and runs away may turn and fight day."

"He that

67. The future is sometimes used to express what will hereafter be proved or be recognised as a truth. Compare the use of the imperfect in 40. E.g.
he will prove to

<PtA6a-o<f>os 1Jft01 a-T a L JkEAAwv KaA6s dya()6s Ea-Ea-()at rpvAa~, ~e a philosopher. PLAT. Rep. 376 0.

68. The future is sometimes used in questions of doubt, where the subjunctive is more common (287). E.g. T[ O~Ta opw0Ev; 0'YJ'TEp' ?j rp ovEVIJ00Eil; what shall we do? shall
we kill our motlted EuR. El. 967: so Ion. 758. IIoL' TPEtfo}hcJ.L. whithenhall I turn? Id. Hipp. 1066. Eh yw ~Jov <jodtJo}hat Art, T' 'i' / . A C . ;312. h L OVJl 'JT"OL'Y)IJOftEV j 'JT"O'TEpov ELS 'T'Y)V 'JT"OI\LV 7rUVTo.s 'TOV'TOVS 7rapaoe~6JkE()a; what then shall we do? Are we to receive all these into the state ? PLAT. Rep. 397 D.
I ). \ ,, / }

69. The second person of the future may express a concession or permission; and it often expresses a command, like the imperative. E.g.
IIp?.s Tavra 7rpU~HS oiov &v eavs, you may act as you pleast SoPH. 0. c. 956. IIavTWS 0~ 'TOV'TO opatJELS, but by all means do this: AR. Nub. 1352. So in the common imprecations, d7roAEtiJ0e, oi}hw~EIJ()E, may you pe1ish, etc. XELpt 8' ov tf;a vo-Hs 7ro-rf.. Ena. Med. 1320. Compare the Latin facies ut sciam, let rne know; abibis, depart. 70. In a few instances the future indicative with 0~ expresses a prohibition, like the imperative or subjunctive with 0~ (259). E.g. TaV'T'Y)V, liv JkOL xp~a-()E 1JVftf3ovAr.p, rpvAa~ETE 'T~V 'lr[IJ'TLv 7rpb>




Tovrov rov 8piKa, Kat p.~ f3ovA~tJ"HT8 d8vat, K.T.A., if you follow rny advice, hold fast to this secU1ity (69), and do not wish to know, etc. DgM. xxiii. 117. 'Ectv 8~ EfJ q>poFljn, Kat J!VVL TOVTO <PavEpoJI1TOt~IJ"T, Kat fJ/I]DEp.[aJ! avroZ> liDHaJ! OcbiJ"ET. Lrs. xxix. 13. AEJ!OV d.DtK~ tJ"Et> 1'-')0E'lTOT Katp(w A.a(JcbJ!. MEN. Mon. 397. So probably ov tJ"tya; /~1JO~V TWJ!O' epEL!> KaTa 1TT6Atv, silence I say nothing of all this in the city. AEsca. Sept. 250. (See 279.)

71. The future sometimes denotes a present intention, expectation, or necessity that something shall be done, in which sense the periphrastic form with p.f.A.A.w (73) is more common.

T l:na<PpovtJ"t rwu f_~ dJ!ayw']> KaKo1Ta8ofwrwF, Er YE 1THl''JtJ"OVIJ"t Kat 8tf~IJ"OVIJ"t Kd /Hy~&ot>IJ"t Ka2 d.ypv1TJ'{JtJ"OVIJ"t; if they aTe to endun hunge1 and thint, etc. XEN. l\Iem. ii. 1, 17. (Here El p.f:A.A.ovtJ"t 1THJ'ljv Kat ou{ljF, etc., would be more common, as in the last example under 7 3.) Aip 1T AljKTpoJI, El p.a X Er, mise youT spur, if you are going to fight. AR. Av. 759, The distinction between this and the ordinary future (63) is important in conditional sentences (see 407).
72. A still more emphatic reference to a present intention is found in the question r Ae~Et>; what do you mean to say? often found in tragedy; as 6\p.ot, r{ A.f: ~Et>; ?j ycip yyvdrn 1rov; Eun. Hec. 1124. So Hec. 511, 712; Hipp. 353; Ion. 1113; SOPH. Ph. 1233. For the future in 11rota~is, see 44 7 and 407 ; in relative clauses expressing a purpose, 565; wit.h &1', 196; with ov 1'-'J, 294-301.

73. (M~:\A,w with tkc Infinitive.) A periphrastic future is formed by fLEA-'Aw aml the present or future (seldom the aorist) infinitive. This form sometimes denotes mere futurity, and sometimes intention, expectation, or necessity. E.g. J\HA.AH rovro 1T(HtTnw (or 1rpa~w'), he is abott to do this, or he intends
to do this. So in Latin, jacturus est for faciet. Mf:A.A.w 1',flJ1s ot8a~HJI Mhv J'-0' 'l owf3oA1J yyoJ!. PLAT. Ap. 21 B. OvKOVJ! OE'iJrTEt TOV TOtoiiTov TtvO~ de2 i11urrci:rov, el JL~AAEL ~ 7roAt.TEa cr0(Ecr8at; if the constitution is to be pTeserved. PLAT. Rep. 412 A. (See 71.)

74. Although tl1e present and tl1e future infinitive were preferred with p.f.A.A.w (7 3), the aorist was still used by some writers, as by Euripides. See AE~CH. Prom. 625 (fLEAAw 1Ta~O,r:v); Eun. Ion. 80 (p.f.A.A.w TVXEZv), 760 (8a.VEtJI p.EAAw), El. 17 (fLEAAOJ!Ta 8aJIEtJ1), Phoen. 300 (p.EAAH> OtyEZJ!) ;-where the metre allows no change.
75. The future infinitive with p.EAAw forms the only regular exception to the general principle which restricts the use of the future infinitive to indirect discourse (see 86 ; 112). 76. The imperfect (seldom the aorist) of p.f:AA.w with the infinitive expres:;es past intention, expectation, or necessity. E.g.




KvKAwtj;, oflK ap gp.AA') dvaAKtOO') dvopo> -ra[prn>> E0fLVUt Jv yAa<{>vpri}, so you wen not aftw all to eat, etc. (cf. 39). Od. ix. 475. See Il. ii. 36. "EtuAA6v a-' lipa ktv1)a-w iyw, I thought I .<hould start you off. AR. Nub. 1301. 'E1Tta--raTlJV Aaf3E'i:v, p.EAAV 'EfLeAAYJa-av 4 vTw KaAw TE Kdya.8r~ 7rat~a-v. PLAT. Ap. 20 A. ~p.f3aUm. THee. i. 134.



77. The future perfect denotes that an action will be already finished at Rome future time. It 1s thus a perfect transferred to the future. E.g.
Kat fLE aJ' ~EAy~vs, ovK dxBa-8~a-ofLat a-ot, &>.>.a fL-yunos dEpyETYJS 1Tap' EJLOL d V arE y p a if; u, you will have been emolled as my greatest benefactor. PLAT. Gorg. 506 0. ''Hv o fLry Y~V1)Tat, fLUT1)JI EfLOL KEKAava-ETat, U"V 8' iyxavwJ' TE()v,)~Et>, I shall then have had my whippings joT nothing, and you will have died g1inning. AR. Nub. 143 5.

78. The future perfect often denotes the continuance of an action, or the permanence of its results, in future time. E.g.

1f er a,, pou;m,
THUC. ii. 64.



s d8wv Tofs E7rtytyvofLEl'OtS fLVfJfL') KaT a AE Ad-

the nwlnoTy of which will be left to ou1 posterity jo1 eve1. (Compare 105.)

79. The future perfect sometimes denotes certainty_or likelihood that an aetion will immediately take place, whid1 idea is still more vividly expressed by the perfect (51). E.g. El. 0~ 1TapAf)o)v [.,; OU'TtU'O;-;V OVJIO.tTO otOJ.~at, 0 r.apwv <{>6f3o>
AEA.VU"ETat, all the present feaT will be at once dispelled. DE~!. xiv. 2. (Here the inferior Mss. have AEAvTat, which would be like oAwAa, quoted in 51.) 'Ppa(E, J<at 1Te1Tpa~eTa.t, speak, and it .shall be WJ ~ooner suid thun done. An. Plut. 1027. E1\8is- 'Apwias dcpEa-n)~, WU"TE <{>Aos ~p.'iv ovoets Ae'Aetj;eTa.t. XEN. An. ii. 4, 5.

80. The future perfect can he expressed hy the perfect participle and f.rro/;,at. In the active voice this is the only form in use, except in a few cases (chiefly EU"T1j~w and Te8v,)~w). E.g.
''Av TavT' el8wfLev, Kat Ta ooJITa Ja-6fLe8a yFwK6TE> 1w.l Aoywv fLO.Ta[wv ct7r1)AAay/Lvot, we shall have al1eady 1esolved to do our duty and shall /wve been j1eed j1om vain 1epoTts. DEM. iv. 50. (See 45 and 831.)

81. A similar circumlocution with the aorist participle ana EU'OfLat is sometimes found, especially in the poets. E.g. Ov U'tw1Tf]U"as f.U"tt.; SoP H. o. T. 1146., U"tt. SorH. o. 0. 816. (See 47 and 831.)
82. \Vhen the perfect is used in the sense of a 1)resent (49), the




future perfect is its regular future; as KEKA~rrof"a, f"EfLV~fTOf"a, dcpEO"T~~w, I shall be named, I shall1emember, I shall withdraw, etc. 83. In many other verbs, the future perfect differs very slightly, if at all, from an ordinary future. Thus is the regular future passive of 7rt7rpdrrKw. Still, where there is another future, the future perfect is generally more emphatic. 84. It must be remembered that, in most caseR in which the Latin or the English would use a future perfect in a dependent clause, the Greek uses an aorist or even a perfect subjunctive. (See 90 and 103, with the examples.)


85. Tl1e distinctions of time which mark the various tenses in the indicative are retained when the optative and infinitive represent the indicative in indirect discourse, and usually in the participle. But in other constructions these distinctions of time disappear in the dependent moods, and the tenses here differ only in their other character of denoting the continuance, the completion, or simply the occurrence of an action (20). The infinitive with av is not included in this statement (see, Chap. III.) The tenses in these two uses must, therefore, be discussed separately.

86. In the subj'unctive and imperative, and also in the optative and infinitive not in indirect discourse (666; 684), the tenses chiefly used are the present and the aorist. The perfect is used here only when the completion of the action is to be emphasized (see 102-110). For the occasional future, see 111-113; 130-132.

87. The present and aorist here differ only in this, that the present expresses an action in its du,ndion, that is, as going on or 1epeated, while the aorist expresses simply its occu?'?'ence, the time of both tenses being otherwjse precisely the same. E.g.
'Ectv 1r o 'ii 'TOvTo, if he shall be doing this, or if he shall do this (habitually); Jav 7rO t~rrv TOvTo, (simply) if he shall do this; El 1r o to { T) TOVTo, if he should be doing this, or if he should do this (habitu-




ally) . Et 71"0 t~<THE 'TOV'To, if he should do this j 71"0 [o 'TOV'To, do this > ' h' ' \ \ (habitually) j 71"0t'Y}<TOV 'TOV'TO, d0 tU. 0'1 V'TW VtK'Y}<TatjU 'T> E")'W Kat v 0 1 ( o [!" 'YJ v a-ocp6s, _on this condition 1nay I gain the vi;t01y (aor-) and u be considered (pres.) wue. AR. Nub. 520. BovAE'Tat 'TOV'TO 71"0tHv, he wishes to do this (habitually); f3ovAE'Tat 'TOV'To 71"0 tqtrat, (simply) he For other examples see below. 1cishes to do this. This is a distinction entirely uuknown to the Latin, which has (for example) only one form, si faciat, corresponding to El 1r0 w ['I) and El 7rO t?]a-oEv, and only facere to correspond to both 7rOtEi:v and. 1rotqa-at (as used above). 88. It is sometimes difficult here, as in the corresponding case of the imperfect and. the aorist indicative (56 ; 5 i), to see any decisive reason for preferring one tense to the other ; and it can har<lly be doubted that the Greeks occasionally failed to make use of this, as well as of other fine distinctions, when either form would express the required sense equally well, although they always had the distinction J'eady for use when it was needed. Compare the present and the aorist subjunctive and optative in the following examples : 'EUv yelp -r O"E !f;a vW KaKOv 7rE7rOiYJKW~, OJLoAoyW d8tKEv Ed.v

fLEvTot f1-'YJ8f.v ~a [ VWfLa



p:qSE (3ovA~fJE~r;, oV


. trv Of"o.Aoy,)a-w; f"'YJOEV {nr' fpov dotKE"Ur&at j if I shall appear (aor.) to have done you any qcrong, and if I shall appear (p1es.) to have done you no wrong. XEN. Cyr. v. 5, 13. El JLEV yap 1rpoa-8~at'TO <PwKas<TVJLf"axovs .. Et o JL'l 7rpoa-8xot'TO, K.'T.A. DEM. xix. 318. Er 'TtvE> r.oAAWJ' &avaTwv 1)a-av ainot, (Zva) 7ral'TWV 'TOVTWV OEKa7rAaa-[asdA:y'Y}86vas 1'mep EKaO"TOV KOJL{<TatVTO, Ka! aV E[ TLVES' EiJEp'}'E<TLaS' Ei<p'}'E'T1)Kb'TE<; ElEJ', (Zva) KaTa 'TGI.BTCI. T1JV d~{av K o0Co tVTo, if any had ca1tsed many deaths, that they might 1eceive (aor.) suffming for all these, tenfold fm each; and again, if they had done kind se1vices to any, that they might in like manner receive (pres.) theiT due Tewa.rd. PLAT. Rep. 615 B. In the last example, it is olJVious that the change from KOJLfa-awTo to KOJ"(otvTo is connected with the change from El ~a- a v to El cV<fJ'}'ETYJKb'TE> ElEv; but it is questionable whether the latter thange is the cause or the effect, and it is also quite as hard to see the reason for this change in the protasis, wl1en both conditions are equally general, as for that in the final clause. Probably no two scholars would agree in the reasons which they might assign for the use of the tenses in these examples. It is certain, however, that either present or aorist would express the meaning equally well in all these cases.

Subjunctive and Imperative.

89. The present and aorist subjunctive and imperative are always future, except that in general conditions (462; 53 2) .:;he subjunctive is. general in its time. In all final constructions the subjunctive is future relatively to the




leading verb. The following examples will show the distinction of the two tenses : IIHeti>fLeea 1ravns <f>e-Dywt-HY uvv vryvu2 <J>Dvqv Js 1ra-rptoa yaZav, let us all be 1Jersuaded; let us fly, etc. Il. ii. 139. T q>w; -r 0p w; what shall I say ? what shall I do ? ITws ovv 7rpt TOVTWV 7r 0 LwfLEV; how then shall we act about this? PLAT. Phil. 6:3 A. 'A vu A.o y L<TW fLE ea 'TU Wfl-OAoywd:va ~fLtY, let us enumemte the points which have been conceded by us. PLAT. Prot. 332 D. .M?)OEv <J>of3YJ efjs, feanwt. But f'-ryov <f>o(3ov, be not timid. T 7roo]uw; what shall I do (in this case)? But ,.[ 7rOLW; what shall I do (generally)? Oil!'-~ TovTo Er7rYJS, you shall not say this. Ov !"'~ yevYJTat, it will not happen. So in the Homeric ovo fowfLaL, noT shall I ever see (6). . ''Av OE 'TLS d Ve L<T'T~Ta t., 7iELpwr6jLEea XELpovueaL, but if any one shall stand opposed to us, we will tTy to subdne him. XEN. An. vii. 3, 11. Kav 7r6AEfLOS i}, ws av E7r' O.A.Aov l X w fLEV <TTPUTElJE<TeaL, <TOV TE Kat 'TWV <TWV a<f>e~6fLeea, and if theTe shall be waT, 80 long as 'We shall be able, etc. XEN. Hell. iv. 1, 38. 'AAX if av ')'L')'VW<TI<W (3EA.ncr-ra., but I will speak as I shall think best. THUC. vi. 9. av f3 0 V A. Yl 7rotfJcrau8a.t q,[A.ov>, dya86v n A.eye 1rept ailrwv 1rp0> rovs d7iayyEA.A.ovras, whomsoeve1 you. shall wish, etc. Iwc. i. 33. "A1ra> A.6yo>, av d1rfj -ril 7ipayfl-a-ra, jLcfrau)v n <f>averat Ka2 t<ev6JI, all speech, if (whmeve1) deeds a1e wanting, appeaTs vain and empty. DE)L ii. 12. }:;IJfLfhUXElY TOVTOL<; JeAoVITLl' a7iUJI'TE'>, oils aJI 6p/o(J'L 7iapE<TI<EVU(TfLE vov<;, all are willing to be allied to those whom they s~e 1nepared. DEM. j.v. 6. 'its &1' d'1r w 7iELewpAJ.a, let 1S obey as I shall direct. Il. ix. 7 04. "Hv yyvs EA. er; eaJ!aTos, OVOEL<; f3ovAenu el']l<TI<Hl', if death comes neaT (the rnmnent that denth comes near), no one wants to die. EuR. Ale. 671. "Hv T1JV elpfJvqJ! 7i0'1)CTWfl-<()a, p.e-ril 1roAA.ijs dcr<f>aA.etas -r0v ?TOAtv olt<fj<Tof.I.<JI, if we (shall) rnalce the peace, etc. Isoc. viii. 20. "Ov p..ev av roll dyJ!wTa. (se. 6 /(VWJI), xaA.oralVE'' QV o' av ')'Vti>ptp.ov (se. LOIJ), au7icf(erat, i.e. whomsoeve1 the dog sees (at any time). Pr.AT. Rep. 3 7 6 A. LlOt<EL p.o< KUTUKUV(T(LL rils>, l'Jia fh'l TU (<vyrJ 'lflWV crrpa'T'IJYiJ, dA.A.a 7iOp<VWfLea 07i]) av T]J crrpan~ ITVfL<f>Ep)), it seems good to me to burn the wagons, that our /1easts of burden rnay not be anT generals, and that we may go on whithersoever it may be best fa?' the anny. XEN. An. iii. 2, 27. Kai yap f3acrtA.e:Us a1pE'i:rat, oilx Z;a avTov Ka'Aws E7iLfJ-EA~TaL, d'J...A.' Zva Kat o1 f.A.6p.1'0L oi avr6v Ell ?TPUTTW<TL. XEN. Mem. iii. 2, 3. ilf.ooum ft'l E7itAaB01'-eBa T0> otKaoe 6oov, I fea? lest we rnay forget the road home. XEN. An. iii. 2, 25. LltaJ!oe'i:ra<]_, A.vcrat, WS ""~ ow(3ijn d.\A.' a7iOA1J<J>Bij-re, i.e. he intends to dest?oy the bridge, that you may not pass ovm hut be caught. lb. ii. 4, 17. iP<vy<, beg one; )(at.p6vTWY, let them 1ejoice; j.L1J J!Oft{(ET~. do not believe. El1r p.ot, tell me; B6re f'-OL rovro, give me this. 'i.q,ev06v'l}v rfs fLOL B6rw, let some one give me a sling. AR. Av. 1187.




90. ,When the aorist subjunctive depends on E7T"<t0riv (or Jmiv, J1r0v), after that, it is refened by this meaning of the particle to time preceding the action of the leading verb, so that E1T"<tod.v Towo



means after I (shall) have seen this, I will come/ and

E1T"<t0UV TOVTO tOw, a1T"~PXOfhat, after I have seen this, I (always) depaTl. In such cases it may be translated by our future perfect

when the leading verb is future, and by our perfect when the leading verb denotes a general truth and is translated by the present. As the subjunctive here can never depend upon a verb of simply pTesent time, it can never refer to time absolutely past/ nnd we use the perfect indicative in translating such an aorist after a verb expressing a general truth, merely because we use the present in translating the leading verb, although this is properly not present hut general in its time. In like manner, after lws, 7rp{v, and other particles signifying until, before that, and even after the relative pronoun or eav, the aorist subjunctive may be translated by our future perfect or perfect, when the context shows that it refers to time preceding that of the leading verb. E.g.
Xp'J o, 07"(1,]1 fi>EV T Jh)cr 8e TOll<; v6fi>OV<;, 01T"Oto{ TtVE<; elcrt U'K01T"Eiv, E1T"Et00.11 o 8~cr8e, cpvActTTELJI Kat XP~cr8at, while you aTe enacting laws, you must look to see of uhat lcind they a1e; but after you have enacted them, yon 1nust guard and use thern, DEllf. xxi. 34. (Here the present n8~cr8e with bTav, while, refers to an action continuing through the time of the leadiug verh ; but 8~cr8e with E1T"Hoav, after that, refeJs to time past relatively to the leading verb.) TavTa, E7T"HOO.l' 7rEp2 rov yvovc; E Z1rw, r6n, av f3ovAYJ<T8E riKOlJEtJI, Jpw, when I shall hcwe S]Jnken abo1tt my bi1th, then, if you desiTe to heaT, I will speak of these things. DEM.lvii. 16. (Here the aorist d7rw, though absolutely fnture, denotes OfOjhUi, time past with refereuce to Jpw.) 'E7TEt0a]l Ota7rpct~Wjhat "J~w, when I shall have accomplished whcit I desi1e, I will come. XEN. A11. ii. 3, 29. 'E1r<toav !S~ ~<pvtj;wcrt Y1J, J.v,)p 1IP'fJfh~vos v1ro T~> 1r6Aews A.y<t 7i aDTol:c; <i1rawov Tuv 7rpE7rovTa, when they have coveTed them with eaTth, etc. THC:C. ii. 34. ''Ewe; av cr{l('}Tai TO U'Kctcpos, T6n XP'J 7rpo81JfLOV<; dvat' E1T"EiQaV o 'J 8aAarra V7r~pcrxu, fhclTUW> 'J U'1T"OV00, as long as the vessel Temnins in safety (present) ; b1d the rnoment that the sea has ovmwhelmed it (aorist). DEM. ix. 69. ''Ewe; av El<jha8us, <!x' eA1roa, until you, lwve learnt fully, have hope. Soru. 0. T. 834. Ma o KAv'7 KEV?J cp~perat TWV dcpavwv, ol' av fh'J Ellpe8wcrtv E<; dFa{pecrtv, and one bier is always carried empty, in honouT of the 'rnissing, whose bodies a1e not (have not been) found. THUG. ii. 34. l:.wvoei:Tat, Cl av aAAoi T1J dpeT1J KaTa7rpa~(J)<Tt, TOVTOJ]I lcrojhOipELV; i.e. he thinks of ha1;ing an equal shaTe in those things which othen by theiT valour have acquind? XEN. Cyr. ii. 3, 5. llav8' oclllv EK 1T"oAp,ov YLYVOfh~V'}'> <ZpfJvYJ> 7rpo<8ii, ravra TOtS dfL<A'Ijcracrw d7r6AAvTai, all things which are (or have been) abandoned when pea.ce is made are always lost to those




who abandoned them. DEM. xix. 151. "H V o' tipa Kat TQV 7rdpq. a-<jlaA.wcnv, dvTEA7r{cravns aAAa J7l"A'I)pw<rav T~V XPE{av, if they have been disappointed in anything, they always supply the deficiency, etc. (154 and 171). THUC. i. 70. Ovxi 7raDCTOfLat, 7rpiv tiv (T TWV crwv dpwv rrn)crw T~Kvwv, I will not cease befme I have (shall have) made you master of your children. SoPH. 0. C. 1040. 11~) crTEva(< 7rpiv paBv>, do not groan until you have heaTd. SorH. Ph. 917. 91.. This use of the aorist subjunctive (90) sometimes seems to approac11 very near to that of the pe1.fect subjunctive (1 03); and we often translate both by the same tense. But in the perfect, the idea of an action completed at the time referred to is expressed by the tense of the verb, without aid from any particle or from the context; in the :wrist, the idea of relative past time can come only from the particle or the context. (See 103 with examples, and 1 04.) The Greek often uses the less precise aorist subjunctive and optative (see 95) where the perfect would be preferred but for its cumbrous forms ; and we sometimes give the aorist more precision than really belongs to it in itself hy translating it as a perfect or future perfect. (See the last six examples under 90.) The following example illustrates the distinction between the perfect and aorist subjunctive:''Ov p~v &v rov &yvwTa (o Kvwv), xaAE7ra[vw 3v o' &v yvc!Jptpov (tov), dcrmfC<Tat, Kav fL'YJOEV 7rtiJ7rOT V71' aBTOV dyaBov 7l"E7l"OV8TJ, 1chomsoever he sees whom he knows, he fawns upon, even if he has hitheTto ?'eceived no kindness/Tom him. PLAT. Rep. 376 A. Compare this with eav &yaBov Tt 7l"U Bv V7l"O Ttvos, &crm5.(Tat, if he eveT happens to ?'eceive any kindness fmm any one, he always fc~wns upon him; and e7r<toav &ya86v n 71'U 827, &cr7ra(<mt, ajteT he has nceived any kindness, he always fa;wns upon him. ' 92. The present subjunctive with p.~] or 07rlJJ> p~ after verbs of feaTing, though it generally refers to a future object of fear, may also denote what may hereafter prove to be an object of fear. E.g. 6.8otKa p,~} dA~78s if, I fea?' it nwy ]JTOVe t?u.e. DEM. ix. 1. 6.EtvW> d8vfLW, fL~) j3AE7rWV o p,avns i], lest the prophet may JJ?'OVe to have his sight (cf. the following od~ELS o fLfiAAov). SoP H. 0. T. 7 4 7 ; 80 Ant, 1114. "Opu p~ 7rpi Tol.'s <jltAniTots Kvf3dTJ>, beware lest it may JJ?ove thctt you are staking what is dew est. PLA'r. Prot. 314 A. '' 0 pa. 07!"W<; ,.~} 7rapa oo~av OJLOAoy{Js. Id. Crit. 49 c. In all these cases the present indicative would be required if the object of fear were really 1)resent (369, 1). Compare the examples of the perfect subjunctive in 103. 93. In a few passages of Homer the aorist subjunctive with p~ seems to express a similar fear that something may prove to have already hal)pened; as odootKa fL'l CT 7rUpd7rYJ, I fear it may prove that she peTsuaded you, IL i. 555. So Il. x. 98, p~} Kotp~crwvTat aTap A.aewvTat, and X. 538, S.tSotKa P-0 Tl -n-J.Bwcrt, I fear lest it may prove that they have met some hann. The reference to the past here cannot come from any past force of the aorist subjunctive itself,




but is probably an inference drawn from the context. As the later language would use a perfect subjunctive in such cases, these aorists seem to be instances of an earlier laxity of usage, like the use of &.1r6AmT6 K for both would have pmishe1l and would perish (44 0). In Il. x. 537 there is a similar case of the aorist optative in a wish: ar yop 0~ il>o' /l<f;ap iK Tpwwv U..a(Ja{aTO fi>Wvvxa<; t7T1TOV<;, i.e. may it prove that they have d1iven the horses away front the T1ojans (95).

94. The present and aorist Ol)tative in independent sentences (in wishes and with llv), and ii1 all conditional sentences except past general conditions (462; 532), express future time, the relation of wl1ich to the future expressed by other moods is explained in 12, 13, and 16. (Some Homeric present or past unreal conditions and present wishes are exceptions: see 438-441.) In all final constructions the optative (which is used only after past tenses) represents the subjunctive after primary tenses, and is future relatively to the leading verb. E.g.
Elth TOVTO dry (utinam sit), 0 that this may be. El8 fi>?J mvm may they not s4fer these things (with a view to the progre~s of their suffering). But Et8E fi>?J TavTa 7Ttl8otv, may they not svjfer these things (viewed collectively). (JtJ TowvTo<; c:\1' <f;{Ao<; ?jfi>Zv yf.vo w, may you become a jTiend to us. XEx. Hell. iv. 1, 38. M?) yf.votTo, <Jnay it not happen. See examples of the 01)tative with /lv below. Ov yap av E1TO.LVO{'Y) fi>, El f.~d.aVJIOLfi>L TOtJ<; EVEpyf.m<;,fm he would not pmise me, if I should banish 1ny benefactors. XEN. An. vii. 7, 11. Et'Y)> <f;opYJTO<; OVK av, el 7rpa(J(J0LS KaAw<;, ynn would not be endumble, if you should be in )JTOsperity (at any time). AEo;CH. Prom. 979. ITw<; yap Ci.v n<;, d. YE fi''l E1TiffTO.LTO, TUVTa (J0cf)o<; EtYJ; for how CMdd any one be wise in that which he did not unclentand (i.e. d nva fL?J f.7r{<TTaLTo) 1 XEN. :Mem. iv. 6, 7. 'A'AX et n fi>?J <f;f.potp.ev, w'Tpwev 4>f.pHv, but if we neglected to bring anything, he always exhorted us to bring it. EuR. Ale. 755. OvK rl7TEAeL7TETO ETL avTov, el /1>~ TL dvayKatov dYJ, he never left him, unless there was some necessity fo1 it. XEN. Mem.
1r tl(J X o LE v,


iv. 2, 40.
El E'AIJot, 1T<ivT' av tooL, if he should go, he would see all. El A.8o t, 1rav8' ~wpa, 1j ever (whenevm) he went, he (always) saw all. Ollll' d 1ravTE> 'A8o,Ev ITf.p(JaL, 1rA?}IJEL -ye ovx v7rEpj3aAo[fi>EIJ' av TotJ<; 7TOAEfJ>ov<;, not even if all the Pe1'Bians should come, should we surpass the enemy in numbers. X1m. Cyr. ii. 1, 8. ''OTE ~w Tov oELvov yf.votvTo Kat f.Ee[?J 1rpos &A.A.ov<; llpxovra<; d7TLEJ'aL, 7roAAot av'Tov U7rEAH7rov, but 1vhen they we~e come o1tt of dangm and it was in their power (prese11t) to go to othe1 commanders, (in all such ca:aes) many left




him. Id. An. ii. 6, 12. "Avev yctp apx6JITWV ov8~v <'iv oilTE KaA6v oilTE dya80v yevo ~To, nothing could be done, etc. Ih iii. 1, 38. OvK ol8a 6 n liv n> XP>)cra~TO avTOt>, I do not know what use any one could rnalce of them. lb. iii. 1, 40. TovTov E7re.8v1J,e.~, Zva e.D 1rp<iTTo~, he desired this in 01der that he might be in p1ospeTity. 'E<f>of3e'iTo fJ-?J TOVTo 7rOt.o'iev, he feared lest they should do this (ha.bitually). t:.fjA.os >jv E7r~fJVjJ,WV apxew, b7rWS 7rAdw AajJ,f3 6. vo ~, E7r~eVjJ,WV 8 Ttfl,acrea~, rva 7rAdw Kepoa{vo~. <f>Ao<; TE

f3ovAETO elva~ TOt<; fJ-EY~CTTa SvvajJ,EVO~<;, rva do~KWV jJ,~ s~sov 8KT)Vo An. ii. 6, 21. (Here the aorist optative would have referred to single acts of receiving, getting gain, and suffering punishment, while the 1)resent refPrR to a succession of cases, and to a whole course of conduct.) "Hv 6 <I>tA~7r7rOS EV <f>6f34! f1,1J EK</>vyo~ Td 7rpdyjJ,a! avT6v, Philip was in fear lest the contTol of aifc~i?'S might escape him. DEM. xviii. 33.

95. The aorist optative with E7rH8>) or E7re{, after that, is referred by the meaning of the particle to time preceding that of the leading verb, like the aorist subjunctive in 90 ; so that E7rH01} [oo ~ a7r1JH means ajteT he had seen he (always) went awny. This gives the aorist in translation the force of a pluperfect. So after words meaning until, and in the other cases mentio11ed in 90. E.g. OVs- fL'Ev i'Oot. cl,niKTWS l6vras, T[l'fS TE cl'Ev t]pWTa, Kat 1rEl 1rV8ot.TO
E7r1JVE~, he asked nny whom he snw m'!Tching in good ordeT, who they wen; a.nd after he had ascmtained, he J ndsed them. XE~. Cyr. v. 3, 55.

llep~EfJ-EVOfJ-EV avo~xee,),

El<dcrTOTE i!w<; d vo LX Cd 17 Tu 8ECTfJ-WT'fJp~ov E7rEdl?J St e?crdEt.fUV mLpd T~JI ':::WI<(lflTY), jOe waited eCICh ?nO?'ning

until the pTison u:a.s opened (or h cul ueen opened); nnd afteT it was opened, we went in to S0c1cttes. PLA'l'. Phaed. 59 D. In PLA'r. Hep. :331 C, Er Tu; A f3 0 ~ 7rapa <f>{A.ov dv8pl><; crw<f>poJ!OVJJTO<; /hA.u., el jJ,aJ!et> U7rfUTO i:,

is thus given by Cicero (Offic. iii. 95): Si gladium qnis apud te sanae mentis deposueTit, npetat insaniens; and there can be no donht that cl.A.ry<f>,;,., dry (the equivalent of de?JOS1WTit) would have been more exact than A.aj3ot in Greek (see 9 J ). For a peculiar aorist optative in Il. x. 53 7, see above (93, end).

Infinitive. 96. A present or aorist infinitive (without &v) not in indirect discourse is still a verbal noun so far that it expresses no time except such as is implied in the context. Thus, when it depends ciu a verb of wishing or commanding or any other verb whose natural object is a future action, or when it expresses purpose, it is future without regard to its tense; as, in f3ovA,of1-a~ vucav (or vucf]<Ta~), I wish to be victoTious (or to pain victory), the infinitive expresses time only so far as the noun vK1JV would in f3ovA,of1-a~ v/1C7JV. Likewise,




when the present or aorist infinitive (without &v) has the article, except in the rare cases in which it stands in indirect discourse (794), it has no reference to time in itself; as in r6 ryvwvat hrurr~.VirJV A-a{3E'iv l.o-rw, to learn is to obtain knowledge, where ryJJ&)IJa.t expresses time only as the noun ryvwo-t> would in its pla:;e. E.g.
"EEtO'TL !LivHv, it is possible to Temain. 'EEEO'Tat TOVTO 7/'0tftv, it will be possible to do this. !::..iofLat VfLWV fLEvH v, I beg you to remain. 'l'i TO KWAVOV ET aVTOJ! EO'Tat f3 a 0[ t Et V 07/'0t j3o{;ATat, 'What will there be to prevent him from going whither he pleases ? DEM. i. 12. 'EKL\.nxra aV-rov TOVTo '"o t t 'iv, I commanded him to do this. 'Ef3ovAtTo O'ocp0s E iva t, he wished to be wise. !::..Hv6s EO'rt A.iytt v, he is skilled in spealving. ''rlpa f3a8{(ttv, it is time to be going. IIav 1rotav<rtv WO'TE O[K'YJV !L~ St8ovat, they do evMything so as to avoid being pu.nished. PLAT. Gorg. 4 79 c. To fLEJI oilv E7r LTtfLUV tcrws cjYq<rat ns av Pcf.ilwv l'vat, -rO 0' 6 rt OEZ -;rpd.TTEtV &.7rocpavEcrf3at, -roVT' eivat crvp..{3ot'iAo1J, some one may say that finding fault is easy, but that showing what 01tght to be done is the duty of an adviser. DEM. i. 16. ('E7rtTLfLav, d7rocj>a[vt0'8at, and 7rpaTntv belong here ; but lvat in both cases is in indirect discourse, 117.) Ov 1rAWVE~as EJIEKEJI TavT' E7rpa~tv, dA.,\d. "P OtKat6npa TO VS' 8r;f3aovs ~ VfLUS dEw v v, he did this not from lore of gain, but becanse of the Thebans ?naking juster dema1llls than you. Id. vi. 13. 'Entx<r817 o 'ATaAriv.,-r; v1)(]'os, Tov !L~ AYJ<rTd.s KaKovpyEiv T1JV Ev{3otav, in ordm to prevent pimtes from ravaging Euboea. THVC. ii. 32. IloAEtb> ern BrivaTOS' dvricrTaTov yn0'8at, it is death fm a c1:ty to ue laid 1Vaste. LYCURG. 61. "Q0'7rEp TWV dvilpwv TOts KaAoZs Kuya.BoZs alpETwnp6v E(J'n 1< d'"o8ctvfv i] (~v al(]'xpws, ovTw Ko,l Twv 7rOA<wv TCJ.ts V7rEPEXo{;(]'ats A. vO' tTE AE'iv (~yovvTo) f.g dvBpwll'Wv dcj>avt0'81)Vat fLUAAov i] ilo{;Aats 6cpB~vat ywop-f.vats, as it is prefemble for honoumble men to die (aor.) nobly 1ather than to contimte living (pr.;s,) in disgmce, so also they thought that it was bettm (pres.) fm the pre-eminent among states to be (at once) 11wde to disappeaT (aor.) from among men, than t<~ be (once) seen (aor.) to fall into slave?y. Isoc. iv. 95. IIfL'"OVO'tV f.s T1JJI KpKvpnv '"P(]'f3Hs, OEOfLEVot fL1l crq>as 7rtptopO.v cj>BHpofLivovs, d.\,\d. To{;s T c/>n!yovms ~vva.\A.ri~aL (]'q>[crt Kac TCJv Twv f3apf3dpwv 7r6AEfJ.OV KaTaXv(]'at, asking them not to allow them to be dest?oyed, bnt to bring their exices to tmms 'with them, and to put an end io the uwua?ians' wa,r. Tnuc. i. 24. T0 yap yvwvaL E7rl<rTf;fLT)V 7rOV A.nf3 dv E(]'TlV, to leam is to obtain knowledge. PLAT. Theaet. 209 E. IIavTe; To KaTaAL7rttv a&a 7rriv.,-wv fLaAurTa cj>EvyofL<V, we all try most of all to avoid leaving them behind. XE:'l. Mem. ii. 2, 3. 0-& yd.p TO fL~ A.o..f3<tV .,-dyaBd. OVTW "f xaA7r0v (tl(J'7rp TO A.~f3ovTa O''!'P'YJe~vat AV7rYJp6v. Id. Cyr. vii. 5, 82. 'l'ov 7rtttv J'"tBvfL[a, the desire of obtaining drinlc. THuc. vii. 84. K<A<-8Et a.D-ri.v i AB< Zv, he eowmands him to go. 'EKEAVO'EV ai>T6v JA.Bd:v, he commanded him to go. K~-




AEVcTet a~'TdV HBdv, he will cornrnand him to go. IIp6s -rcjj fLYJ8'f.v JK 'Tqs 7rpwf3<os Aa f3 ei v, -rovs alxfLaAcb-rovs iAvrya-ro, besides nceiving nothing from the ernhassy, he ransomed the captives. DEM. :x:ix. 229. El 1rp'O -rov -rovs .PwK~as choA~<YBat tfYJ~i<Yat<YB< fJoYJB<Zv, if before the destruction of the Phocians you should vote to go to their assistance. Id. xviii. 33. Ta> al-rla> 1rpovypalj;a, -rov fk~ nva {YJr~rya 1roT i~ o-rov -ro<Yovros 7r6A<fLD> Ka-:-~<YT?), that no one may ever ask the reason why, etc. Trruc. i. 23. T6v v1r~p -rov fl~ yEVf.<YBat -rav'T' dywva, the contest to prevent these from being done. DEM. xviii. 20 l. No account is here taken of the infinitive with lfv (204).

97. Tl1e distinction between the present and aorist infinitive is well illustrated by Aristotle, when he says of pleasure, Eth. x. 3, 4, ~<YB~vat fLEV yap e<Yn -rax~ws i!J<Y7r<p 6pyt<Y8~vat, 1)ow8at 8' oil, ov8'f. 1rpos ETpov (3a0{etv 8~ Kat avt<<Y8at KO.l mfvra Ta TOtaVTa. fl<Ta(3 cfAA 1 v fLEV. ovv ds n)v 1]8ov1)v -rax~ws Kat {3pa8~w<; e<Ynv, iv<py.Zv 8 Ka-r' a-&n)v ovK eryn -rax~ws, Aeyw 8' 1)8E<Y8at. TVernay fiECOME pleased (1]<Y8~vat) quickly, as we rnay get ang1y quickly; but 'We cannot BE pleased (1)8E<J8at) quickly, even as cornpand 'With another person, although we can thus 'IJJallc and grow and do such things. FVe rnay then change into a state of pleasure quickly m slowly, but we cannot actually enjoy the pleasure, I mean BE PLEASED (1JOE<J8<u), quickly. So in FLAT. Theaet. 155 c, Socrates says, avw 'TOV yyJ!<<J8at y<FE<Y8at dovva-rov (se. EfLE EA.cfr-rw), i.e. without going thro1tgh the pmcess of becomin:; (yiyvE<Y8at) srnalleT, it is impossible for me to get (y<ve0'8at) smalle1.
. 98. Xpcfw, dJ!atpf.w, B<cr7r[(w, and other verbs signifying to give an oraculaT response, gi;Jnerally take the present or the aorist infinitive, expressing the command or warning of the oracle, where we might expect the future in indirect discourse (135). These verbs here take the ordinary construction of verbs of commanding, advising, and warning. E.g.
Ayerat 8~ 'AAKfLa[wvt r6v 'A1r6A..\w -raVT1]V T~JI y1)v XP~O'aL olKdv, it is said that Apollo gave a Tesponse to Alcrnaeon that he should inhabit this land (warned him to inhabit it). Trruc. ii. 102. XpwfL~V'f! E -rep IG.\wvt dvEZAEV 0 8EoS ev -rif 'TOV Lltos -rfi fL<y[(J''T'(! Jop-rfi Ka-raA.a(Jdv T?JV 'A8YJ1'aiwl' dKp61roAtJ1 1 that he should seize. Id. i. 126. 'E K XPYJ TO yiip -ro'i:a-t L7rap-rt~T'{/<Yt, ~ AaKEOa[flova dFcfryraTOJI yHe<Y(Jat i] Tov (3acrtAea O'~~wv a1roA~0'8a~. HDT. vii, 220. 'E(Jf.<Y7rt<YE KafL{O'at Kat d<JtOei:v. EuR. I. T. 1014. '.Qs XPYJO'fLOV ovros n)v 1r6Atv 8w~8ap~va~, as if then 'IJJere an omcle doorning the city to perish. FLAT. Rep. 415 C. IloAAaKt ycfp o1 EH7rE voVO''f! il1r' &pyaA.~rt ~ {(}'()a l ~ {11rQ TpcbE<JO" t 8 a fL ~V a t, the diviner told him that he nwst either die by painful disease, or perish at the hands of the Trojans. IL xiii. 667, But we find aVEGAEV EO'E<Y(Jat, THUC. i. 118; XP~O'av TOS KpaT~O'ElV 1 LYCURG. 99; ~K~XP?/0'1'0 j3a.O'oA<V<r<ov, HDT. ii. 147; as indirect discourse.




99. Even verbs of saying and thinking, as A.f.yw when it signifies to c1 ymmwdui, and ooKet, it seems good, may take the present or aorist infinitive not in indirect discourse, like other verbs of the same meaning. El"IToV seldom takes the infinitive, except when it signifies to command (753). The context will always distinguish these cases from indirect quotations. E.g. Tovrots ~A.eyov "IT AE t'v, I told them to sail. DEM. xix. 150. (T DVTO VS V .. eyov "ll"Aet'v would mean I said that they were sailing.) El"IT~W p:rJf) 1Tap tevat lis r~v dKp61ToALV, having given orders that no one should pass into the citadel. XEN. Hell. v. 2, 29. "'.Q <j>O..ot, ~O'Y] pev KEY Jywv d"ITotJH Ka2 llpptv fJ-VT)UT~pwv Js OptA.ov dKovr[uat, now I would cornrmand you to join nw in hurling, etc. Od. xxii. 262. IIapaoovvat A.f.yet, he tells us to give her up (he says, give her up). AR. A v. 16 7 9. t.oKEL ~p.r:v rovro "ITD H LV (or "ITOt~uat) it pleases us to do this. (But OOKEi: p.ot TOVTO "ITOtEi:v (or "ITOt~(Tat) generally means it seems to me that you are doing this, or did this.) "Eoo~e in the sense it was resolved, introducing a decree, is followed by the present or aorist (not future) infinitive.

100. Verbs of hoping, expecting, promising, and swearing form an intermediate class between those that take the infinitive in indirect discourse and other verbs (136). When they refer to a future object, they naturally take the future infinitive, but may also have the present or aorist infinitive (not in indirect discourse) like verbs of wishing, etc. Thus he promised to give may be V"ITEUXETO 8tli6vat (or oovvat) as well as V"ITEUXETO OWUELV.
To facilitate comparison, the examples of the present and aorist infinitive thus used are given with those of the future in 136.

101. The present al'rt6s ~lp.t, I am the cause, is often used with reference to the past, where logically a past tense would be needed ; as atn6s f.urt roVT<f! 8ave'iv, he is the cause of his death, instead of ahtos ~~~ TOVT<(l 8avet'v, he was the cause of his death. This may make an ordinary aorist infinitive appear like a verb of past time. E.g.
AZrwt oDv tlut Kat "ITDAAwv :)8'1) lf;evuB~vat Kat 81} d8Kws
they a?e the cause why you were deceived and some even perished (i.e. they caused you to be deceived and sonw even to pe?ish). LYS. xix. 51. TEI9vautv ot 8 {wvres afnot 8avdv, they an dead; and the living are the causes of their death. SOPH. Ant. 117 3. "H p.o' p.T)rpl p.v 8avei:v p.6v'Y] p.eTanos. Id. Tr. 1233.

y nvas d1ToAeu8at,


102. As the perfect indicative represents an act as finished at the present time, so the perfect of any of the dependent moods properly represents an act as finished at




the time (present, past, or future) at which the present of that mood would represent it as going on.
103. The perfect subjunctive and optative are very often expressed in the active, and almost always in the passive and middle, by the perfect participle with 6.> and t?JV; and this combination of a present and a perfect makes the time denoted especially clear. Where the present would denote future time, the perfect denotes juture-pe1ject time. E.g.

To xp6vov yeyev~a-Bat 'lrOAVV OEOOtKa ""'~ nva A?)BYJV BfLtV 7r7rO t~ KTJ, I fear lest the lapse of a long time that has occurred may (when you come to decide the case) prove to have caused in you some forgetfulness (see 91). DEM. xix. 3. (M1) 'lrotfl would mean lest it may cause, the time being the same as before.) Xp?J avrd. [a TfAEVT~a-avra EKctnpov 7rptfLEVH] d.Kova-at, Zva TEA.EWS EKanpos avrwv d.7rclA~1)7J Td ocpctA6fLl!U, we must hea1 what awaits each of the1n afte?' death, that (when we have .finished) each ?nay have fully nceived his deserts. PLAT. Rep. 614 A. Tov> fL"Ev a.\.\ov>, K&v 00WKOTS" ala- t V EVBVJta>, n)v d.etA.oy[av op(~ 7rponwofLEVOV<;;, I see that other men, even if they have already 1endered their accounts,-i.e. 1j they aTe (in the state of) persons who have nndmed thei1 accm<nts,-always offer a perpetual ?"eckoning. DE~L xix. 2. 'Av8peZ6v y< 1ravv VOfL[(ofL<V, oc; &v 1rnr )u) {7] 1rar!pa, we always consider one very manly who has (may have) beaten his father. AR. Av. 1350. NofLOV B,ja-ELv fL')O<vt rwv 'E.\.\~vwv -&pJis f3o,)BEZv 8,. &v fL?J 1rp6npo<; f3f3oYJKWS -&pXv fj, to ,enact a law tlwt you shall assist no one of the Greelcs who shall not previously have assisted you.. DEM. xix. 16. ('Os &v fL1J 1rp6npos f3o1JBiJ would mean who slwll not previously assist you.) "E8eta-av fL?J Ava-a-a ~pXv ifL7r<1T"Tt.iJKot, they fewed lest madness might prove to have fallen upon us. Xn:N. An. v. 7, 26. (M1} EfL7rf.7rrot would mean lest it might fall upon us.) 'Eoc,)Bryv rwv OtKao-rwv JLYJO~v rowvrov 1rpa~at, Zv' ~yw f1-YJSva 'ABYJva[wv U1T"EKrovw> dY)J', that I might not he in the position of having put an Athenian to der;,th. DEM. liii. 18. "Hv yap EvpcBfi A.!ywv o-ot ro.1ir', ywy' C!.v eJ<1r<cpEvyolYJv 1ra8os, I should (in that case) have escaped ha1m. SOPR. 0. T. 839. IIws OVK &v olKrporara 'lrUVTWV eyw 7T"1T"OJI8w, dYJV, cl EfL" fYJ</J[CTalJITO <tvat ~,,ov; how should I not have su.ffend the most pitiable of all things, if they should vote nw to be a.n alien? DE~L lvii. 44. (This could have been expressed, with a very slight difference in meaning, 7rw> oi 'lr1T"OJ18ivs ECTOjJ-D.t, Jrlv r'7</J{a-wvrat; how shall I not have suffered, etc.) El onovv 1T"c1T"ov8ws Kr1Hpos 1lfLWV d1), ov Kat dfL<fJ6npot &]! rovro 7r 1T" 0V8 0 t fh JI ) if each of US shou.ld have S?~fJe?ed anything whatsoever, would not both of us have su,ffered it? PLAT. Hipp. M. 301 A. OvK !Lv 8td. Tovr6 y' d<v ovK <v8vc; Sd:iwKor<'>, this, at least, cannot be the reason why they did not pay it at once: lit. the11 would not (on inquiry) prove t11 have not paid it at once on this account. DEM. x.xx. 10,
104. The perfect subjunctive in protasis corresponds exactly to the




Latin future perfect indicative; but the Greek seldom uses this cumbrous perfect, prefening the less precise aorist (91). The perfect optative, in both protasis and apodosis, corresponds to the Latin perfect subjunctive; lmt it is seldom used, for a similar reason (95). The perfect optative can hardly be accurately expressed in English. For when we use the English forms wonld have suffered and should have sujfe1ed to translate the perfect optative, these are merely vaguer expressions for will and shall have suffmed. (See the examples above.) I should have suffmed is commonly past in English, being equivalent to bm8ov &v; but here it is future, and is therefore liable to be misunderstood. There is no more reference to l)ast time, however, in the perfect optative with &v, than tl1ere is in the future perfect indicative (77) in such expressions as fkrfn)v Jfko2 KKAa.!Ja-Tat, I shall have had my whippings jo1' nothing (referring to those received in his boyhood), AR. Nub. 1436.

105. The perfect imperative is most common in the third person singular of the passive, where it expresses a command that something just done or about to be done shall be decisive and final. It is thus equivalent to the perfect participle with gaTw. E.g.
Ta.vTa. fkEV 81) Ta15TV dp~a-8w, let so much hwve been thus said, dpYJfkEVO. ~rTTW), i.e. let what has been thus said be sufficient. FLAT. Crat. 401 D. But OfkWS OE dp~a-&w on, K.T.A., still let as much as this (which follows) be said (once for all), that, etc. Id. Rep. 607 C. IIEpt TWv l8wv Ta.vni fkOL 7rpOLP1Jif8w, let this have been said (oncefo1 all) by way of intToduction. Isoc. iv. 14. 'Ta.vTa 7T'7T'a[a-8w T Vfk'i:v, Ka.~ l:a-ws iKavws EXH, let this be the ,end of the play, etc. PLAT. EuthycL 278 D. TET0:x8w ~fktV KaTa Ch)fkOKpaTtO.V 0 TOLOVTOS av~p, let S'Uch a man 7'emain (where we have placed him), corresponding to dernocmr:y. Id. Rep. 561 E. 'A7rttpyrfa-8w ory ~p.'iv afm) 1) 7ro)uTEa, let this now be a sufficient desc1iption of this fonn of rJovernment. lb. 563 A. M<xpt Tov& wp[a-8w VfkWV ~ f3paOVT~>, at this point let the lirnit of your sluggishness be fixed. THUC. i. 71. The third person plural in the same sense could be expressed by the perfect partici1)le with ila-Twv, as in PLAT. Rep. 502 A, oilTot Tovvv TovTo 7r7rLo- fkEvO t Ea-Twv, grant then that these have been persuaded of this. 106. On this principle the perfect imperative is used in mathematical language, to imply that something is to be considered as proved or a~sumed once jo1 all, or that lines drawn or points fixed are to remain as data for a following demonstration. E.g. ElA~<f>&w E'lT' Tijs AB TVXOV (]'YJfJ-LOV TO 6., Ka. a<f>vp1)a-8w U7r0 Ar T)J AD. ra-'Y) ~ AE, let any point 6. be assv,med as talcen in the l~ne AB, and AE equal to AD. as cut off j1om Ar. EucL. i. Pr. 9. 107. The perfect imperative of the second person is rare ; when it is used, it seems to be a little more emphatic than the present or aorist. E.g.




'H~ uv Tovo M8.~o. Il. v. 228. M~ 7rlf:>of37Jrr8, do not be afraid. 'fnuc. vi. 17. M6vov rrv ~p.l.v 7rUT'Ta 8ewv 7rf7rOL7JirO Kat s-~~av Ob>, only malce us (immediately or once for all) solemn )Jledges and give the right hand. XE:s'. Cyr. iv. 2, 7. II1ravuo, stop! not another word! DEM.

uiv. 64.

108. In verbs whose perfect has the force of a present (49) the perfect imperative is the ordinary form ; as p.f.p.v'r)rro, KEKA~(J"Ow, l(J"Tal1t, E(J"'Ttt'Tw, Tf.Bva8t, n8vd.Tw, rUTw. So KEX~JVU'T, AR. Ach. 133; p.~ KKpayaTE, Vesp. 415. The 1)erfect imperative active seems to have been used only in such verbs. Occasionally we find the periphrastic form with the participle and lp.{, as uTw ~vp.f3f3'YJKv'ia, PLAT. Leg. 736 B.

109. The perfect infinitive not in indirect discourse generally represents an act as finished when the present would represent it as going on (96). E.g.
01}8~ f3ovAdJE(J"8at en wpa, dA.A.d. (3e(3ovAV(J"8at Tijs yd.p i.7rLOV(J"'r)S VVK'TGS 1rUVTa TUVTU ()';: 7r7rpax ea~, it is no longm time even to be deliberating, but (it is tirne) to have done delibemting; for all this must be finished within the coming night. PLAT. Crit. 46 A. Kat p.~v 7rEpt Giv y 7rpo(J"fTd~a.TE 7rpo(J"0Kn Ot<pK'YJKEVa.~, and it is his duty to have attended to the business about which you gave him instr1wtions. DEM. xix. 6. (This refers to an ambassador presenting his accounts on his return.) Avv<Tvyxav 1roAA.axov Std. T0v (J"T<voxwp{av Td. p.f.v fi.U.o~s J p.(3 ( (3 AT) Keva. t 'Ta 8' a&ov> J p. (3 E (3 A.iju eat, ovo T 7rEpt p.{ay ~V V7J p 'T ij ()" 8 a,~, it often befell them to have made an attack 011 One side and (at the sam,e ti?l,w) to huve been attacked themselves on the othe?', etc. THuc. vii. 70. 'AvayKT) yd.p ra p.v p.eytuT' aVTwv i)87J KaTaKEXPiJ(J"Ba.~ p.tKpd Ttva. 7rapa.AEAetlf:>8a.t, for it must be that the most important suhjects have been used up, and that only unim;portant ones have been left. Isoc. iv. 7 4. OiiK ~8EAov Jp.(3avEw &d. T0 KaTCL7rE7r A.ij X l Tii ~(J"(J")], they were unwilling to embaTk on account of having been tmTijied by the defeat. TRue. vii. 72. To yd.p 7roAAd. <hoA.wA.eK v at Ka.TO. 'TGV 7r6Aep.ov 'T~S ~}J-ETEpa.s dp.cAe[a,c; av 'Tl<; Bd'YJ OtKa[w>, TO of. }J-~Tf 7rUAa~ TOVTO 7rE1rOJ18Eva.t 7rElf:>"f)VEVctt TE nva. ~)p.'i:v (J"VjJ-p.ctxtav TOVTWV avTppo7rov, T~S 7ra.p JKdvwv .~voa> dJcpyeTwl llv ~ywye BE[T)v, for our hat-ing lost many things du~ing the wa1 one might justly cha1ge upon our neglect; but our never having suffered this before, and the fact that an alliance has now appea1ed to us to make up f01 these losses, I should consider a benefaction, etc. DEM. i. 10. (Compare ycy<vij<rBat in the first example under 1 03.) "Ecp8a<rav 7ra.potKo'BoJLfr (J"aVTE>, w<rn p.YJKE'Tt p.~TE aDroi KWAV()"0at {nr' avTwv, JK<{vovs T Ka.t 1rO,V'Tcf7rct<TlV ci.7rE(J"TEp'YfKEVU~ . . (J"cpas d1roTnX{<rat, i.e. they canied their own wall .fint beyond that of the Athenians, so as no longer to be themselves interfmed uith by the?n, and so as to have effectually prevented them from walling them in. THUC. vii. 6. 'E1rep.EA1)fh] Kat TWV >..ot7rwv, Wu'Tf Twv 1ra.p6vTwv Tars &v8p<!J1ro~> &ya8wv p.YJ8f.v p).v livev T'~S 1r6A.ew>






El vat, Td o~ 7TALO"'Ta otd. TaVT'Y}V YYV~o-Oat. Isoc. iv. 38. TotavTa Kat 'TOtTavTa KU'TlO"KlVao-av ~plv, wa-n fL'YJOlVt 'TWV J7TtytyvofLvwv trrrlpf3oA.~v AEAE'i<j>Oat, they rrutde such and so great acquisitions as to have no possibility of surpassing them left to any one who should come after them. DEM. iii. 25. t:.loofLEV avTots 7rpoKa a-vyKK6q,Oat, we allow them to have cut us up for nothing (i.e. we make no account of their having done so). AR. Nub. 1426. See [ARIS!I.'OT.] Eth. Nic. vi. 2, 6 : OVK ;<T'TL 0~ 7Tpoatpl'TOV ovo~v yyovo>, oiov ovoll> 7rpoatplt'Tat "IA.wv 7T7Top8'YJKEvat, but nothing past can be purposed j for example, nobody purposes to have sacked Ilium, i.e. the expression 7rpoatpovfLat ''IA.wv 7TE7Top0'Y)KEVat would be nonsense. This illustrates well the restricted use of the perfect infinitive.

110. The perfect infinitive sometimes signifies that the action is to l>e decisive and perrrutnent (like the perfect imperative, 105); and sometitUes it seems to be merely more emphatic than the present or aorist infinitive. E.g. Ei1rov n)v Ovpav KK A'i:a-8at, they ordered that the door should be shut (and remain so). XEN. Hell. v. 4, 7. BovA6fLEVO> &ywV Kat OtKaUT'YJp[cp fLOL 0 twpa-8at 11Up' VfLtV on T&vavr[a JfLOL Kat 'T01J7'0!> 1l"E7rpaKTat, i.e. wishing to have it once for all settled in your minds. DEM. xix. 223. 8Aova-as 7rpOs 7TVAats 7Tl7TTWKevat, eager to fall before the gates. AESCH. Sept. 462. "HA.avvv e1rt roils Mevwvos, wa-T' eKdvovs EK7Tl7TA~x0at Kat TPEXHV e1rt Td o1rAa, he manhed against the soldiers of Menon, so that they were (once for all) thmouqhly f1ightened and mn to arms. XEN. An. i. 5, 13, (Here EK1l"E7TA.~x0at is merely more emphatic than the present or aorist would be.)

111. The future is used in the dependent moods only in the optative and the infinitive, and in these it is never regular except in indirect discourse and kindred constructions and in the periphrastic form with fLEUw (73). For the future optative in indirect rliscourse see 128-134 ; for the future infinitive in indirect discourse see 135 and 136. 112. In constructions out of indirect discourse the present and aorist infinitive can always refer to future time if the context requires it (96), so that the future infinitive is here rarely needed. Therefore, after verbs which naturally have a future action as their object but yet do not introduce indirect discourse,-as those of commanding, wishing, etc. (684 ),-the prt>sent or aorist infinitive (not the future) is regularly used. Thus the Greek expresses they wish to do this not by {3ovA.ovTa.t TovTo '(:Ot~a-Hv, bnt by {3m',A.ov-ra< -rov-ro .,-oterv (or .,-ot-Yja-at). So the infinitive in other future expressions, as after wa-Te and in its final sense, is




generally present or aorist. fEAAw, see 7 3.)

(For the single exception after

113. On the other hand, when it was desired to make the reference to the future especially prominent, the future infinitive could be used exceptionally in all these cases. Thus we sometimes find the future after verbs signifying to be aUle, to wish, to be unwilling, and the like ; sometimes also in a final sense or with wrrT and Jcf} ~n; and sometimes when the infinitive with the article refers to future time. This use of the future is a partial adoption of. thr~ forni of indirect discourse in other constructions. It was a particularly favourite usage with Thucydides. E.g.
'ESEo)B'1rrm1 8~ Kai'TWJI 'Meyapewv vavO't O'q>a> ~Vf7rpo7rffHv, they asked the Jjfeyareans also to escort them with sh1:ps. 'l'auc. i. 27. 'E(:!of..\ovro 7rpOHfwpo)O'E0'8at. Id. vi. 57. So E1f'tXHP?JO'Etv JBEA?JO'Et<;; AESCHIN. iii. 152. To O'T6fa a'llrov OtEVOOVJITO KAJJO'HV. THUC. vii. 56. 'EqnefWOt fEV 'T'ry> 7r6.0'o)S tip~Etv, (:im]OEI:v /l ap.a E'll7rpE7rW'> (:!ovA6fEVot 'ToZs ~avrwv ~V'JYEVErn Kat ~Vfl-fl-6.xot<;. ILl. vi. 6. (Here f3o'18EI:JI is regular.) ToiJ 7'aZs vavO't fl-ry dOvfl-Ei:v h 'X Etpr/ O'Et v, to prevent them, from being without spi1it to attack .them in shirs. Id. vii. 21. Ollr' d7rOKWAVO'EtV ovva'T'Ot OV'TES. Id. iii. 28. El O'E -.; EV .\6yots 1rdO'HV OVV1']rr6fl-Err0a. SorH. Ph. 1394. EE ns Ei> rovTo dvaf36.AAE'Tat 1r o to) 0' Et v 'Ta llf.ovTa, if any one JJOstpones doing hi8 duty as fa? as this. Dl,M. iii. 9. (The orclinary construction woulrl be dvaf36.AAETat 7fOtEtJI or 1f'Oto)O'at.) 01lTE 7'WJI 1rpoy6vwv f'-Ef'-l'{jO'(}a, " , , , ,, I I , , [OH OVTE 'TWV AE')'OVTWV aVEXEO' 0at, JIOfl-OV 7'E 01']1J'EtJI Kat, ypa1fELV, K.'T.A. DEM. xix. 16. '(Here we have OE;; e~O'HV.) IloAAov ll~w EfLaVTbV YE diltK~O'EtV Kat Kal Efl-aV'TOV epdv a0r6s. PLAT. Ap. 37 B. Tovs 6!'-o)povs 1rap~!loO'av 'T!p 'ApyEwv oqfrp ll1il 'TaV'Ta ll w, XP'10'E0'8at, that they might put them to death. THUC. vi. 61. So 7f'E1JO'E0'8at, Id. iii. 26. 'Ecp' 0TE f3oo7e~O'HV. AESCHIN. iii. 114 (see 610). 'A1roOE~w a07'0v 'To)v 1rpoZ1<a o.J OEO<uK67'a ovrw fl-E')"lAot> TEKfl-1']pots WO''TE Dfl-OS a1f'aV'Tar;; dO'EIJ eat. DEM. xxx. 5 : su xxix. 5. 'EA1r[Ot 'TO dcpav<; 'Tov Ka'TopewO'EtV J1f';a11 TE>, having committed to hope what was ~"nce?tain in the JJrospect of success. TRue. ii. 42. (Here Ka'Top8WO'HV is more explicit than the present KaTopOovv would be : 7'0 dcpav'Es 'T'Ov Ka'TopOovv would mean simply what was uncertain in regard to success.) Tov ,, XE2pas eABEtV 1f'tiJT6TEpov 7'0 hcpof:i'lO'HV ~f'-OS dKtvbvvw> ~yovv'Tat, they feel ?JW?'e confidence in the }J?'ospect of jrighteniny us without risk than in meeting us in battle. Id. iv. 126. To fh~JI ovv ~EAey~Etv a{,r0v Oappw Kat 1r6.vv 1f't0'7'E1Jw, Ihavecourage and g?'eat con.iidence as to my convicting him. DEM. xix. 3. (Here most of the ordinary Mss. read J~EAyxnv.) See also THUC. iv. 115 and 121, v. 35, vii. 11, viii. 55 and 74; and Kriiger's note on i. 27, where these passages are cited. In several of these there is some Ms. authority for the aorist infinitive.
~J ~




114. The future perfect infinitive occurs only in indirect discourse (13 7), except in verbs whose perfect has the sense of a present (82).

115. \Vhen the optative and infinitive are in indirect discourse, each tense represents the co?Tesponding tense of the direct discourse; the present including also the imperfect, and the perfect also the pluperfect. See the geneml1windples of indirect discourse (667). The optative
is included here only as it is used alter past tenses to represent an indicative or snlJjnnctive of tl1e direct discourse. No cases of the optative or infiniti Ye with IJ.v are considered here : for these see Chapter Ill. For the meaning of the term "indirect discourse" as applied to tl1e infinitive, see 684.

116. The present optative in indirect discourse may represent the following forms of direct discourse : 1. The present indicative of a leading verb. E.g. ITEptKA~> 7rpoytyopv, on 'Apx 8afL6> ol ~EVO<; L1), PeTicles announced that Anhirlamus was his fn:end (i.e. he said gvo> fJ>O[ Jcrnv). TBUC. ii. 13. "Eyvw<Tav on /(Vu> 0 q)tSf3o> di], they leamed that their fear was groundless (i.e. they learned KV6<; Jcrnv o q)6f3o>). XEN. An. ii. 2, 21. 'E1fvv8dvero L olKOLTO 1) xt!Jpa, he asked whetheT the country was inhabited (i.e. he asked the question OLKLTat 1) xt!Jpa ;). XEN. Cyr. iv. 4, 4.

2. The present indicative or subjunctive of a dependent verb. E.,q. Et1fV on av8pa ayo 3v /[p~a OEO t, he said that he was b?inging
a 1nan whom it was necessaTy to confine (he saicl livopa li yw 8v Eipgat 0t:). XEN. Hell. V. 4, 8. 'H YELTO a1faV 1fOOJCTHV aVTUJI 1: T<; dpyvptov 8t8o['Y), he believed thcbt the 1nan would do anything if one weTe to give hirn money (he believed a1faV 1f0~CTH Mv T<; dpyvpwv ouStji). LYS. xii. 14.

3. The present subjunctive in a question of appeal (287).


\I 'f3 \ I I I ,, I ,, K 1\.Eapxoc,; 01JAEVE'l'O, L, 1ffL7r'OtEV TWUS 1) 7r'UVTS 0V 1 Clearchuo wao deliberating whether they should send a Jew 01' should all go. XEN. An. i. 10, 5. (The question was, 1fEfJ>1fWfJ>EV TLJias ?) mivrS




lwfLH; shall we send a few, or shall we all go 1 See 677.) The context will always make it clear whether the optative represents a subjunctive (as here) or an indicative (1).

4. The imperfect indicative of a leading verb.


'A7rKpvavTo on o-& fLdpTvs 1rapd'r), they replied that no witness

had been present (when a certain payment was made). DEM. xxx, 20. (They said ovo2s 7rapijv.) This is the rare imperfect optative (673). The imperfect indicative is regularly retained in such cases, and is always retained in a dependent clause of a quotation (689, 2).

117. (As Present.) The present infinitive in indirect discourse generally represents a present indicative of the direct form. E.g.
P~O't yp<i</>v, he says that he is writi'flg; ~</>') yp<icpHv, he said that he was writi'flg; </>1JO'fl r p d </>Et v, he will say that he is (then) writi'flg. (In all three cases he says yp&<f>w.) 'AppwO'u'iv 7rpo</>a.O"[(Tat, he pretends that he is sick; J~wp.o0'V dppWO'Tdv TovTov[, he took his oath that this man was siclc. DEM. xix. 124. OvK E</>YJ a.vTils d>.X iK'ivov O'TPO.TYJYE'iv, he said that not he himself, but Nicias, was general; i.e. he said o1JK yw a.1hils dA.X EK'ivos O'Tpa.TI)yel:. THua. iv. 28. See other examples under 683.

118. Verbs of hoping and swearing may thus take the present infinitive in indirect discourse. This must be distinguished from the more common use of the present and aorist infinitive (not in indirect discourse) after these verbs, referring to the future ( 100; 136). E.g.
'EA.7T[(wv ivo.L avepw7rwv oA.tj3WTO.TOS, TUVTa E7TLpWTa, he asked
this, t1usti'flg that he was the most happy of men. HDT. i. 30.

So i. 22,

A.7rt(wv <nToOE[YJv u dvat lO'xvp~v Ka2 Tilv AEwv T7pvO'ea~. IB:vva 8' f.A.7T[(w A.eyLv, and 1 hope I speak for the common good. AEscH. Sept. 76. 'Op.vvvus f3 A.e7rt v Tov ovKe,.' ovro. (wvT' 'Ax~A. A.a 7T<iAtv, i.e. sweari'flg that they saw Achilles alive again. SoPH. Ph.
357. Compare the first two examples with EA7T[{L ovvo.Tos eiva.t, he hopes to be able, PLAT. Rep. 57 3 C ; and the last with OJLOO'a.L l vat p.f.v TYJV apx~v KOI.V~V, 7rdVTO.S 0' l>fLtV a7rOOOVJ'a~ T~V xwpav, to swear that the dominion shall be comnwn, and that all shall surTender the land, DEM. xxiii. 170. (See 136 and the examples.)

119. (As Irnpe1ject.) The present infinitive may also represent an imperfect indicative of the direct discourse, thus supplying the want of an imperfect infinitiv~. E.g.




Tlvas ovv nlxa> v7roAaf'-f36.vef vxw8at Tots Owl:> TOv CI.>A.t'11'7TOV oi' W'11'VOV ; what prayers then do you suppose Pkilip made to t!M Gods when he was pouring his libations-? DEM. xix. 130. (Here the temporal clause 6i' ECT7TVOv shows that EVXCT8at is past.) II6np' or(]"() 7TAEOV ii>wKEa> 81)(3awv ~ CI.>A.t7T7TOV Vf'-fiJV Kpanl:v T~ 7roAEf'-'f; do yo1t think that the superiority of the Phocians over t!M Thebans oT that of Philip oveT you was the greater in the war (the war being then past) 1 DEM. xix. 148. (Here the direct discourse would be JKp6.Tovv and JKp6.TEt.) Ilw> yfip otECT()E ovCTxpws iK o H v 'OA.vv8ovs, Er T[s n A.eyot KfLTa CI.>tAt7T7TOV KaT' EKvovs TOVS xp6vovs iJT' 'Av8Ef'-OVVTa aVTOt> a</J[H; . apa 7rpoCTOOKUV avTOV> TotaVTa ' ,, .. " ,, \ ' '1''(3 \ \ 7rLCTU" () at (se. ouCT () ) ; ap' OLEO" () E, OT TOV!> Tvp.avvov> " aAA, (ToV> 8ETTaAoVs) 7rpOCTOOKav, K.T.A.; for how unwillingly do you think the Olynthians used to hear it, if any one said anything against Philip in those tirnes when he was ceding Anthernus to thern, etc. 1 Do you think they were expecting to sujfeT such things ? Do you think that the Thessalians, when he was expelling the despots, were expecting, etc. 1 DEM. vi. 20 and 22. (The direct questions were 7rwS ~Kovov d Aeyot; and 7rpoCTECOKwv ;) Kai yap ToVs E7Tt Twv 1rpoy6vwv ~f'-WV A.eyovTa> UKOVW TOVT'f T'i e8H XP0CT8at, I heaT that they used to follow this custorn. DEM. iii. 21. Ta f'-EV 7rp0 "EAA1JVOS ovof. ElVat ~ E'lr[KA1)U"t<; aVT1J (se. OOKt), in the ti1nes before Hellen this na1ne does not appeaT to have even existed. Tauc. i. 3. Again, in the same sentence of Thucydides, 7rapEXECT8at, to have fumished. M1Jof.v oZov &A.A.o f'-1) Xa vaCT ()at ~ d7TW> o_e~OLVTO, Zva y[yvotTO. PLAT. Rep. 430 A. METU TaVTa e</>1) CT</Jas f'-EV 0Et'lrVi:v, TOV of. LWKpaT1) OVK dCTtEVat' TOV ovv 'Ay6.8wva 7TOAA6.Kt> KEAEVHV f'-ETa7TEf'-lfaCT8at TOV LwKp6.T1J, 2 of. oiJK iav. PLAT. Symp. li 5 c. (He said, J8El7TV0Vf'-EV, 0 (if. 2,, 'oiJK dCTfJH' 0 oi'!v 'A. iKEAVV' eyw OE oiJK Efwv.) LVVTVXEi:V yap (E</>1)) 'ATpECTT[bif Trapa ii.>tA[TrTrov TropEVOf'-EV'f, Kat f'-ET' avTov yvvaw Kat 7rat06.pta (3ao(ov,for he said that he had rnet (aor.) Atrestidas earning frorn Philip, and that there weTe walking with hirn (impf.), etc. DEM. xix. 305. Tovi' Jyw <I>1Jf'-L odv ~f'-E f'-1J Aa8E'iv, I say that this ought not to have escaped rny notice. DEM. xviii. 190. (The direct form was Tovi' EOH Ef'-E !'-~ ,\a8EZv, 415.) The imperfect infinitive is found even in Homer; as Kat CTe, yepov, TO 7rptv f'-EV aKOVOf'-V tJA.(:Jwv dvat, we heaT that you were once prorperous. IL xxiv. 543. SoIl. v. 639; Od. viii. 181, 516. For the imperfect participle, see 140.

120. This use of the present infinitive as an imperfect mu&t be carefully distinguished from its ordinary use after past tenses, where we translate it by the imperfect, as in f.</J1) TO CTTp6.TEVf'-a I'-6.XECT8at, he said that the army 1cas fighting. This has sometimes been called an imperfect infinitive; but here p.&. XECT8at refers to time present relatively to t</>1); whereas, if it had been used as an imperfect, it would have referred to time past relatively to ii<fl1J, as in i!<fl'YJ TO CTTpaTEVf'-a Tij 7rponpa['f 1'-6. XECT8at, he said that the arrny had been fighting on the day




before. In the former case the direct discourse was 1-ufxerat, in the latter it was Jftrixero. Such an imperfect infinitive differs from the aorist in the same constructipn only by expressing the duration or repetition of an action (as in the indicative) ; it gives, in fact, the only means of representing in the infinitive what is usually expressed by .Ayn on J'l/"o[n, he says that he was doing, as opposed to AEYH /)n E'li"O{Y)rr<v, he says that he did. (For the similar use of the present optative to represent the im1Jerfect, see 116, 4.) This construction is never used unless the context makes it certain that the infinitive represents an imperfect and not a present, so that no ambiguity can arise. See the examples. So sometimes in Latin : Q. Scaevolam memoria teneo bello Marsico, cum esset summa seuectute, cotidie facen omnibus con veniencli potestatem sui. Ore. Phil. viii. 31. So Q. Maximum accepimus facile celan, tacme, dissimulare, insidiari, praeripere hosti um consilia. Ore. de Off. i. 108.

121. The perfect optative in indirect discourse may

represent1. The perfect indicative of a leading verb.


"EAEYE orra &ya86. Kvpos ITprras 1/"1/"0t?)Kot, he told how many services Uyrus had done the Penians. Hm. iii. 75. (ITmot~Kot here repre,ents 71"7/"0[Y)KE.) Oi'!rot EAE"jOV ws 1/"EVTaK6(]'tat avToi\; dYJ (]'a V h Tov IIetpauils- oeoeKa(]'ftEVOt, LYs. xxix. 12. (Here the direct disc<,mrse was 'li"EVTaK6rrw[ Etrrtv oeoeKaaptvot.)

2. The perfect indicative or subjunctive of a dependent verb. E.g.

El'li"EV l}{ft'l/"1/"0V oiJK E'l/"atvo{r; el TavTa 7r1/"0t'Y)KW> etY) (he said ovK i'll"atvw el TavTa 'li"E'li"O{YJK<, I do not appTove-him if he has done this). X1m. An. vi. 6, 25. 'E.AyoftEV EJia EKatTTOV v OEOt J'll"tTY)OEVELV, Els 8 aVTOV ?J cpVtrL> Jr.tTY)OEWTdTYJ 'll"ecpvKvta e t'Y) (we said EJ(a(]'Tov v OE2 E'li"LTYJO<V<w, els 8 &v 'li"EcpvKW> if, each one is to pmctise one thing, fm which his nature is best fitted; though this might l'e -rrcpvK, like 71"E'li"OYJK< in the first example). PLAT. Rep. 433 A.




122. The perfect infinitive in indirect discourse generally represents a perfect indicative of the direct form. E.g.
<Prwi ToVTo r.orpaxvat he says that he has done this; cf>r; rovro 7r11"paxvat, he said that he had done this; cf>~(]'H Toiho 71"1/"paxhat, he will say that he has done this (the direct form in each case being




'fr{rrpaxa). ''ErpYJ XP~f1-a.8' ~a:vrcjl rov~ 8YJf3a[ovs t7rLKEKYJPVXEJ'at,

he said that the Thebans had offered a rewaTd for his seizuTe. DEM. xix. 21. In An. Nub. 1277, 7rpoa-KEKAfja-8a{ fl-OL ooKe'is (according to Mss. Rav. and V en.), you seem to nw to be suTe to be summoned to couTt (to be as good as alTeady summoned), the infinitive represents a perfect indicative referring to the future (51). There is probably a regard to the perfect of the preceding verse, a-ea-e'i:a-8cd fl-OL ooKe'is. So TRue. ii. 8 : iv ro-6r'f re KEKwAva-8at iMKn ~Kaa-r'f rd. 1rpayfl-ara < fl-~ rts ai>r6s 1rap~rrrat, and each man thought that things were the same as stopped in that matteT in which he was not himself to tcdce paTt. After a verb of swea1ing: IJtJ-vve fl-YJO~v elpYJKEvat 1repl ai>Tov <f>avA.ov, DEM. xxi 119. After JA.1rCw: f.A.rrCwv rov A.e0v rerpva-8a,, HDT. i. 22 (see liS, above).

123. The perfect infinitive rarely represents a pluperfect of the direct fcrm. E.g. A~yerat llvopa EK7rE7rA1)x8at 7roA-6v TLVa xpovov r.l r</i
rov Kvpov, it is said that et num had been st1uclc 'with a?JW>!ement for some tinw at the beauty of CyTus (i.e. J~e7r~rrA>JKTo). XEN. Cyr. i. 4, 27. 'AvrEA.eyov, A.yovres fl-1) E7r'Y) yy EA &a L 7rW TO-S' fT7rOJ!Od.s EfT~7rf1-tfUV 'ToVs o7rA[..,.as, saying that the t?uce had not yet been proclaimed ( THUG. v. 49.


124. The aorist optative in indirect discourse may represent1. The aorist indicative of a leading verb. E.g.
''EA.e~av on 7rEtJ-lfELE a-rpas (3aa-tA.e-6s, they said that the king had sent them (i.e. they said Z7r<f1-tf11 1}tJ-&s o (3amA.evs). XEN. Cyr. ii. 4,


ToT Jyvwa-e'Y) on OL /3ap(3apot TOJ! avepwrrov V7r01fEfJ-tfULEV,

then it became known that the baTbwians had sent the man. XE:\1, An, ii. 4, 22. 'Er6Af1-a >..yew <ts rroA.A.d. TWI' Efl-WV A.a(3otev, he daTed to tay that they had taken (EA.af3ov) much of my pTopeTty. DE~L xxvii. 49. 'Hp<vrwv avr6v el avarrAeva-Eev exwv apy1~pwv, I asked him whetheT he had set sail with money (i.e. I asked him the question, avrrA.eva-as ;). DEM, L. 55. (This form is rare; see 125.) 'Errnpwra rtva Zoo,, he asked whom he had seen (i.e. r{va flOf>, whom did you see?). HDr. i. 31. So i. 116; dpero Ko8ev A.a(3ot.

2. The aorist subjunctive of a dependent verb.

Ev~avro a-wr~pw 81~a-ELv


f.v()a 1rpwrov els rptAtav y~v drp{Kotvro,

they vowed that they wottld make than!; o.ffeTings fm theiT deliverance wheTeveT they should fint enter a friendly land (i.e. ev()a &v arptKWfJ-E()a, 8-6crofev). XEx. An. v. 1, 1 (see iii. 2, 9).
An aorist indicative in

dependent clause of a quotation is regularly

retained (689, 3).




3. The aorist subjunctive in a question of appeal (287).

Oi 'Emoa_~Lvwt Tov 8Eov ~7r~povTo El 7rapaoo tV Kopw8tots T~v '11'6.\.w, they asked whether they should deliver up their city to the Oorinthians (i.e. they asked the question, 7rapaOWfLEV TIJV 7!"6.\.tv; shall we deliver up our city?). THUC. i. 25. 'Ea-K67rovv 87!"w!i Kd.AA.ta-7 f.veyKatJL' a-&r6v, I loolced to see how I could best endure him (i.e. I asked, 7r<;:;> v y Kw avr6v; how can I end1tre hirn?). EuR. Hipp. 39:3. Ateo-L<fJ7n)'TE O"K07TWV 0 'Tt a7TOKpli!Ul'TO, he continued silent, thinking what he shmdd answer (i.e. thinking T[ &7roKp[vwfLat;). XEN. Mem. iv. 2, 10. (See 677.) 125. The context must decide whether an aoriRt optative in an indirect question represents the aorist subjunctive (as in 3) or the aorist indicative (as in the last examples untler 1). Thus the first example ui1tler 3 might mean they asked whether they lwd given up their city, 7rapeOOfLEV r~v '11'6.\.w; But in most cases the aorist subjunctive is the direct form implied, and an aorist indicative used in a direct question is generally retained; El dva7rAnJo-tEV in 1 is, therefore, exceptional.

126. The aorist infinitive in indirect discourse represents an aorist indicative of the direct form. E.g.
P'l)<Ttv Tovro '11'0 t?j<Tat, he says that he did this (i.e. he S(lys Tovro ; ~</>'l) Tovro '11'0 t1ja-a t, he said that he had done this (i.e. he rovro E'1r'ofJo-a); <f>~a-Et rovTo '/f'Ot~a-at, he will say that he did tfois (i.e. he will say rovro E'1r'o'l)a-a). '0 Kvpos Af.yErat yH~a-8at KaJLf3Va-w, Oy1us is said have been the son of Cambyses. XEN. Cyr. i. 2, l. ITaA.at6ra'Tot Aeyovrat EV fLEPEL Ttl't Tij> x!hpa> KvKAW'/f'E!i o 1K1j a-at, the Cyclops are said to have settled most (tnciently in a part of the country. THUC. vi. 2. "'H(J'av V1r01I"TO< a-&roi:s JL~ 7rpofhiJLw> <T</>[<Tt 7reJLtfat a E1rEJLtfav, they we?'e suspected by thern of not having sent to thern with alacrity 'What they did send. Tau c. vi. 75. 127. Although the usage of the language is very strict, by which the aorist 'iufiniti ve afte1 verbs of saying, thinking, etc. is past, as representing an aorist indicative, still several passages are found, even in the best authors, in which an aorist infinitive after such verbs as VoJL[(w, o1ofLat, and even <f>'lJfL[ refers to future time. Many critic.s, especially Madvig, 1 deny the existence of this anomaJy, and emend the offending aorists to the future or insert av. If they are allowed (and most of the passages still ~tand uncorrected in many editions), they must he treated as strictly exceptional ; and no principle, and no consistent exception to the general principle, can l)e based on them. E.g. <l>dTo yap T[a-ao-8at J.A.dras, for he said that he should punish the


1 See Madvig's Bemerkungen uber einige Puncte der griechischen Wortfugwngslehre, pp. 34-44 : Griech, Synta;x, 172 a, Anm.




offenders. Ocl. xx. 121. (In Il. iii. 28, we have in most :M:ss. and editions 4orf:ro yd.p T[uw8at &.A.dr'Y}v, in precisely the same sense; but Bekker has rtrrarrOat.) So f.4ocfJ1-1JV T[rrarrOat in Il. iii. 366. Kal a1rrcp oB fL~fLtpau8at 'A7rptTJV (se. d7rKpvaro) 7raptuu0at yd.p Kat avros Kat lLU.ov.- at<tv, and (he answered) that Apries should not have reason to blarrw hirn ; joT he not only would be present himself, but would bring others. HDT. ii. 162. (Notice the strange transition from the aorist ('?) to the two futures.) <l>1JUtV ovo~ T~V ~LOS ''Eptv 7rEO<tJ UK~t{arrav EfL1rOOWV ux<8'iv. AESCH. Sept. 429. OifLO.L yap VtV tKTVUa l raO, I think of imploTing. Eun. I. A. 462. (Hermann reads iKrdrrLv by conjecture.) 'Ev6fLL<TaJI E7rt8EfLVOL pq.Ofwr; Kpar~ uat, they thought they should gain the victory. THUC. ii. 3. No!'-l(w, ~v i1r1rEils yevwfLat, &J18pw1ros 1rT'fJJIOS yEVtu8at. XEN. Cyr. iv. 3, 15 OvK i!4oauav E1r LT p Etf a ravra y<JI~u8at, they said they would not permit this to happen. LYs. xiii. 15; same in xiii. 47. Tovro o~ oi'ra oi fLaAuna y<ve0"8at, l O"Ot rrvyyevotro, and he thinlcs that this would be rnost likely to happen to him if he should join hirnself with you. PLAT. ~rot. 31,6 C. (Here we should expect yve0"8at aJI, to corresponu to H UV)')'JIOlTO.) An. Nub. 1141 is commonly quoted in this list, as having OtKauau8a [ <f>arr fLOL in all Mss.; but in the year 1872 I found OtKa0"<0"8at in Cod. Par. 2712 (Brunck's A) and by correction in 2820, so that this emendation (as it is commonly thought to be) is confirmed. It may be thought that the aorist is less suspicious in the Homeric passages than in Attic Greek, where the uses of indirect discourse are more precisely fixed.

128. The future optative is used chiefly in indirect discourse after past tenses, to represent a future indicative of the direct form. Even here the future indicative is generally retained (670, b). E.g.
'Y1rH1f"WJI TaAI..a oTL aDroc; rdKEZ 1rpato t, 0xro, having suggested as to what remained, that he would hirnselj attend to things there, he departed. THUC. i. 90. (Here 7rpafot represents 1rpcftw of the direct discourse, for which we might have 7rpafH in the indirect form. See, in the same cha1)ter, d7roKplJiaf'-VOt oTL 7r~f'-fovrnv, having replied that they would send, where 7rEfLtfOV might have been used.) Et rwa <f>-6yovra A1JtfOLro, 7rpOYJ)'DpEV<V OTL ws 7rOAfL[cp XP~Uotro. XEN. Cyr. iii. 1, 3. (Here the announcement was Et rtva A~IfOfL<u, w> 1rOAfLL<f' "EA)'JI OTl ETOlfLO> t'f} ~yw8at ailro'is ls r6 ~Et..ra, v8a 1roA.A.d. A.~if;otJITo. XEN. An. vii. 1, 33. (He said rotfLO> dfLt . ~v8a A.~tffT{).) Here belongs the rare use after

in THUO~ vi. 30, /.LE~ JA7rSo~



o' c:r


d.fta Kai JAocj,vpfL::Jv, T<i p~v ~~


/JtpotVTO, i.e. (they Saileil) with hope and




lamentations at once,-hope that they might acquire Sicily, lamentations at the thought whether they should ever see their friends again (6fofl'e8a.;). 129. The future optative occurs first in Pindar, in an indirect question, EKEAWIJEV owKp'i:va.t <lv-rtvo. IJX''JIJOL ns ~pwwv, to decide which maiden each of the heroes should take (T[va. IJX~IJEL)), Py. ix. 126. It is used chiefly by the Attic prose writers, as the correlative of the future indicative, that te11se having had no corresponding optative form in the older language, as the present, perfect, and aorist indicative and subjunctive ha<l. It is never used with av.

130. Apart from its use after verbs of sa.ying and thinking, the future optati1re is found in object clauses with o71"w> after verbs of striving, etc. (339). Here its use is closely akin to that in indirect discourse, as it always represents thought which was originally expressed by the fdture indicative. E.g. 'E7rEJ1-EAEtTo o1rws fl'~TE aiJLTOL fl"'JTE d71"oTot cro tv-ro, he took care
that they should be neitheT withrmt food nor without d1inlc (his thought was o71"ws fl''JTE i!crovTa.t). XEx. Cyr. viii. 1, 43. 'E7rEJ1-e.A,)8o7 07rWS oi CTTpo.nwTU.L TOD> 7J'C:lJ!ovs ovv-,)crotvTo V7ro<jopetv. XE:-<. Ag. ii. 8. M?)o~v oEov aA.Ao fl'')XUVaiJ8at, i) 01rW> 1Jf1-tV OTL KaAAL<TT(J, TOD<; vofl'ovs 8~otv-ro wcr1rEp f3a<j>-,]J!. PLAT. Rep. 430 A. See Tim. 18 C, J1-1)XaVWf1-EVOVS 07/'WS Jl-?)8ds yvi!JIJOLT0 1 VOJl-lOVIJl DE 7rrf.JITES (where yvwcrotTo represents yv<!JcreTa.t, while the next word JIOfl-WVCTt is retained in the in<licative). ,EIJK07rEL MeJ'EKA?)s 01rws fl'YJ EIJ"o tTo a1ra<s, d.AA' tcrotTo a.imfi ocrns (wvnf TE {?)poTpo<jo-,)IJot Kat TEAEvTojiJavTa ea fo t aBT6JI, Kat eis 'TCJV E71'ELTU XPOJIOJI TU VOfuC6J1-EVa avT~ 1r0 L~(J"O t, Menecles toolc thought tha.t he 'Jm:ght not be childless, bu.t rnight have some 01~e to support his old age U(hile he lived and to buTy hin1 when he died, etc. IsAE. ii. l 0 (see 134). Other examples are XEN. Cyr. viii. 1, 10; Hell. vii. 5, 3; Oec. vii. 5; PLAT. Ap. 36 C; Isoc. xxi. 13; !SAE. vi. 35; DEM. xxvii. 40 (o1rws fl''crBwcrotTo, in the J\.fss.) In XmN. Hell. ii. 1~ 22 we ,hav~ with th~ f;tt~re optative: 1rpoei7rev Jl-?)OELS KLJI?)CTOLTO EK T'f}S Ta~EWS Jl-1)0 aJI(J,~OL'TO. In all sueh cases the future indicative is generally retained (340).



131. The future optative is found in four passages after verbs of fearing, three times with fl'~, and once with 01rws fl'~ : KaTf3a.Ae ,-6 'HpaKAew-rwv -reixos, ov TOVTo <joof3ovfl'EJIO>, /'-~ nves 7rope.,;CTOLJITO J1ri TYJV JK<,,ov 8vvaf1-tV1 notfeMing this, lest any should rnarch into his dominions. XmN. HelL vi. 4, 27. So XEN. Mem. i. 2, 7. 'A.A.\ii /Cat 'TOV<; 8eovs &v EOELCTa.<; 7CapaKWOVJIEVELV, fl'YJ OVK op8w<; avTo 1ro "J(J"O t<;. Pr,AT. Eutbyphr. 15 D. Ov f1'6vov 1r<pt T~> f3a1Javov /Cat T~S o[K?)> EOEOO[KEL, dA.Aa Kat 7rpt TOV ypafLJl-aTEtov, 07CWS fl'YJ v1ro Tov .MeJ'E~EI'OV IJvAA?)<f;{I,]IJOLTO. Isoc. xvii. 22. (Here the fear was expressed originally by I'J7rWS fl'''J crvAA?)</>8~cre-rat, 370.) .As fl''l with the future in<licat.ive is rare after verbs of fearing (367), it is still rarer with the future optative after auch verbs. 132. No case is quoted of the future optative in a pure final



clause, except a peculiar one with 1'-~ in PLAT. Rep. 393 E: 'Ayai'-~ P-vwv 'Jyp!atv<v, f.vr<AADfk<Vo~ vvv T< a1nevat Kat a:Oe,, 1'-~ f.A8iiv, 1'-'J aim:{! r6 r< a-K'fj7rrpov Kat ra rov Bwv a-TEfkfkaTa ovK e1rapKf.a-ot. (Another reading, e7rapKea-H<v, of inferior authority, is adopted by Bekker.) If e1rapKeirot is l'etained (as it is by most editors), it can be explained only by assuming that Plato had in his mind 1'-'J ovK e7rapKEa-H as the direct form. M0 final with the future indicative occurs in Aristophanes, Homer, and Theognis (see 324) ; there is therefore no objection to 1'-'l e1rapKea-oL as representing 1'-'J e1rapKea-H. We must remember that Plato is here paraphrasing Homer (Il. i. 25-28), but by no means litemll;-. The Homeric line is Mt} vv rot oll XP a. {a- 1'-II a-Kry7rTpov Kat a-TEfkfka 8wZo (see 263). 133. As !vo. never takes the future indicative, it can never have the futme optuti ve. 134. A future optative rarely occurs in a relative clause of purpose after a past tense; as alp<Bf.vr<s f.q>' cjS-rE a-vyypatj;at VOfkOV~, Ka8' ovO'nvas 1ro ALr<va-o 'vro, having been chosen for the p11-rpose of rnalcing a code of laws, by which they were to g01;ern. XEN. Hell. ii. 3, 11. (Here we have an indirect expression of the thought of those who chose the Thirty, of which the direct form is found in ii. 3, 2, g13o~< rpL15.Kovra avopas Af.a-8ru, oi' TOVS 7rarp[ovs VOfkOVS ~vyypatf;ova-L, KaB' oils 1r0ALT<:V(J"0Va-L.) See lsAE. ii. 10 (quoted in 130).

135. The future infinitive is regularly used only in indirect discourse (111, 112), where it always represents a future indicative of the direct form. E.g.

r pcttj;EtJI </>YJa-v, he Sa.ys that he will write; ypatfELJI f</>YJ, he Said that he would write; -ypa>fEw </>~a-EL, he will say that he will write: all representing ypafw, I will write. lioAAous YE ga-Ea-Ba, ~AEyov rovs f.BEA?]a-oJiras, they said that the1e would be many who would be willing. XEN. Cyr. iii. 2, 26.
136. Verbs of hoping, expecting, ]J1'ornising, swea1'ing, and a few others of like meaning, form an intermediate class between those which take the infinitive in indirect discourse (with the time of its tense preserved) and those which do not. When these refer to a future object, they regularly take the future infinitive in indirect discourse ; but they also allow the aorist and even the present infinitive (uot in indirect discourse), like verbs of wishing, etc. Examples are given of different verbs of this class with both constructions : Tpwa-iv 13' EA7rETO ev,.,_'u, vqas fvt7rp?)a-LV KTEJIEEtJI 8' ~pwas 'Axawvs. Il. XV. 701. 'EeA7rETO KVOOS apea-8at, he was hoping to obtain glory. Il. xii. 407. ''HA7rt(ov yap fkUX?JV a-w8a,, for they




expected that there would be a battle. THUC. iv. 71. 'Ev V .:rr[St wv Tct -rdxTJ alpr}<rEtV. Tauc. vii. 46. 'EA.7Tt(Et Mva-ro~ ELvat apxEtv, he hopes to be able to rule. PLAT. Rep. 573 C. (Compare Elvat in HDT. i. 22 and 30, quoted in ll8.) IIaA.w ~p.o>..: a 7Tapos OV7TOTE ~A7Tt<TV 7Ta8e'iv. EuR. H. F. 746. El yap Kpan}<rHav -rep vavnKcp, To 'Pr}ywv ~A7Tt(ov pq.Uws X<tp<!J<ra<rBat, they hoped to subdue Rhegium. THuc. iv. 24. Ovo' &v EA7TtS ~V av-ra (3eA.Tw yevf.<rBat, there would not be even a hope of their becoming better. DEM. iv. 2. Besides these constructions, EA7T(w (or EA7T[s) has the infinitive with liv in TRue. vii. 61 ; ~s with the future indicative in EuR. EL 919, with the future optative in TRue. vi. 30 (see 128), with the aorist optative and liv in Tauc. v. 9 ; 07Tws with the future indicative in SoPH. EL 963, EuR. Her. 1051. Tov <rTpa-rrry0v 7Tpo<TOoKw -rav-ra 7Tpa~ov. XEN. An. iii. 1, 14. MeveA.ewv 7Tpo<r00Ka p.oAeiv, expect M. to come. AEsCH. Ag. 675. IIpo<TOOKWV pq.Siws ~ 0.7T a-r{j (TO. t. lSAE. xi. 22. 'Y7T6 -r' e<rxe-ro Kai Ka-revev<rev Sw<rep.<vat. Il. xiii. 368. 'EK -rov-rov v7Te<Txe-ro P-TJxav~v 7Tap~etv. XEN. Cyr. vi. 1, 21. };l, yap we<Txov (YJT~<Tetv. PLAT. Rep. 427 E. 'Y7To<Tx6p.evos p.~ 7Tp6<r8ev 7TO.V<Ta<T8at 7Tptv aVTOlJS Ka-raycfyot orKaOe, having promised not to stop until he had restored them to their homes. XEN. An. i. 2, 2. 'Y7TE<TXETO p.ot (3ovA.d<Tw8at. Ib. ii. 3, 20. 'flp.oA6n<Tu. els -rr}p.epov<re<T8at. PLAT. Symp. 174 A. 'Op.oA.oyr}< 7Totr}<rov TO KEAev6p.evov. Id. Phaedr. 254 B. So ANT. vi. 23 ; AND. i. 62. Compare <f>ap.~v Tov-rov .1!p.oA.oy1)KEVat -rav-ra 7TOtrJ<THv with <f>a<rKovTes <TE .1Jp.oA.oy1)KEVat 7ToAtTEVe<rBat, PLAT. Crit. 51 E and 52 D. See Crit. 52 C; and compare ~vvWov 7ToAtTEved8a.t, ib. 52 D. 'E7Te[<TB'l}V T~V <Tvvo.Sov Tii oyoou op.oA.oy{j<Tat 7TO tr}<Ta.<TBa.t. DEM. xlii. 12. ' 'Hyyva-ro fL1Jb~v a.vTovs KaKoV 7Td<rw8at, he pledged himself that they should suffer no harm. XEN. An. vii. 4, 13. Ilpo<Taya.ywv eyyv'l}Tas ~ p.~v TropevwBat, having given secu1ities as a pledge that he would go. Id. Cyr. vi. 2, 39. Kai or} p.ot yepas ailTos d<f>a.tpr}<rwBat d7TEAE'is. Il. i. 161. So xv. 179; Od. xi. 313; HDT. vi. 37; EuR. Med. 287. 'H7TetA1]<TEV v{ja.s' {AKEfJ-EV. Il. ix. 682. 'H7Te[A'I)<Ta.v a7TOKTE'iva.t fi7Ta.vTa.s Tous ev -riJ olK[q.. XEN. Hell. v. 4, 7. Taxa. ELKOS <TVV avT</) (3ovArJ<rE<rBat Eiva.t, it is likely that soon nobody will want to be with him. XEN. Cyr. v. 3, 30. 'EK p.(v Tov Ka.Kws 7TpaTTetv Tas 7TOAES p.eTa.(3oA.{js -rvx E'iv e7Tt TO (3A.nov E1Kos


0~ ToV 7ravT&:rraa-t yvEa-8at &.vJo-TaTov Ka2 -rWv Kot.vWv

EA7r{Owv <rTEp') B{jva t. LYCURG. 60. "Op.o<r<TOV ~ p.~v p.ot apr}~EV. Il.i. 76 j sox. 321. 'Op.o<Ta.S <ha~fV orKa.o', E) Tpoav p: fiyE, SOPH. Ph. 941; cf. Ph. 594, 623. 'Op.6<ra.vTE> TavTa.ts Jp.p.everv. XEN. Hell. v. 3, 26. 'Ava.yKaCE Tov Kep<To/3MrrTYJV ofL6<rat dvat fL~V T~v dpx~v Kow~v, 7TavTas 8' Vfl-LV <hooovvat T~V xwpav. DEM. xxiii. 170.





137. The future perfect of the dependent moods is rare, except in verbs whose perfect has the meaning of a present (49), where it is an ordinary future (82). "\Vhen it occurs in other verbs, it is only in the infinitive of indirect discourse. E.g. Tavra (e</J'YJ) 1re1rpa~w8at ovoi:J' 1) Tptwv ~f-Lpwv, he said that we
should see these things alnady accomplished within two or tlwee days. DEM. (Here the direct discourse was 1rE1rpa$erat Tavra, these things will have been already accomplished.)

xix. 7 4.


138. The tenses of the participle generally express time present, past, or future relatively to the time of the verb with which they are connected.
The uses of the participle with liv are not included here. these see Chapter III.


139. The present participle generally represents an action as going on at the time of its leading verb. E.g.

Tovro 1rOWV(TtV vopJ{ovTS 8Kawv elvat, they do this thinking it is 'E1roiovv vopJ{ovT<S, they were doing it in the thought, etc. 'E7ro['l}(Tav VOf-L[{ovTs, they did it in the thought, etc. llot~(TOV(TLV vopi{ovTs, they will do it in the thought, etc. Tavi 1rpax81 K6vwvos
ix. 56.
(TTpaTYJyovvros, these things we?'e done when Canon was geneml. Isoc. ('2TpaT'l}yovvros is present relatively to ~7rpax8YJ.) KaTot -ra.vra 1rparrwv r 1roet; now in doing this what was he doing? DEM. ix. 15. Tavra 1repu8ev ytyv6f-Leva, to see this go on. DEM. xviii. 63.

140. The present participle is also used as an imperfect, like the present infinitive (119). With the participle this use is not confined (as it is with the infinitive) to indirect discourse. E.g.
0( U"Uf-L1rpe(T{3EVOVTE'> Kat 7rap6VT'> KaT<Lf-L<LpTvp~(]'OllfrtJI, those
who wen his colleagues on the e1nbassy and who we?e present will testify.

xix. 129. (Here the eml,assy is referred to as a well-known event in the past.) .Pav<-rat yap 1) vvv ' KaAovf-LEVYJ o-6

{3e(3a{w<; olK01JjLEV'f], JA,\a f-LETaVa(TTa(]'HS T ofJ(J'Ut 1rp0Tpa, Kat pr:;.8w; EKa(TTOt T~v avTWI' &7roAEL7rOVTE>, i.e. the following things are evident, 'E,\ ov 1ra/ (3e(3a[w<;; <jJKEtro, 6./.../...a f-LTava(]'Ta(THS





~<rav, Ka~ lKacrrot T~v iavrwv d7r~AH7TOV. THuc. i. 2. OZOa r6v "' ' ~ ' " t ~ ' ' '' ' .iJWKpO.T1)V OHKVVVTO. TOtS t,;VVOV<TtV eaVTUV KO.II.OV KO."'(O. ()' OVTO.' VV " oU3a. o~ KJ.Kdvw ITWcppovovvTE EU"T l:wKpdrf.t <TVV~U"TrJV. XEN. Mem. i. 2, 18. (The direct discourse was EOdKVV and ea-wcppovdT1)V.) In THUC. iv. 3, ~ Il-6..\os ~U"TtV ev rii Mea-U"1)V[~ 11"0T~ OVU"TJ yiJ, Pylas is in the country which was once Messenia, oVU"TJ is imperfect, and denotes time absolutely past, as is shown by 7TOTe, without which it would be the country which is (now) Messenia. 141. An attributive present participle (824) occasionally refers to time absolutely present, even when the leading verb is not present. This is always denoted by vvv or some other word in the context. E.g. T~v vvv BotwTav KaAovfl'eV1JV ~K1J<rav, they settled in the country now called Boeotia. TRue. i 12. '0 Tovvv tl>[A.tmros i dpx_~s, oiJ7rw Aw7rdeov<> ITTpO.T1J"(OVVTOS, oi>Be rw~ OVTWV Ell Xeppov~<TCfl vvv d7rEITTO.Afl'evwv, l:eppewv Kat D..op[a-Kov l..\J.f-Lf3ave, Philip then in the beginning, when Diopeithes was not yet general, and when the soldiers who ARE NOW in the Ohenonese had not yet been sent out, seized upon Serrium and Doriscus. DEM. ix. 15. (Here <rrpar1)yovvros is present to the time of A.d.JLf3a.vE, while &vrwv is present to the time of speaking.) For a corresponding use of the aorist participle, see 152.


142. The perfect participle in all its uses represents an action as already finished at the time of its leading verb.

. E.g.
'E7Tawova-t Tovs dp'r)KoTas, they praise those who have spoken. 'E1ryv.:<Tav ToVs dp1]Koras, they praised those who had spoken. 'E'lTatvea-ov<Tt TOV'ii dp1] KOTas, they will praise tlwse who will (then) have spolcen. 'E7rEOEL~a ovo~v J.A.1J()~s &7T1J'Y'Y" AKO TO. (Ala-xGJ11]V), I showed that Aeschines had announced nothing that was true (i.e. I sho1ved, ovov dAl)8s d'lT~yyeA.Kev). DEM. xix. 177. Tovs 8ea-fl'<~Tas f-LETEf-LEAovro a7rooe8wKores, they repented of having restored the captives. THuc. v. 35. T~s AloAtoos xaAE7rW'ii ~cppV a7rEU"TEP1JJLEVOS, he took it ha_rd that he had been deprived of Aeolis. XEN. Hell. iii. 2, 13.

143. The aorist participle generally represents an action as past with reference to the time of its leading verb. E.g.
Tavra 1r0 t~<Ta VT'> a7rEA8otV (3o.UAovrat, ha-ring done this, they (now) wish to go away. ' TavTa el7ro VTES a7rqAeov, hating said this, they went away. Ov ?ToAA.ot <f>a.[vovrat [11V ..\86 VTE'>, not many appear to have joined in the expedition. THuc. i. J 0. BotwTot g 'ApV1J'> a VO.U"T<f!lres T~V BotwT{av iJKYJU"av, Boeotians who had been driven



from A rne settled Boeotia. TRue. i. 12. "E<fJafLEV o&e ETrtU'T'!JfL'YJV oi5Te liyvotav i_.ff al!Tif EU'EU'8at, aAAa TO fLETagv av </Javev dyvo[a<; Kat f.TrtU'T~fl-1)>, i.e. we said that it would be the p1ovinee of neither knowler1Je nor ig1wmnce, but of that which should hctve appeared (</JavEv) in diw course between these. PLAT. Hep. 4 7 8 D. (Here <f>avEV is past to Eunr8at, though absolutely future; see 22.) , A<f>[KETO oeDpo TO TrAowv, yv6vTW V TWV Ke<f>aA.A.~vwv, aJirt7TpdTTOYTO<; TOVTov, EVTav8a KaTa11' AELY avT6, the vessel arrived here, the Oephallenians having determined that it should return to this port, although this man opposed it. DEM. xxxii. 14. (Here yvovTwJI denotes time past relatively to a<f>[KeTo, and avrt7rpttTTOVTOS time present relatively to yvovTwv, which is its leading verb.)

144. When the aorist participle is used with any form of

A.av8ttJiw, to escape the notice of, Tvyxavw, to happen, and <fJ8avw, to a-nticipate, except the present and imperfect, it does not denote

time past with reference to the verb, but coincides with it in time. Thus EA.a8ov a1reA.86vTes means they went au;ay secretly ( = a1r~i\.8ov
i\.a8pa) ; OVK (cp8'Y]crav a7rei\.86vres, no sooner were they gone ( = ov 1rponpov a1r~A.Bov); i!rvxov elueA.86vns, they came in by chance, or they happened to come in ( = elcr~A.Bov Tvxv) E.g. Toils 8' Ei\.a8' elueA.8&!v ITp[afLos, and Priam entered 1tnnoticed by them. Il. xxiv. 4 77; so xvii. 2 and 89. ''EA.a8~v (avT0v) a</> 8vTa 1rttvTa Kat KaTacpi\.ex8hTa, everything took fire and was cunsumed before she knew it. TRue. iv. 133. Aav8avet (historic !)resent) U'T~A'YJv 1ra [uas. SoPH. El. 7 44. "E<f>8'YJ ope~ttfLEYOS, he aimed a blow fint Il. xvi. 322. AvTot <f>{}~(JOJITO.t aUTO opaua YTES, they will do it first themselves. PLAT. Rep. 375 c. Ov yap cp8Yj fLOL uvp.{3affa ~ dTvxta, Kat ev8vs iTrEXe[p'Y]uav, K.T.A., joT no sooner did this misfortune come upon me, than they undmtuok, etc. DEM. lvii. 65. LTpaTLa ov 1roAi\.1) enxe fLEXpt 'lu8fLov 7rapEA8oDcra, an anny' of no g1eat size had by chance marched as jar as the Isthmus. T11uc. vi. 61. "EnxE o Ka.Ta TOVTo Tov KaLpov A8 v, and he happened to come just at that moment. Id. vii. 2. 'Oi\.iya 7rpus TU fLEAi\.ovTa TVXE2v 7rpcl~aJITS (se. ~yovvTa.t), they think that it was their fortune to accomplish ouly et little in compurison with their expectations. Id. i. 70. So TovT' ETvxov i\.a(3wJI, I happened to take this, A it. Eccl. 3 7 5. '07r1rOTEpOS KE </J8iJcrLV opegttjLEYOS xpoa KaAov, v;hi,-Jwver shall first hit, etc. Il. xxiii. 805. Bovi\.o[J.L'YJY ch i\.a{)e/:y avT~V a7rei\.8wv, I should like to get away without his lcnowin'J it. XEN. An. i. 3, 17. Tot>> dv8pw1rov<; i\.~(]'OfLEV e7rt7reuoYTe<;. lb. vii. 3, 43. Evi\.a(3eZcr8a.t 7rapKEAvecr8 ill~AoLs-, 81rws p.~ 1rtpa Tov 8iovTos uocpt!JTEpot yv6/l-EVOL A~U'ETE fna<f>8aplvT>, you exhorted one another to take care not to become wise ovennuch and so get corrupted unawares. PLAT. Gorg. 487 D. (Here yev6p.evot is an ordinary aorist, past with reference to the future phrase A~crT 8wc/l8apevTE<:.) The last four examples show that this use of the aorist participle is allowed even when both participle and verb refer to the future.


145. The aorist participle has the same use with uvp:ll't-rr'Tw, to happen, in Herodotus (890). E.g. Ka2 'TOOE l'TEpov uvv~7rEUE yEvo JLEVov, and this other event occurred (as it chanced). HDT. ix. 101. So uvyKvpew in HDT. viii. 87 (see 889). 146. An aorist participle with the present or imperfect of any of the above verbs (144) cannot coincide with the verb in time, and retains its own reference to past time. This combination seldom occurs.1 E.g. ''07rEp A.a{Jovua 'Tvyx<fvH fi-~'T'f)p xpow, which, as it happens, the mothm has taken in her hands (happens to have taken). EuR. Bacch. 1140. ''Apuna 'Tvyx<fvovcn -rrpa~avns, it happens that they fared the best. ' I '' \ > \ > 1 > \ '(3 ,1 I soc. IV. l 0 3. LA>.tKatw<; av 'T'f)V av'T'f)V EvEpyEutav a7roAa OtfLEV, T)V7rEp .. av'TOt 'Tvyx<fvOJLEV Els VJLaS v-rrap~a vns, we should justly 1'eceive back the same kindness ?JJhich it is owr own fortune to have .forst shown to you (we happen to have begun). Id. xiv. 57. Ilpos 'TL 'Tovi' d1rti>v KvpE'is; wherefore did you chance to speak thus (does it chance that you spoke)? SoPH. EL 1176. lloii Kvpii: h'To7rtos uv8Es; Id. 0. C. 119. "Opa Ka8' v1rvov JL~ Ka-ravA.tu8ds Kvpfl, see lest it may chance that he has retired to sleep within. Id. Ph. 30.. Compare uvvEK{;p'f)UE -rrapa7rEcrovcr;L, happened to collide. HDT. viii.. 87 (889). M~,, JL{a A.~s n Ka2 ~oov~s ~VJL7r{7rTEt YEVOJLEVT), i.e. happens to have occurred (Badham proposes ytyvoJLEVT)). PLAT. Phil 4 7 D. Ovo' llpa KpKT)V J~'A8Ew A8ov'TES ..\~Bofi-EV,norwasitunknown to Girce that we had 1eturned from Hades. Od. xii. 16. "Ocrot -r{;yxavov oi!-rws a&poot ~vvE~E..\Bov'TE>, all who happened to have thus come out together. THuc. iii. 111.. EZ -r 1rov alywv -rrEpt..\Hcp8ev -r{;yxavE yevos, if any race of goats happened to have been left. PLAT. Leg. 677 E. , Aptcr-ray6pYJ 0~ CTVVETrtTrTE 'TOV avTOV xpovov TrttV'Ta CTV VE ,.\ 86 V'T a, and it was the fortune of A. that all these came to him at the same time. HDT. v. 36. (Here it is difficult to distinguish the doubly past time; but the analogy of the other examples, and the difficulty of conceiving an imperfect and aorist as coincident in time, seem decisive.) 'Op8ws crept ~ cp~fh'f) uvve(JawE Bovua, rightly, as it happened, had the report come to them. Id. ix. 101. Just below: T~S av'T~S ~fhEPYJ> crvv(3awE yvEu8at, i.e. they (the battles of Plataea and Mycale) happened to fall on the same day. In LYB. xii. 27 we have the aorist and perfect participles together with i'T{;yxav, each expressing its own time: ocrns civTHTrWV YE E'T{;yxavE Kat yvWfh'f)V cl'll'oOEOtty}Lhos, who chanced to have spoken in opposition and to. have shoum his opinion. It appears from these examples that the aorist participle can coincide in its time only with forms which have a similar aoristic or complexive meaning, while in other cases the verb and participle are distinct in time . 1 For the examples of Tll/'xavw here given I am indebted to an unpublished


paper on this construction by Dr. James R. Wheeler, in which notice of this peculiarity is taken for the first time (so far as I am aware).




147. l. The perfect participle can always be used with the verbs of 144 to denote an action which is completed at the time of the leading verb. This is the most common way of expressing past time in the participle here. E.g. 'ETvyxavov apn 1T'apetAT)<j>6Te<; T~V dpx~v, they happened to have just received thei1 authmity. TRue. vi. 96. 'Edv ns ~OtKTJKW<; n Tvyx<fvv T~v 1r6A.w, if it ever happens that one has wronged the city. DEM. xviii. 123. So TBuc. i. 103 (see 887). 2. The present participle with these verbs is regular, representing an action as going on at the time of the verb. See PLAT. Crit. 49 B and the four following examples (with others), in 887.

148. ln many constructions in which the aorist participle follows a verb in the sense of the ordinary object infinitive (not in indirect discourse), it does not refer to past time, but differs from the present participle only as the aorist infinitive in such a construction would differ from the present (96). This applies especially to the participle with 1repwpw and J<j>opw (1repte!8ov, ~7T'etoov), in the sense of allow, not interfere with, and 6pw (eioov) permit and see (cf. 884 and 885). E.g. IIpo<Toex6t-tevos ToVs 'A8'Y)valovs KaToKv~<Tetv hpttoeZv aDT~v [n~v y~v] T fLT) t<Ta v, dveZxev, expecting that they would be unwilling to see their land ravaged, etc. Tauc. ii. 18. But in ii. 20 we find the aorist infinitive, ijA.'ll't(ev T1JV yi)v oiYK ilv 'll'epttoeZv TtJ-1)8~vat, would not let their land be ravaged, referring to precisely the same event from another point of view (see 903, 6). M~ 1T'EptiOTJTE 1ltJ-Ea<; ow<j>8aplvTa<;, do not look on and see us destroyi!d. HDT. iv. 118. Oii fL1J <T, ey<il 71'pt6tjotJ-U' d'll'EAeovTa, I will by no rneans let you go. AR. Ran. 509. "ETAT)<Tav E71',0tV p~tJ-'Y)V /)-~V T1)V 1T'6Atv '}'EVOJl.EVT)V T~V OE xwpav 7!'op8ovfLEVT)V, a'll'UVTU 0~ TOV 1T'6AEtJ-OV 1T'pt T~V 'll'UTp[oa n/v ai>Twv 'Y''Yv6fLE v o v. Isoc. iv. 96. (Here the aorist participle denotes the laying waste of the city (as a single act), while the presents denote the continuous ravaging of the country and tl1e gradual coming on of a state of war. This is precisely the difference between the present and aoriBt infinitive in similar constructions.) 'E'll'doov T1JV avTwv 'll'arp[oa dva<TTaTov yevotJ-EVTJV. A::-;T. v. 79. Ei KE~l'OV '}' LOOt/)-' KaTEA86vT' ''Atoos eiirw, if I shoul,l see him go down and enter Hades. Il. vi. 284. l\1-t] Jll8ELV eav6v(J' v'll'' d<Trwv, not to see me killed by the citizens. EuR. Or. 7 46. .6.,a Til <Tw</>povEtv Trf 71'W7roT' eloes i}o'YJ dyaeov n yev6tJ-EVov; AR. Nub. 1061. ''OTav aVTOv rov E~aic/Jvqs 7r'TO. [a- a VTU 7rp0s Tfj 7r6AH Kai f. K X Ea V'Ta Td TE aVToV KaL Eav-rUv, . . . ?} d1ro8av6v-ra i] K7re:a-6vTa ij dTtftW(JV'ra Kat T~V oD<Tiav a'JT'a<TaV d'JT'of3a/...6vTa.. PLAT. Rep. 553 A. So Rep. 498 D, Prot. 324 B; AEBCH. Supp. 423; SoPH. Ant. 476. So after dKoVw; as arK' Jeav<T' d'JT'6VTOS dKOVCfLEV, in case he will hear me speak, ll. vi. 281. To<TavTa <j>wv1)<TavTos El<TTJKOV<Tapw, so much we herd him say. SoPH. O.C. 1645. So also 7rpaBEvTa TAfjva,, endured to be sold, AESCH. Ag. 1041 ; <T1rdpas ETAa, Sept. 754;




for TActw with the regular infinitive, see Isoc. iv. 96, quoted above. So fJ-El!tl! J!O<TT~<TUJ!Ta avaKTa, to await the king's return, Il. xiii. 38.

149. The aorist participle loses its reference to past time also in the peculiar construction in which the participle with its noun has the force of the infinitive with its subject; as p.<Td. 'L-vpaKo-6uar;; olK~u8d uar;;, after the founding of Byracuse ( = p.<Tct TO 'L-vpaKovuar;; olK tu8fjvat), THUC. vi 3. See examples in 829 (b). 150. An aorist participle denoting ~hat in which the action of a verb (generally aorist) consists may express time coincident with that of the verb, when the actions of the verb and the participle are practically one. 1 E.g. N<vu' E7rt or KUAE<Ta<;;, he called him to him by a nod. Od. xvii. 330. Bfj dl~aua. Il. ii. 167.. Ei:; y' E7rOL?)<Ta<;; dvap.vf]uar;; p.e, you did well in reminding me. PLAT. Phaed. 60 0. Mf] TL f.~ap.ctpT'Y)T ep.ov KaTatf'Y)~ t<TafJ-evo t, lest you make any mistake in condemning me. Id.Ap. 30 D. ITaZoa KaTaKavwv ~vf]A.v 7raTct~ar;;, having killed a child by the st?"olce of a daggeT. XEN. Au. iv. 8, 25. ''Ho'Y) 7rW7roTe ovv ij oaKovua KaKov T <Tot i!owKev ~ AaKTl<Ta<Ta; did your mother eve?' do you any ha?m by biting oT kicking you? Id. Mem. ii. 2, 7. IlefJ-7rt ws TOJI 'Aa-Tuox1)]1 Kpv~a E7rWTELAas OTL 'AAKt{3uio'Y)> avTC)]I Tll 7rpcty{J-aTa q>BelpEL, i.e. he sends a p1'ivate message, etc. THuc. viii. 50. After a perfect: oa-' ~!'-as dya8d. of.opaKas elp~F'Y)JI 7rotf]o-as, what blessings you hwve done us 1:n making a peace! AR. Pac. 1199. The following examples among many in the New Testament illustrate the usage : 'A7roKpt8ds EC1rev EJI7rapaf3oA.atr; a1jToZs, A.eywv, he answe1ed and spake to them in J?ambles,, and said. MATTH. xxii. 1. (Aeywv is the ordinary present, less closely connected with et1rov than d7roKpt8e[r;;.) ITpoo-ev~af'-eJioL et1rav, they pmyed and said. Act. Apost. i. 24. E7rO['Y)o-as 7rapayev6fJ-El'OS, thou hast well done that thou a-rt conw. lb. x. 33. 151. In such passages as WfJ-oA.oy'Y)o-aJI TOtS 'ABYJvaCots TE[X'fJ TE 7r<pteA.6vTEs Kat Fav<; 1rapa86J1T<S ~6poF TE Ta~afJ-evot, THuc. i. 108, the aorist participle is past with reference to the time of the beginning of the peace to which WJLoAoy'Y)o-av refers, and the meaning is, they obtained terms of peace, on condition that they should fint (before the peace began) tem down theiT walls, etc. Such passages are Tauc. i. 101, 108, 115, 117. See K1;iiger's note on i. 108, and Madvig's BemeTkungen, p. 46. Madvig quotes, to confirm this view, LYS. xii. 68: V7rE(]"XETO elp~F'Y)V 7rotf]o-ew {J-f]TE OfJ-YJpa oovs f'-~TE Tct TELX'YJ Ka8eAWJI {J-f]T< Td<; vavs 7rapa.8ovr;;, i.e. lw p1'omised to make a peace without giving pledges, etc.

152. An attributive aorist participle occasionally refers to


See the discussion of this, with especial reference to the New Testament,

where examples of this kind are frequent, by Professor W. G. Ballantine, in the Bibliotheca ,Sacra for October 1884, p. 787.




time absolutely past, without regard to the time o its verb.

'H yp.6va 7rapx6p.Evot MEya7ravov TOY Ba(Jv>..wvos VO"TEpov ToVTwv J7rtTpo7rE00"aYTa, i.e. they had as their leade?' Megapanus, who after this was made govemor of Babylon. HDT. vii. 62. (Here the aorist participle is past at the time of writing only; it is even future compared with the time of 7rapEx6p.EYOt.) So in vii. 106: KaTEAt7I"E 8~ avopa Tot6v8E MaO"KUfkYJV ')'EYOp.Evov, and he left M. (in authority), who (afterwa?ds) proved h'imself such a man (the evidence of his later merits follows in a relative sentence). For the corres1)onding use of the present participle see 141. For the use of the aorist infinitive and participle with av, see 207 and 215. For the aorist participle with exw and Eixov as a circumlocution for the perfect and pluperfect, as 8avp.aO"as i!xw and Eixov, see 4 7 and 48. For the rare use of the aorist participle with EO" for the future perfect, see 8L For the aorist participle in protasis, see 472 and 841.

153. The future participle represents an action as future with reference to the time of its leading verb. E.g.
TovTo 7I"Ot?)O"wv f.pXETa.t, he is coming to do this; TovTo 7rot~O"WY ljA8Ev, he came to do this. ITEf.L<f>(j~O"ETa.t Ta.vTa pw v, he will be sent to say this. Oi8a a~Tov TOVTo 7rOdO"ovTa., I.know that he ?oill do this; o2'8a. TOVTO 7rot~O"WJI, I know that I shall do this; fj8EtY a~TOY TOVTO 7r o 1. 1) 0" o vT a, I knew that he would do this. For the various uses of the future participle, and examples, see Chapter VI.





154. The aorist and sometimes the perfect indicative are used in animated language to express geneml t1uths. These are called the gnornie aO?ist and the perfect, and are usually to be translated by our present. 155. These tenses give a more vivid statement of general truths, by employing a distinct case or several distinct cases in the past to represent (as it were) all possible cases, and implying that what has occurred is lik~ly to occur again under similar circumstances. E.g. KaT&av' 6p.w~ or dEpyos d1'~P o TE 7roAH opylhs, the idle man
and he who has labou1ed ?nuch alike 'must die. Il. ix. 320. "OuTE Kat aAKLfkOJI av8pa <f>o (3 El: KO.L d<f>dA7'0 v[KYJY, who terrifies even a valiant




man and snatches his victory away. Il. xvii. 177 (see 157, below). B[a Kal p.ey6)..a.vxov EO"</>a A.ev EV xp6vp. PIND. Py. viii. 15. ~ocpol D~ p.EA.A.ovra rp~ral:ov liVEJI-OV EfkaBov, ovD' 1nr6 KEpDH Id. Neni. vii. 17. Kat oi] cp[Aov Tl') EKTav' dyvo[a') V7rO, and now one ?nay kill a friend through 1:gnomnce. AEscH. Supp. 499. 'AA.A.O. ra ro~avra d, f'-EJ! a?rat Kat f3paxvv xp6vov dvrex<~, KaC (jcp6opa I'' ~VBYJ<TEV E1rt

rat> EA?r[a-~v, av TVXTJ, v-1{) xpovtp DE cf>wpara~ Kat 7rpC al'>ra Karappl, DEM. ii. 10 (see 157 and 171). "Hv apa a-cpaAwa-tJ!, avnA7r[a-avns tlA.A.a 1rA1)pwO'av r0v XP'tav, they s1~pply the deficiency (as often as one occun). THlcC. i. 70. "Hv De ne; ro1!ruw n 1rapaf3avv, (YJJ~-[av avrol:<; hre8(T(J,JI, i.e. they impose a penalty upon every one who tmnsgresses. XE:-<. Cyr. i. 2, 2. .6.WWJI tlYJI'-a 7i'VVfkciTWV i.Ko1'-~(T O'Tevovra 1rovrov. SoPH. Aj. 674. l\1[' 1Jf1-Epa r~v fkEV KaB<'iA.Ev vlj;68cv, TOV D' 1jp llvw . . Euu. Fr. 424. "Orav 6 ''Epw<; yKparea-npos YEVYJTa~, owcpBdpH T 7rOAAa Kat 1JO[KYJO"V. PLAT. Symp. 188 A. ~orav ne; wa-1rp oi'iroc; LO')(lYlfJ), 1) -rr-p0nl r.p6<paa-~<; Kat fkU<pov ?r'Ta'iap.a a1ravra ,iJ!,xanO'E Kat o~Ava-Ev. DEM. ii. 9. 'E1rHOUV ns r.ap' Jp.ov fkafJ17, Jav p.'Ev f3m!AYJTat, a7TOOfOWKl' o cyw 1TpcirTOJI-Ul apyvpwv 0.v OE f'-1J, J.AfJwv EL<; ,,p6v dJ~-6a-a<;, OIJ011 av <Pt7 fltta <tva~ ra 1'-afJ~JI-ara, Toa-ovrov J<.areBYJKEv. PLAT. Prot. 328

B. (Here the perfect and aorist, according to the Mss., are used in nearly the same sense, .he pays. But SaUl)P reads a7rEOWKV for

a1TOOEOWKV.) IloAA.oi Ota oo~aJI Kat 1TOA~TlK1JJI 01JJ!ap.w p.<yd.Aa KaKa 1rE7rov8aO'~JI, i.e. many always have suffered, and many do suffer. XEN. llfem. iv. 2, 35. DE Jl-1l p.1roDwv dJ!aJ!Taywva-Ttp Eiworc- TETp.1)Ta~.


Tm:c. ii .. 45. The gnomic perfect is not found in Homer.

156. The sense as well as the origin of the gnomic aorist is often made clearer by the addition of such words as 1roAM.Kt<;, 1)8;), or ov1rw. Such examples as these form a simple transition frum the common to the gnomic use of the aorist : IIoA.Aa <nparo?r<oa ~017 ~7rfa-<JI v-rr' Vwa-a-6vwv, i.e. many cases have nlready arisen, implying it often happens. Tm:c. ii. 89. .MEA.AwJI y' iarpo<;, Ti] JIOIJ'f O~OOV<; xpovov, ld.O'a'T' i)01) JI-UAAov ~ Tfi-WJ! xpoa, the slow physician, by giving the disease time, may work nwre cu1es than he who cuts too deep. EuR. Fr. 1057. lloAAaKt<; i!xwv ne; ovo ni.vayl<ata J!VV avpwv f1T AOVT'Y}IJ'' WIJT xhf:pov<; TpE<{><tv, i.e. cases have often occu1'Ted in which such a man has become ?ich the next day, etc. PHih Fr. 120. 'AfJvl'-ovvns llvopE> ov1rw Tpo1rawv EfTT1)a-a v. PLA'r. Criti. 108 c. Ozlot<; f1rAOVT1)0'V Taxf:ws oKaws wv, no man ever became rich suddenly who was just. MEN. Fr. 294. Compare DEM. iv. 51. (See Kriiger, 53, 10, A. 2.)

157. General truths are more cohunonly eXJlressed in Greek, as in Engli811, by the present. The present and aorist appear together above, in nearly the same sense ; the gnomic ltorist is, however, commonly distinguished from the })resent by referring to a single or a sudden occurrence, while the present (as usual) implies duration.




Thus in DEM. ii. 10, above, the aorist ~vO'YJO"<V implies a sudden blossoming out with hopes, as opposed to the continuance or repetition expressed by dvTEXH, hold out, cpwpaTat, a?'e detected, and KaTapp<~ fall in ruin.

158. An aorist somewhat resembling the gnomic is very common in Homeric similes, where it is usually to be translated by the present. E.g. ''Hpt11" 8' <1Js OT TtS opvs 1Jpt1HV, and he fell, as when an oak .falls,
(literally, as when an oak once felT). I!. xiii. 389. This can better be seen in the longer and more complicated examples which are quoted under 54 7 and 548.

159. The gnomic aorist is found in indirect discourse in the infinitive and participle, and even in the optative. E.g.
(a) ''Orrov 8' Df3p(etv opav (}' il f3ovAETat rrapfi, TUVT'Y)V J16fLt(E T~V 11"6AtV xp6v<p ?TOTE e~ ovp[wv OpafLOVO"UV ES f3v8uv ?TEO"Ei:v, but where ?nan is permitted to insult and to work his own will, lJeliet'e that that state, though it may ?'un before fair breezes, must in time ~ink to the depths. SoPH. Aj. 1082. (Here ?TEO"EW represents e?TEO"EIJ of the direct fonn, which can be only gnomic.) Er O"ot oos rrapE(]"T'/KEv ~yovfLtil'<p xaAE?TOV elvat cptA[av (]"VfLfLEVHV, Kat Otacpopas '}'EVOfLEV'fJS KOtV~V d.fLcpoTf.pot> KaTaO"T'ljvat T1)v O"VfLcpop&v, if you .fea?, thinking that it is hnrd for friendship to abide, and that when a quarrel occu?'S the cala.mity that arises is common to both (the direct form would be xa-\m6v e(]"TLV, Kat KOtv1) KaTEO"T'YJ ~ O"VfL<f>op&). PLAT. Phaedr. 232 B. 'HyovfLEV'Y]> 01) d.AYJ8Eas OVK av ?TOT <f>a'ifLEV avTfj xop'Ov KUKWV aKoAov8'ljO"nt, now when truth leads, we neve?' could say that a chorus of evils accompany he1 (~KoAov81JO"ev). PLAT. Rep. 490 C. (b) '.2fLtKp(j) xaAtJI(j) o' oloa TOllS 8vfLOVfLEVOVS t11"?TOVS KaTapTV8EJITUS, and I know that high-spi1'ited horses an tamed by a small bit. SoPH. Ant. 478. 01'0a TOVS TOt01JTOV> ev fLEV T(j) KaT' avTovs /3['-1! AV?T'YJPOVS OJ!Tas, TWV o E?THTa d.v8pwrrwv rrpoO"?TO[YJO"LV ~vyyeve[as TtO"t Kat fL~ oD(]"aV KaTaAt1r6vTas, I know that such men, although in their own l?jetimes they are o.ffensive, yet often leave to some who come afte? them n desi1e to clnim connexion with them, even whe1e then is no g1ound for it. TRue. vi. 16. (c) A clear case of the gnomic aorist in the optrttive is seen in PLAT. Rep. 490 B, in the peculiar omtio obliqua introduced by d.rroAoyYJ(]"6fLe8a oTt (in A), which implies a philosophic imperfect (40) and thus takes the optative. We have 11"<</>vKws et'Y), EfLfLEJiot, l'ot, etc., representing 7rE<f>vKe, EfLfLEVEL, el(]"t, etc.; and afterwards yJ!o[rJ TE Kat dkYJ8ws C0YJ Kal TP~<f>otTo (representing eyvw TE Kat d.AYJ8ws Cii Kal Tp~cpeTat), i.e. he attains knowledge (aor.), and then truly lives and is nourished (pres.), where the gnomic force of the aorist is plain, (See 676.)

160. The gnomic perfect is found in the infinitive of indirect discourse iu DE~!. ii. 18: el Of TtS O"W</>pwv i) o[Katos, 7rapewa-8at Ka~




Ev ovDEv6s EtvaL fLEpH T6v Towv-rov (1>rp[v), such a 1nan (he says) is always thrust aside and is of no account. 161. The imperfect was probably never used in a gnomic sense, except where the form is aoristic in other respects, as lKAvov in I!. i. 218, ix. 509 ; cf. xiv. 133.

162. The imperfect and aorist are sometimes used with the adverb to denote a customary action, being equivalent to our narrative phrase he would often do this or he ~~sed to do it. E.g.


f,L1JPWTwv &v a1>To-Ds T{ A.yoLEv, I used to ask them (I would ask them) what they said. Pr"AT. Ap. 22 B. Er TWES rooLEV 7rJ) To-Ds a-1>ETEpov<; ETnKpaTOVVTa<;, av8rip<r1J<rav aJI, whenever any saw their friends ' in any way victorious, they would be encouraged (i.e. they we1e encouraged in all such C{LSC8). THUC. vii. 71. IIo.A.AdKL<; 'JKOV<rap.EV av TL KaKW<; vpJis fJovAEv<Tri}l-EVOV<; Jl-E')'a 7rpfiy}J-a, we used ve1y often to hea1 you, etc. AR. Lys. [j 11. Er Ti<; avT<[j ?rEp[ TOV avnAE')'OL Jl-1JDEV ~xwv a-a<f>'Es .Ayw', J7Tt TfJV vm)8Ea-w e7Ta v~ y EV ltv 7TriJITa Ti'JJI .A6yoF, he always brought the whole discussion back to the main point. XEN. Mem. iv. 6, 13. '07T6TE npoa-(3J-..tfa TLVO'S TWV EV Tat<; nf.~E<Ti, TOTE Jl-EV d?rEV a V. (Jj avDpE<;, K.T.A. TOTE o' ail EV aAAOL<; &v EA~]1. Id. Cyr. vii. 1, 10. S<J Hm ii. 109, iii. 51 and 148. This construction must be distinguished from the potential indicative with ctJI (243). See, however, 249. For the iterative imperfect and aorist with av transferred to the infinitive, see 210.

163. The Ionic iterative imperfect and aorist in -a-Kov and -a-K6fL'JV express the repetition of such actions as the ordinary imperfect and aorist express. E.g. "A..\.Aovc; fLEl' yap 7Tat8ac; f.fLovs 1r6oas w1<vc; 'AxtAAEvs 1rpvaa-x', ov nv' i[J-..urKE. Il. xxiv. 751. ''OKw<; ..\Bot o Nd.Aos J1rl OKTW 7r~XEas, lipoE<rKE ArylJ'ffTov T?JJI ~vEpBE MfL1>ws. Hm. ii. 13. 164. Herodotus sometimes uses the iterative forms in -o-Kov and -a-K6fLYJF with av in the construction of 162. He uses this form of the aorist in only two passages, in both with av. E.g . .PotTEova-a KAa[<a-KE &v Kat 6ovpa-K<To. iii. 119. 'Ec; TovTovs OKW<; e.AfJoL 0 2:KVA?JS, T~V f"~V <rTpaTi~V KaTaAdna-KE f.v T<ji 7rpoa(J'TtL4}, avT6s 0~ OKWS A8oL f.s T6 TEtXO>, .Ad(JE<TK ctJI 'EAA?Jv[lla {a-8~Ta. iv. 78. So A.ri(J{a-Kov liv, iv. 130. See Kriiger, II. 53, 10, 5.





165. In dependent sentences, where the construction allows both the subjunctive and the optative, the subjunctive is used if the leading verb is primary, and the optative if it is secondary. (See 21.) E.g.
IIpaTTO'l!O"tV iiv (3ovAwvrat, they do whatever they please; but l7rpaTTov Cl (3ovAotvro, they did whatever they pleased.

166. In like manner, where the construction allows both the indicative and the optative, the indicative follows primary, and the optative follows secondary tenses. E.g.

AyovO"w on rovro (3ol)Aovrat, they say that they wi.~h for this; on rovro (3ovAotvro, they said that they wishedjoT this.

167. To these fundamental rules we find one special exception. In indirect discourse of all kinds (including sentences denoting a purpose or object after Zva, o1r~s, fL~, etc.) either an indicative o1 a subjunctive may depend upon a secondary tense, so that the mood and tense actually used by the speaker may be retained in the indirect form. (See 667, 1.) E.g.
Ei1l'EV (3ovAerat, for ei'7rEJI (3o{JA.otro, he S(~id that he wished (i.e he said (3ovA.ofLat). 'E<fo(3el-ro f-L?J rovro yE:vTJTat, for f.tjJo(3el:ro fL?J Tovro -yf.vo tTo, he feaTed lest it should happen (i.e. he thought, tfof3ovf-La' fL?J yf.vTJTat). (See 318.)



168 . .An only appa1ent exception occurs when either a potential optative or indicative with av, or an optative expressing a wish, stands in a dependent sentence. In both these cases the original form is retained without regard to the leading verb. It is obvious that a change of mood would in most cases change the whole nature of the expression. E.g. 'Ey~ OVK o2'8' 01l'WS av TtS O"atjJeO"TEpov E7rtOei~EHJI, I do not know how any one could show this more cleaTly. DEM. xxvii. 48. 6.el: -yap EKE[v<t' rovro iv ri/ yvWfLYJ 7rapaO"T~O"at, ~> -&f-Le'tc; iK T~> df-LeAeas ravTTJS T~> a-yav [O"w> iiv opfL~O"atTE. DEM. iv. 17. El 8' -&f-Le'is &A.A.o 'Tt yvWO"EO"fit, 0 !L0 yvo LTO, rva oteO"(h avn}v 'fvx0v E~ELV; Dmr. xxviii. 21. .A few other unimportant exceptions will be noticed as they occur. 169. It is therefore important to ascertain which tenses (in all the moods) are followed, in dependent sentences, as primary tenses by the indicative or subjunctive, and which as secondary tenses by the optative.

170. In the indicative the general rule holds, that the present,




perfect, future, and future perfect are primary, and the imperfect, pluperfect, and aorist are secondary tenses. 171. But the historical present is a secondary tense, as it refers to the past ; and the gnomic aorist is a primary tense, as it. refers to the present.
See HDT. i. 63 (umler 33), where the optative follows an historical present; and DEM. ii. 10, TRue. i. 70, XEN. Cyr. i. 2, 2 (under 155), where the subjunctive follows gnomic aorists.

172. The imperfect indicative in the protasis or apodosis of an unfulfilled condition (410) and in its potential use (243), when it refers to present time, is a primary tense. E.g. "Eypacf)Ov &v ~AtKa BfJ-as ED ?Tot~a-w, El ED i}oetv, I would tell you
in my letter how great services I would render you, if I knew, etc. DEM. xix. 40. ITavv &v ~<Pof3ovfJ-1JV, fJ-1J 1l?Top~a-wa-t A6ywv. PLAT. Symp. 193 E. 'E,Po{3oBfJ-YJV &v a-,P68pa A.y<Lv, f-~ 86~w, k.r.A., I should be very much afmid to speak, lest I should seem, etc. PLAT. Theaet. 143 E. Tavr' Clv ~OYJ Ayav ~?TEXdpovv, Zv' do?) rE. DEM. xxiii. 7 (for the' construction here see 336). See XEN. An. v. 1, 10; DEM. xvi. 12.

173. On the other hand, the aorist indicative in the same constructions (17 2), and also the imperfect when it refers to the past, are secondary tenses. E.g. 'AAAa Kat rovs Oeovs &v ~/Jaa-as ?TapaKtvovvEVELv, fJ-'J oBK 6p0ws
aBr6 7TOL~CTOLS. PLAT. Euthyph. 15 D. 'AAA' o0/J~ fJ-ETa 7TOAAwv fJ-aprvpwv a?TOOLOOVS ElK'iJ TLS 8.11 J?T{CTTEVCTEV, i'v' d TLS /' {yvo LTO ?w,Popa., KofJ-[a-aa-Oat pCf/J[ws ?Tap vtJ-'iv ovvYJTaL. DEM. xxx. 20. (Here the subjunctive 8vvq,rat is properly used after a past tense (318), but the optative shows that the leading verb is secondary.) See Zva y[yvowro, after an imperfect with av, PLAT. Men. 89 B. Xprjv ~?TE[pEa-Oat K6TEpa r1)v ewvrov i] r~v K!Jpov Ayo t &px~v, he ought to have asked whether the oracle rneant his own or Gyrus's empire. HDT. i. 91.

174. All the tenses of the subjunctive and imperative are primary, as they refer to future or to present time (89). E.g.
''E?TEcr&' 07TrJ av TU,; ~y'ljrnt, follow whithersoeVB?' any one leads the way. THUG. ii. 11. 'L,Ko?Twp.Ev El 7Tpf1TEt i} oll. PLAT. Rep. 451 D. 175. But when a subjunctive depends upon a past 'tense, as often happens in final clauses (318), it may be followed by an OIJtative; as in XEN. Hell. vi. 5, 21, .ryyE n)v rnxta-TYJV ELS T~v Ellratav, f3ovA6tJ-Evos d?TayayE'iv -rovs 6?TA[ras 1rptv Ka2 ra 1rvpa -rwv ?ToAEfJ-[wv loE'iv, Zva 0~ TLS d 7T V GJs ,PEvywv J 7T ay a')' 0 L, he led on, wishing to lead off his soldiers before they even saw the enemies' fir-es, that no one might say that he had led them off in jli,tht (187). \Vith the other reading, Zva f-tl Tt<; d'.rot, the example would illustrate l76 A (below).





176. As the optative refers sometimes to the future and sometimes to the past, it exerts upon a dependent verb sometimes the force of a primary, and sometimes that of a secondary tense. A. When it refers to the past, as in general suppositions with El and relatives after past tenses, or when it takes its time from a past verb (as in a final clause), it has the force of a secondary tense, B. When it refers to the future, as in future conditions, in i~s use with flv, and in wishes, it is properly to be considered primary. In many cases, however, a double construction is here allowed. On the principle of assimilation the Greeks preferred the optative to the subjunctive in certain clauses depending on an optative, the dependent verb referring to the future like the leading verb, and differing little from a subjunctive in such a position. A dependent indicative is, however, very seldom assimilated to a leading optative. Such assimilation of a dependent verb to an optative takes place (I) regularly in protasis and conditional relative clauses depending on an optative of future time; (2) seldom in final and object clauses after tva, otrws, fL?J, etc.; (3) very rarely in the case of the indicative in indirect quotations or questions, but (4) more freely in the case of the subjunctive in indirect questions. These four classes of sentences which depend on an optative referring to the future are treated separately below (I.-IV.) 177. I. (a) In protasis and conditional relative sentences depending upon an optative which reje1s to the future, the optative rather than the subjunctive is regularly used to express a future condition. E.g. El'?)s <f>op?)TU<; oilK &v, El trpauuot> KaAt~<;, you would be unendurable,


you should be p1osperous. AEsc:s:. Prom. 979.

'Avopi {3 K' oilK

d~H fLEyas '1\AafL~vws A'tas, 8s &v?)TOS T' Et?) Kat 8ot !:1YJfL~Tpos dKT~v. Il. xiii. 321. Ilws yap flv Tts, d. J' 1-'-0 J-rrCTTatro, Tavra

uo<j>us LYJ; for how could any one be wise in thnse things which he did not undeTStand? XE:~. l\iem. iv. 6, 7. fl.otTo llv o.vTov fLEJ!etv, ECTT uV dtrA8ots. Id. Oyr. V. 3, 13. El dtro8v{JCTKOt fLEV 7r<ivTa oua TOV (0v fLTo.Aa(3ot, E7rt01J o dtro8avot fLEVOt EV TOVT~J, J.p' QV 7r0AA>J dv<iyKYJ TEAEVTWVTa 7rctVTO. -u8vavat; if all things partaking of life should die, and after dying should remain dead, must it not very certainly follow that all things would finally be dead ? PLAT. Phaed. 7 2 0. 'Hs d1roAotTo Kat &A.A.os ns TotavTa J' p~(ot, may a11y othe1 111an also perish who shall do surh things. Od. i. 47. TEBva['l)v, on fLOL PYJKETt ravTa fL~ Ao t, rnay I die, when I (shall) no longer cure jO?' these!

MrMN. Fr. i. 2.



(Here 8-rav p.ry~<en p.eAy might be used without change of meaning. See the second example under b.) 178. (b) On the other hand, the dependent verb is sometimes in the subjunctive or future indicative, on the ground that it follows a tense of future time, especially when the leading verb is an optative with &v used in its sense approaching that of the future indicative (235). E.g.
"Hv oi'iv fl'rf.By> fl'OL -rov-rov, ovK &v d7rooofryv, if then you should (shall) learn this for me, I would not pay, etc. AR. Nub. 116. "Hv U' J.cpE>..wfl'aL, K<iKtU'-r' 6:1ro>..op.ryv. Id. Ran. 586. 'Eyw o~ -ra-6rryv fl'EJI 7"~JI elp~JI'fJV, ws av eis 'ABryva[wv Ad7r7JTat, OVOE7r0T av U'Vf1'(3ovAevU'atp.L 7rOL~U'aU'8at -rfj 7r6AH, J would never advise the city to 1nake this peace, as long as a single Athenian shall be (should be or was) left. DEM. xix. 14. (Here ws Ae[7rot-ro would be the common form.) "Q(]'7rp av DfLWJI EKa(]'TOS al(J'xvvBd'f} n)v -n5.fw AL7rEW 1JV av TaX 8ii ev r<ji 7roA(pcp, as each one of you would be ashamed to leave the post at which he may be (might be) placed it~ wa1. AESCHIN. iii. 7. (Here ~v -rax8[ry would be the more common expression.) Twv J.-ro7rw-rrf.-r~v &v <t'f}, el -rav-ra ovv'f/Bet> f1-1J 7r p ri.f H, it would be nne of the strangest things ij, when he gets the power, he jails (shall fail) to do this. DEM. i. 26.

179. It will be understood that no assimilation to tl1e optative can take place when the protasis i>< pre~ent or past, as a change to the optative here would involve a change of time. See 561. 180. II. (a) In final and object clauses with t'va, ,1J,, 87rws, ocppa, and fl'~, the subjunctive (or future indicative) is generally u~ed after a potential optative with av or after an optative in protasis referring to the future. E.g. "'Hp& l<f. JIVJI ajl ~fl'tV oiKao' E7rOLO, ocppa ioy, K.T.A. Od. XV. 431. So OJ. vi. 57, xvi. 87; I!. xxiv. 264. 6.t WTO<; av7ravpa (J'Vfl'cpf.pot, ws opoll(J'IJ SoPH. El. 1439. Tts awov av KaAE\TELf.V, tils tov p.e;
Elltc. Hacch. 1258. 'OKvO['fJJI av ds Ta 7rAoZa Efl'f3avnv, !-'-~ Ka-raOV(J'U' cpo(3op.ryv o' av -r<ji ~f'EfLOJit E7f'f.(J'Bat, f-'-1J ~p.a<; ciyrf.yv i)(jf.V ovx o[6v 7" E(J'TaL JfeA.8eZv. XE:->. An. i. 3, 17. '}'[,. OVI< av cpdyo t, tJia fl'~o' liKwv avTfj 7rf.pt7T'f(J'IJ; DEM. XXV. 33. &v VfLUS fl'Ef'a OJI~(J'at TO \Trprf.nvp.a, f.L ~7rLfL<A'fJ8dYj7"f. 87rWS' dvTt -rwv &7roA.wA.6-rwv w<; TaXL\TTa \T-rpa-r'f}yo2 ~<a~ A.oxayo2 &vnKaTa(J'-raBw(J'tv. XEN. An. iii. 1, 38. El oe Kat 87rw<; dp~Y'f/ E(J'Tat cpavf.pot eLY)Tf. E7rLf1-f.AOVfLEVOt.. Id. Vect. v. 10 (see 180, b).

(b) ThP- only examples of the optative here are one in Aristophanes, one iu Plato, and six in Xenophou 1 : - 6.Lct TOW ELKOTWS (3 ovAo LJIT' O.v 1Jfl'OS ~~oA.wA.evaL, tva Td<;

7"Af.7"aS A rf. (3 0 Lf. Jl. AR. Pac. 411.

OvK av


d {3ovK6Aov<; 7rpO(J'8i:fLH, tva OL ywpyot

1 See Weber, Absichtssatze, pp. 220, 221 ; 245-247. Weber's collection of examples is complete.

7rdvv '}'E p.f.ya 7"L dYj, ~7rt 7"0 dpovv exouv

I have assumed that




f3ovs. PLAT. Rep. 370 D. ITetp<{lp.'f)v (&v) p.~ 1rp6a-w ip.wv Eivat, Zva, d 1rov Katp'Ds et'fj, htcpavd'fjv. XEN. Cyr. ii. 4, 17. So Cyr. i. 6, 22; An. ii. 4, 3, iii. 1, 18 (with various readings in last two). 'H cpvAaK0 YEAo{a TtS &v cpavotTO, d 11-~ <TVYE E7rlJJ-EAOtO 01rWS e~w8v n d<rcpepotTo. XEN. Oecon. vii. 39. El o~ Kat o1rws To {v f:J.Acpo'Ls lep'Ov avT6vop.ov yevotTO <j>av<pot d'l]TE E1rtfLEAOVfLEVOt. XEN. Vect. v. 9 ; but in the next sentence, o1rws dp~v'l] ea-Tat (see 180, a).

181. (c) After an optative in a wish twelve examples of tl1ese clauses with the optative and ten with the subjunctive are cited from Homer and the lyric and tragic poets. These are
T&xta-ra fLDt voov f.Tal:pot lv, tv' Jv KAta-v A.ap'Ov TETvKofL<8a o6p7rOV. Od. xiv. 407. So xviii. 368, XX. 79. (Subjunctive in Il. xvi. 99, xxiv. 74; Od. iv. 735, xviii. 202.) So THEOG. 885, 1119; PIND. Py. v. 120 (?). (Subj. PIND. Nem. viii. 35.) ''EA.8ot 81rws yevo tTo rwvo' EfLOi Avr~pws. AESCH. Eum. 297. rcvo[fLaV tv' vAaEV E7rCT7"t 1r6VTOV 7rp6f3A1JfL' aAK'A.VCTTOV1 TdS lepdS 01f'WS 7rpoa-Ef1f'OtfLEV 'AB~vas. SoPH. Aj. 1217; so Ph. 324 and Tr. 953. (Subj. SoPH. Tr. 1109.) Et fLOt yevotTo cp86yyos Jv f3paxo(n, ws m5.F8' OfLaprfl Twv exotJITo yovv(hwv. EuR. Hec. 836; so Hipp. 732. (Subjunctives in EuR. Hel. 174, Suppl. 621, I. T. 439, Ion. 671.)

182. No case of either subjunctive or optative after an optative in a wish in prose is cited by Weber. Perhaps one may be found in DEM. xviii. 89, where Cod. ~ reads, ifw OtafLaprot<v, Kat fLETduXOtEv
wv VfLEL'> oi ra f3Ana-ra f3ovA6fL<Vot Tovs 8wvs alr<tT, fL~ fLETaoof<v VfLLV wv avrot 7rpDrJPYJVTat, which can best be translated, in which (hopes) may they be disappointed; and 'rnay they (rather) share the blessings for which you, who wish for the best, pray the Gods, lest they involve you in the evils which they have chosen for themselves. M~ with the subjunctive in this sense occurs twice in Demosthenes, xix. 225, xxxviii. 26. The alternative, if we keep this rea(ling, is to make fL~ fL<TaDot<JI an independent wish, as if it were fL1JDE fLETaC.oi:ev, the usual reading.

183. In relative sentences expressing a purpose the future indicative is regularly retained after optati ves and even after past tenses of the indicative (566). For exceptional cases of the optative in this construction see 57 3 and 57 4, with 134. 184. III. In indirect quotations and questions depending upon an optative which refers to the future, the indicative is the only form regularly used to represent an indicative of the direct discourse. E.g. Ov yap UJI TOVT6 i et1f'Ot>, WS eA.a8ev. AESCHIN. ii. 151. 'EKELVO AEYHJI av E7rtXHP~CTH .A7rTVYJS, WS ai AHTDvpylat ds 7f'EIJY)TUS avBpw7rO'V'> epxoJIT(l,' (187). DEM. XX. 18; .SO xvi. 4. El a7rOOHX(;h'1 T[Jia XP~ ~yew8a, rov Aawov. XEN. An. iii. 2, 36. 185. But in DEM. xvi. 5 we find the OlJtative in an indirect quotation: ov yap EKEtv6 y av ef1f'OtfLEV1 C:>'> dvraAAa~aa-8at {3ovAofLe8' There are no other 'avnmiA.ovs AaKEDa'JLOV{ovs avr~ 8YJf3awv.




readings, and we must call it an exceptional ca.~e of assimilation (we could not say this, that we wished, etc.) unless we emend it either by reading {3ovA.6p,E8a (as proposed by Madvig, Bemerk. p. 21) or by inserting liv. In PLA'P. Rep. 515 D, we find in the best Mss. r &v orn avrov El-irli:v, Er ns avrc{! A.eyot on r6n fl-EJI Eti!pa ~A.vapas, JIVJI OE op86rEpa (3 A~11" 0 t j what do you think he would say, if any one should
tell him that all that time he had been seeing foolish phantoms, but that now he saw more correctly? (Some Mss. read

In Il. v. 85, Tv3E[(l'I)V OVK &v yvo'/)S rrorepoun JUTd'l), the optative represents p,ena-rtv in the direct question ; but ovK &v yvoYJs here refers to the past, meaning you would not have known (442). 186. IV. In indirect questions depending on an optative, the optative may represent an interrogative subjunctive (287) of the direct question. E.g. OvK &v ~xots JgA.8wv 8 Tt xr0o a-avr0, if you should withdmw, you would not know what to do with you1selj. PLAT. Crit. 45 B. OvK <iv ~xots 0 n XP?)a-a to a-avr<iJ, dA.X lA.tyyt<{l'I)S &v Kat xaa-p,c{!o OVK ~xwv o n d rr o t s. Id. Gorg. 486 B. The direct questions here were r xpwp,at ;-r xp~a-wp,at ;-r drrw; The subjunctive can always be retained in this construction, even after past tenses (677).

187. The present, perfect, and future of the infinitive and participle, and the aorist infinitive when it is not in indirect discourse, regularly denote time which is relative to that of the leading verb. They the}efore merely transmit the force of that verb, as primary or secondary, to the dependent clauses. E.g. BovAErat A.eyEtv r rovro Ja-nv, he wishes to tell what this is. 'Ef3ovAEro A.eynv ri rovro d'l), he wished to tell what this wus. <P'I)a-tv aK'I)KOEVat ri EG"rtV, he says he has heU?d what it is. "E~YJ aK'YJKOEJ!at r EZ'I), he said he had heard what it was. P'l)a-t rrot~a-Hv o n <'lv {3ovA'I)a-8e, he says he will do whatever you may wish. ''E<f;'l) rrot~a-Etv on (3ovA.ota-Oe, he said he would do whatever you might wioh. Mvova-w (3ovA.6p,Evot eloevat r Ea-rt. "EfLEvov {3ovA.'~f1-EJIOt El3vat r[ d'l). Mevova-w ctK'I)Ko6ns r Ea-rtv. "Ep,evov ctK'I)KOores r[ d'l), they waited, having hea?d what it was (r Ja-riv ;). Mevova-tv aKOVG"Ofl-EJIO r Ja-nv. "Ep,EVOV aKOVG"Ofl-EVOt r[ d'lj BovAErat YFWVat r rovr6 J(TT V' he wishes to lea.m what this is. 'E(3ovA.ero yvwvat r[ rovro e'1), he wished to learn what this was. Ovoevl 7rW7r07" TOVTWJI 3EOWKaT T~JI 3wpav TflVT'I)l' 01io' <lv oo['JrE, J~Ei:vat rovs lS[ovs Jx8pots v(3p(Hv mirwv EKU(m.p, 6rr6r' &v (3ovA'I)rat Kai !lv llv O<;V'I)Tflt rp6r.ov. DE)L xxi. 1 iO. 0{18' VJJ-tV ovr< 8'1)(3aOt> OVTE AaK3atfLOVotS OVOE1rW1rOTE G"VVEXWP'le'l) rov8' V1r.d TWJI EA.>.:>]vwJI, 71'0 ttV o n {Jo-DA.otfJB<, never 'Was this g1anted you, etc., to do whatever you pleased. Id. ix. 23. Here rrotEw denotes a habit,




and is followed by the optative (532); if the leading verb were o-vyxwpL-rat, we should have '1J"OLELV 8 7t &v f3o-6A1J<r8e. Compare the two subjunctives in the preceding example. 188. The present infinitive and participle representing the imperfect (without lfv), and the perfect representing the pluperfect, are secondary tenses in themselves, without regard to the leading verb. E.g. Ilw<; yap oreu8e ovuxepw> QKOVE L v, r -r<; TL Aeyo t; how unwillingly do you think they heard it, when any one said anything f DE:M. vi. 20. So PLAT. Rep. 430 A. See these and other examples under 119. For the perfect see XEN. Cyr. i 4, 27, and THuc. v. 49, under 123.

189. The aorist infinitive in indirect discourse is a past tense in itself, and is therefore secondary. E.g.
<i>1]1J'C ")'VWVat -r 'TOVTO ei1J, he ~E</>'1) yvwvat T Towo E i'l), he said

says that he learned what this was. that he had learned what this was. <P1JuC yap OJ-LaAoy~ua fL Tov KA~pov T'f 'ITatot TO ~fLLKA~pwv p.e-ra8wuetv el VLK~IJ'<lLfLL -rovs ~xov-ra<; al'>r6v (he says I promised, fLETaStiJuw Jav VLK~ITW). ISAE. Xi. 24. 8aA~V 8p(j.T-ra Tt<; 8epa'ITatvi<; a'ITOIJ'Kwfat Aeyerat, w<; 'Td fL~V Jv ovpavtj) 7rpo8VfLOLTO el8vcu, Ta 8' ~fL1rpou8ev mhou Aav8avo aw6v. PLAT. Theaet. 174 A. ,.Ap&. IJ'O OOKW ov fL<lVTLKW') & vvv 8~ neyov el'ITetV, OTL 'Ay6.8wv 8avp.(l(]'TWS ~po'i ~')'ti! o' d7rop~CTOLfLtj Id. Symp. 198 A. In all these cases the optative depends on the aorist infinitive as a past tense.

190. The aorist participle properly refers to time past relatively to the leading verb. It is therefore secondary when the leading verb is past or present, so that the participle refers to time absolutely past; but it may be primary when the leading verb is future, if the participle refers to time absolutely future. E.g. "Iu'T ~{I-US f. A.fJo v-ra<; rva TOVTO tOO !fLV, you know that we came
that we might see this.
"l!'~<j>wv 8~ 8dua<; p.~ Oe'IJ6e1] 'TrOT~

i:v' X ot OtKJCew, alyw.A6v v8ov -rpe<j>E, and once he took fright lest he might sometime lack pebbles (for votes) to enable him to be a judge, and so he lceeps a beach on the premises. AR. Vesl1. 109. IIpos dpy,]v ~K~Ep(c, /L(ed:<r<i JLOI. H"jHV 8. xrnCotfLt, you rush into a passion, after you gave me leave to say what I wished (i.e. & &v XPrit'n>). Sopfi. El. 628. 'Y7ret71'WV TlfA.Aa O'Tt avT6s rdK~r 71'pa~ot, <Px~To. THUC. i. 90. Tii 0 &.<rnyt rv'IT-ru8w 1rA'IJYas v1ro K~pvKos v -rii dyop(j., K1JpvgaVTos c'Sv eveKa fLE AAEL TV'ITnu8at, i.e. let the crier flog him, after proclaiming (having proclaimed) fo1 what he is to be flogged. PLAT. Leg. 91 7 E.

191. The tenses of the infinitive and participle with &v are followed, in dependent clauses, by those constructions that would follow the finite moods which they represent, if these stood in the same position. See Chapter Ill.

THE PARTICLE "AN. 192. The adverb &v (with the epic ICE, Doric d) 'has two uses, which must be distinguished. 1. In one use, it denotes that the action of the verb to whidt it is joined is dependent upon some condition, expressed or implied. This is its force with the secondary tenses of the indicative, and with the optative, infinitive, and participle : with these it belongs strictly to the verb, to which it gives a potential force, like our would. . 2. In its other use, it is joined regularly to El, if, to relative and temporal. words, and sometimes to the final particles W';, 07TW<;, and oif>pa, when any Of these are followed by the subjunctive. Here, although as an adverb it qualifies the verb, it is so closely connected with the relative or particle, that it often coalesces with it, forming Mv, ~V, l:iv, gmv, o1romv, m:toav, 1rav or 1r~v (Ionic E7rEav).
These statements include only the constructions which are in good use in Attic Greek. For the epic use of K~ or Civ with the subjunctive in a potential sense (as with the O}Jtative) see 201, 1; for K~ or Civ with the future indicative see 196.

193. There is no word or expression in English which can be used In its first use (192, 1) we express it by separately to translate the form of the verb which we use; as A8ot he would go; 1jA8ev he would have gone. In its second use, with the subjunctive, it generally has no force that can be made perceptible in translation. The peculiar use of can be understood only by a study of the various constructions in which it occurs. These are enumerated below, with references (when it is necessary) to the more full explanation of each in Chapter IV.









194. No theory of the origin o either liv or K~ has yet helped to explain their meaning, however valuable the discussion o the question may have been to compar~tive philology. It s~ems ~o be clear that KE is the older partiCle ; It occurs 6 21 tnnes m Homer while &v occurs 155 times; in Pindar the two are nearly balanced; &v has a preference for negative sentences, being very often attached to the negative; &v is more emphatic, as appears indeed from its fixed accent, while Ke is enclitic ; KE is much more frequent than J.v in relative clauses in Homer. 1 But, practically, it is still safe to assume that the two particles are used in substantially the same sense in all epic and lyric poetry. In Herodotus and Attic Greek only J.v is used.


195. The present and perfect indicative are never used

with /lv. This seems to occur chiefly when Plato and Aristotle use Kllv d ( = Kat liv, d) like Kat d, without regard to the mood of the verb which is to follow, to which Kav really belongs. See PLAT. Men.
72 C, K&v d wo>..>..a{ daw, lv y n Etoos TaVTOV waam ~X 0 V en, i.e., even if they we many, still (it would seem to follow tltat) they all have one and the same form. So Rep. 579 D, Soph. 247 E. So A1uSTOT. Pol. iii. 6, 1, Kllv d w>..dovr;, with IJ"K7rTEOV u-r[v. Examples of a different class (without Kllv d) have now almost disappeared from our texts. One of the last relics, PLAT. Leg. 712 E, ytil ile ov-rw vvv ~a{cfwYJ> &v pwn/Ms ovrws 6rrp Eiwov, ovK ~X w dwci:v, is now simJJly emended by reading avEpwr"f)(Jd,.

196. The future indicative is often used with Ke or /lv by the early poets, especially Homer. The addition of &v seems to make the future more contingent than that tense naturally is, sometimes giving it a force approaching that of the optative with &v. E.g. 'AA.A: re', Jyw OE /(~ TOG Xap[Twv p.av orri\onpawv 8wuw, O'TrVL~
tjLVaG Ka1 CT0v KEKA~8at uKotnv, I will give you one of the younger Graces, etc. Il. xiv. 267. Ka KEn> &8' pH Tpwwv -&rr~p1)VOpE6v-rwv, and some one will (or may) thus speak. Il. iv. 176. '0 ilE KEJ! laxoi\wO""ETat 8v KEV tKWJLaG, and he may be angry to whom I come. Il. i. 139. El 8' &y, TO~> av Jywv' OL 8~ 7TG8EuBwv. Il. ix. 167. ITap' ~JLOG y< Ka1 &>..i\ot, ol: K~ JL< nJL00'ovut, others, who will honour

See J\fonro, Homeric Gramnuzr, pp. 265-267.


For Pindar, see Gildersleeve

in Am. Jour. Pkil. iii. pp. 446-455, where Juay be found a complete enull!eration of the passages in Piudar containing either <i.v (30 cases) or KE (33 cases).





me. Il. i. 174. El 8' '08V<Te~s V.. Bot Kal rKot! is 7ra-rpl8a yat'av, altfa K crvv ~ 7rat8t (3Eas a7rOT[crera L avopwv. Od. xvii. 539. Here a1T"OT{creTa{ K<, Which may be aorist SUbjunctive (201, 1), is used neatly in the sense of the optative, corresponding to the optatives in th~ protasis. K is much more common with the future than &v. 197. The use of &v with the future indicative in Attic Greek is absolutely denied by many critics, and the more careful revision of the texts has greatly diminished the number of examples cited in support of it. Still, in several passages, even of the best p1ose, we must either emend the text against the Mss., or admit the construction as a rare exception. E.g. Alyv1T"Tovs 8 ovx opw 1T"o[q, ovvaJLEt CTVJLJLUX<' XP"')CTUJLEVOt pJiAAov liv KoAacrw& T{js vvv crvv lJLot ovcrYJs. XEN. An. ii. 5, 13. E~YJ OVV r6v Jpwr<!JJLEVOV El1T"EtV, OVX ~KEG, ~civat, ovo' {iy {J~Et 0Vpo, he
said that the one who was asked replied, "He hasn!t come, and he wq_n't come this way." PLAT. Rep. 615 D. (The only other reading is ~~ot. The colloquial style here makes av less objectionable; see SoPH. Ant. 390, quoted in 208.) "E~YJ A.ywv 7rp6s iJJLUS ws, El Ota~EV~O[JLYJV, ~OYJ av VJLWV o1 vlds 1ravT<s 7ravra1racrt ow~Bap~croVTat. Id. Ap. 29 C. KCJ.v ~T' ~n ~6vwv 8 tfo JLa t aTJLa (so the Mss.). EuR. EL 484. See 208 and 216, on the future infinitive and participle with &v.

198. The most common use of llv with the indicative is with the secondary tenses, generally the imperfect and aorist, in the apodosis of an unfulfilled condition (410) or in ~potential sense (243). 199. The imperfect and aorist indicative are sometimes used with lfv in an iterative sense (162), which construction must
not be confounded with that just mentioned (198).

200. In Attic Greek llv is regularly used with the subjunctive in protasis and in conditional relative sentences, and sometimes in final clauses with W'> and 07TW'>, being always closely joined with the particle or the relative ; but never in independent sentences. See 325, 381, and 522. 201. 1. In epic poetry, when the independent subjunctive has nearly the sense of the future indicative (2 84), it sometimes takes KE or llv. This forms a future potential expression, nearly equivalent to the future indicative with " or &v, and sometimes approaching the optative with " or







El KE Jk~ OW'l](J'tV, ~?'~ KEV o.irr6s lAw }kat, and if he does not give her up, I will take her myself. Il. i. 3 2 4 ; see also i. 13 7. See 285 and 452. For the variety of nearly equivalent future
potential forms which the Homeric language presents, reduced tO one in .Attic Greek, see 235. 2. The epic language has ,d or llv with the subjunctive in the constructions of 19 2, 2 ; but its use of "~ or &v in conditions is less strict, and that with final particles is more free, than the Attie use of llv. See 325-328; 450-454; 468-471; 538-541.



202. The optative with &v forms the apodosis of the less vivid future condition (like the Euglish form with would or should), or has a potential sense. E.g. El 7'0V7'0 7rOt-lj<J"EtH', llBA.w<; li V E r'f/, if he should do this, he would be wretched. 'Hoews liv epo[}J,TJV avT6v, I should like to ask him. (See 233 and 455.) For construction of liv or KE with El or the final particles and the optative, see 460; and 329, 330, 349, 350, 351. 203. As the future optative came into common use after the future indicative with liv (196) was nearly extinct, it was never used with &.v.


204. The infinitive can be used with Clv in all cases in which a finite verb would have &v if it stood in its place. This is found chiefly in indirect discourse, in which each tense of the infinitive with ~v represents the corresponding tenses of the indicative or optative with :iv in the direct form. The context must decide whether the indicative or optative is represented in each case. 205. (Present.) The present infinitive, which represents also the imperfect (119), when used with ~v, may be equivalent either to the imperfect indicative with av or to the present optative with ii.v. It can represent no other form, as no other form of these tenses has joined with the verb in a finite mood. E.g. "" ' ' ' ,, " .,. ~ " t: ' ""'f/(J'LV avrovs EAEV 8' epovs o.v t vat, t, Tovro e1rpa<;av, 1!e says that they would (now) be j1ee, if they harl done this (etvat liv representing ~<Tav O.v). <l>1)<T~V a-6rous ~Aev8epovs a dva t, EL TOVTO 7rprl~Hav, he v


say.< that they would (hereafter) be free, if they should do this (etvat av representing ,}1Jrrav &.v). Ote<J"Be yctp TiJv o/JK llv <f>vA.aTTov Kat T~V TLfJ-1JV Aap.f3avHV 7'WV ~VAWV i do you think he would not have talcen ca1e and have eceived the pay for tho titrnber ? DEM. xlix, 35. (Here the direct discourse would be ~</>vAUTTEV av Kat JA6.ftf3av(Y.)





MapTvp[tp expwvTo, p~ av 'TOVS ye lcrotf!J<f>ovs &Kovras, et p~ TL ?JOKoVV ols e1riJecrav, ~vcrrparevetv, they used us as an argument, that people who had an equal vote with themselves (like us) would not be seTving with them against their uill, ttnless those whom they attacked uere guilty of some wrong. THL'O. iii. 11. OTpat ylip av oilK dxapcrrws pot x Et v, jiJt I think it would not be a thankless labour (oilK av i!xot). XEN. An. ii. 3, 18.

206. (Pe?fect.) The perfect infinitive, which represents als;) the pluperfect (123), when used with c'i.v, may be equivalent either to the pluperfect indicative with c'i.v or to the perfect optative with av. E.g. El fl-~ ras dperas l>7r~p allrwv EKdvas Ot Mapa!:lwvt Kal LaAaf-tVL
'1Tapcrxovro, . . . 1ravra Tav!:l' v1ro rwv f3apf3apwv O.v f.aA.wKevat ns), if those at Marathon and Salamis had not exhibited those deeds of valour in theiT behalf, any one would say that all these would have been captund by the ba1barians. DEJ\f. xix. 312. (Hel'e JaA.wKevat &v represents f.aAWKecrav &v.) 'AA.A' ovK O:v ~yov{J-at avroV> OiwfJv d~[av OeOWKEVat, el dKpoacrapevot avrwv Karm)n)<f>craur8e, but I do not believe they would (then) have suffered sufficient punishment, if you after hearing them should condemn them. LYS. xxvii. 9. (Here the protasis in the optative shows that OeOWKEVUt av represents 00WKOTS av EiEv (1 03); but if the protasis were el Kare>fryfcracr!:le, if you had condemned them, 00WKEVO.t av would represent EOEOcbKECTO.V &v, they would have suffered.) See also, in xxvii. 8, o1lK av d7roAwAi:vat, dA.Aii O[K'YJV oe8wKEVaL, representing perfect optatives with &v. ,Av8pa7roOWOELS liv OLKaws KEKAfjcr8at (-Y)yefro). XEN. Mem. i. l, 16. (Here KEKAfjr:rOru av represents KKAYJf-EVOt liv eiev.)

(se. <f>1}rrnev


These constructious are of course rare, as are the forms of the finite moods here re})l'esented. 207. (Aorist.) Tl1e aorist infinitive with av may be equivalent either to the aorist indicative with av or to the aorist optative with av. E.g.
o~K av ~yei:cr8' a~TUll KQV J7rtOpafhi:V; do you not believe that (if this h~rd berm so) he wo1dd even have run thither? i.e. ollK liv J7ri:Spaf-ev; DEM. xxvii. 56. ,, Avev OE CTELCTf-OV OVK a V fhOt 00Kt TO Towvro ~vfl-f3i]vaL yevEcrOat (ollK liv ~vfhf3fjvat representing ovK tlv tvvi:f3?J), but unless there had been an earthquake, it does not seem to me that such a thing could by any chance have happened. TRue. iii. 89. Taus 'A!:Iryva[ov;; ~A.1rt(ev (crws liv J1re~eA.!:Ie'iv Kal T~v yfjv ovK liv 7rp Gt 0 E rv Tfl-1)!:17jvat (i.e. i'a-w;; liv J1fe~i:A.8otev Kal OVK &v 7repdootev). Id. ii. 20. 0-Do' av KparYjcrat avrov;; TYJS yfjs ~)'OVf-O.t (i.e. Kpar~ cretav av). Id. vi. 37. 208. (Future.) The future infinitive with av can be equivalent

only to the Homeric construction of the future indicative with av. But as av is not found in Homer with the future infinitive, this construction rests chiefly on the authority of passages in Attic writers, and is subject to the same doubts and suspicions





as the future indicative with in those writers. (See 197.) Unless we exterminate the latter, there can be no objection to this as its representative. In the following passages it is still retained on the best Ms. authority. Nopi(ovus, EL Ta1rrryv 7rpJmJV A.a(3otV, pr,:Ows aV O'<j>[O't TaAAa 7rpocrxwpfJO'ttV. TRue. ii. 80. (Here the direct discourse would regularly have had either the future indicative without or the aorist optative with liv.) The same may be said of 'rHuc. v. 82, VOJJ-[(wv JJ-EYtO'TOJ! &v O'<j>as w<j><AfJcrov (where one J\1s. reads by correction w<j>EA~O'at). See also THUC. vi. 66; viii. 25 and 71; and PLAT. Orit. 53 D; Orat. 391 A. 1ro8' ?J~Hv ovp' ltv E~'Y)v xovv eyw, I decltLred that I should be very slow to come hither again. SoPH. Ant. 390. (Here the colloquial style may account for 0~<w av, as for {)~tt iiv in PLAT. Rep. 615 D, unless we take liv with ef'YJvxovv. See 197.) In PIND. 01. i. 108, we have El o Jl-TJ Taxv A7rot, ETt yA.vKVTEpav KEY i!A7rOjJ-Ut O"VV & eo~ KAdfov. As the future optative is never used with (203), this can never be represented Ly the future infinitive with av. 209. The infinitive with 3.v is rare in the early poets, occurring but once in Homer, IL ix. 684 (quoted under 61l3), and three times in Pindar, Pyth. vii. 20 (present), Pyth. iii. 110 (aorist), and Ol. i. 108 (future, quoted in 208). 210. The infinitive with sometimes represents an iterative imperfect or aorist indicative with (162). This must be carefully distinguished from the potential use. E.g. 'A KOVW 1\UKWULfWVWV') TOT Jl-f"aii.OVTaS ay Kat KUKWO'UVTU') T'Y)V , A ~ Q \I " ' xwpav dvaxwp<'iv J.,f orKoV 7rUAW, I hear that the Lacednemonians







at that time, after invading nnd rnvaging the couut1y, used to retuTn home ngnin. DEM. ix. 48. (Here dvaxwpE'iv &.v represents al'<xwpovv &.v in its iterative sense, they used to return.) .PaO't JJ-F yap avr~v Ep7rT6fL110V




-rWv Ex6vTwv dvEpwv oVK ltv E~EABE'iv dtrO T~s (J"t7r{nr-; avTt,BoAE'iv IJ.v Oj)-Otw>, they say that, when he was feeding on

men of wealth, he neveT would get away from the nwal-t,ub; and they all alike used to imploTe him (ovK ltv J~~,\()Ev, ol o ?)vn(36A.ovv d.v). AR. Eg_. 1295.

211. The infinitive with in the cases already mentioned, stands in indirect discourse after a verb of sayi/l{f or thinking. Sometimes, however, it is found in other coustructions, where the present or aorist infinitive (without 3.v) would be expected. In such cases there is an approach to the usage of indirect discourse, so far at least that the infinitive with dv has the force of the corresponding tense of the indicative or optative. E.g. Ta 0~ EVTO> OVTW> EKaTO, WO'T ~OtO'Ta &v ES vowp :fvxpov O'<j>os avrovs pL7rTt.V, so that they would most [Jladly hnve tlwown themselves into cold wateT !tJ{rr-r<w <1.v here being equivalent to <ppt7rTOV <1.v). TRue. ii. 49. Mtas Tpefn 7rpo~ vvKTOs, wO'u JJ-fJr' f.JJ- JJ-fJT' &AA.ov,






ocrns ~ws ~pi., f3 >.&fat 1/'oT' l1v, so that you could harm (f3>.&fuas l1v) neither me nor any other who beholds tlte light. SoPH. O.T. 374. So Tr. 669. "E~8aa-aJJ 7rapd.86wrcs T~JJ TbJJJ 'A8rwawp olKooopav, if>cru J.I!Y)KETt }L~T avro~ KWAVcr8at aVTwv, EKdvovs 7' Ka~ 1/'aVTd1/'aCTLJ! a1/'CTTP'fJKEJ!at, d Kat KpaTOLEJ!, f.l,~ &v En O'~fis d1/'0TH X er at, so as to be no longer themselves obstructed by them, and so as to have deprived them absolutely of the power of ever again walling tloem in, even if they should be victorious. THUC. vH. 6. "Ycropov T~JI J!VKTa 1/'acrav wcrr' i:crws f3ovA~creTat K &v EV AlyfnrT<p r v X e t'v l:Jv f.l,UAAov ~ Kpwat KaKws, we will rain all night long, so that perhaps he will wish to have the luck to be (that he might by chance find himself) in Egypt mther than to judge unfairly. AR. Nub. 1130. (Here Tvxev &.v follows f3ovAof1-at like the future infinitive in THUC. vi. 57: see 113.) We have 11'tw followed by the infinitive and l1v in THUC. vii. 61, TOT~> TVX'fJ> Kctv fl-<8' ~p,wv EA11'lO'avns O'T~vat, hoping that fortune may take sides with us (O'ra17 ll.v). See also SoPH. El. 1482, dA.>.& f.l,Ot 11'dpE'ii Kctv crptKp'Ov el11'dv, but permit me at least to say a little (that I might say even a little, Ebr-Ot}Lt li.v). See the corresponding use of the future infinitive in similar expressions, where there is the same approach to indirect discourse (113). 212. Even the infinitive with the article occasionally takes ll.v, as in ANT. V. 8, TOVTO vp,as (, OV Ti! ~e{;rew ctJJ TV 11'Afj8os TV Vf.l,ETepov, this I will teach you, not because I would avoid your people. In SoPH. Ant. 236, T~S V..7r8os TO p,~ 11'a()dv &v &.>.A.o, the hope that I could not suffer anything else, the construction is practically that of indirect discourse (7 9 4).




213. When the participle is used with &v, each tense represents the corresponding tenses of the indicative or optative with l:1v.
The participle with ll.v is not, like the infinitive with ll.v, found chiefly in indirect discourse; but tiv is more frequently added to an attributive or a circumstantial participle (822) to give it a potential force equivalent to that of the indicative or optative with ll.v. The participle with l1v is not found in Homer or Pindar. 214. (Present.) The present participle (like the present infinitivt) with tiv represeuts the imperfect indicative or the present optative with &.v. E.g. Oi8a a1hovs >.ev8epov<; &v ovTas, El rovTo l7rpaav, I know they ttould (now) be free, ij they had doue this. Ol8a avTovs JA.ev8epovs &1! <JvTas, el rovTo '11'patetaJJ, I know they v>ould (hereafter) be free, if they should do this. (In the former lJvnL<; liP represents 1jcrav ll.JJ, in the latter Ei:'f]crav tiv.) Twv Aap.f3a.v6vTwv 8tK''7V (j vTES &v Ot.Kafws (i.e. ~fl-U' l1v), whereas we should justly be among those who inflict punishment.





"01T'p. Ecrx /L~ KaTU 1r6AEtS a-6-Tdv ~17"t7rAEovTa. -rq., lieAmrovv'IJa-ov wopBew, dovvarwv ~~~ Svrwv ({Jpwv) lm{Jo'I]Be'iv, when you would huve been unable to wing aid (do-6vaTOL ltv ~re;. THUC. i. 73. Ilo,.\,A' liv i!xwv i!np el7rEtJI 7rEpt avT~S 7rapa,.\,e7rw, although I might be able to say many other things about it, I omit them. DEM. xviii. 258. 'A7ro 7raVT~S av ~pwv ,.\,6yov OLKa[ov /L'IJXUV'I]fl-a 1rOLKAov (ie. os Ci)l <f>pots), thou who wouldst derive, etc. SOPH. 0. C. 761.
DEM. lvii. 3.

215. (Aorist.) The aorist participle with ii.v represents the aorist indicative or the aorist opta.tive with U.v. E.g. 01JTE 8vra olJTE ctV ')'EV0f1-EVa A0')'011"0WVCTLJJ, theyrelatethingswhich
are not real, and which never could happen (i.e. o-DK <'iv yevotTo). THUO. vi. 38. 'Ecp ~pwv oil yeyov6s o-&8' o?oa el yevopevov liv, (a thing) which has not occurred in our day, and I doubt whethm it ever could occur (yevOLTO av). PLAT. Rep. 414 c. 'A,.\,.\<i prcotws av d~eBds, el Kal P-ETp{w<; TL roVTwv hrof'I]<JE, 1rpodAero d1ro8avew, whereas he might easily have been acquitted, etc. XE:-.. :Mem. iv. 4, 4. Kat el dmjxB"la-Be WU'7rEp ~)pets, EV t<Jf1-EV p~ <'iv ~i<Ja-ov vpfi.s A111r1Jpo:Vs ')'EVOJl-EvOV s TOts fvppaxou;, Kat dvayKa<J{)Ev'l'a'> aV ~ apxHv, K.T,,.\, (i.e. oiJK av E')'Elle<J8E, Kat ~ivayKO.a-8'1]7' av), if you had become odious as we have, we are sure that you would have been no less oppressive to your allies, and that you would have been forced, etc. THuc. i. 76. 'Opwv T6 7rapa1'dxurpa a7r Aovv Bv Kat, 1 E'lrLKpaT~ITHE TLS Tf)s dva(JO.a-ews, pf(-0ws l1. V aiJ'rO ,.\'I] 1> e EV (i.e. />rc-8ws av A'IJ~ee'1]), seeing that it would easily be taken, etc. Id. vii. 42. So ws TaX li\1 a-vp.(JO.vrwv, DEM. xxiii. 58 (see 918).

216. (Future.) A few cases of the future participle with l!.v, representing the future indicative with l!.v, are found in Attic writers. These rest on the same authority as those of the future indicative and the future infinitive with av (197 and 208). E.g. 'A~f.ETE ~ Jl-?J d~f.ETE, WS ~fkOV OVK liv 'l'rOL?JU'OVTOS a,.\,,.\,a, o~o e paAw 1rOAAt5.KLS TEBvavaL (i.e. OVK av 1r0LYJITW O.A.A.a) : so .all Mss. PLAT. Ap. 30 B. To:Us OTLOVV llv EKdv<p 1rOL~<TOVTaS avup'rfKOTES EK Tfjs woAEw> ea-e<JBE. DE~r. xix. 342. (Here 1nost l\lss., including};, have 1rot~a-ovras, but A has 1rOL~<TavTas.) Da,.\,at TLS ?)ows liv wws
lpwrfJa-wv Kt5.8'1]Tat, many a one has long been sitting here uno perhaps would be very glad to ask (so all Mss.). DEM. ix. 70. 217. The participle with can never represent a protaBis, because , there is no form of protasis which could be represented by a participle, where ll.v is separable from the conditional particle. (See 224.)




218. I. When av is used with the subjunctive, if it does not coalesce with the relative or particle into one word (as in Mv, Jrav, etc.), it is generally separated from it only by such monosyllables as pf.v, <if., -re, yap, Ka[, v~, 1rep, etc., rarely -ris. See examples under 444 and 529.





2. In Homer and Hesiod two such words may precede Ke; as 7r<p yap KEV, d yap YV KE, l yap -rts KE, os fl-EI' yap K. This is rare with av in prose; see DEM. iv. 45, 67rot fl-EI' yap av. Exceptional are 87ro ns <lv, olfka.L, 1rpou8fj, DEM. ii. 14; 6 n /lA.A.o <lv OoK'ji Vfk'iv, XEN. Cyr. iv. 5, 52. The strange Ka(J' wv fl-YJV{;TJ /lv ns, ANT. v. 38, is now correctetl to <lv fl-YJV{,'[h but still stranger is 61rocrov ~ <f>apvy~ &v ~fl-WV xavoavv 0) AR. l' 259. 21~ When av is used with the optative or indicative, it may
either stand near the verb, or be attached to some other emphatic word. Particularly, it is very often placed directly after interrogatives, negatives, adverbs of tirne, place, etc., and other words which especially affect the sense of the sentence. E.g. 'AA.>ca ds ory Bewv Bepa7rea. eZry &v ~ OO"LOTYJS j PLAT. Euthyph. 13 D. 'AA.X OfkW> To Ke<f>aA.a.wv a.DTwv p<;;J5ws &v d1rots. Id. 14 A. OvK &v ory TOvo' /lvopa fhaXYJS Jp-6craLO fkTeA8wv, Tvoe[OYJV, /ls VVJ! ye
llv Ka~ Llt~ 7raTp~ fhaxotTo; Il. v. 456. IIws &v Tov aifkvAtfJTa.Tov, x8p'Ov UAY)fl-O., To-6s T OLO"O"apxa.s 6A.uuas (3a.uLA~s, TEAOS eavOLfkL KO.DT6s. SoPH. Aj. 389. IIoA.A.a dv UKWV eopwv. Id. 0. T. 591. TaxtcrT' /lv n 1rOALY oi ToLOvToL ~Tepovs 7r<uavTEs d1roA.euELa.v. TRue.


ii. fi3.

220. 1. By a peculiar usage, :;,v is often separated from its verb by such verbs as otofka.L, ooKw, <f>YJfk, oiOa., etc. In such cases care must be taken to connect the :l.v with the verb to which it really 'belongs. E. g.
Kat vvv ~oews llv fkOL ooKW Kotvwv~ua.t, and now I think I should gladly take paJ"t (av belonging to Kotvwv~ua.t). XEN. Cyr. viii. 7, 25. So AESCHIN. iii. 2 (end). Ollo' llv VfkEts olO' 8n i1ra-6uau(JE 7rOAEf1-0VvTEs, nor would you (I arm sure) have ceased fighting. DEM. vi. 29. II6Tpa yap <i.v ot0"8 p(jov ilvat; DE~!. xlix. 45. 'EKAE~O.VTa a P.,~T 7rpor)OL fl-Y)Oets ""~r' (}_V 0~8YJ T~fLEpov PYJ e~ Va L, selecting what nobody lcnew bej01ehand and nobody thought would be mentioned to-day. DEM. xviii. 225. (Here pYJBfjva.L av = PYJ(JE[YJ av. If av were taken with 0~8YJ, the meaning would be, what nobody would have thought had been rnentioned.) T oDv av, E</>YJV, dYJ 0 ''Epws j PLA'l'. Symp. 202 D. 2. Especially irregular are such expressions as oDK olOa llv El, or OVK ().V oiOa El, followeJ by an optative or indicative to which the ay belongs. E.g. OvK olo' V el 7r d (J" a. L, I do not know whether I could persuade him. EuR. Med. 941. (The more regular form would be ovK oloa cl 'lrd(J"a.Lfl-L av.) So Ale. 48. OvK ll.v olo' El ovva[fl-YJY. PLAT. Tim. 26 B. 0-&K olo' llv cl EKT1)0"afl-YJV 7raZoa. TOLOVTOV. XEN. Cyr. v. 4, 12. So 01JK &V olo' 6 TL aAAo El X0 V lfYJ<f>uau8aL, I do not lcnow what other vote I could have given (T aAAo Etxov <lv fry<f>(J'a0"8at ;), DEM. xlv. 7.


221. (Tal :l.v.) Among the words to which av is very frequently joined is Taxa., perhaps (i.e. quidly, soon), the two forming Tax' av, which expression is sometimes supposed to





mean perhaps. But riLl /J.v cannot be used unless the /J.v belongs in its ordinary sense to the verb of the sentence. Thus nix' liv y~votro means it might perhaps happen, and rrf.x' Clv ~yvETo means it n~ight perhaps have happened; but the latter can never mean perh'aps it happened, like rU"W<; ey~vETO. T<fxa alone often means perhaps, as in XEN. An. v. 2, 17. Adstotle writes r<fxa and CJ.v separately in the same sense as -r<fx' CJ.v; as r<fxa o~ Kat 1'-0.>..A.ov liv Ta{m]V 1nroAa.f3ot, Eth. Nic. i. 5, 6.

222. "Av never begins a sentence, or a clause before which a comma could stand. But it may directly follow a parenthetic clause, provided some part of its own clause precedes. E.g. 'AA.>.: 6l !'-~>.: /J.v f'-Ot U"tr[wv Ot'li"Awv 8Et, AR. Pac. 137. So r?. 1'-~A.A.ov, E'l!"El y~votr', tiv KA-6ot<; (or without the commas), the future you can hear when it comes, AESCH. Ag. 250.


223. Av is sometimes used twice, or even three times, with the same verb. This may be done in a long sentence, to make the conditional force felt through the whole, especially when the connexion is broken by intermediate clauses. It may also be done in order to emphasise particular words with which av is joined, and to make them prominent as being affected by the contingency. E.g. > \ '(3 ~ \ >I '1'> > ' ,!_ " LU"T> av, EL(]" 8' EVO<; II.U OLfU, O'fjii.WU"ULf'- > UV Ot UVTOLS 't'POVW. SoPH. El. 333. Ov rliv A.6vrE> afiet> dv8aA.oi:Ev CJ.v. AESCH. Ag. 340. ''AA.A.ov> i liv oi5v ol6"'e8a ra ~p.~npa A.af36vra<; OE'i~at liv p.aAtU"ra Ei n p.erpta(op.Ev. THee. i. 76. (See 220.) Ovr' liv KEAEDU"atfl, ovr' &v, El 8EA.ot<; En 1TpaU"U"EtV, ep.ov i tiv ~o~w> Bp~YJ> p.era. SOPH. Ant. 69. A~yw Ka8' EKaU"TOV OoKdv a V p.ot TOV a~TOV avopa 'l!"ap ?]pl:Jv E7r2 rrA!'iG"T~ aV trOIJ Kat fl-'Ta xap[Twv JLd.AurT' aV EVTparrAw~ T0 lrWp.a aVrapKe;; 1rapExea-OaL. TRue. iL 41. (Here tiv is used three times, belonging to 1Tap~xeU"8at.) 'Yp.wv o~ epTJp.O> t>v o~K li v iKavo> olp,at Etvat o{(-r' av <f>A.ov <h<f>EA~U"at oll-r' liv exBpov dA.~aU"8at. XEN. An. i. 3, 6. (Here /J.v is used. tlJree times, belonging to dvat./ OvK liv ~ydcr8' avrov Kav E'll"tOpap,eZv; DEM. xxvii. 56.

224. A participle representing a protasis (472) is especially apt to have an emphatic ::.v near it. This, by showing that the verb is to form an apodosis, tends to point out the participle as conditional in an early part of the sentence. E.g. '.T ; ' rl. '"'' ' ' ' ' ' ' c 'Df"tU"aTE TO TE 't'auii.OV KUI' TO f"EU"QV Kat TO 'li"UVV UKpt f3'E<; UV VY K p a eh p.aAtU"r' li V l (]"X -6 Et V' believe that these, if they shonld be united, would be especially stJong. THee. vi. 18. (Here ~vyKpa8f.P, not with <'iv, is equivalent to .,l gvyKpa&dq.) 'Ay.:Jvas <'iv Tts p.ot OoK<L, E</>7], tJ 'll"aTEp, 1TpOEL'll"iJJV ~KaU"TOt<;. Kal J.8.\a 1Tpon8d<; p.aAtU"r' av





1TOtEtV ev duKe'iuOat, it seems to me, said he, father, that if amy O'M should proclaim contests, etc., he would cause, etc. XEN. Cyr. i. 6, 18.
(Here the protasis implied in the participles is merely emphasised by l!v, which belongs to 1l'Otetv.) See also A.eyovTos l1v nvos 1l'tuHvuat ofeuOe; (i.e. d ns ;A.eyev, E1l'[O'TEVO'av tf.v ;) do you think "they would hO!IJe believed it, if any one had told them? DEM. vi. 20. (Here l1v stands near A.eyovTos only to point this out as the protasis to which its own verb 11'tO'TEVO'at is the apodosis, with which liv is not repeated.) 225. (a) Repetition of K~ is rare ; yet it sometimes occurs. E.g. T<il K( JLdA.' ~ K( V f-JLHV Kai EO'O'VJLV6s r.<p oooo, ~ KE JLE TE6V'YjVtav ;V JL<yripoww a7l'EV. Od iv. 733. (b) On the other hand, Homer sometimes joins l1v and KE in the same sentence for emphasis. E.g. KapTepal, lls oi;/ tf. v Ke v '' ApYJ> 6v6uatTo JLETeA.Owv ovT K' 'Aer,va['YJ<T0'6os. Il. xiii. 127.

226. When an apodosis consists of several co-ordinate clauses with the same mood, (fv is generally used only in the first and understood in the others, unless it is repeated for emphasis or for some other special reason. E.g. ~ " ' ' ' ' 0 ~ ' \ ' '(3 \ ' '\ ~\ 0 VO a V EJL> 'f)VtKa VVp0 U1!'01l'/\HV OV/\OfLYJV, KUTEKW/\VEV, OVVE TotavTa A.eyew TovTcp 7rpoO'haTTev, J~ <1lv ~Kt0'8' vp.eZs Jp.eAAeT'
E~tEVU.t. DEM. xix. 51. (Here av is understood with 7rp00'ETUTTV.) OvTW OE opwv oDoev a. V Otricpopov TOV ETEpov 71'0 tor, d..\A' Jr.l TUDTbV touv &pcp6npot. PLAT. Rep. 360 c. 0DKOVV Kliv, el 7rpbs avrb Tb cpws dvayKaCot avTbV {3Ae7l'HV, aA.yef:v T v Td. IJJLp.aTa KalcpevyHv &.7roO'Tpecp6p.evov (otH); lb. 515 E. (Kav belongs to the infinitives; 223.) See also XEN. An. ii. 5, 14.' IIrivTa iJP" <PA.t7r7l'os, r.oA.A.d. A.eyovTos ip.ov Kal 8pvA.ovvTos &.el, Tb p.ev 7rpwTov <1>s &v els Katvbv yvchJL'f)V d7l'ocpatvoJLEI'OV, JL<Td. TavTa 8' ws dyvoovvTas 8t8ri<TKOVTos, TAEVTWvTos 8(. WS a. V 7rpos 1l'1l'paK6T(JS aVToVs Kal dvoO'tWTriTOVS dv8pch7rovs ovoev vrrouuA.Aop.f.vov. DEM. xix. 156. The clauses with tils represent (1) eAEyov av d ecpcuv6p.YJV, as I should have spoken if I had been rnerely informing my colleagues; (2) ws EA.eyov (&v) el &.yvoovvTas i8alaO'Kov, as


I should have spoken if I lwd been instructing ign01ant men; (3) tils A.eyotp.t &v, as I slwuld spealo to men who had sold themselves, etc. In

the second clause, the construction remaining the same, l1v is omitted ; but in the third, where an optati,"e is implied, l1v reappears. In PLAT. Rep. 398 A, we find l1v used with two co-ordinate optatives, understood with a third, ancl repeated again with a fourth to avoid confusion with a dependent optative in a relative clause. "Av may be understood with an optative even in a separate sentence, if the construction is continued from a sentence in which l1v is used with the optative; as in PLAT. Rep. 352 E: ''EO'B' OTCf> llv IJ.A.Acp f8ots ~ 6cp0aAp.os; OD o~Ta. T oe; &.Kovuats l1Ucp ~ tiJO'[v; So with 1rp<iTTOL after y<ip, lb. 439 B.






227. "Avis sometimes used elliptically without a verb, when one can be supplied from the context. E.g. Ot olKemt peyKovaw dA.X oilK &v 1rpo Tov (se. ;ppeyKov), the slaves
are snoTing; but they wouldn't have been doing so at this hour in old times .AR. Nub. 5. '12> ovT' &v dO'TWV Twv8' &v Jdrrotp. T<p, OVT' &v TEKVOLO'L Tots ~fl'ol:s (se. Jd71'oLfi'L), O'Tepywv DfhWS. SoPH. 0. 0. 1528. T [ &v 8oKe O'OL Ilpfap.os (se. 11'pacu), el Trio' i)vvO'ev; but what think you P1imn would have done if he had accomplished what you have? AEscH .Ag. 935. L.wcppwv Jl-EV oDK &]/ p.aAAov, eilTvx~ o' rO'wS (se. OllO'a). EuR. Ale. 182: cf. .An. Eq. 1252. (See 483.) So 71'<0S -yap tiv (se. dq); how could it? 71'<oS o!JK lJ.v; and similar phrases; especially W0'11'ep &v el (also written as one word, W0'11'epuvd), in which the aJ' belongs to the verb that was originally understood after cl; as cpof3ovfl'evo> W0'11'ep &1' el 71'a'is,jeaTing like a child (originally for cpo(3ovfl'evos (},0'7/'ep &v Jcpo(3iiTo el 7/'a'is 'ljv). PLAT. Gorg. 4 79 .A. See DEM. xviii. 194: T[ XPlJ 7/'otew; Wu11'ep &v d ns vavKA:qpov 71'd.vT' J71'2 O'WT'fjp[a 11'priavm . , , T~S vava-y[as aln~To, what a1e we to do? (TYe aTe to do) just what a shipowner would do (11'ow'i lJ.v) if any one should blame him for the wreck of his ship, etc. See cp~uELev lJ.v, which explains the omitted verb, just afterwards.

228. K~v in both its meanings (as Ka[ with the adverb (;.v, and as Ka[ with iv = J~v) may stand without a verb. E.g.
'AA.A' /J.v8pa XP~ ooKetv 71'eO'etv &v K&v &71'6 O'fi'LKpov KaKov. SoPB' .Aj. 1077. (Here Ktiv, for Kat /J.v, which we may express by even or though it be, belongs to 7/'eu<'iv understood.) 'lKav<os oi'iv TOVTo ~XOfht:V, KiJ. v el 7/'AeovaxiJ O'Ko7ro';uv; me we then satisfied of this (and should w~ be so) even if we we1e to look at it in various ways? PLA.T. Rep. 477 .A. (We must supply ~Kuvws ~XOGfhEV with Kliv.) See different cases of Kav din 195, in whicl1 a verb follows to which Clv cannot belong. Ka.1 d7/'0L TLS &v, olfhat, 11'po0'8iJ KiJ.v JLLKpav OVVO.JLLV, 7/'avT', and, I think, wheTever we add.even (though it be) a little power, it all helps. DEAf. ii. 14. (Here Kllv =Kat tiv ns 11'po0'8iJ, even though we add.) MeTp'fJO'OV <lplJV'fJS T[ fi'OL, K&v 7/'evT' ETYJ, measuTe 'IIW out sonw peace, even if it be only fo1 five years (Kal &v fi'ETP~O''[I>) .AR. Ach. 1021. 229. ''Av may be used with a relative without a verb, as it is with .l (in av =El av) in the last examples (228). So in XEN, An. i. 3, 6, WS pov oi'iv l6vTOS d11''{/ &V Ka2 vpds, OVTW T~V yvwp.YJV ~xen (i.e. d7T1J &v Kat -&JJ,e'is r'fjT<), be of this 'lltind, that I shall go wherever you go.


230. This cha1)ter treats of all constructions which require any other form of the finite verb than the simple indicative in absolute assertions and direct questions ( 2). The infinitive and participle are included here so far as either of them is used in indirect discourse, in protasis or apodosis, and in other constructions (as with 7rp{v and &ScrTe) in which the finite moods also are used. 231. These constructions are discussed under the follow:lng heads:I. The potential optative and indicative. II. The imperative and subjunctive in commands, exhortations, and prohibitions.-Subjunctive and indicative with p,7] and p,1) in cautious assertions.-''07rw<; and o7rw<; p,~ with the independent future indicative or subjunctive. III. The subjunctive (like the future indicative) in independent sentences.-The interrogative subjunctive. IV. Ou P-0 with the subjunctive or future indicative. V. Final and object clauses after Zva, we;, b7TW<;, IJcppa, and p,ry. VI. Conditional sentences. VII. Helative and temporal sentences, including consecutive sentences with &ScrTe, etc. VIII. Indirect discourse. IX. Causal sentences. X. Expressions of a wish.






The Potential Optative and Indicative.

232. We find fully established in the Home1\ic language a use of the optative and the past tenses of the indicative with &v or IC, which expresses the action of the verb as dependent on circumstances or conditions; as e"AOo~ &v, he rnight (could or would) go; ~"AOEv liv, he n~ight (could or would) have gone. Such an optative or indicative is called potential.



233. It has already been seen (13) that Homer sometimes uses the optative in a weak future sense, without Ke or av, to express a concession or permission. Such neutral forms seem to form a connecting link between the flimple optative in wishes and the optative with aJ', partaking to a certain extent of the nature of both. (For a full discussion of these forms and their relations, see Appendix I.) Such expressions seem to show that the early language used forms like i!A.&o~Jl-~ and too~Jl-~ in two senses, I may go and I rnay see, or may I go and may I see, corresponding to i!A.&w and tow in their two Homeric senses I shall go and I shall see (284 ), or let me go and let me see (257). 234. The neutral optatives like Jl. iv. 18 are rare even in Homer, the language having already distinguished the two meanings in sense, and marked them in most cases by external signs. The optatiYe expressing what may happen in the future took the particle KE or av, and was negatived by ov, denoting the relations which we express by our potential mood with may, can, might, could, would, and should. Thus if.A.o~Jl-[ KE 1} KEV o.A.o1Jv, I may sla.y or I mny ue slain, Il. xxii. 253; dv~p of. K]1 ov n .6.~6s v6ov dpD<r<ra~ro, CL mnn cnnnot contend agninst the will of Zeus, Il. viii. 143.1 On the other hand, the simple optative (without Ke or av) was more and more restricted to the expression of a wish or exhortation, and was negatived by Jl-~ ; as Jl-~ 'Yf.votro, may it not hctppen, 7fBot6 Jl-D~, listen to me (Od. iv. 193), as opposed to ovK dv ")'EVO~TO, it could not happen. The potential forms i!A.&o~Jl-~ av
1 When the idea of ability, possibility, or necessity is the exr1ression, and is not (as above) merely auxiliary, special verb like 0<\vap.a<, iii, or xpfJ. Especially, the generally expressed by ilfi or XPiJ with the infinitive ; h:i1n we must obey, SoPH. Ant. 666.

the chief element in it is expressed by a idea of obligation is as ToDil xpi) KAVEw,




and l8otf1-t 6.1' differ from the more absolute future indicative and the old subjunctive forms ~A.Bw and lBw, I shall go and I shall see, by expressing a future act as dependent on some future circumstances or conditions, which may be more or less distinctly implied. The freedom of the earlier language extended the use of the potential optative to present and sometimes even to past time. See 438 and 440.

235. In most cases the limiting condition involved in the potential optative is not present to the mind in any definite form, and can be expressed in English only by such words as perchance, possibly, or probably, or by the auxiliaries could, would, should, might, etc. with the vague conditions which these imply (like if he should try, if he pleased, if he could, if what is natuml should happen, etc.) Sometimes a more general condition is implied, like in any possible case; as oi!K Clv 8 EX o [fl-'IJ v TovTo, I would not accept this (on any terms); here the expression becomes nearly absolute, and may often be translated by our future, as OlfK Clv ?- E d ?-'I] V TOV p6vov, I will not give 11p the throne ( AR. Ran. 830), or (in positive sentences) by must, as 'll'<fvTEs BavfJ-d.Cot<v <iv TovTo, all must admire this. The optative thus used with no conscious fe~ling of any definite condition, but still implying that the statement is conditioned and not absolute, is the simplest and most primitive potential optative. It is equivalent to the Latin potential svbjunctive, as credas, dicas, cemas, putes, etc., you may believe, say, perceive, thinlc, etc. The , Homeric language has six forms, all expressing futurity with different degrees of absolutene~s and distiuctness; as Stfof-l'at, Stfof-l'a[ K<, Z8w}J>at, l8wf-J'a[ K<, l8o[f-J'YJV, l8o[f-J''7v KE (or &v), containing every step from I shall see to I should see. Of these only the first and the last (with a tradition of the seconrl) survived the Homeric period, and the others (especially the fifth) were already disappearing during that period (240), being found unnecessary as the language became settled, and as the optative with Ke or &v became more fixed as a future potential form.

236. In the following examples of the potential optative no definite form of condition is present to the mind:'Ef-J'ol ()~ T6i Clv 'll'OAV KEp8wv E rYJ, but it would at that tirne (be likely to) profit me far nwre. Il. xxii. 108. 'PEvyw}J>v en y&p K<v dA.v~at /J'f!V KaK6v ~f-J'ap, let us flee; for perchance we rnay still escape the evil day. Od. x, 269. IT.\"JO"[ov dA.A.~A.wv Ka KEV 8to i'O"TEVO"oas, the roclcs are close together: you might perhaps slwot an arrow across the space. Od. xii. 102. So Od. xxiii. 125. 0-JKOVV 'Tf'O potS av T~VO<: 8wpeav ~pool; would you then grant me this favour? .AESCH. Prom. 616. So




1rav yd.p &v 1r~ Bo t6 p.ov, for you can learn anything (you please) from me. lb. 617. T T6v8' av d?ToL<; O.A.A.o; whatelsecould yousayof this '11Uln? SOPH. Ant. 646. So Ant. 552 and 652. IIoA.Ad.s av e!Jpo LS p:'Jxav&s, you can find nwny devices. EoR. And. 85. "Et(op.a 'TOL Ka~ oilK av Aet4>8dYJV, I will follow you and in no case will I be left behind. HDT. iv. 97. Oi p.~v (se. AE')'OV'T<;) <ils oil8ev~ av 7p67Tip a 0 L( V oi 'ABYJva'iot. THUC. vi. 35. HEvBa 7TOAA~v p.f.v crw1>pocr~VYJV Ka'Tap.6.8ot av 'TLS. XEN. An. i. 9, 3. So Mem. i. 3, 5, iii. 5, 1 and 7. A' ' \ ' ' ' ' '' ' ' .ut> es 'TO V av'Tov 7TO'Tap.ov ovK av e p. (3 a LYJ s, you cannot step tmce ~nto the same river (saying of Heraclitus). PLAT. Crat. 402 A. Ov p.~v ECT'TL KaAAtwv ~8os ovo' av y h 0 L'To, there is none and there could be none. Id. Phil. 16 B ; so 64 B. 'AKoVot<; &.v, you can heM. Id. Rep. 487 E. /j.n~d'Tw <ils oi 8e'T'TaAoi vvv oilK av JA.ev()epot yvotV'TO li.crp.evoL, let him show that they would not now gladly become free. DEM. ii. 8. 'Hoew<; 8' av i!ywy' Jpolp.YJV Ae7T'TlVYJV, but I would gladly ask Leptines. Id. XX. 129. El 1J')'V6YJCTE 'TaV'Ta, ')'EVOL'TO ydp av Kat 'TOV'TO, if he did 1wt know this,-and it might easily so happen. lb. 143. OiJ,.' av OV'TOS i!xot A.ynv ove' ilpe'is 7TELcr8dYJ'T. Id. xxii. 17. Ilo'i ovv ,. p a 1r o p e 8' av in what other direction could we possibly turn ? PLAT. Euthyd. 290 A. OvK av pe8elp.YJV 'TOV ()p6vov, I will not give up the thTone. AR. Ran. 830. So OVK av oexo [p.YJV, AESCH. Eum. 228. Tts oiK av ay &.era L'TO 'TWV avopwv JKelvwv -rfjs ape'Tfj<;; who would not admire the valour of these men ? (i.e. every one 1nust admire their valour). DEM. xviii. 204. BovA.olp.YJv &v, I should like, is used like velim. For J(3ovA.6p.YJv CJ.v, vellem, see 246.

en ;

2;iJ f-EV KOtJ-[(ots av CTEaV'TOV fi BeAns, you may take yourself off whither you please (a milder expres~ion than K6pt(e creavr6v). SOPH. Ant. 444. So Ant. 1339. K AVo L<; av ~ory, i:f>o'if3e 7rpocr-ra-r~pte, hear me now. Id. El. 637. Xwpo'is aJ' el&w. Id. Ph. 6i4. So probably Il. ii. 250 : ,.0 OVK av f3acrtA.fjas avd crr6p.' ~xwv a')' 0 p E 1~ 0 L S, therefore you must not take kings upon your tongue and talk (Or do not take, etc.)

237. The potential optative in the second person may have the force of a mild command or exhortation. E.g.

238. Occasionally the potential optative expresses what may hereafter prove to be true or to have been true. E.g. IIov ofjr' av elV o1 ~vot; whe1e nwy the strangen be? (i.e. where is it likely to turn out that they me ~) SoPH. El. 1450. 'H yd.p p~ (se.
cro4>a) 4>a.VA1J 'TLS av E t1J, for it may htrn out that my 1L"isdom is of a mean lcind. PLAT. Symp. 17 5 E. 'EA.A.~vwv nvds 1>(L(TL ap?Tdcrat Evplb7T1JV dryer a v 8' <'iv oil-rot Kpfjns, and these u;ould prove to be Cretans (or to have been Cretan-~). HDT. i. 2. Av'Tat 8~ OVK av 1rOAAai Et1JCTa v, and these (the islands) would not prove to be many. TRue. i. 9. This l1as nothing to do with the Homeric use of the optative with 1<~ or av in a present or a past sense (438; 440). See th.e similar use of the subjunctive with p~ after verbs of fearing (92).




239. The potential optative may express every grade o potentiality from the almost pure future ouK ~v p.e8dp.'l]v, I will . . rw t gwe up ( der any ctrcumstances)to'"~''' un , ovK av otKatws es KaKov 'li'EU'otp. n, I wuld not justly fall into any trouble, SoPH. Ant. 240, where &Kaws points to the substance of a limiting condition, if fustice should be done. From this the step is but slight to such cases as OVTE EU'8[oVU'L 'li'Ae[w ~ ovvavTaL <jlepHV. Otappaydev yd.p av, they do rwt eat more than they can carry/ for (if they should) they would burst, XEN. Cyr. viii. 2, 21, where el EU'8[otev is necessary to complete the sense and is clearly understood from the preceding words. A final step in the same direction is taken when the condition is actually stated as part of the sentence. As ~'AOot av means he would go (under some future circumstances), if these limiting circumstances are to be definitely expressed it is natural to use the corresponding form of condition, el with the optative, as cl K<AEVU'Ha> ~'AOot av, if you shmtld command he would go. The protasis is thus assimilated to the apodosis in form, as it conforms to it in sense and general character. So when a conclusion is to follow such a condition as cl K<AEVU'Has, the corresponding optative with i!.v, i.e. the potential optative, is naturally chosen, although nothing but regard to harmony and symmetry makes either if you should command he will go or if you com1nand he would go, or the equivalent Greek forms, objectionable. In fact, these very forms are far more common in the more fluid Homeric language than in the fixed and regular style of Attic prose. There is, therefore, no necessary or logical bond of union between two forms like cl k<A<vU'eta> and ~'AOot fiv. This connexion is, indeed, far more the effect of assimilation in form, as appears especially when the apodosis contains an optative in a wish; as in W> a11'6AotTO Ka2 ({)..A.os OTLS row.:vni y< pf.(ot, may another perish also who shall do the like (Od. i. 47), where if a1I'oArr8w had been used we should naturally have had pf.(YJ.
For examples of the optative with /lv or Ke with a definite protasis expressed or implied in the context, see 455 and 4 i2.

240. The use of r'lv or KE with the potential optative had already become tlxed in the Homeric language. A few cases of "neutral optatiYes" in Homer, which seem to show an en.rly 1)otential use without KE <Jl' av, have Leen given above (13). Besilles these, a few more distinctly l)Otential optatives without av or KE occur in Homer, but they are exceptions to the general usage even there. Such are the following : OlJ n KaKwnpov aAAO 1I'a8otp.t. Il. xix. 321. TovTOV ye U''li'OJLEVOLO Ka2 JK 'li'vpdS al8op.evow ap.</Jw VOU'T~U'a LJLEV. Il. X. 246. 'P<fa e.6, y' WeA.wv dfJ>dvovos 8wp~<TatT0, IL X. 556! see Od. iii. 231. Xepp.a0wv 'Aa/3, 0 o.J y' avOp <flepotev. Il. v. 302: so XX. 285.






Oll n<; 1rduE u yvvai:Ka. Od. xiv. 122. So also Il. vii. 48, xiv. 190, 45, 197. See, further, HEs. Theog. 723 and 725; PIND. 01. x. 21, Py. iv. 118.

241. Some cases of the optative without !1v oc~ur with the indefinite
O<; in Homer, and with unv oun<;, ~UTtV OTrW<;, f.unv O'TrOt, in the Attic poets. These form a class by themselves. E.g. OiJK u0' os u~> yE dv.a> KE<j>aA~> dTraAaAKot. Il. xxii. 348. Ov yd.p ryv os Tis u<j>tv i1rt uTixa> ~y~ua tTo. Il. ii. 687. OvK f.uO' o1rws A.f.~a'JL' Td. tfwoq KaM.. AEscH. Ag. 620. OiJK f.uO' oT<p JL<[(ova JLO'ipav v<[fLatJL' ~ uot. Id. Prom. 292. Oi~K f.unv oun~ TrA~v EJLOV KdpatTO vw. Id. Cho. 172. "EuT' ovv o1rws ''AA.wquns is y~pas JLOAot; EuR. Ale. 52. "EuO' o1rot ns TrapaAvuat tfvx&v; Ibid. 113. 242. On the other hand, a few other cases in the Attic poets are mere anomalies, even if we admit that the text is sound. E.g. Ted.v, Zev, 01;vauw Tis dvopwv vTr<p(3au{a KaTaux o,; what transgression of rnan can check thy power? SOPH. Ant. 605. 'AA.>..' V7rEpTOAJLOV avopos <j>povrwa T<; A.f.yot; AESCh. Oho. 594. llw> ovv T&o', ws d1ro t n>, E~YJJL&pTavE>; i.e. as one rnight say. (?) EuR. Andr. 929. BUo-uov ~ A.f.yot ns 7TWAOV> eun)uaJLEV. Id. Hipp. 1186. ''flf1'TrEP d1rot Tt> TOTros, as 'one would say roTros. 0) AR. Av. 180. The cases cited from Attic prose are now generally admitted to be corrupt. See Kriiger, ii. 54, 3, Anm. 8.



243. As the potential optative represents a future act as dependent on future circumstances (234), so the potential indicative originally represents a past act as dependent on past circumstances. Therefore, while ?jAOEv means he went, .fjA.OEv ;!,v means he wou.ld have gone (under some past circu.mstances). It is probable that no definite limiting circumstances were present to the mind when this form first came into use, so that 1jA0Ev av naturally signified merely that it was likely, possible, or probable that he went or (as we express it) that he mi,qht have gone or wonld hnve been likely to go, sometimes that he lwve gone. In this sense it appears as a past form of the potential optative, e.g. of in the sense he might perchance go or he wou.ld be likely to go (in the future). The same relation appears in Latin, where credas, pnte.>, cemas, dicas, you would be likely to believe, think, etc., are transferred to past time as e1ederes, putaTes, cerneres, diceres, you. wou.ld have believed, tlwught, etc. 1 Here putet and

neo, av

1 We are probably justified in assuming that the past meaning w hi eh here appears in crederes, etc. is the original meaning of the Latin imperfect sub. junctive in this use, as it cert11inly i; that of the Greek imperfect iudicative with 6.v. Sec 43:i.




putaret are precisely equivalent to ofot'TO :iv, he would be likely to think, and ~E'To :lv, he would have been likely to think. 244. We find the potential indicative in its simplest use (last mentioned)-with no reference to any definite condition, but merely expressing past possibility, probability, or necessityin all classes of Greek writers. E.g.
Ovo' &v ~n <f>pdottwv 7rep dv~p '2:ap7r'Y)06va oZov ~yvw, no longer would even a shrewd man have known Sarpedon. Il. xvi. 638. 'Y1r6 KEV 'TaAarrl<f>pova 7rEp oeos e TA.ev, fear ?rvight have 'seized even a man of stout heart. Il. iv. 421. See other Homeric examples below. 'AA.X :q A E JLEV 0~ 'TOVTO rollvetoos 'Td x' &V 6pyfl f3 tarr8ev ttfiA.A.ov ~ yvWJLU <f>pevwv, but this repToach may pe1haps have come from violence of wrath, etc. SoPH. 0. T. 523. (Here TaX: &v :qA.8e expresses past possibility, with no reference to any definite condition, unfulfilled or otherwise.) 8eo/:s yap :qv ovrw <f>A.ov Tax' d. v n JL'YJV{ovrrw Els yevos 7raAat, for perchance it may have been thus pleasing to Gods who of old bore some wrath against onr race. Id. 0. C. 964. (According to the common punctuation Tax' &v would be taken with JL'YJVovrrw, = oi' rdx' d.v 1't (pryvwv, who may perch111nce have bome some wrath, see PLAT. Phaedr. 265 B, below; but the analogy of 0. T. 523 favours the other interpretation.) Ilpos 7r'OLOV &v r6vo' avr6s OVOVtT<TEVS ~?TAEt; i.e. who might this man have been to whom Ulysses was sailing ? Id. Ph. 57 2. "0 8wrrdttevos ?Tas av ns dv1)p ?jparr8'YJ Mws elvat, every rnan who saw 'this dranta (the "Seven against Thebes ") would have been eager to be a warrior. AR. Ran. 1022. (This is the past form of 7ras d.v ns Jparr8e ['Y) Saws e'tvat, every one would be eager, having no more reference to an unfulfilled condition tha'n the latter has.) D. t {3'Y)rra v, ws JLEV EGKdS Kat A.eyE'Tat, E7r't tTXEDtWV, raxa &V OE Kat & ?TWS EtT7rAE{J<TaV'TES, i.e. while they probably cTossed on rafts, they may pe1haps have crossed in some othe? way by sailing (8t{3'Y)rrav with Taxa &v in the latter clause meaning they ??taY have, or ?night have, pe1haps crossed under other (possible) ci?'Cumstances). THUC. vi. 2. 'E7reppw<T8'Y) o' av TtS EKEtVO lowv, and any one would have been encouraged who saw that. XEN. Hell. iii. 4, 18. 8aTTov i) ws ns <'i.v 1Je'To, sooner than one would have thought. Id. An. i. 5, 8. ''EvBa o~ ~yvw av ns ocrov d.gwv et'Y) 'TO </>tA.e'ia-8at apxovTa, there any one might have leamed, etc. Id. Cyr. vii. ' ' '\ I \ I ' ' ~ , .,. I\ 1 38. 'E V Ta11T'[J 'T'[J '1)1\tKt<f 1\/'0VTeS 7rpoS Vf'aS EV TJ av fta/\t<TTa ', E?Tt<TTE{J<Ta'Te, tallcing to you at that age at which you would have been most likely to have put trust in them. PLAT. Ap. 18 C. ''l<Tws JLEV dA.'Y)Bovs TWOS f?TU.7r'TOJLEVot, raxa o' &v Kai ctAAO<TE ?Tapa<f>EpOJLEVot, JLV8tKOV TtV<l vp.vov 7rp0CTE1ra[<Taf'EV "Epwra, while perhaps we were clinging to some truth, although perchance we may have been led aside into so?ne error (?Tapa<f>p6JLEVOt d.v = 7rape<f>Ep6j-te8a /J.v), we celebmted Eros in a mythical hymn. Id. Phaedr. 265 B. T yd.p Kat f3ovA.6ttevot ftETE7r'EJL7r'Err0' av avrovs EV T01JT<P "r<tJ Katp<j); jo? with what wish even conld you possibly have been summoning them at this time? DEM. xviii. 24. Ilws av ~ ftlJ 7rap6w JL'YJO' E7rtO'Y)j-tWV f.yw 'T[ <TE ~O[K'Y)<Ta; i.e. how was I




likely to do you any wrong? Id. xxxvii. 57. ToY xopoy uwD..e~a i:xnrtp lJ.y ~8tu'Ta Ka~ JmT'I}8et6TaTa dp.cpoT~pots iy[yYeTo, I collected the clwrus in the way wlvich was likely to be most agreeable and convenient to both. ANT. vi 11. Two Homeric examples are peculiar in their reference to time : ' AA.Ai!. TaX una 1retpa o1rws KEY 8~ ~Y TraTp8a yal:ay tK'I}at ~ yap (w6Y ye KLX~ITEat, ~ KEY 'Op~IT'T"IJS K'TEtYEY ilTrocpBap.eYos, uv 8~ KEY Ta<j>ov dyn{3oA~U"ats, but strive with all speed to come to your fatherland; for either you will find him (Aegisthus) alive (and so can kill him yourself), or else Orestes may have already killed him before you come, and then you can go to his funeral. Od. iv. 544. (Here ~ KEY K'TEtYeY, by a change in the point of view, expresses what will be a past possibility at the time of the arrival of Menelaus, to which time the following optative is future.) Ka~ yap Tpwas <j>aU"t p.aX'IJTas f.p.p.evat aY8pas, oZ Ke TaXLU"Ta EKptvav p.~ya YEtKos,for they say that the Trojans are men of war, wlw would most speedily lwve decided a mighty strife (implying that they would therefore speedily decide any impending strife). Od. xviii. 261. (This was said by Ulysses before he went to Troy. See 249.)

245. In most cases of the past tenses of the indicative with

aY there is at least an implied reference to some supposed circumstances different from the real ones, so that ~A.IhY aY commonly means he would have gone (if something had not been as it was).

When we speak of a past event as subject to conditions, we are apt to imply that the conditions were not fulfilled, as otherwise they would not be alluded to. This reference to an unfulfilled condition, however, does not make it necessary that the action of the potential indicative itself should be unreal, although this is generally the case. (See 412.) The unfulfilled past condition to which the potential indicative refers may be as vague and indistinct as the future condition to which the potential optative refers (235); as if he had wished, if he had tried, if it had been possible, in any case, and others which are implied in our auxiliaries might, could, would, slwuld, etc., but are seldom expressed by us in words. Compare ov8eY av KaKOJJ 7r0t~(]"EtaJJ, they C(fUld do no harm (i.e. if they should try), with ov8ev ll.v KaKOJJ eTrolTJU"a.v, they could have done no harm (i.e. if they had tried). E.g.
Ov yap Kev 8vvap.eU"Ba 8vpawv 111{"1/A.awv rhrwU"a.U"8at A.t8oY, for we could not have moved the stone from the high doorway. Od. ix. 304. Mf.votp.' &v. ~e. AOJJ 8' &v EK'TbS ~~~ 'TVXtV, I will remain; but I slwuld have preferred to take my chance outside. SoPH. Aj. 88. TovTov T[s /J.v <rot Tdv8pos &p.,[vwv vp~B"IJ; who could have been found, etc.? lb. 119. 'EKAVDJJ &v Jyw ov8' {ly ~ A7r L(]"' avoav, I hea7d a voice which I could never even have lwped to hear. Id. El. 1281. 6.v' Jtf.>..e~a.s, oTv eyw ryK<a--r' &v ~(}~),_'YJrr' JA.wA6Tow KAVv. Id. Ph. 426. KA.vetv &v o~8' a1ra~ Jf3ovA.6p.'fJv, I should have wished not to hear it even once. lb. 1239. OvK E(]"B' 07rWS ETKEY lJ.y TJ 6.tos oap.ap A'l)'TW 'TOU"aV'T'l)Y




dJJ-a8[av, under no circunMtances would Leto have been the mother of so great ignorance. EuR. I. T. 385. 0lKta 1rpayJJ-aT' cl<raywv, i~ illv y av E~TJAE"'fXOtLYJV, by which I might have been exposed. AR. Ran. 959. ToTE 6ft ~v, Kat TOS X.Etpas OVK av Ka8wpwv, it was then dark, and they would not have seen the slww of hands (in voting). X1m. Hell. i. 7, 7. IIo!wv 8' av ~pywv ~ 1f0VWV 1J Ktv86vwv <l1fE(T'TYJ<Ta V; frorn what acts, etc., would they have shrunlc back (i.e. if they had been 1equi1ed of them)7 Isoc. iv. 83. IIpo 1fOAAwv JJ-EV av XPYJJJ-rf.Twv iTtJJ-TJ<Trf.JJ-1JV TO<TOVTOV 86va<r8at 'T~V <jnAo<ro<jJ[av. L<TW'> yap oiJK av 1JJJ-ds 7r AEW'TOV arrAd</J8'Y}JJ-EV, ov8' av eArix_t<TTOV fJ-EpO<; U7T'EAav<raJJ-EV avT~<;' irr<dJry 8' ovK ovrw<; gxa, f3ow\o[JJ-'fJV <lv 7ra-6(ra<r8at Tots </>AvapovvTas. Id. xiii. 11. Ot' E11"0 [YJ<TU. V JJ-~V ov8f.v av KW<uv, P-~ 7T'afhi:v 8' J<j;vAa~av'T' av i'<rw;;, TO{;Tov;; e~a"/T'a'TUV atpcw8at, these who could have done him no hcmn, but wlw might JJe?haps have gnarded themselves ngainst suffering any. DE~L ix. 1:3. ToTE 8' a1ho 7'0 7rpaytl av f.Kp[1'7'0 i<:p' avTov, b1tt the case wmdd then have been decided on its own merits. Id. xviii. 224: so 101. l1ws Ill' ovv vf3purnKWTp011 av8pw7ro<; l>fJ-ll' ixp~<raro; Id. xix. 85. Ov jJ-(t(ov ov8f.v av KQTE A t1l'V livt8os. Id. xlv. 35. ''A 8' 1JJ1-'i:v 8tKatw<; al' V1f~fJXEV EK -r~;; dp1JVTJ'>, -rav-r' dv8' illv d7rE00l'TO a-&Tol Aoy!(c<r8aL. dJ...J...a TQUTQ [J-EV 1l V av OfJ-OlW';; 1Jf1-L1', EKE'i:va of. -ro-6Tots ill' 7rpo<r~v d fl-1J 8ta TOVTOVS, bnt (it is not 1iuht) to set off against what they themselves sold what would justly have been ou1s by the peace; but these W01Lld huve been anTS all the sarne (in wny case), while the othe?'S would have been added (or would now be added) to them, had it not been fm these men. Id. xix. 91. (Here V7T~PXEV av and ~v av refer to an actual fact, the possession of certain places; the apOUOSiS 1rfJOfT~V a11 refers tO SOlllething Which WaS prevented from becoming a fact, This passage shows the natural steps from the potential form to the apodosis. See 2 4 7.)

246. When no definite condition 1s understood with the potential indicative, the imperfect with dl' regularly refers to past time, according to the older usage (435), like the aorist; a~ in the examples above. The imperf(~ct referring to present time, which is common in apodosis a.fter Homer ( 410), appears in these potential expressions chiefly in a few simple phrases, especially in i:f3ovA6Jl-'7v vellem, I should wish, I shm;,ld like (also I should have lilced). Even in Homer the construction with <J<j;eAov and the infinitive ( 424 ), which includes a form of potential indicative (4-15 ; 416), sometimes refers to present time. E.g. 'Eyti.! 8' J(3ovAOJ1-TJV &v O.VTOV<; dA1]8~ J...yav JJ-E'rryv yap av Kai Jp.ol TOVTOV Tdya8ov OVK i.:Acixt<TTOV f'EfJO'>. vvv of. OVTE np6;; T~v 7r0At]l U.VTOLS TOtavTa vrrapxEL OVT npb> Efi-E, and I should like it if they spoke the truth; for (wC?e that so) no small JJa?t of this advantage would be mine: but this is not true of them, etc. LYs. xii. 22. Mno[av, oF E.f3ovA6p'fjv &1' :>roAAwv El'EKEv Nl', Midias, whornfo? many reasons I





should like to have alive. AESCHIN. iii. 115. See LYCURG. 3, (For i.f3ovA6fL'YjV av as past, see SoPH. Ph. 1239, quoted in 245.) See also .AR. Nub. 680, iKEtvo 3' i)v &v Kap367r'YJ, KA.EwVVfL'YJ, and this would be Kap367r'Yj, etc. For f3ovA.o{fL1JV tlv, velim, see 236. For IJcpEAov and the infinitive as 1)resent in Homer, see 424.

247. It is but a slight step from the potential forms quoted in 245 and 246 to those which form the couclusion to au unfulfilled condition definitely implied in the context. After Homer the imperfect with av may here refer to present time. E.g. 'A;\Aa KE J<EtJ'a fLrLAUTTa l36Jv oA.ocp1;p(w 8vp.~), but you uould hnve lamented most ~n yon1 heart if you had seen this (iowv = d EiOEs). Oll. xi. 418. ODo J<EV o.Dros V7rEK</JV{'E ~<~pa p.EAall'av, d,\A.' 'H<f,a.wros pvro, nor would he uy himself lwve escaped, but Hephaest?<s nscued him. n. v. 22. 'AJ...).: dKJ.(]"(}.t fLEl', 1)o15s ov yd.p &1' t<apa 1roA.v&TEcp1)s <So' Eip7rE, but, as it seems, he has good news; jm (otherwise) he w01dd not be coming with head thus thickly C?"OW1Wd. SoPH. 0. T. 83; so 0. C. 125, 146. IIoA.A.ov yd.p av Tct ?Jl' a~ta, fm instrunwnts would be worth much (if they had this powe1). PLAT. Rep. 3 7 4 D. ''H )'ETE n)v tlp,]v?)JI DfLws ov ya.p i}1' on O.v J1rotEtTE, jo1 the1e was nothing that you could have done (if you lwd not kept tlw peace). DE~L xviii. 43. "2YJfLEZov 3 ov ya.p C!.v 3Evp' i) tc o v los 1~fLfis, for (otherwise) they n:ould not have come hitlwT to yon. Id. xix. 58. T6TE if>t,\[7r7r4J 7rpo3E3wJ<vat 7T'Ul'Tos C!.J! (]"X E Jl air[a,JI, in that case she (Atl1ens) would have had the bla,me of having betrayed all to Philip. Id. xviii. 200. See other examples in 4 7 2.

248. The final step is taken when an unreal condition i.s expn;sse<l as part of the seutence, forming the protasis to which the potential indicative is the apodosis ; as 1JA8El' a:l' El fKEAEVtTa, he would have gone if I had commanded him. The dependent protasis, by a natural assimilatiop, has a past tense of the indicative corresponding to the form of the apodosis. On the other hand, when an unreal condition has been expressed, as El fKEAEV(]"{J., the potential indicative is the natural form to state what would have been the result if the condition had been fulfilled. (See 390, 2; and 41 0.) The potential indicative does not change its essential nature by being thus made part of an unreal conditional expression, and it is not necessarily implied tbat its action did not take place (see 412). Although the latter is geJJcrally implied or inferred, while the reverse selclom occurs, still it is important to a true understanding of the nature of the indicati,e with llv to remember that it is not esRential or necessary for it either to refer to an unreal condition or to denote in itself what is contrary to fact. For a periphrastic form of potential indicative with EOEt, XP1)J!, etc:, with the infinitive, see 415.




For the Homeric use of the present optative with K~ or d.v as a present potential form (like the later imperfect with ilv), see 438. For the rare Homeric optative with K~ in the sense of the past tenses of the indicative with kf or &v, see 440.

249. From the primitive use of the past tenses of the indicative to express what was likely to occur under past circumstances, we may explain the iterative use of these tenses with ilv (162), which is generally thought to have no connection with the potential indicative with llv. Thus -i})Jhv d.v, meaning originally he would have gone (under some past circumstances), might easily come to have a frequentative sen~e, he would have gone (under all circumstances or whenever occasion offered), and hence to mean he used to go. See SoPH. Ph. 443, Ss o-DK av d.Ae-r' Elcrd:rra.~ el'TrEtv, 01rov p:ryoEi<; f.pYJ, (Thersites) who used never to be content to speak but once when all forbade him (lit. when nobody permitted him). Originally ovK EtAETo would mean he would not have been content (under any circumstances), hence he was never content. The optative f.i[>YJ (532) shows the nature of the expression here. See the examples under 162, and the last example under 244. This construction is not Homeric ; bnt it is found in Herodotus .and is common in .Attic Greek. There is no difficulty in understanding it as an offshoot of the potential indicative, when it is seen that the latter did not involve originally any denial of its own action.



The Imperative and subjunctive in Commands, Exhortations, and Prohibitions.-Subjunctive and Indicative with 1-'-~ and ~-'-~ ov in Cautious Assertions.-''07rw'> and g7rws- 1-'-~ with the Independent ,_(, Future Indicative, etc~

250. The imperative is used to exhortation, or an entreaty. E.g.


a command, an

A.fyE, speak thou. ihvyE, begone! 'E.AB.iTw, let him come. XatpovTwv, let them rejoice. ''EpxEcrBov KAtcrEYJv IT'JAYJuf.oEw 'AxtM)os. Il. i 32~. ZEv, BEwpos TwvoE 1rpo.yp.aTwv yEvov. AEscH. Cho. 246. For prohibitions, i.e. negative commands, see 259 and 260.

251. The imperative is often emphasised by &yE or &ymc:, </>epE, I:Bt, lkvpo or oEvTE, come, loolc here/ or by El o' &yE (474). ".AyE, cpepE, and I:Bt may be singular when the imperative is plural, and in the second person when the imperative is in the third. E.g.




EZ'If' fOG Kal Tovoe, <f>[Aov T~Ko~, 8~ n~ i:lo' iO"T[v. Il. iii. 192. 'AU.: liye fL[fLveTe 7ravns, EvKv~fLtOe~ 'Axawf. Il. ii. 331. Ba(J"K'


Uh, ov>.e lJvHpe, fJodS ~7Tt v~as 'Axatwv. 11. ii. 8. Aye 80 &.Kowan. XEN. A.p. 14. "AyeTe OH'lTV~a-au. XEN. Hell .. v. 1, 18. <Mp' el'lT 8~ fLOt. SoPH. Ant. 534. <f>fp< 8~ fLOG To8e eL'lT~. PLAT. Crat. 385 B. 80 A.f.~ov TJfLtv 7rpWTov TovTo. XEN. Mem. iii. 3, 3. "IfJ, vvv 7rap[O"Ta0"8ov. AR. Ran. 13 7 8. I Gvvv A.t(3avWTOV bevpo TLS Kal 71'iip OOTW. Ib. 871. Ka{ fLOt 8evpo, Jj MA1)Te, el7re; PLAT. Ap. 24 ~Evn, A.e[71'ETe O"T~yas. EuR. Med. 894.


252. The poets sometimes use the secorul person of the imperative with 7ras in hasty commands. E.g. "AKove 7ra~, hear, eve1y one! AR. Thes. 372. Xwpet 8<vpo 7ras
V1T"YJPETYJ~ TD~EV<, rratE' O"<f><vo6v'Y)V r[s fO, 8oTw. Id. Av. 1186.
8~ O"tW'JTa 7ras dv~p.


Id. Ran. 1125.

253. The imperative is sometimes used by the dramatists after ot<r8' oand similar interrogative expressions, the imperative being really the verb of the relative clause.1 The difficulty of translating such expressions is similar to that of translating relatives and interrogatives with participles. E.g. 'A>..>..' ot<rfJ' oopaO"ov; T<f O"KEAH fJevE TTJV rrhpav, but do you krww
what you 'lltust do?-strike the 1oclc with you1 leg! AR. Av. 54. OtO"fJ' U'VfL7rpa~ov; do you know what you 'lltust do fm 1ne? EuR. Her. 451. Ot(J'8a vvv iJ. JW' yev0"8w; b<O"fLa ToZs ~f.vota-t 7rpoO"fJEs, do you know whrd must be done for me ?-put bonds on the stmngers. Id. I. T. 1203. 0tU'8' et,. 'lTOl'YJ<TOv; do you know how you must act? SOPH. 0. T. 543. (Compare EuR. Cyc. 131, oicrfJ' ovv il fipaU'os; do you krww what yo1i a1e to do?) The English may use a relative with the imperative, as in which do at your peril. See HDT. i. 89, KanU'ov <f>vA.aKovs, oi' >..ey6vTwv 6Js &.vayKa{ws lxet. So SoPH. 0. C. 4 73. A peculiar interrogative imperative is found in JI-TJ ~~~O"Tw; is it not to be allowed? PLAT. Polit. 295 E; and J:7ravEpwTw d Kda-fJw, I ask whethm it is to stand, Id. Leg. 800 E. (See 291.)


254. The imperative Rometimes expresses a mere assumption, where something is supposed to be true for argument's sake.. E.g. IIA.ovTet TE yap Ka7.' oiKov, el {301!Aet, fLEya, Kal (~ nlpavvov (J'xq;J xwv, i.e. grant that you an rich and liw in tyrant's state (lit. be rich, etc.) SOPH. Ant. 1168. IIpoO"Et71'<ITW nva </>tAtKWS 0 TE apxwv Kat lotWT'Y)>, suppose that both the ruler and the private address one in a friendly way. XEN. Hier. viii. 3.


255. The want of a first person in the imperative is supplied

1 See Postgate in Transactions qf the Cambridge Philological Society, Ill. 1, pp. 50-55.




by the first person of the subjunctive, which expresses both positive and negative exhortations and appeals (the negative with JL~). ''Ay<, aym, El o' fly., cpep, ret, oevpo, and OeV'T (251) may precede this subjunctive; .so sometimes a, pe1'1nit, let.

VEWJ1-o8a, r6v8e o' f.wp,ev, let us sail homeward with OU?' ships, and leave hirn. Il. ii. 236. 'AA.X aye P,t]KETL Tavra A.eywveea, but come, let us no longe?' talk thus. IL xiii. 292; so ii. 435. 'AAX aye 01J Kai VWt p,e8wp,8a eovptOOS aAK~S' 11. iv. 418. El 8' flyer' av<f>t 1T'OAW <TVV revxe<Tt 1T'Hp1)8WJLev. Il. xxii. 381; so 392. ihvTe, cp[Aot, r6v ~eZvov f.pwp,e8a. Od. vi1i. 133. M1) 81) 1T'W A.vwp,eea t?T?Tovs, aA.X 16vns II<hpoKAOJI KAa[wfheJJ, Il. xxiii. 7. 'A.\X d OOKeZ, 1T'AEWJLev, 6pfLau-Bw raxv>. SoPH. Ph. 526. 'E1T'<TxeroJI, p,a(Jwp,!1, Ib. 539. < o~ ow?T<pavwfl'<JI A.6yovs. Euu.And. 333. 6.evp6 0'01.1 O'TEfw Kapa. Id. Baccb. 341. 'E?T<Tx<>, f.p,(3aA.wp,ev ds aAAov 'A6yov. Id. El. 962. IIapWJLEV TE ol'Jv W0'1T'EfJ KvpoS' KeAdJEl, a(]"KWfJ>Ev Te o.' il>v JhaAt<TTa OlJliYJ(]"OjJ>eea KaTEXEll' & oeZ, 7Tapf.xwp,v TE ~p,as aVT01JS, K.T.A. XEN. Cyr. viii. l, 5. M~ 1T'OT cpwp,V EVEKa rovTwv p,ryof.v J1-a'AA.6v 1T'OTE fvx~v J.1T'6A.A.w&at. PLA1'. Rep. 610 B. ''Ea 01) vvv f.v <Tot <TKefwluea. Id. Soph. 239 B.

256. The first person plural is most common, and generally expresses an exhortation of the speaker to others to join him in doing or in not doing some act. E.g. "IwfJ-ev, let us go; f1-1J EwJLev, let us not go. 0 l'Kao 1T'ep <n'w Vt]V(]"i

257. TlJe less common first person singular is, in affirmative exhortations, g~merally preceded by a word like &ye, etc. (251 ), or by some other command, and the speaker appeals to himself to do something or to others' for permission to do it. In negative appeals with v!J the first person singula,r is rare and poetic; the speaker may call on others to avert some evil from himself, or he may utter a threat or a warning. E.g.
'AA.'K &ye 01J Ta XP1JfLai' apl,()fh~<TW Kat Eowfhat, come, let me count the things and see. Od. xiii. 215. 'AA.A' aye&' VfJ-CV rn!xl JvdKW &wpt'fxe~vat. Od. xxii. 139. 8a7T'TE /Le OTT! raxo<TTa, 7T1JAas 'A[oao 1T'<p~<Tw, buTy me as quickly as possible; let rne pass the grttes of Hades. Il. xxiii. 71. 'AAA.' aye JIVJI E7TfhHJIOJI, J.p1]ta TEVXEU 8v w. Il. vi. 340. <Ppe aKov<Tw, corne, let me hea1. HDT. i. 11. l:Zya, 1T'voas fhaew cpp< 1T'p6s ovs (3a'Aw. Euu. H. F. 1059. 'Ih<Txer', av81)v rwJI e<Tweev iKp,aew. Id. Hipp. 567. Af.y< 81), Zow. PLAT. Rep.


M~ <Te, yepov, KoA'(I<TtV Jyw 1T'apa v17 v<Tt K'x dw, let me not find you at the shi]'IS! Il. i. 26. M~ <Tev d.Kov<Tw dxop,f:vov. Il. xxi. 475. 'AA.A.a "'' E/( ye T~O'OE ylj> ?Top&fJ-EVO'OV ws raxtO'Ta, p,~o avTOV 8avw.

SOPH. Tr. 801.

'"D t<Zvot, fJ>1J 80T' a0lKt]8W. Id. 0. C. 174.


258. In the first person (255-257) botl1 present and aorist sub-

junctive are used with

the distinction of 259 applying only to the




second and third persons. In affirmative exhortations the second and third persons of the ~ubjuJJctive are not regularly used, the imperative being tl1e only recognised form. But in SoPH. Ph. 300, <j;f:p, iJj TEK!'OV, vvv Kal T~ r-qs v1jO'ov 11-a8vs (if the text is sound), the positive 11-U.()y]> Heems strangely to follow the analogy of the negative fL~ JLU.8vs. Nauck reads JLaf)E here. Stee also T~ fa<j;tO'fW. dvaTE81j. in an inscription quoted in .Appendix I. p. 385.

259. In prohibitions, in the second and third persons, the present imperative or the aorist suLjunctive is used with 11-n and its compounds. The distinction of tense here is solely the ordinary distinction lJetween the present and aorist (8 7), and has 110 reference to the moods. E.g.
M1] 1roEt rovro, do not do this (habitually), or do not go on doing this (or stop doing this); fL'l 1rat1)uvs rovro, (simply) do 1wt do this. 'E~av8a., fL'l KV8 vo<p, Zva d8oJLEV &JL<f>w. I!. i. 363. 'ArpE8ry, fL'l fd8/ E7rt(J"TafL!10S ua<f;a Z1rEZv. Il. iv. 404. 'ApyEZot, JL0 7r(O Tt 1u8fETE 8ovpt8os dAK'J'> ll. iv. 234. Elmi fLOI dpoJLEVI.f! VYJfLEprf:a, p.')8' i7rtKVUTJ> Od. XV. 263. ''Ho') vvv u<f 7rat0t ~7r0S <f;ao, fL')0' i7r{KEvBE (compare the last example). Od. xvi. 168. T<f vvv 11-~ 11-ot 2aAA.ov v li.>..ycut Bv/1-0V optvn> Il. xxiv. 568. See n. xxiv. 778. Q\1~ 8>] fLE i!>..wp Llavaoi:utv EaO'))'> KEZ(r8at. Il. v. 684. KAvBt !1-1)8 !1-EY>lPrJ'> Ocl. iii. 55. l\1,) 7r'WS av8paut 8v0'!1-EVE0'(J"tV EAwp Kat Kl!p/1-a yev>JU(}E, do not become prey and spoil to hostile men. I!. v. 487. M,i 7rOT d7r~ 7raO'aV oAEO'O'TJS dyAatryv. Orl. xix. 81. 'YttE'S 8 rfj yij rij8E 11-~ (3apvv Korov ulo)f1Ju8E, 11-'l 8vfLovu8E, 11-'78' dKap1rav TEV~1JTE. .AESCH. Eum. 800. ''Ov 11-'Jr' OKVELTE, fL~T' d<j;,)r' e1ros KaKov. SoPH. 0. 0. 731. l\11) B~u8E v6fLov 01J8f:va, dAAct Tov~ (3Aa7rrovras VfLaS Avuan. DEM. iii. l 0. (Here 80'8E would not be allowed; but AvO'arE, an affirmative command, is regular.) M~ Karct rovs VOfLOVS OtKaU1)T' fL~ (3o1)81)U1JT ri) 7r7rOV8on 8Ewa 11-'J EDopKELT. Id. xxi. 211. M1) 1rp17> 1raZ, 8q8a .AR. Nub. 614. Kal JL1)8ds v1roAaf3TJ fLE (3ovAEu8at J\.a(}EZv. Isoc. v. 93. Kai JLry8els olEu8w Jl dyvoeZv. Id. iv. 7 3.

260. The third person of the aorist imperative is sometimes used with 1-'-~ in prohibitions ; but the second person with fL~ is very rare and only poetic. E.g. :M,78' ~ (3a <rE 01JOafLW> vtK1J<rarw. SOPH. Aj. 1334. M1Jof: uot JLEA1Juarw. AESCH. Prom. 332; so 1002. Kai JL'18ds VfLWV 7rpou8oK1JUaTw aAAws. PLAT .Ap. 17 c. TiJ fL~ fLOt 1rarpas r.o8' OfJ-O[TJ i!v()EO TtfLV I!. iv. 410; see Od. xxiv. 248. M~ 1rw Kara8vueo 11-wAov "Ap1)oS. Il. xviii. 134. M~ tfEV<TOv, .(Jj ZEv, fL''l JL' EATJS UVEV oopos in SoPH. Peleus, Frag. 450, is parodied in .AR. Thes. 870, fL'l fEvuov, iJj ZEv, rqs J1rwvu'ls f.A.1r8os.





261. In the following Homeric examples the independent subjunctive with p.~ expresses apprehension, coupled with a desire to avert the object of fear, both ideas being inherent in the construction. The third person is the most common here.
M~ o~ vrjas l.Aw(J't Kal o'llK~n <f>evKra 7rtAwvrat, nwy they not (as I fear they may) seize the ships and make it no longer possible to escape. Il. xvi. 128. M~ 0~ p.ot TEAE(J'W(J't ew~ KaKd. K08m Ovfl-0, may the Gods not bring to pass (as 1 fear they may) bitter woes for my soul. Il. xviii. 8. M~ Tt,uEvos ,H~v KaKbV vias 'Axatwv, may he not (as 1 fear he may) in his wrath do anything to harm the sons of the Achaeans. Il ii. 195. "!), ,uot ~yw, p.~ rts fi-Ot i!cpatvv(]'tv MAov aDn &OavaT<tlV, Od. v. 356. M~ 1rl!!s fl'' JK(3alvovra f3ci>..v A[8aKt 1rpor~ 1rerpu Kvp.a p.y dp7rci~av, fJ-EAE>) 8 p.ot l(J'(J'ETat 6pp.0, I fear that SO'fiW great wave may dash 'I1W against a solid roclc, and my effort will (then) be in vain (the expression of fear being merged in an assertion). Od. v. 415. See also Il. xxi. 563 ; Od. v. 467, xvii. 24, xxii. 213. Twv d KEV 1ravrwv dvr~(]'op.Ev, p.~ 7roA{J1rtKpa Kal alvO. (3as &7roT{(]'Eat iAOwv, i.e. 1 fear you may punish their violence only to our bitter grief (and may you not do this). Od. xvi. 255. M0 n KaKbv pegw(J't Kat ~p.eas JgEAci(J'w(J'tv, liA.\wv 8' d<f>tKwp.E8a ya!av, may they not (as I fear) do us so'fiW harm and drive us out, and may we not co'IJW to SO'fiW land of others. Od. xvi. 381. M~ fl-LV E{'W ,u~v lwv; 6 <le p' ovK J.\~(]'L, I fear I may approach him as I come, while he will not pity 'llW, Il. xxii. 122 (see Od. v. 415, above). M~ TOL KarO. 7rUVTa: <f>ciywuv KT~,UCJ.Ta OQ(]'(]'ci,uEvot, (J'D ()~ T'Y)V(]'[T)V 68ov uevs. Od. XV. 12. ' The present subjunctive occurs in Od. xv. 19, fl'~ n </>Ep>)Tat, and in xvi. 8 7, ,u~ p.w KEpTOfJ-EW(]'tV. See also 1rAwvrat in Il. xvi. 128, above. (See 258.) In these exalllples sometimes the fear itself, and sometimes the desire to avert its object, is more prominent.

262. (a) By prefixing oELow or cpo(3ovf1-at to any of the subjunctives with ,u0 in 2 61, we get the full construction with verbs of fearing ; as <ldow p.~ v~as .\W(J't, I fear they may seize the ships, in which p.~ EAW(J't represents an original construction which at first followed odow paratactically-! fear: may they not seize the ships-and afterwards became welded with it as a dependent clause. So if o<;{ow were removed from a sentence like 8E{8w p.0 n 1ra&vo-w, Il. xi. 4 70, we should have an independent clause like those quoted above. See ,u~ &.p.a(]'(]'Tl and oEiow p.~, Od. v. 467 and 473. (b) In like manner, by prefixing other verbs than those of fearing to such clauses, the original negative final clause with fL~ is developed; as p.axovp.e8a ,u~ v0as EAW(J't, we will .fight that they




may not seize the ships. Again, if the leading clause were removed from a sentence like aVTOV fL[/Lv' E?rt ?rvpytp, fL0 ?rat8' opif>avtKOV 8~17> X~PTJV TE yvvatKa, remain here on the tower, lest you rnake your child an orphan and your wife a widow, Il. vi. 431, there would remain /L~ . 8~y>, do not malce, or may ym~ not rnalce, in the originally independent form, like the clauses with /L~ in 261. (See 307.)

263. (M~ oi> with the Subjunctive.) The clause with /L~ expressing desire to avert an object of fear, in its original simple form as well as in the developed final coustruction, may refer to a negative object, and express fear that something may not happen. Here fL0 ov is used with the subjunctive, like ne non in Latin. Thus /L~ vqas EAW(TL being may they not seize the ships, /L~ ov vqas
EAWin would be rnay they not jail to seize the ships, implying fear that they may not seize them. Holller has one case of fL?J ov after a verb of fearing : OELOW fL?J oil TLS TOt {nrOCTXTJTat TOOE epyov, Il. X. 39. He has several cases of fLlJ ov in final clauses and one in an object clause (354). 11. i. 2~;--~ VlJ TOt ov xpaLCTfLU (TKq?rTpov Kai CTTEfLfLa 8wZo, is often cited as a case of inuependent ftry ov, meaning bewan le.~t the staff and fillet if the God shall pove of no avail to you. So Del briick (I. p. 119), who nevertheless quotes Il. i. 565, d.\A' aKeovO"a K6,8,JCTO ifL0 8' E?rt7fel8w fLV8tp, fL-1 vv TOt ov xpa[CTfLWO"tV O(}"Ot 8eo lcr' EV '0.\vfL?r<p, as containing a dependent final clause. In the two other cases of p.~ ov with the subjunctive in Homer, Il. xv. 164 (an object clause, see 354), and xxiv. 569 (final), tl1e dependence of the clause with /L~ ov is even more obvious ; and in Il. xxiv. 584 we have in fL~ ovK p!JcratTo the decisive proof that this clause is felt to be dependent in the change froi:n the subjunctive to the optative after a past tense. It is therefore more than doubtful whether fL~ ov xpaiCTfLU in Il. i. 28 is not dependent on fL?l cr Ktxdw in vs. 26. Plato in paraphrasing this passage (Rep. 393 E) takes the clause as final and dependent (see 132). But, whether we have a case of independent fL?J ov with the subjunctive in Homer or not, there can he no doubt tl1at this i~ the original form from which came the dependent final clause with /L~ ov.

264. After Homer we have the independent clause with /L~ in Aeschylus, Ag. liLt and 341 ; in Euripides we have independent p.~ in Ale. 315 (fL~ crovs 8ta<f>8dpYJ "fUfLOVS), Orest. 776 0~ .\0.(3wcrt ,r aCTfLEVot), H. F. 1399 (aifLa fL?J CTOts igo, 7fE7fAOtS), and fL~ ov in Tro. 982 (fL?J ov 1rd0"17'> croif>ovs), besides Rhes. 115 (01) ov fLOAvs). Aristophanes, Eccl. 795, has a doubtful p.~ ov Aa(3y> (Heindorf and Meineke, for Mss. .\a(Jots). Besides th(se few cases, we have in Plato three of !LfJ with the subjunctive implying apprehension in the Homeric sense (261): Euthyd. 272 C {fL1) ovv ns ovdlcrv), Symp. 193 B (p.~ fLOt V7fOAaf3v), Leg. 861 E (fL'1 Tt<; ol?JTat). Euripides and Herodotus are the fi1st after Homer to use /L~ ov in dependent clauses of fear (306).








too rude a thing to tell. Gorg. 462 E. M0 ti.>s cA.,/Bws Tavm r.rKfpf<aTa -rwv p'fo[w; dr.oKnvvVvTwv, I s;;,spect these nwy JWove to be considerations for those, etc. Crit. 48 c. M 'l <Pa.vAOJI !i Kal ov Ka8' o06v, I think it will be bad and not in the 1ight W(ty (i.e. f<1J ov {i). Crat. 425 B. 'AA.Aa. 1-"'l ovx o1hws EX[), J.U.' fj d86Ta -rBEr.r8at (i.e. ~J-0 iJ). Crat. 4:l6 B. 'AA.A.a f.OJ ov Tovr' i) xaA.Er.~v, BJ.JI<LTOJI JK<PvyEI:v, but I suspect this may not be the hnrd thing, to escape death. Ap. 39 A. 'Hf<tV 1-"1 ov8v aA.A.o (TK1rTEOIJ -a, I arn inclined to tln:nlc we have nothing else to eonsider. Cl'it. 48 C. l\;f ?J ov OEYJ inroA.oyCwBat, I think there will be no need of taking into acco1mt, etc. Crit. 48 D. l\1?) ovK i'i btOaKTOV ap<T~, it will p?obably turn out that virtue is not a thing to be taught. Men. 94 E. 'AA.A.d. fi-?J oi>x oVTot ~f<Ets iJJf<Ev, but I think we shall not prove to be of this lcind. Symp. 194 C. 1 See also Aristotle, Eth. x. 2, 4, 00 m)8~v A.eywr.rtv (v. l. A.yovr.rtv), there' can hardly be anything in what they say. (See 269.) In DEM. i. 26 we have 1-"~ "-av r.tKpOv Elr.Ei:v iJ, I am afraid it mny be too lwrsh a thing to say. The present subjunctive here, as in dependent clauses of fear (92), may refer to what may pTove true.

265. In Herodotus v. 7 9 we have dA.A.a f<O.A.A.ov 1-"~ oil -rovTo y f<av~wv, but I suspect rather that this may pmvt not to be the meaning of the OTacle. This is the first example of a construction, very common in Plato, used also by Aristotle, and found once in Demosthenes, in which 1-"~ with the subjunctive expresses a suspicion that something may be (or may prm;e to be) true, and p.~ ov with the subjunctive a suspicion that something rnay not be true ; the former amounting to a cautious assertion, the latter to a cautious negation. Examples from Plato are : Ml) dypotK6upov !f -r6 dA.qBs <ir.ELI', I am afmid the t?uth may be

266. In these cautious assertions and negations, although no desire of the speaker to avert an object of fear is implied, there is always a tacit allusion to such a desire on the part of some person who is addressed or referred to, or else an ironical pretence of such a desire of the speaker himself.

267. The subjunctive with 1-"!J in this sense is sometimes found in dependent clauses. E.g.
"Opa. f<?J ilA.A.o Tt -.6 yEvvaZov Ka.l -ro dyaBov !f -rov r.r~(nv Ka~ r.r.f(<a-Bat, see to it lest (it prove true that) these may be different things, etc. PLAT. Gorg. 512 D. The common translation, see whether they may

1 Other examples in Plato are Pl1aed. 67 B, 69 A ; Theaet. 188 D ; Crat. 429 C, 432 A, 432 B, 435 C, 438 G, 440 C ; Men. 89 C, 94 B ; Lys. 209 A, 219 D, 220 A ; Symp. 214 C; Parm. 130 D, 132 B, 134 E, 136 D ; Leg. 635 E; 'l'he<tg. 122 B; Amat. 137 B. See Weber (pp. 191, 192), who gives these example" in Plato, with HDT. v. 79 and DBM. i. 26, as the only cases of indepeudent I"TJ or !"1J ou in this peculiar sense before Aristotle.



f"n AND f"~



not be different, gives the general sense, but not the construction, which is simply that of p~ &A.A.o n V(265) transferred to a dependent clause, . 268. In a few cases Plato has p~ with the subjunctive in a cautious question with a negative answer implied. As pry &A.A.o n fJ TovTo means this rnay possibly be SO?JU!thing else, so the question p.ry aA.Ao T f; Tovro / means can this possibly be something else ? The four examples given by Weber are:M~ n O.A.A.o 'll'apa ravra; can there be any other besides these? Rep. 603 c. '!' Apa pry aA.A.o Tt iJ 8avaros ~ TOVTO / is it possible that death can ]Jrove to be amything b'Ut this ? Phaed. 64 c. So p~ Tl aA.A.o f; ~. K.T.A..; Parm. 16:3 D. 'A.Ua pry ip1) 11'eptepya f; KaL TO ipwr~a-al rre 7rEpt roDrov ; but can it be that even asking yo1; about this is inquisitiveness on my purt? Sisyph. 387 C (this can be understood positively, it muy be thut it is, etc.). In XEN. Mem. iv. 2, 12, the same interrogative construction occurs with /LTJ ov: p?J O'OV ov (!1!vwpat iytiJ TU T~S OtKawrrvV?JS iipya igY)y~rrarr8a t ; do you suspect that I shall be unable to expluin the wo1lcs of Justice? , In PLAT. Phil. 12 D we have 'li'WS yap ~oov~ ye ~ooviJ p.ry ovx 6p.ot6TaTov l!.v dYJ; for how could 01U! pleasure help being most like another? Here et1j av takes the place of fj, and 'li'WS shows that the original force of p.~ is forgotten.




/L~ or ov may express a similar cautious assertion or suspicion about a present or past act. As <foof3ov/Lat P-?7 1rarrxet (or E'll'a(hv) means I fear that he is suffering (or suffered), so P-0 'll'aO'xn or P-0 E'll'a8<v may me~tn I suspect he is SU;fjering or I suspect he suffered, and P-0 ov 'll'(lfTXEL or p.ry oiJK E'll'a8<1' may mean I suspect he does not (or did not) suffer. (Of. :265.) E.g. J\1 ?J yap 'TOVTO /LE v, T~ (~v 61I'O(TOVOYJ xp6vov, r6v ye WS dAYJ8ws avopa Jarov JfTTt KO.L Ol> </>tAofvXYJTEOV (i.e. Kat P.'l 0~ </>tA.), joT I am of the opinion tlwt this, mRTely living jo1 a ce1tain time, is what one who is truly a man should disngard, and that he should not be fond of life. PLAT. Gorg. 512 D. (This }Jassage is often strangely emenJ.ed and explained.) 'AA.X apa P.'l ov TOWVTYJV vr.oA.ap.f3dvets O'OV T0v /La81Jrrtv iirreO'Bat, I suspect that you do not think your leurning v:ill be like this. Id. Prot. 312 A. 'AA.A.a tL?J TovTo ov w/LoAo'Y~rra/LEV, but perhaps we did not do well in assenting to this. Id. Men. 89 C. (This may be interrogative (268) : can it be that we did not do well, etc. ?) So Al'istotle, Eth. X. l, 3, tL!J 'li'OT o ov KaAWS AyeTat, but it rnay be that this is not well said : compare x. 2, 4, quoted in 265. 270. Apart from independent sentences with P-0 oiJ (263-26 9), this double negative occurs chiefly in ordinary clauses after verbs of fearing where the object of fear is negative (305 ; 365).

269. The present or past tenses of the indicative with



' 0?rro~ AND 5?rro~ /-'~ WITH FUTURE INDICATIVE, ETC. (271

"0?rro~ AND g?TCil<; /-'~ WITH THE INDEPENDENT

~1rws or with the future indicative to express either a positive exhortation or command or a prohibition. Thus o1rws rovro pf.ts, see that you say this, is a familiar way of saying el1r~ rovro. So 81rws P.0 roiYro EpEts is equivalent to P.0 rovro Er7TI/'> This expression was probably suggested and certainly encouraged by the common Attic construction of o1rws and the future after verbs of striving, taking care, etc. (339); so that it is common to explain this form by an ellipsis of CTK67TEt in CTK61rn o1rws Tovro Epli:s, see to it that you say this. But we may doubt whether any definite leading verb was ever in mind when these familiar exhortations were used (see 27 3).

271. The Athenians developed a colloquial use of

~ws p.~

272. The earliest example is AESCH. Prom. 68, o1rws p.~ CTavrov olKnf.'i:<; 7rOTf., beware lest at some time you may have yourself to pity, which conveys a warning, like p.~ CTE Ktxelw, Il. i. 26. In AESCH. Ag. 600, we have the first person singular with o1rw<; (used like the subjunctive in 257): 07TW<; 0' f1ptCTTa r6v ~p.{)v alooi:ov 7r6CT!V CT7Tf.-6CTW OE~aCTf1at (not mentioned by Weber). In Sophocles there is only one ea~e, 0. T. 1518, yq> p! o1rws;f.tS &.7rotKov, send me forth an exile from the land (like;ov p.E). Five examples in Euripides are simple exhortations, as &,A.,\' o1rws &v~p ~CTH, but see that you are a man, Cycl. 595; so also Cycl. 630, H. F. 504, I. T. 321, Or. 1060 (with doubtful construction): one conveys a w.arning, Bacch. 367, llE:v(kils 8' o1rws p,~ 1rf.v8os dCTo {CTH 86p.ots rot'> CTotCTt, beware lest Pentheus bring sorrow (1rf.veos) into your house.

273. We find the greater part of the examples of 271 in the colloquial language of Aristophanes,l who often uses the imperati,,e an<l o1rw<; with the future as equivalent constructions in the same sentence. E.g. KanfBov CTD -rii CTKf.VYJ, x07rwS ipe'i:s JvTavBa p.ryo~v tfevoos, put down the paclcs qniclcly, and tell no lies here. Ran. 627.
'AA.A' p.(3a x&J1rws ape 'is n)v 'iwmpav. lb. 377. See also Eq. 453, 495, Eccl. 952, Ach. 955. ~i!v oDv o1rws CTWCTH> p.~, so now save me. Nub. 1177. ''01rws 1rapeCTo p,ot Ka2 CTV Ka1 ra 1Tat8a, be on hand, you and your children (an invitation). Av. 131. "Aye vvv o1TW> v<j>ap7rrfCTE<. Nub. 489. 274. (Examples from Prose.) o7rWS oiiv ~(Tf./j ()e avope; a~W! r{js EAev8epas, pmve younelves men worthy of freedom. XEN. Ait. i. 7, 3.

See Weber, pp. 85, 95, 113,124, for the history ofthis usage. Weber cites 41

examples from Aristophaues, besides Ach. 343; 18 from Plato, whose extra-

ordinary use of the independent sentence with J.L-IJ has been noticed ; 7 from Xenophon, 9 from Demosthenes, 2 from Lysias, aud one from lsaens.

280] ' 07rroc;- AND 57rroc;- #~ WITH FUTURE INDICATIVE, ETC.

A , ,. .,.


o1TWS }LOt, .. aY Bpw1T, (J!') EpH S OTt (TTt Ta OWOEKa OLS E> see that you, , , ,

do not tell rM that twice six a1e twelve. PLAT. Rep. 337 B: so 336 D. if!f.pE o~ o1rws JLEJLVYJiTOJLEBa TaVTa. Id. Gorg. 495 D. q0'11"ws '}', /J.v n TOVTWV y[yvryTaL, TOVTovs ~1TatVEiTE!T8E Kat Tt}L~iTET Kat iTTE</>avd:JiTETE, ~JLE o JL~ Kat JLEVTOt Kctv Tt TWV ~vavTwv, o7rws TOVTots 6pyHtiT()E, DEM. xix. 45. ''07rws To[vvv 7r~pi Tov 1TOAEJLOV JLYJOEv ~p<'is, see therefore that you say nothing about the war. lb. 94. One case occurs in Herodotus in iii. 142. (See also 280, below.)

275. Althongh the st>cond person is naturally most common in these expressions, the first and third persons also occur. E.g. ''OJrws o To ITVJLf3oA.oJ' A.af3ovTEs E7THTa 1T A1]iToF Ka() Eo o 1! JL E()a. AR. Eccl. 297. Orp.ot T<tAas, 6 ZEVS chrws JL~ / tnf;Ta ., don't let Zeus see me! Tel. Av. 1494. Ka2 07rws, i/JiT7rp ipwTwiTt 7rpo8vJLws, oVTw Kat 1Totdv JBcA.lj~ToviTLV. Dmr. viii. 38. (See also 278.) 276. ''AyE and </>ep< (251) sometimes introduce this construction. See examples above (2 73 and 27 4). 277. In a few cases the prol1ibition with 07rWS p.~ takes the form of a warning. Besides AESCH. Prom. 68 alld EuR. Bacch. 367, quoted above, see XEN. Cyr. i. :3, lli, OJT"WS ovv fL~ d7roAcl'')'OlJfLEvos, look out that you are not flogged to death. So PLAT. Prot. 313 C, quoted in 283. 278. ''07rws p.lj with the future indicative or the subjunctive sometimes occurs in independent sentences implying a desire to avert something that is not desired, like fLfJ with the subjunctive in Homer and sometimes in Attic Greek (261; 264). E.g. ''01TWS fL?J al!Txpol p.f.v </>aFovp.EBa a!TeEIIEtS of. EiTOp.EBa, let us not appem base and be weak (as I fea? we may). XEN. Cyr. iv. 2, 39. "01rws tJ-?J dvayK<i~Twp.<F (so most Mss.) a,}Tovs, Kllv 11-0 f3ovAwnat,. dya.8o1s '}'El'E!T&at, the1e is danger of our compelling them to becorM bmve, even against their will. lb. iv. 1, 16. Ka2 01TWS 'Y' f143 TO xwpov ?JGEWS opW(TlJI vea l<aTEI<aVOV ~fLWV TOVS iTVP,fLaxovs, and let us not
allow them even to enjoy the sight of the place where they slew our allies. lb. v. 4, 21. "Or.o;s /;,?] <PfJr.rv ns ?JfLUS ~8vJra8eZv, take ca1e lest any one say of us, etc. Id. Symp. iv. 8. 'AA.A.' 07rWS fJ-?J o1\x oios T, E!, '11"poevp.ovp.VOS of. yf.AwTa 6 <P A fJ (TW, but I am afraid that I shall not have the power, but tliat in my zeal I shall make myself ridiculous. FLAT. Rep. 506 D. So Men. 77 A.

279. These cases (278) are analogous to those of 07rwS p.lj with the future indicative or the subjunctive after verbs of fearing, in place of the simple p.lj (:370). They are also a connecting link between the subjunctive with p.~ in prohibitio11s and the rare future indicative with p.~ in the same sense ; as TaVTYJV </>vAa~<T T0v 1r1Tnl', Kat tJ-?J f3 0 V A~ (T E (TeE dof.vaG, hold fast to this security, and do 1Wt wish to k?ww, etc., DEM. xxiii. 117 (see other examples in 70). 280. In a few cases o1rw; p.~ with the subjunctive expresses a cautious assertion, where the simple p.lj is generally used (265). E.g.







Kai 01rws p,~ iv p,~v TOtS (wypa<f>~p,a<rw TovTo, and it may be that this will prove true in the case of pictures. PLAT. Crat. 430 D. Weber (p. 264) quotes HDT. vi. 85 for this sense : OKws e~ v<rT~PTJS p,~ n r. "' )\ ,.. I' 1'\ \ ' \ I ' ~I'\ vp,w, TJV TavTa 7rOT]<TTJTE, 7raVWII.E () pov KaKov ES TTJV XWPTJV Ep,,..,a~tw<rt, it is 1wt 'llnlikely that they will turn about and bring some fatal harm on your country; but this can be understood like the examples in 278.

281. 'fls &v <rKo7ro2 vvv fJ T E Twv dpTJp,~vwv, mind now and guard what I have said (i.e. be watchful to do it), SoPH. Ant. 215, must be brought under this head (271). In the early stage of the Attic construction of 01rws with the future, of which only two cases occur in Aeschylus and one in Sophocles (272), ws &v flu was here used like o1rws E<r<<r8E. Compare E7rtp,EAEf:u8at ws &v 1rpax81], XEN. Hipp. ix.. 2 (see 351). 282. In AR. Ach. 343 is the single case of 01rws p,~ witl1 a present tense, expressing a suspicion and apprehension concerning a present 1 \ \! " I ' I{.J ' I() ground 0 f f'ear; a11.11. 01T"WS fTJ I V TOS TP/"<O<TV EyKa 7JVTaI 7r0V A.leot, but I am afraid they 1ww have stones hidden somewhere in their cloaks. This bears the same relation to the common owws p,~ with the future (272) that <f>o{1ovp,at f~ wduxov<rtv, I fear that they are sujfe1ing (369, 1), bears to <f>of3ovp,at p,~ wduxwutv, I fear that they rnay suffer (365) ; a11Cl the same that p,~ TovTo eaTeol' e<rTt (269) bears to f~ <TKE7rTEOV V(265). 283. Positive independent sentences with fhrws all have the future indicative, the regular form in dependent object clauses of this nature (339). Among the 33 independent clauses with 01rws p,~ which are cited (excluding An. Ach. 343) ten have the subjunctive, and four others have more or less Ms. support for the subjunctive. Of the ten, the two quoted in 280, aHd the three from Xenophon quoted in 278, are either in cautious assertions or in sentences implying fear or the averting of danger, where the subjunctive is the regular form. The other five express warning, and are quoted here : q07rwS o~ TovTo p,~ otod~vs p,YJoiva, but be sun that you teach this to nobody. AR. Nub. 824. Kai 01T"W<; YE p,i) 0 <ro<f>t<rTry<; e~a7raT'IJ(TYI ~pas, and do not let the sophist cheat us. PLAT. Prot. 313 C. 'AA.A.' 01rws p,~ n 1)p,a> u<f>~"Av To d<l TovTo. Id. Euthyd. 296 A. "07rws p,7Jo<s iH 1r [(]'y), do 1wt let anybody persuade you, etc. Id. Charm. 157 B. Kal 07rW'> f1J 77'0/.~<TYJT 8 7rOAAdKS vp,as f3A.w.fEv, and see that you do 1wt do what has often ha1rned you. DEM. iv. 20. Four of these subjunctives are of the <r- class, easily confounded with the future indicative, and the judgment of scholars on these has depended to a great extent upon their opinion about the admissibility of the subjunctive with 07rWS and 07rW'). f~ in dependent object clauses (339). This question will be discussed in 364. But it may fairly be claimed, independently of the main question, that these cases of owws p,~ with the subjunctive in prohibitions are supported by the analogy of p,fJ with the subjunctive in the same sense. Thus JL~ otod~'[}>, do 1wt teach, makes o1rws p,ry fMd~vs in the same sense much more




natural than the positive 07rWS 8dM.glls would be, for which there is no such analogy and little or no Ms. authority. On this ground the examples are given above as they stand in the Mss.


Subjunctive, like the Future Indicative, in Independent Sentences.-Interrogative Subjunctive.


284. In the Homeric language the subjunctive (generally the aorist) may be used in independent sentences, with the force of a future indicative. The negative is E.g.


Oti yap 7T"W To[ovs i:8ov


otio U>wp.a t, for I never yet saw nor

shall I ever see such nUJn. Il. i. 262. 'Yp.Zv iv 7raVTuut 7rpu<.Avrd. owp 6 v o p. ~ v w, I will enumemte the gifts before you all. Il. ix. 121. is 'A[8ao Kat iv VK1)uut <j>advw, I will descend to Hades and shine among tlUJ dead (said by the Sun). Od. xii. 383. (Here the future and the subjunctive rpadvw do not differ in force.) 1\ o-&8 'Am5A.A.wvos EKarow, I will renwmbm and will not forget the far-shooting Apollo. Hymn. Ap. 1. A-&rov oi 86.varov p.r;, ov8~ V11 TOY yvwro[ Tf yvwra Tf 7r11pos AAaxwut 8avovra, i.e. they shall not give his dead body the honour of a funeral pyre. Il. XV. 349. El o K T8vr;wros aKOVtJW, ufwa rf: oi X fVW Kat E7rt Krpm KTpdgw, I will 1aise a mound f01 him, and pay him funeral honours. Od. ii. 222. Ov yap Ts fL (317 Y JK<iJv dKovra Otr;Tat. Il. vii. 197. Ka1 1ror n<; t7r1JUtv, and sorne 01w will say. Il. vi. 459. (ln vs. 462, referring to the smne thing, we have Cf>s 1ror TtS EPEft.) OvK Err8' OVTO<; av~p ovo' f(]"(]"ETat ovo Y~JJr;Tat, os Kv Tr; u~ v1t Xtpas E7ro[ut. Od. xvi. 437. 0-68 dvur~ns 7rptv Kat KaKoV aAAo 7rU81wea, nor will you bring him back to life; sooner will you suffer sonUJ new evil besides. Il. xxiv. 551 (the only example of the second person).


285. This Homeric subjunctive, like the future indicative, is sometimes joined with KE or av in a potential sense. This enabled the earlier language to express an apodosis with a sense between that of the optative with O.v and that of the simple future indicative, which the Attic was unable to do. (See 201 and 452.) E.g. El 0~ KE !'-~ owvuLV, ~y0 o KfV avr6s, but if he does not
give her up, I will talce her myself. Il. i. 324. (Here EAWfta[ K<JJ has a shade of meaning between V ..otp.r;v Kv, I would take, and alpfJuop.a,, I




will take, which neither Attic Greek nor English can express. See 235, end.) Ti]v 1'-tv 7fEJ'-tfW, Jyw oe K' O.yw Bpur7J8a, her I 'will send; but I shall talce Briseis. Il. i. 1tl4. Nvv 8' &v 1roAA.a 1rd.8y(]"t <j>.\ov d:rr/i 1ra-rpos Uf'-ap-rwv, but now he must suffer much, etc. Il. xxii. 505. "Hts 1nr<po7rA.y(]"t rcfx' O.v 1ron 8vl'-6v 6..\e(]"crv, by his own insolence he may perchance lose his life. Il. i. 205. 286. In tl1e following cases the subjunctive and the optative with KE or tl.v are contrasted : "AA.A.ov K' Jx8a[p1)(]"t {3porwv, ll.A.A.ov K< <j>d. o-ry, one 'llwrtal he (a king) will hate, and another he may love. Od. iv. 692. Ef -rs (]"E f8otro, avrK' av J~d7fot 'Ayaf'-~fWOJJL, Ka K<V d)!Jf3A.YJcrt> AJJcrws V<Kpo'i:o yev'I)Tat, if any one should see you, he would stmightway tell Agamemnon, and there might (may). be a postponem,ent, etc. IL xxiv. 653. El f"EJJ 01J dvr{3to)J (]"i>JJ nvx<crt 1r<tp7]8ELYJS, Olll< tl.v TOL xpa(]"/"TI(J'L p,;:,, Kat rap<j>EES lo. IL xi. 386. ComJ;are ~JJ x' Bf"LJJ (]"(J.<j>a E t7fW OTE 7rpor<pos )'E 7rv()of"'I)V, (a message) which I will (would) tell you plainly so soon as I shall (should) hea1 it, Od. ii. 43, with )v x' )f"Zv cra<j>a d'7ro' OTE wponpos )' 7rV80tTO, ii. 31,-both referring to the same thing.

287. The first person of the subjunctive is used in questions of appeal, where the speaker asks himself or another what he is to do. The negative is fl-1 In Attic Greek this subjunetive is often introdueed by fJovA.H or fJovA.ue, sometimes in poetry by 8eA.t<; or 8hT. E.g.
E i1r w rovro ; Bhall I say th1:s ? or {3otJAEt d1rw rovro ; do you wish that I should say this? M ] rovro 1ro twp-<v (or 7rot~crWfMJJ); shall we not do this? T <<7fw; or r f3ovAE(]"() Et:1rw; what shall I say? or what do you want me to sa.y ? II i7 r' tl.p' Jyw, <j>AE 7'EKJJOJJ, ;: w ; TEV OWJLa8' [Kw fW t; ~ l8vs (]"~<; f"YJTpOs t:w Kat (TOto ODf"OW; whithe1 shall I go? to whose house shall I come ? etc. Oc1. XV. 509. "'H a-Jros K V() w; <f>acrea, o f" 8t>f"OS dvwyH. Od. xxi. 1 ~,4. "'52, Z~il, -r__ A.~~w; ;:or ~PEJ'W~ ~ () w, 1r& TEP; SOPH. 0. C. 310. .!21'-ot <yw, 1ra {3w; 1ra (]"'f'Wj wa KEAcrw; Eun. Hec. 1056. IIoZ -rpa7rWf"at; 7f0t wopevew; lb. 1099. Et7rW Tt rwv Elw86rwv, iS OE(]"7rOTa; shall I make one of the ngular jokes ? An. Rau. l. TJ,a yap f"aprvpa l"d(w 1rap&(]" X Wf"a t; i.e. how shall I b1ing jo1waTd a gnater witness? DE~f. xix. 240. l\11]8', Jv Tt wvwf"at, l!pwf"a' 61r6crov 7fWAEi:; may I not aslc, etc.? l\1,78' dwoKpvwJLai ovv, av r[s f" epwrlj. veos, av dow ; and '!nay I not answer, etc. ? XEN. Mem. i. 2, 36. Mi] dwoKpvwf"at, dA.A.' ETEpov ct:ww; Pr.AT. Rep. 337 B. So l"'l <f>wf"EV; lb. 554 B. Mtcrew(J'WfLE8a OVJJ K~pvKa, ~ avros dvd1rw; lb. 580 B. ME()vovra tl.vopa 1ravv (]"<j>68pa 8~cr8E (]"'Uf"7fOTYJV, ~ a7rwp.EV; will you receive him, or shall we go away? Id. Symp. 212 E. ''Apa 1'-'J alcrxvvewfL<V r6v IIEpcrwv f3a(]"tA.a 1'-'fL+




a-ar:r8a ~ ; shall we then be asharned to irnitate the king of the Persians ? --we shall not be ashamed; shall we? XE:-~. Oec. iv. 4. Ilov 81} (JovA.o Ka8~C6p.evo~ &; whe1e wilt thou that we sit down and nad? PM.r. Phaedr. 228 E. (So ib. 263 E.) BovA.o oDv E7l'~<rK01T'WfLEV 07/'0l! ~81 TO 8vva.T6v errn; XEN. Mem. iii. 5, l. BovAH A.a(Jwp.a.~ 8qTa K(J,~ 8-yw T[ rrov; SoPH, Phil. 761. BovAerr8' E7l'tt<r7T'Errwp.ev; Eua. Hec. 1042. 8EA.<ts ttdVWJ.tEV avTov<rWJ.tV y6wv; So!'H. El. 81. T rro~ eaos 8~/ dKaew; Id. 0. T. 650. 8..\eTe 8YJpa.rrwtte8a 11EV8ws 'Aya.vYJV {tTJTEP K {3a.KXVftUTWV, xap~v T' aJI(J,KT 8wp.ev; EU!t. Bacch. 719. BovA.err8e TO oAov 7/'payp.u dcpwp.ev K<Li ~ tYJTWfLEV; AESCHIN, i. 73. So with KEAEVETE: 'AAAa 7l'WS; Et7l'W KE AEVETE Kai OVK opy~e'i:a-8e; do you cornma.nd rne to speak, and will you not be angry ? DEM. ix. 46. In PLAT. Rep. 372 E, we find el 8' aD (3ovAw8e Ka.i cpA.eyp.a.vovrrav 7!'6A~v 8ewp~<rWj.tV, ov8v 6.7!'oKwAvtt, but if, aoain, you urill have us examine an inflamed state, the1e is nothing to prevent. This shows that (3ovA.err8e is not pareuthetical, but is felt to be the leading verb on which the subjunctive dep~nds (see 288). Iu Phaed. 95 E, .. , ~ "' , ~ , R ,, tva fLYJ Tt O~(J,'t'1!')'1/ YJp.<tS, H, T Tt /"'Ol!I\.H 7l'pO<r 1/S 1) (J,'t'1\.YJS, t 1 SUble junctiVeS may derend on l:va..

e- , '"''

288. El {3ovAerr8e 8ewp1J<rWj.tev, if you wish us to examine, quoted in 287, shows that we have in {301!Aerr8e with the subjnnctive a parata,cis not yet developed into a leading and a dependent clause. It is probable that nothing like this was felt in the simple subjunctive as it is found in Homer. The original interrogative subjunctive is probably the interrogative form corresponding to the subjunctive in exhortations (256) ; ..\Bwp.ev, let us go, becoming A.8wp.ev; shall we go? (See Kiihner, 394, 5.) When {JovAEL and (3ovA.err8e were first introduced in appeals to others, the two questions were doubtless felt to he distinct ; as {3ovA<rr8e; d7rw; do yCYU wish ? shall I speak ?which were gradually welded into one, do you wish that I speal.? Compare in Latin cavejacias,-visne hoc 'Videamus? etc. Ko conjunction couhl be introduced to connect {JovAet or 8AHS to the suhjunctiYe in classic Greek, as these verbs could have only the infinitive ; but in later Greek, where Zva. could be used after 8Aw, the constructiou was developed into 8EAETE l:va. d7rw; do you wi.~h me to speak ? See mJ.vTa o<ra. UJI 8EA7)T tVrL 7l'O~W<r~V l>p.l:v Ot ilv8pw7ro~, whatsoever ye W01tld that men should do unto you, N. '1'. MATTH. vii. 12. So eaw l:va o<{is p.o~ T1JV KecpaATJV 'Iwavvov Tov (3a.7l'TL<rTov. ~I Altc. vi. 25. 'l'hese forms appear in the New Testament side by side with the old construction without Zva ; as T rro~ 8..\ns 7l'O~~rrw; what wilt thou, that I should do unto thee ? with the answer, Zva ava(;w, that I rnay receive my sight, Luc. xviii. 41. So (3ovAerr8e ip.l:v d7roAvrrw TOJI (Jarr~A.Ea Twv '1ov8<tLWJ'; loH. xviii. 39. From 8AETE tl'(J, Ei7l'w; comes the modern Greek 8EAETE va Ei7!'w; W?.:ll J/OU that I speak? and probably also the common future 60. d7rw, 1 sh(J,ll speal; (if ()d represents 8,\w va).




289. The third person of the subjunctive is so~r..etimes used in these questions of appeal, but less frequently than the first, and chiefly when a speaker refers to himself by Tis. E.g.
IToTepov (TE ns, Alux[v'f), T~<; 7TOAEWS x8pov ~ Ef-OV eivat <Pii; i.e. shall we call you the city's enemy, or rnine? DEAL xviii. 124. EiTa Tav8' ovTot 7Tetu8wutv iJ?Tep avTwv ue 7Totei'v, Ka1 Ta ~> uf]> 7TOVT)p[as ~pya cp' eavTovs d11ao~w11Tat; i.e. me these men to believe, etc.; and are they to assume, etc. 1 Id. xxii. 64. T ns ei11at TOVTO cpfj; Id. xix. 88. Ilws T[<; Tot 7Td8T)Tat; how can any one obey you? Il. i. 150. 81iyaTEp, ?Tot ns cppo11T[oos EA.Bv; SorH. 0. C. l 70. IIoZ ns oii11 cp{;yy; Id. Aj. 403. IT68ev oii11 ns Ta{;Tl)S lip~T)Ta t p.cix'f}>; PLAT. Phil. 15 D. Ilws oiiv ifT' et?TT/> on KaKoZs; EuR. H. F. 1417, the only case of the second person, is probably corrupt. Dindorf reads &11 ei7Tots. 290. The subjunctiYe is often used in the question T[ ?Tci8w,what will become of rne? or what harm will it do rne? literally, what shall I undergo? E.g. n p.ot Jy&!, T[ ?Tri8w; T[ 11{; p.ot p.~KtuTa yev17Tat; Od. v. 465. SoIl. xi. 404. T ?Ta8w; T 8 opw; T 8 p.0<; AE><CH. Sept. 1057. T 7Td.8w TA~p.w11; Id. Pers. 912; AR. Pl. 603. T[ ?Ta8w; T[ o; o[p.ot. SOPH. Tr. 973. To f-EAAov, l xpry, T[ yap 7T a w; I shall sujeT what is to come, if it must be; for what harm can it do me? EuR. Ph. 895. (The difl"erence between T[ ?Ta8w; and ?Tciuxw in its ordinary use is here seen.) 'l2p.oA.6yT)Ka T[ yap 7Ta8w; PLAT. Euthyd. 302 D. So in the plural, T[ yap 1Ta8wp.e11 p.ry (3ovA.op.e11wv -&p.wv Ttp.wpeH11J HDT. iv. 118.

291. (Negative p.~.) The negative p.~ of the intermgative subjunctive is explained by the origin of the construction (288). If ~A.8wp.e11; shall we go? is the interrogative of ~A.8wp.e11, let us go, then p.~ A.8wp.eJI; shall we not go? is the interrogative of p.1) ifA.8wp.e11, let us not go, and implies (addressed to others) do you wish not to have us go? This is still more evident when (3o{;A.eu8e is lJrefixed to the subjunctive (288). Similar to this interrogative form of the subjunctive of exhortation is the rare interrogative imperative (also negatived by p.~); as &11 o p.eTa TEX11'YJ> ypafas &.cpKrJTat, p.~ ~euTw 01) eTEpa 7TporynJ.Trew; i.e. is he not to be allowed to give otheT orders? PLAT. Polit. 2!J5 E, where p.~ J~euTw; is the interrogative of fi-'J ~euTw, let him uot be allowed, as p.ij eA.8wJ1EV; (above) is that of p.1) V.. 8wp.E11, let us not go. See also the. indirect question in PLAT. Leg. 800 E, E7Ta11epwT<7J, TWV EKp.aydwv Tat<; <il8as el 7rpWT011 EV Tov8' ~f1-G11 apElTK011 K [ (]" w, I ask again, whether first this is to stand approved by us as one of our models for songs. We cannot express such an imperative precisely in English ; and there is the same difficulty with oiu8' o O(laO"ov; etc., in 253. See also wuTE with imperative forms (602).

292. 1. When the future indicative is used in the sense of the





interrogative subjunctive (68), it may be negatived by p.fJ; as 1rws ovv p.fjTE lj; cf>avepws; how then shall I escape telling an open lie? (where there is some Ms. authority for lj;e{)(, DEM. xix. 320 (see Shilleto's note). 2. A similar use of fJ-fJ is found with the potential optative (with av) in questions, if the idea of pTevention is involved in it; as r{ OVV oB U'KO?rovp.ev mos llv aBTwv p.1) fhap.apTrivotp.ev; why then do we not consideT how we can avoid mistalcing them? (the direct question here would differ little from 1rws fL~ 8 tafloapTri vwp.ev ;). XEN. M em. iii. 1, 10. So ?r<OS av 'TtS Jl-1J 8vp.ii> AEyo t ?rEpt 8ewv; how can one help being excited when he speaks of Gods? PLAT. Leg. 887 C. Sometimes such an optative with p.fJ is in a second clause, preceded by a positive question, so that the harshness of flo1J av with the optative is avoided; as 'T[ av A.yov'TES ELKOS i) av'TOt d?roKvo'i:p.ev i) 7rpus 'TOVS JKe'i: ~vp. p.rix01>S U'KYJ1rTDf1oEVot p.~ f3oYJ8o'ifloev; i.e. what good g1ound can we give joT holding back ouTselves, or what decent excuse can we make to our allies there for withholding ou1 aid from them? THUC. vi. 18. So Tfva av 'Tp07rOV Jyw p.ya ovva[/L')Y Kat f.L1)0efs p.e dotKOLj in what way can I have great power and pnvent any one from doing me wrong? PLAT. Gorg. 510 D. See also Isoc. v. 8, xv. 6. In DEM. xxi. 35, 'lrDTEpa P-~ oii> Ota 'TOV'TO MKYJY 'l fJ-E[(w OO[YJ OtKa[ws; shall he escape punishment for this, OT would he mther deserve a still gTeatm penaUy ?-oo[YJ is used as if Ol.JK ltv oo[J) had preceded (Schaefer inserts Kav). In PLAT. Phaed. 106 D is the singular expression, U'xoA.if yap llv n llA.A.o <f>8opav 11-'J o x o tTo, eZ ye To d8aJ'aTov dtowv <lv <f>8opav 8~eTat, for haTdly can anything else escape j1om admitting destTUction if the imrnoTtal, which is eternal, is to admit it. This (liffer~ from the preceding interrogative examples merely in the substitution of U'xo>..ij, hardly, for 1rws or Ttva Tpo?rov. '

293. As oB cannot be used with the interrogative subjunctive, p.~ here sometimes introduces a question which expects an affirmative answer. See XEN. Mem. i. 2, 36, and PLAT. Rep. :337 B, 554 B, quoted in 287; aud compare XEN. Oec. iv. 4 (ibid.), where a negative answer is expected. In PLAT. Hep. 552 E, we must read p.~ oiwfJ-e8a (not ol6p.e8a, Herm.), shall we 1wt think? as the answer must be affirmative (see Stallbaum's note).


Ov ft~ with the Subjunctive and the Future Indicative.

294. The subjunctive and the future indicative are used with the double negative ov p.~ in independent sentences, sometimes expressing a denial, like the future indicative




with ov, and sometimes a prohibition, like the imperative or subjunctive with p.iry. The compounds of both ov and p,~ can be used here as well as the simple forms.
For a discussion of the origin of this construction, and of the relation of the sentences of denial to those of prohibition, see Appendix II.

295. (Denial.) The subjunctive (usually the aorist), and sometimes the future indicative, with ov p,1) may have the force of an emphatic future with ov. Thus ov p,~ rovro "f~V'Y}Ta~, sometimes ov p,~ 'TOV'To ryEv1)a-E'Ta~, means this su1ely will not happen. E.g.
(Aor. Subj.) Kai Twvo' dKo1~<ras ov n JL?J A'YJcpl:)w, I shall not be caught by any trick. AESCH. Sept. 38. So Sept. 199, Snpp. 228. Ov JL~ 1!"[1:)rrrat,he will not obey. SoPH.Ph.l03. Ov yap <r JL~ yvw<rt. Id. EL 42. Kat ov n JL~ A.axw<rt 'TOVO <rVJLJLcixov. Id. 0. C. 450. Oil'Tot tr' 'Axatwv, oloa, 11-~ ns vj3p[<rrJ. Id. Aj. 560. '0 o' ov 7rap<rnv, ovo JL?J JLO AI! 1l"OTE, but he is ?Wt hm, and he never will come. EuR. H. F. 718. Kov JL~ 1ro8' &A.w. AR. Ach. 662. Twv ~v Kpa'T~<rWJLEV, ov JL~ n> ~JLZV aAAOS <rTpaTdS dvn<r'TU KOT dvl:)pcfJ7rwV. Hnr. vii. 53. So i. 199. Ov JL~ 1roTE f.trj3ci"Aw<rtv. THuc. iv. 95; cf. v. 69. Ov JL-IJ <r Kpvtfw 7rp0S ovnva f3ovAOJLat dcptKE<r8at. XEN. Cyr. vii. 3, 13. '!ls oi 'ApJLEVtO ov JL?J ogwvTa 'TOVS 1l"OAJL[ovs. Ib. iii. 2, 8 (see 296, b). ''Av Ka8c!JJLea atKo,, OVOE7rOT' ovoev ~JLtV ov ,_,~ yev'YJTat Twv o<ovTwv. DEM. iv. 44; so ix. 75. 0-iJT yap y[yvETat OVT yeyovV ovo OVl' JL~ YEV'Y]Tat dA.A.owv 01:Jos 7rp'Os dpET~v,. fo?' there is not, 1wr has there been, nor will the1e eve1 be, etc. PLAT. Rel) 492 E. (Here ovo JL~ yev'Y]Tat seems merely more emphatic than the ordinary OVOE y<V~<rTa t.) (Aor. Subj. 2d PeTS.) Ov yap Tt JLfiA.A.ov JL?J cpvyu> T0 JLop<rtJLov, for you shall none the 1nore escape your fate. AESCH. Sept. 2 81. 'AA.A.' oil 7roT' f.g f.JLo:V y JL~ 1ra fJu s TOO. SoPH. El. 1029. Ov JL?) 1rol f.s T~v l:Kvpov eK'ffA<V<rYJ> lxwv. Id. Ph. 381. 'AA.X oil n JL?J <f>vf''YJTE Aa.tfryp<f 1roo. EuR. Hec. 1039. Kovxl JL?J 1rav<rry<rOE, you will never cease. AR. Lys. 704. (Pns. Subj.) "Hv iap ct7ra~ OlJO ;} Tptwv 'JJL<pwv oo~v d7ro<rXWJLEV, ovKn JL~ OVV'Y]Tat f3a<rtA<vs ~p,as KaTaA.aj3EZv. XEx. An. ii. 2, 12. So ov JL~ O>JVWVTat, Id. Hier. xi. 15. ITp0s TavTa KaKo>)p'}'<t Ked <rVKOcpavT<t, d n ovvaCTat d"AA' ov JL ~ alas T' i)s, but you will not be able. PLA1'. Rep. 341 B. Ov. -yap JL?J ovvaT'Os iJj. Id. Phil. 48 D. In the much-discussed passage, SOPH. 0. c. 1023, aAAOt yap o1 <r7rVOOVTS, oils QV JL?) 1l"OT xwpas cpv-yovns T~<r8' E7rVXWVTat ewt:>,for the?e are others in eager pursuit; and they (the captors) will neve1 (Le in a condition to) be thankful to the Gods for escaping these and getting out of this land, the chief force is in cpvy6vTEs, as if it were ov JL~ cpvyw<r W<r'T E7rVXE<r8at fJEots, the present subjunctive expressing a state of thankfulness.


Ov fUl



(Fut. Ind.) Ov rrot p.1) p.8elj;op.n[ ?roTe. SoPH. El. 1052. Oil p.~rrorf. er' EK TWv ~Opd.vwv, W yEpov, liKovT&. 'TtS lf~ t. Id. 0. C. 176; so OVK oDv p.~ 000t7rOp~O'etS, 0. c. 848. Ma TOV 'A7ToAAw ov p.~ rr' Jy<17Tept61j;op.d7TeA8ovT' (i.e. 7Tept61j; a7TeA8ovTa). AR. Ran. 508. Toils 7TOV7Jpoils ov p.~ 7TOTE (3eAT[ovs 7TO t~rreTe. AESCHIN. iii 177.

296. Ov p.~ with the subjunctive or the future indicative can stand in various dependent sentences : (a) Especially in indirect discourse; as ev yiip oto' rracpws OTt Tav8' .. ov p.~ '7TtAa8n. AR. Pac. 1302. So XEN. Cyr. viii. 1, 5, Hell. iv. 2, 3; PLAT. Rep. 499 B. See also THUC. v. 69. We have ov p.q with
the future optative after ~>, representing the future indicative of the direct form, in SoPH. Ph. 611: Tar U.A.A.a 1ravT' J8f.rr1rtrrev, Kat Ta7Tt Tpo[as 7T~pyap.' ~s ov p.~ 7TO'TE 7TEpo-otev et p.~ TOVOE ayotVTO. (The direct discourse was ov p.~ 1roTe ?repo-eTe Jav p.ry Tovoe &.y7]rree.) In a similar construction in XEN. Hell. i. 6, 32, the future indicative is retained after a l)ast tense: el7reV OT! ~ L7TapT7J ov8v fJ.1J KUKtOV 0 tK te i:Tat w'Tov d1ro8avovTos. In EuR. Phoen. 1590, we have the future infinitive of indirect discourse with ov p.~: el7Te Tetperr[as ov p.~ 1rou, o-ov T~VOE y~v OtKOVVTOS, <:D 7Tpa~etv 7T0Atv, representing ov p.~ ev 1rpa~et 1roA.ts. (b) In causal sentences with ~s; as AR. Av. 461: > 8app~<Tas, ~s Tas 0'7Tov8a> ov p.~ 7TpOTepov 7T a p a f3 wp. e v, for we will not break the truce befoTe you have spoken. So XEN. Cyr. iii. 2, 8 (see 295). (c) In consecutive sentences with wrru; as PLAT. Phaedr. 227 D:

OVTWS 7TtT8vp.Y)Ka aKovrrat, wrrT', Jav 7TOtfj TOV 7Tep[7TaTOV Mf.yapaoe, ov p.~ O'OV d1roAetcp8w. In AESCH. Ag. 1640, 'TOV OE p.~ 7ret8avopa. (ev~w f3a.pda.u; oifn r~ rretpacpopov KptewvTa 1rwAoJ1, and I will yolce hirn who is not obedient
under a heavy yoke, (and I will let him run) by no rneans as a wanton colt in t?"aces, ovn p.~ belongs grammatically to (el)~w, though its position makes it affect the following worrls in sense: cf. Ka' p.ryv TO~' EL7TE P.0 yvwp.ryv p.o{, Ag. 931, where the fo!'ce of fL1J falls on the words that follow it. See Paley's note on Ag. 1640 (1618).

297. (Prohibition.) In the dramatic poets, the second person singular of the future indicative (occasionally of the subjunctive) with ov ft~ may express a strong p1oh,ibition. Thus ov ft~ "A-a"A-~O'El~ means you shall not prate, or do not prate, being nearly equivalent to ft~ "A-a"A-Et or t-th "A-a"A-~O'TJ~ .E.g.
"Q ?rat:, T[ 8poe/:s; ov fL0 wap' ox>..<t> TUO YYJPV<Tet, do 1Wt (I beg you) speak out in this way before the people. Eun. Hipp. 213. "Q 8vyanp, ov p.~ fLV8ov 1rl woA.A.ovs pei:s. Id. Supp. l 066. Ov fL1J yvva.tKwv ottA.uv elo-o [rrets A.oyoJI, do not adopt the cowmdly language of wornen. Id. And. 757. Ov fL1J f.~eyepe'is ToJI V7TV'f! KaToxov KdK-




Ktvf]ut<; Kdvau-rf]uHS rpot-r&oa OLv~v v6uov, 6i TtKvov, do not (Here ov p.f] belongs to three verbs.) T 7rOtE'is; ov p.~ Ka-rafJ~uH, don't come down. AR. Vesp. 397. Ilo'ios Zn)s; OV p.~ )vqpfJuvs. ovo' E(}"Tt ZVS, Zeus indeed! Don't talk nonsense; there isn!t any Zeus. Id. Nub. 367. (Here all Mss. have A:qp~~s. See Nub. 296, quoted in 298; and section 301 below.)
wake him and arouse, etc. SOPH. Tr. 978.

298. A prohibition thus begun by ov p.~ with the future or subjunctive may be continued by p.TJOt with another future form. An affirmative command may be added to the prohibition by a future or an imperative with dA.A& or oE. E.g. Ov p.~ KaA.d:s p:, 6iv8pw<jl, ~KEn-6w, p.TJOE KanpEtS -rovvop.a, do not call to ?ne, I implore you, nor speaJc my name. Au. Ran. 298. 011 /)-~ 7rp0(}"0[(}"tS X'ipa /)-1)0' afH 7rt1TAWV, do not bring your hand neanne nor touch my garments. EuR. Hipp. 606. 011 p.~ 7rpO(}"O[(}"tS X'ipa, f3aKXll(}"ftS o' l&.!v, P.TJO' e~op.6p~H p.wp[av n)v (}"~V ep.o,
do not bring your hand near me; but go and rage, and do not wipe off your folly on me. Id. Bacch. 343. (Here p.TJOE continues the original

prohibition as if there had been no interruption.) Ov /)-~ AaA1J(}"EtS, dA.A.' dKoAov8~(}"tS ep.o[, do not prate, but follow me. Au. Nub. 505. Ov p.~ ow-rptt{Ets, dA.A.a y-6(}"t -r~s 8-6pas, do not delay, but taste of the dom. Id. Ran. 462. Ou p.~ rf>A.vap~(}"HS fixwl', W fE:av8a, J)._)._' dpap.EVOS Ot(}"HS 7rriAW Ta (}"Tpwp.a-ra. lb. 524. Ov p.~ 0V(}"/)-1'~S (}" rf>O... ots, 7r(J.ll(}"ft o Bvp.ov Kai 7rd.AtV u-rp'etj;HS, OE~ft o owpa Kai 7rapatT~(}"ft 7raTp6s, be not inimical to friends, but cease your rage, etc. Euu. Med. 1151. Ov p.~ (}"KWtfTlS /)-l)OE 7rOL~(}"TlS (so all the Mss.) a7rp o1 -rpvyooa[p.ovS OVTot, &A.A.' Urp~p.t, do rwt scoff, nor do what these wretches do; but keep silence! Au. Nub. 296. (Here the imperative is used precisely like the fulmre with dA.Nf or ot in the preceding examples.) The clause with tJ-TJOE is here a continuation of that with ov p.~, ov belonging to both. The future in the clause with dU.d. or of. is like that in 7rd.VTW<; TOVTO op&(}"LS, by all means do this, .AR. Nub. 1352 (see 69). A single o11 p.~ may introduce a prohibition consisting of several futures connected by Ka[, as in SoPH. Tr. 978 (quoted in 297).

299. Sometimes ou with the future indicative in a question implying an affirmative answer (thus equivalent to an exhortation) is followed by p.~ or p.TJOE with the future in a question implying a negative answer (and thus equivalent to a prohibition). Here there is no case of ov p.f]. E.g. Ov (}"ty' dve~H, tJ-TJOE OtA.[av dpd:s; will you not keep silence, and not becmne a cowaTd? SOPH . .Aj. 7 5. (Here /)-~ onA.[av dp'i<;; is an independent question, will you be a coward ?=do not be a coward.) Ov ()fU:r(}"oV oZ(}"Et<;, p.TJO' d7rt(}"T~(}"t<; ep.o; will yon not extend your hand and not distTust rne? Id. Tr. 1183. OvK E'l (}"V -r' orKovs, (}"V n Kp~wv KaTa (}"Teyas, Kal p.~ TO fl-TJOEV ll.A.yos 1> p.ey' or(}"T; Id. 0. T. 637.




300. All the examples under 297 and 298 are usually printed as interrogative, in accordance with Elmsley's doctrine, stated in his note to EuR. Med. ll20 (ll51) and in the Quarterly Review for June 1812. He explains ov p.~ AaA-l]crw;; as meaning will you not stop p1ating? (lit. will you not not prate?); and when a second clause in the future with p.7JOE or &A.A.ri follows, he extends the interrogative force of oil also to this. But this explanation requires an entirely different theory to account for oil p.-1] in clauses of denial (295), where no question is possible. Moreover, the five examples of the second person of the subjunctive quoted under 295, taken in connection with those in 297 and 298, are sufficient to show the impossibility of separating the two constructions in explanation. One of the examples in 298 (AR. Nub. 296), where the imperative dcp-l]p.n follows in the clause with &,\,\&., seems decisive against the interrogative theory. The examples under 299 are really interrogative ; but they consist l)ractically of an exhortation followed by a prohibition (both being interrogative), and contain no construction with ov fh'll at all. 301. In most modern editions of the classics the subjunctive is not found in the construction of 297; and in many cases the first aorist subjunctive in -crvs has been emended to the future, against the authority of the Mss., in conformity to Dawes's rule. (See 364.) Thus, in AR. Nub. 296 and 367 the Mss. have the subjunctive; and in 296, ov p.~ crKwlfvs could not be changed to ov p.~ crKwfH>, as the future of CTKW"li"TW is Elmsley's emendation crKwfet, which is ado1)ted by most editors, requires a greater change than should be made merely to sustain an arbitrary rule, which rests on no apparent principle. If both constructions (295 and 297) are explained on the same principle, there is no longer any reason for objecting ~o the subjunctive with ov p.-1] in prohibitions ; and it seems most probabie tl1at both future indicative and subjunctive were allowed in both con,. structions, but that the subjunctive was more common in clauses of denial, and the future in clauses of prohibition.


Final and Object Clauses after ''lva, 0cppa, and M~.

1 '

n~. ''071'm~,


302. The final particles are tva, oo~, o71'm~, and (in epic and lyric poetry) 6cppa, that, in order that. To these must be added p;ry, lest, which became in use a negative final particle.




303. The clauses which are introduced by these particles, all of which are sometimes called final clauses, may be divided into three classes:A. Pure final clauses, in which the end or purpose of the action of any verb may be expressed; as gPXETat rva TOVTO tov, he is CO?ning that he may see this j a7rJpxETat rva. J.L~ rovro toy, he is departing that he may not see this ; ~A,OEv tva rovro toot, he canw that he ?night see this. Here all the final particles are used, but with different frequency m various classes of writers (see 311-314). . B. Object clauses with o1rwc; or o1rw> J.L~ after verbs of striving, etc. ; as ud1rE o1rwc; "fEV~O"ETat, see that it happens; O"K07rE o7rW') J.L~ ryEv~uErat, see that it does not happen. These clauses express the direct object of the verb of st?iving, etc., so that they may stand in apposition to an object accusative like TOVTO ; as 0"/CO'Tr TOVTO, 07rWS' J.L~ (}" o'tETat, see to this, viz., that he does not see you. They also imply the end or purpose of the action of the leading verb, and to this extent they partake of the nature of final clauses. 0. Clauses with J.L'l after verbs of fearing, etc.; as fjJo(3ovJ.Lai J.L~ rovro ryJv1]mt, I fear that this ?nay happen; fjJo(3~8?J J.L~ ro.Dro ryJvotro, he feancl that th?:s might happen. These clauses have in use become object clauses, though in their origin they are of a very different nature (2 G2 ; 3 0 7). 304. Although the object clauses of class B partake slightly of the nature of final clauses, so that they sometimes allow the same construction (the subjunctive for the future imlicative), still the distinction between chsses A an<l B is very strongly m:trked. An object clause, as we have seen, can stand in apposition to a preceding TovTo; whereas a final clause would stand in apposition to 'TOY'TOV EVKa, as epx<'Tat 'T01J'TOV EVKa, Zva ?j{htV f3oryfJ~O"YJ, he comes for this purpose, viz., that he nu~y assist us. The two can be combined in one sentence; as 0"7rouoatH o1rws 1r AovT~O"H, i'va To1,s cp~Aw;; dj 1r o t r], he is eager to be rich, that he may benefit his
Care muRt he taken not to mistake the nature of an object clause with o1rws when itH subject is attracted by the leading verb; as o-K67rH T?JV 1r6Aw o1rws o-wfJ?)o-erat for O"K07rt o1rws ~ 1r6Ats a-w&~o-<Tat, see that the city ts saved. So also when an object clause of the active construction becomes a subject clause in the equivalent passive form; as E7rpUTTETo o1rws O"Vfhfhaxiav Etva' >frycpuio-fJ<, it was bTought about that






you should vote to ha1;e an alliance made


iii. 64), which

represents the active COllStruction E7rpaTTOV 01rWS yn')cpttfT8.

305. The regular nerrative after Zva, ws, orrws, and ocppa is!'-~;
but after

lest, m\ is used.


'ArrpXTat, ;:v" fl-'J TovTo iou, he is departing that he may 1wt see this. <I>oj3'i:Tat fL~ ov Tovro yv17Tat, he 1~s afraid that this may not happen.
306. This use of fl-~ oil (305) occurs in Homer in a few final clauses (263) and once after 0[8w (IL x. :39). After this it is confined to

clauses after verbs of fearing, with the exception of XEN. Mem. ii. 2, 14, Cyneg. vii. 10, and the peculiar fl-~ 01\K irrapKffTOt in PLAT. Rep. 393 E (132). This use of ov after p.0 is naturally explained by the origin of the depentlent clause with !'-~ (262) ; but after fL'l had come to be felt as a conjunction and its origin was fOl'gotten, the chief objection to fl-'l . . . fl-''l 'Nas probably in the sound, and we find a few cases of it where the two particles a1e so far ~tpart that the repetition is not offensive. Such a case is XEx. Mem. i. 2, 7: f.8w!t~"b o' d Tt>
<joof3oi:To fL~ yvofLEvos Ka.A.~s J((lyu.8os T0 Ta fLEYtfTTCL fJPYET'fJfTavn fL'J T~V fLEYtfTTl)V xapw ~ 0 t, where we should expect 1'-'l ovx E~Ot. So TRue. ii. 13 : lJ7rOTorr0fTu.> fL'J . . . rrapaA.irr'J Ka2 fL'l 01JWfT1]. So in a final clause, fl-'l ... 1'-'l rrpofTOEXOtTo, PLAT. Euthyd. 295 D.
DEVELOPMENT OF CLAUSES WITH Zva, w<;, orru><;, ocppa, AND fL'l

307. The development of final clauses and of clauses with fL'J after verbs of fearing from an original parataxis, or co-ordination of two independent sentences, is especially plain in dependent negative clauses with the simple ~~- Thus &rro{]'nx, !'-0 n J!01J{]'17 ~Hp'J, withdraw, lest Hem notice conything (Il. i. 522), presents the form of an original paratactic expression, which would mean withdraw :-may not (or let not) Hem notice anything, the latter clause being like fl-'l 8,) v~as EA. wfT t, rnay they not take the ships (Il. xvi. 128), and 1'-'l o~ ~ot TEAEfTWfTt 8w2 KaKa 1oj8m (Il. xviii. 8). (See 261.) Such sentences as these last imply fear or anxiety lest the event may happen which 1'-'l with the suhjunctive expresses a desire to avert; and in a primitive stage of the language they might naturally be preceded by a verb of fparing, to which the (still independent) subjunctive with 1'-'l would stand in the relation of an explanatory clause defining the substance of the fear. Thus o{ow-fl-'l v~a> i!AwfTt would originally be two independent sentences, I fear :-rnay they not take the ships; but would in time come to be felt as a single sentence, equivalent to our I fear that (lest) they rnay take the ships. After <foof3ovfLat fL~ TavTo ?ra8wfTw (for example) was domesticated in the sense I fear lest they may suffer this, the second clause followed the ordinary course, and began to be felt as a thoroughly dependent clause; and when




the leading verb became past, the subjunctive became optative, as cpof3~8YJv JJ-0 TovTo 1ra8otev, I feared lest they rnight suffer this. When this stage is reached, all feeling of the original independence may be said to have vanished and a dependent clause is fully etitablished. As this decisive evidence of complete dependence is constantly found in the Homeric language, we cannot suppose that such an expression as odootKa I-'~ Tt 1r&B(J)(J"tv (II. x. 538) was still felt to be composed of two independent sentences, although the original paratactic form is precisely preserved. Indeed, we have no evidence that the step from parataxis to hypotaxis was taken after the Greek language had an independent existence. 1

308. It was a simple and natural step to extend the construction thus established to present and past objects of fear, although we cannot assume for the primitive language such independent indicatives with p.~ as we find later (see 269). In Homer we find 8e[8w JJ-0 Bea VYJfEPTEa d1rev, I jea,r that the Goddess spoke the truth (Od. v. 300). .This use was greatly extended in Attic Greek (see 369 ). 309. This simple construction of a dependent verb introduced /-'-~ with no connecting conjunction remained the established form after verbs of fearing in all periods of the language ; and occasional exceptions, like JJ-0 cpof3ov 0s d1rop~(J"H>, do not fear thatyou will be at a loss (371), o,j cpof3e'i: o1rw;; p.~ dvo(J"wv 1rpa:yfl-a Tvyxavrr 7rpaTTwv; (370), and o,) cjlof3ovfte8a f.Aa(J"(J"W(J"wBat, we are not ajraid that we shall have the worst of it (372), in place of the regular JJ-0 a11'op~U]>, tL0 'Tvyxavns, !-'-0 EAU(J"(J"WBWfL11, only prove the rule. The original independent sentence with f'-~, expressing an object of fear which it is desired to avert, like fl-0 v~a> Awrn, is well established in Homer and appears oceasionally in the Attic poets (261; 264). But in Plato it suddenly appears as a common construction, expressing, however, not an object of fear but an object of suspicion or surmise (265 ), so that /-'-~ with the subjunctive is a cautious expression of a direct assertion ; as fl-~ dypotKonpov {i To dATJBE> el11'ei:v, I rathe1 think the truth rnay be too rude to tell (Gorg. 462 E). 310. In like manner, the simple negative form of the pure final clause, as d1J'O(J"TtXE, f1J Tt VO~(J"11 ''Hp'l} (quoted above), was already established in Homer, the negative p.~ serving as a connective, so that the wa-nt of a final conjunction was not felt. Here also the feeling of dependence is shown by the subjunctive becoming optative when the leading verb is past; as in cpru~op. aL

I-'~ Uiu and ~</>vyov



-ris p..~ ~So.

But it is obvious that

See Brugmann, G1iechische Grarmnatilc, p. 122.






only negative purpose could be expressed by this simple form, in which p.~ could serve as a connective. We find, it is true, a few positive sentences in which a purpose is implied by the mere sequence of two clauses; as ciA.A.' ayE vvv l8vs KLE Ncrropoc; hr11"0'> "'> <r ~ > ' '8 oap.ota' EWOJJ.EV ( SUb' ) 1JV nva fJ-1)TV EV CTT1) ECJ"CJ"t KEKEV EV, I.e. go )
I I ()

straightway to Nestrfr : let us know what counsel he buries in his breast (Od. iii. 1 7), and 8a11"Te JJ. OTT raxtura. m!Aas 'Atoao 7rEp~uw, bury me as quickly as possible: let me pass the gates of Hades (Il.
xxiii. 71 ). But these disconnected expressions, with no particle to unite them, could never satisfy the need of a positive sentence of purpose. To supply this want, several final particles were developed, and were already in familiar use in Homer. These are i'va, w<;,, and ocppa, which will be discussed separately. 311. ("I va.) '1Iva is the only purely final particle, having and, or of the nothing of the relative character of temporal character of &fpa. Its derivation is uncertain. It appears in Homer as a fully developed final conjunction, and occasionally also in the sense of where (Od. ix. 136) and whither (Od. xix. 20). It is overshadowed in epic and lyric poetry by ofpa, and in tragedy by ws; but Aristophanes uses it in threefourths of his final sentences, and in Plato and the orators it has almost exterminated the other final particles. As Zva is purely final, both in use and in feeling, it never takes O.v or KE, which are frequently found with the other final particles, especially with the relative ws-.


312. ('!.ls.) 1. 'n, is originally an adverb of manner, derived like OVTWS from the stem of from the stem o- of the relative oilTos. As a relative it means originally in which way, as/ as an indirect interrogative it means how, whence comes its use in indirect discourse (663, 2). Since purpose can be expressed by a relative pronoun, which in Homer regularly takes the subjunctive < > I > (568), as 1)y<p.ov > Eu 1\0V or.a<ruov, os KE JJ-E KEt<r > ayayo, send rne a good guide, to lead me thither (Od. xv. 310 ), so can it be by the relative adverb of manner, as KpZv' O.vopas KaTa cpvAa, KaTa cpp~Tpas, ws cpp~TPYJ cpp~Tp1)cpW dp~')'TJ, cpvAa o cpvAots, divide the men in that way by which clan may hel1J clan, etc., i.e. (sa) divide them that clan may help clan, etc. (Il. ii. 362). Here the original force of ws can be seen ; but in Od. xvii. 75, STpvvov p.ov TroT~ owp.a ')'VVatKas, ws 70t 8wp U71'01if.p.'fW, in order that I rnay send you the gifts, the final force is as strong as if we had tva cir.or.p.fw. 2. 'Sls, however, always retained so much of its original relative nature that it could take KE or av in a final sentence with the subjunctive, like other final relatives, which in Homer hardly K JJ-E KEW ever omit K before a subjunctive (568). Compare








&yayu (ahove) with the equivalent I:Js K~ fh< Keur' dyayv. The final clause thus receives a conditional form, with which it must have received originally more or less conditional force. 1 Thus an expression like 7re8eo 6>s &v Kvoos apiJaL probably meant originally obey in whateve1' way you may gain glory, or obey in some way in which you rnay gain glmy, 6>s av &piJaL being chiefly a conditional relative clause (529); but before the Homeric usage was establislwJ, the final element had so far olJliterated the relative, that the conditional force of ws llv must have been greatly weakened. The expression in Homer (Il. xvi. 84) may have meant obey tlud (if so be) you may gain me glory. (See examples under 326.) The same is true of the less common use of Ke or ;{y with ocppa and 07rWS in Homer (327; 328). How far the original conditional force survived in tlw Attic ws ;;J, and orrws av with the subjunctive, especially in &rws llv of Attic prose, is a question which at this distant day we have hardly the power to answer, and each scholar will be guided by his own feeling as he reads the expressions. (See 326; 328; 348.) It certainly can be seen in some of Xenoplwn's uses of ws av with the subjunctive; see Cyr. ii. 4, 28, and Eques. i. 16, quoted in Appendix IV. 3. 'Sls and ws K< with the subjunctive are used in Homer also in object clauses after verbs of planning, considering, etc. (341 ), where orrws with the future indicative is the regular Attic form. 'Sls (with ~- av) is by far the most common final particle in tragedy; it seldom occurs in Aristophanes and Herodotus ; while in Attic prose it almost entirely disappears? except in Xenophon, with whom it is again common, though less so than 07rws or Zva. (See Weber's tables in Appendix III.)

313, ("07rWS.) 1. "07fWS is related tO ~S aS 07r0Te tO OTE, being the adverb of the relative stem o- and the indefinite stem 7rocombined.3 Like ws, it is originally a relative adverb, meaning as/ and it can always be used in this sense, as in o1lrws orrws
See Gildersleeve in Am. Jour. Phil. iv. p. 422. Weber (p. 174) quotes two passages of Demosthenes as examples of final ws with the future indicative, a coustruction otherwise unknown in Attic prose: ws o cra.rpws -yvwcrwOE 6n d.\'fJIJ'f, \')'w, l')'w epw, xxiv. 146; and
2 1

Ws DE Kara4>ct;vEs


On 7rp6repov dvaw-xvvroOvres 1repu:yEvovro,


ro.s p.a.prupla.s, xliii. 42. But compare the common formula of the orators J>s (or /in) d.\'fJIJ'i! \ryw, \a(3 ri)v p.etprupla.v (or KaAEL rous p.ciprupa.s), e.g. in DEl!. xxvii. 28, with the occasional full form, lvet do'l,r retvra. on &:r..'fJO'ij Af')'W, f..a.(3 ri)v p.a.prvpla.v, DEM. xlv. 19 ; so xviii. 305. See also ws <lK6ra 7rOLDvp.<v, Ka.l reil!' evvof}cra.r< (se. tva Eiii'l,TE), XEN. Hell. ii. 3, 33. 'l'his common ellipsis shows that in DEM. xliii. 42 we can easily SU])ply a final clause like tva EloT,n before ws Ketra.rf>a.vs gcrretL, that ym may know how it is to be established, etc. In xxiv. 146 there is no need even of an ellipsis, as we can translate how you are to know that I speak the truth, I will explain to you. 3 See Delbriick, Co?Jj. ~._ Opt. p. 61.






ovvavTat, thus as they can, THUC. vii. 67. Then it is used in indirect questions, in the sense of orif rpfYTI'if, how, in what way, and is followed by the future imlicative; as (]'Ko7re'iv 07rws ~ 11"6Ats (]'w8~(]'era t, to see how the city can be saved. So ro'ts yeyevrt /LfVOLS 11"0V'YJpO'i>, /hws fL0 0W(]'0V(]'L OLk'YJV, oObv OELkVV(]'L, he shows those who have been rascals how they can avoid suffering punishment ( = bTif rp6mp fL0 ot6uov(rt), DEM. xxiv. 106. Then, by a slight modification in sense, it may denote also tlw object to which the striving, etc., is directed j SO that (]'k011"EW (or (]'k011"tV rovro) 071"WS ~ 11"6Ats (]'w80(]'erat may mean to see (to this, viz.) that the city shall be saved. Here, lwwevt>r, the subjunctive is sometimes allowed, as the interrogative force of o7rw> is lost sight of and its force as a final particle, in order that, begins to appear. From this it becomes established as a final particle, and denotes the pu'l'pose in ordinary final clauses. From the original fore~ of o7rw> as a relative, u~ed in indirect questions in the sense of how, we must explain its occasional use in indirect questions in the sense of ws (706). The interrogative force of 011"W'> can be seen from passages in which other interrogative words take its place in the same sense; as DE~L XVi. 19, (]'k011"tV eg OTOV Tp011"0V fL1J yeV~(]'OVTat (4>.\ot), to see in what way they can be pTevented j1orn becoming friends; and THUC. i. 65, E7rpauuev 011"17 <hq)eA.{a r<> yH>)(]'eTat, he negotiated to have some help come (how sonw help should come). So TRue. iv. 128, e7rpa(]'(]'EV ilrif rp67rif raxt(]'Ta ro'is fLEV ~vfJ-f3''l(]'ETat rwv o~ ct11"aA.\.agerat.l 2. Although o7rw> is fully established in the Homeric language,

both in its half-interrogative use after verbs of planning, etc. (341 ), and also in its final sense, it seldom occurs in Homer in either construction. It first becomes frequent in the Attic poets. In Thucydides and Xenophon it is the most con11ncm final particle ; and in these writers, as in tragedy, its final use greatly exceeds its use in object clauses. The latter, however, far exceeds tbe final use in Herodotus, Plato, and the orators; but here Zva has gained almost undisputed possession of the field as a final particle. 3. ''07rws never takes KE or ;{v in pure final clauses in Homer. "011"w> &v with the subjunctive appears for the first time in final clauses in Aeschylus (328), and afterwards maintains itself vigorously by the side of the simple ~11"w>. In object clauses ovw> ke with the subjunctive is found in a few places in Homer, and ~11"W'> in a few in the Attic poPts, while 071"W'> in these clauses in prose is found chiefly in Plato and Xenophon (348). 314. ("0 <j>po..) The most common final particle in Homer



S<e J\laddg's Syntax, 123.




is i)cppa, which is originally a temporal particle, meaning while (so long as) and then until. From the last meaning the final force was naturally developed, as the idea of until, when it looks forward to the future, may involve that of aiming at an object to be attained, as in English we shall fight until we me free. Another temporal particle meaning both uhile and until, i!ws, is used in a final sense in a few passages of the Odyssey (614, 2). Both of the temporal uses of /:lcf>pa appear in full vigour in Homer; but its final character must have been more distinctly marked at an earlier period than that of either t1>s or o1rw<;, so that it seldom took either KE or ::.v before the subjunctive. ''Ocf>pa is found only in epic and lyric poetry. 315. (Negative Final Clauses.) The need of these final particles was first felt, as has been shown (31 0 ), in positive clauses of purpose, as a negative purpose could always be expressed by the simple Jk~, which thus became in use a conjunction. Still the final particles were as well suited to negative as to positive final clauses, and they could always be prefixed to Jk~, which thus was restored to its natural place as a negative adverb. Thus <f>ev~oJkat va Jk~ T<; JLE Zov has the same meaning as the older <f>evoJLat Jk~
T<; JLE Zoy, I shall flee, that 1w one may see me. The hiRtory of the Greek language shows a gradual decrease of final JL0 an<1 an increase of the final particles with /1-~ in negative final clauses. 1 The tendency in this direction was so strong that o1rw> Jk~ sometimes took the place of Jk~ even after verbs of fearing, to express the obje'ct of the fear (370), while it became the regular form after verbs of striving, etc., to express tlie object aimed at (3:39). The different origin of the negative final clauRe (with Zva Ji-0, etc.) and of the clau~e with 1-'-~ explains the fact that, while clauRes introduced by the final particles are negatived by Jk~, tl10se introduced by ft~, lest, are negatived by ov. (See 306.)

316. Finally, the Attic Greek took the last step in developing the final clause, by using the past tenses of the indicative with lva, w>_. and o1rw> to express a purpose which failed of attainment because of the failure of the action of the leading sentence ; as T 1;,' ovK ~KTnva<;, t1!, ft~1ron TovT' i!ona / why did you not kill me, that I rnight never have shown thw? (See 333.)
1 In Homer, Hesiod, and the lyric poets we find 131 cases of simple p.~ and 50 of the final particles with f-L~ ; in tragedy the proportion is 76 : 59 ; in Aristophanes it is 8 : 55 ; in Herodotns, 8 : 53. In Attic pro'e (except in Plato and Xenophon) the simple f-LfJ in final clauses almost vanishes. Thucy, dides has only 4 or 5 cases ; the ten orators only 4 (Demosthenes 2, Isocrates 1, Isaeus 1); Plato 24; and Xenophon 12.






317. Pure final clauses. regularly take the subjunctive if the leading verb is primary, and the optative if the leading verb is secondary. E.g.
Ni!v o' f.pxw8' brl 0t7r'VOV, iva ~vvaywp.H "Ap'f]a. n. ii. 381. 2:ol o' 6Joe JLVYJfTTqpe<; inroKpvovTn.t, iv' dojjs awo<; a-cp 8vp.cp El 0WCT L 0~ 7rdVT<; 'Axaw. Od. ii. 111. E1rw TL oqTa Kll..\X, iv' 6py(v '' ' ' ' ' ' ' ' '' ' ~ 7rii.0V; S OPH. 0 . T . 364. K at yap f3 aCTLII.Ell<; atptTat, OVX Wa EaVTOV KaAw<; E7r!JLAqTat, d..\..\' i'va Kat ol (..\6p.evot ot' ai!TOV di 7rpaTTWCTL. XEN. Mem. iii. 2, 3. floKt JLOL KaTaKava-at TU<; ap.a~a<;, iva p.~ Ta (d!'Y'fJ ~p.wv a-TpaT'fJ'YU Id. An. iii. 2, 27. llpos Tovs twvTas, i'va JL'fJO~v lUX d1rw, Tov (wvTa ~~Ta(e, DEM. xviii. 318. (Here the final clause depends on some implied expression like I say this.) ''flpv.v8', i.V &8avaTota-t cj>6ws cp~pot ~o~ f3poTota-tv. Od. v. 2. .P..\os ~f3o-6AETO /{vat TOt<; JL~'YLCTTa ovvap.~vots, i:va &oLKWV JL~ OLOO['fJ 0K'fJY. XEN. .An. ii. 6, 21. TO f~cpta-p.a TovTo ypacf>w (hist. pres. ), i:v' oin-w yyvotv8' ol opKOL, Kal JL~ dpws Tqs ep~K'fJS KO.TO.CTTO.['fJ DEM. xviii. 27. Bov..\~v o' 'Apyeot<; l!7r08'fJCTOJLE8', ~ TLS 6v~a-EL, 6Js JL~ 7rdVT<; 3..\wvTat 6ova-a-ap.~vow Toto. Il. viii. 36. flwvoEtTat ailn)v (ytcpvpav) ..\va-at, 6Js "'~ ow[3qTE, &..\X EV JLFmp a7rOA'fJcp8qT, XEN .An. ii. 4, 17. llf.cpv o' EvpvTov, 6Js Ai!y~av AaTptov p.ta-8ov 7rpaa-CTOLTO., Pind. 01. x. (xi.) 31. Ka rl f.~~7rEJL7rOV, 6>s p,6v"l KAVots. SoPH. .Ant. 19. "E1rep.l{a 6Js 1rv8otTo. Id. 0. T. 71. TovTo oV7rEp lveKa cp..\wv <PETO OEI:a-8at, 6>s a-vvepyovs f. X o t, XEN . .An. i. 9, 21. Tov 8~ p,v'fJa-T~pes ..\oxwa-tv, o1rws &1ro cf>v..\ov o..\"'Tat f.~ 'l8aK'fJ'> Od. xiv. 181. M8es Too' ayyos vvv, 07rWS TO 7rUV p,a8ys. SoPH. El. 1205. Els KaLpov ~KEL<;, 07rWS T~S o[KTJS aKo-6<rys. XEN. Cyr. iii. 1, 8. ITapaKaAE'is laTpovs, o1rws JL~ &1ro86.vy. Id. Mem. ii. 10, 2. 0~ TO.VTO. yyva-8at, oi>x 07rWS TOtlS awovs xopovs Kp VWCTLV Ot 7rOALTO.L, ovo' 07rWS TOVS awovs avA~JTUS E7ratvwcnv, ovo' o1rws Tovs ai!Tovs 7rOL'fJTas alpwvTat, oi!o' iva Tof:s ai!Tots -lj8wvTat, d..\..\' i:va Toi:s VbJLOLS 7rd8wVTaL. Id. Mem. iv. 4, 16. 'Ev xdpECTCTLV E8'fJKev, o1rws en 1rqp.a cj>vyotp.t. Od. xiv. 312. 'AcptKOJL'fJV, o1rws a-oi! 1rpos 06p,ovs eA.8ovTo<; ev 1rpa~atp. n. SoPH. 0. T. 1005. 'E7rpCTf3 EVOVTO E'YK"-'fJJLO.TO. 7r'OLOVJLVO!, 07rWS CT'f'LCTLV OTL JLE'YW'T'f] I , \ I I tl ...J... I ,, I 7rpbcpaa-ts d"l TOV 7rOAEJLE'iv. THUC. i. 126. KaTaveva-op,aL, Scppa 7rE7rO [ 8y <;. Il. i. 524. "Op<rEO 8~ vvv, ~EI:ve, 1ro..\tv8' ip.Ev, IJcppa ue 7r~p.fw. Od. vi. 255. AvTap f.p.ol yepa<; avTx' ETOLJLUUaT', Scppa JL~ olos 'Apydwv dy~paa-TOS f.w. 11. i. 118. fl6p,ov <l>Epa-Ecpovas f...\8, ocpp' loo! VLUV Ei7!''[) s. PIND. 01. xiv. 20. ''fls 0 JL~V f.v8a KaT~uxEr' E7rL'YbJLEIIOS 7rp ooo'i:o, ocpp' ETapov 8a7rTot Kat e1rl KTepEa KTEpuELv. Od. iii. 284. 'AA.U a-v p.Ev vi!v avn> &1r6a-nxe, p.~ 'Tt vo~a-n ''Hp1J f.p.ol o~ K TO.VTa JLEA~CTETaL ocppa TEA~a-a-w. Il. i. 522. Ov O~T awov lf~w 8ei!po, JL~ TL<; &vap7rauv; SoPH. .Aj. 986. Ava-tTEAEt ~aCTa! ~V T!fi




1rap6wn, p.~ J<al -rov-rov 1roA~p.wv 7rpocr80p.e8a. XEN. Cyr. ii. 4, 12. Atye-ra el7rEW on &7rt~VO. f3ovAotTo, p.~ 6 'lrO.T~p TL ll X&otTO KO.~ ~ m5Ats p.~p.<f>otTo. lb. i. 4, 25. Aovrrat KEAer', ws p.~ ITp{ap.os toot VLOV, JL~ oJL~V &,XVVJLEYTJ Kpa8tTJ x6Aov oiJK JpvcratTO. Il. xxiv. 582. For the relative frequency of the final particles, see Appendix' Ill.

318. As final clauses after past tenses express some person's previous purpose or motive, they allow the double construction of indirect discourse (6 6 7, 1) ; so that, instead of the optative, they can have the rnood and tense which the person himself vtould have used in conceiving the purpose. Thus we can say either lj'ABEV 'tva root, he ca111e that he might see, or /lj'ABEV 'tva rov, because the person himself would have said ~pxof"a 'tva tow, I come that I may see. Hence the subjunctive in final clauses after past tenses is very common, in some writers even more common than the regular optative. E.g.
'E'Il'e KAuSrra VT 0 o' 15/...e&pov &v8puS7rots, i'va VfT Kat lrrcrop.eVOU'tV Od. viii. 579. 'AxAtJv ?>' av TOt &'If 6<f>8aAJLWV Hov, 1) 7rptv E'll'~ev, 6cpp' y ty vuSrr KTJ'i ~p.f.v Beov ?jof. Kat avopa. Il. v. !27.

'A ptU''TEV<; t;VYE f3 OVAEVEV EK'Il'AEVU'at, O'll'WS E'll't 'll'AEOV 0 U'tTOS O.V'TU'X'U ' {' ,, ' \ ~ ., ' ' ~ THoc. i 65. "'H..\8ov 7rperrf3evrr6p.evot, 07rw<; JL~ rr<f>lrrt TO 'ATnKov (vavn-Kov) 7rporryev6JLevov EJL7r68wv yh7]TO.. Id. i. 31. 'ExuSpovv EK nov olKtwv, O'll'W<; JL?J KO.Ta <f>ws 8aprraAewTEpot<; OVU't 7rporr<f>pwvTat Kat .rr<f>a-w EK Tov i'a-ov yyvwvTat, &A..\' ~rrcrov<; iJ)rrt. Id. ii. 3. Ka f.7rTryo<; rre o.JK i}yetpa.v, Zva W> ~otrrTa 8tayvs. PLAT. Crit. 43 B. IT..\o['a KfLT~KaVU'EV Zva p.~ Kvpos owf3fj. XEN. An. i. 4, 18. TavTas lva KwAv7J8' o1 v6p.ot rrvv~yayov -Dp.a<;, oilx Zva Kvp[a<; Tot> dotKovrrt 'll'ot~Te. DEM. xix. l. Kal 7rep~ TOVTwv f.p.v~ rr87Jv, Zva p.~ TailTa 'll'a8YJTE. Id. iii. 6. (Here the purpose was conceived in the form Zva p.1) Tavra 'll'aBwcrtv.) 319. This principle applies also to clauses with o1rw<; after verbs of striving (339) and with p.~ after verbs of feaTing, etc. (365). 320. This iR a favourite construction with certain authors, especially Thucydides, who also, on the same principle, prefers the indicative and subjunctive to the optative in ordinary indirect discourse after past tenses (670). The early poets, on the other hand, especially Homer, use it very sparingly. 1
\I ' ' I

1 Weber, p. 243, gives a comparison of the usage of various writers, showing that the proportion of suhjuncthes to optatives after past tenses in pure final clauses and after verb' of fearing is as follows :-in Homer 35 : 156, Pindar 2: 10, Aesr:hy!us 2 : 9, Sophocles 2 : 23, Euripides 31 : 65, Aristophanes 13: 37, Herodotus 86:47, Thucydides 168: 60, Lysias 22: 19, Isocrates 21: 17, Jsarns 8: 17, Dernostlwnes 40:40, Aeschines 13: 7, Plato 22 : 79, Xenophon 45 : 265. In all writers before Aristotle 528 : 894. In the Attic writers and Herodotus, excluding Xenophon, the two are just equal, 441.




321. The subjunctive thus used for the optative makes the language more vivid, by introducing more nearly the original form of thought of the person whose purpose is stated. As the two forms are equally correct, we sometimes find both in the same sentence, just as we find the indicative and optative interchanged i11 indirect discourse (670; see 677 and 690). E.g. 'EgaKo<T[ovs Aoyd8as JgeKpwav, 01rws Twv n 'E1rt1roAwv d'YJ<Tav rpVAaKES Kat, ~V ES aAAO 'n Bev, Taxi! gvvEtT'TWTES 71' a pay [y V w V'T at, i.e. they selected them, that they might be guards of Epipolae, and that they might be on hand if they slwnld be needed for anything else. THUC. vi. 96. ITapav<Txov OE cppvKro:Us, 071'W<; at:racp'lj ra <T'lJJLELa 'TOt<; 71'0AEJLots fl Ka.1 JL~ (3 o 'l} 8 o E v, they raised fore-signals at the same tvrne, in
order that the enemy's signals might be unintelligible to them, and that they (the enemy) might 1wt bring aid. Id. iii. 22.

A common interpretation of the latter and of similar passages, that "the subjunctive mood indicates the immediate, and the optative the rerrwte consequence of the action contained in the principal verbs, the second being a consequence of the first" (Arnold), manifestly could not apply to the first example. 322. The use of the optative for the subjunctive in final cla.uses after primary tenses is, on the other hand, very rare, and is to be viewed as a mere irregularity of construction. See &gw T'ljA' 'IBaKT)'>, i'va fLOt (:J{oTOJI 7roA:Uv aAcpot, Od. xvii. 250; 071'71'W<; p.,ax~otJI'TO, n. i. 344; and vii. 340, xviii. 88. So SoPH. EL 56, 0. C. 11 ; Hm. ii. 93 (Zva JL~ dJLG.proteJ!). Most of these are emended by various editors ; and no good reason for the anomaly appears in any of them. 323. Sometimes the optative is properly used after a leading verb which implies a reference to the past as well as the present. E.g. TovTov EX rov rp61roJ1 o v6p.,os, tva fLYJbE 7reur87jvat JL'YJO' f.ga7ra'TYJ 87jvat yf.vo tr' f.1rt rp B~fL<t> DEM. xxii. 11. (Here lXH implies also the past existence of the law; the idea being, the law was rnade as it is, so that it might not be possible, etc.) So DEM. xxiv. 145, 147. In DEM. iii. 34 Zva rovB' v1rapxot depends on a past verb of saying to be mentally supplied. In AR. Ran. 23, TovTov 8' 6xw, tva JL~ raAat7rwpo'i:ro p.,YJB' axBos <f>epot, I arn letting him ride, that he might not be distressed, etc., the meaning of 6xw goes back to the time when Dionysus first let the slave mount the ass. 324. (Futwre Indicative.) The future indicative occasionally takes the place of the subjunctive in pure final clauses. It occurs chiefly with orrws, very seldom with ocppa, ws, and fL~ and never with Zva. 1 It has essentially the ,same force as the subjunctive. E.g.

lhrws: AEsCH. Cho. 265, Suppl. 449 ; SorH. 'Aj. 698 (1); EuR. El. 835 ; AR. Vesp. 528, Pac. 309, 431, Lys. 1093, Thesm. 431, 653, 285 (1), Eccl. 783, 997; ANDOC. i. 89; XEN. Hi pp. i. 18, Mem. ii. 1, 1 (1ratOEVHP 81rws f11Tat 1). In XEN. Cyr. ii. 1, 4 and 21 the Mss. vary : in Cyr. iii. 3, 42 ll1rws is probably independent. For orppa: Il. viii. 110 ; Od. iv. 163, xvii. 6. For ws: EuR.

Weber cites the following cases, in addition to those given above. For




Ald ?!~ p,aAaKo'i:(rt Kai alp.vA[oun >..6yotU"t 8~AyH, o1rw~ 'IfM.K1J~

i1T"tA~U"~,'Tat. Od. i. 56. M~ 1rp6U"AVU"U"E, ~p.wv o1rw~ p,~ ~v 'TVX1JV ?ltacp8<pe'i:~. SOPH. Ph. 1068. 'A1rop,vKTf.ov ?!f. U"ol y', o1rws A.~lj;Et 7T"tiiv. EuR. Cycl. 561. 'Ap8w U"' o1rw~ dp.f3AaU"Tave'i:~. AR. Lys. 384. 'En-' avTovs TOV> 7rpo>..6yov> <Tov Tpelj;op,at, o1rw> To 1rpwrov T~> Tpay!p8tas p.f.por; 1rpwn<Trov f3a<Tavtw. AR. Ran. 1120. ITpotf.vat

(8e'i:) TWv To7rwv Jv8vp.ovp,evov, 07T"W> p.~ 8tap.apr~<TE'Tat. XEN. Cyneg. ix. 4. Xp~ dvaf3tf3a(Etv J1ri Tov Tpoxov TOV> dvaypacpf.vra>, 07T"W> JL~ 7rpo7epov vV~ ~U"Tat 1rptv 1rv8f.<T8at TOVS av8pas a7raV'Ta>. AND. i. 43. 8apU"vvov 8 o[ ~Top cppeU"iv, 8cppa Kat ''EKrwp dU"eTat. Il. xvi. 242. 'ls rt pf.~op.ev; that we may do what? SoPH. 0. C. 1724. "lU"'T1 elKOS ljp.fis JL~ f3pa8-6vew EU"'Tt, JLlJ Ka[ ns olj;eTat XlJJLWV LU"WS Kan[1rTJ. AR. Eccl. 495. ~o p.~ KexoA.c!:JU"ETat, Il. xx. 301.

"Av or "' in Final Clauses with Subfunctive.

325. The final particles which have a relative origin,


o7rw>, and llcppa, sometimes have O.v or KE in final clauses with

the subjunctive. They did this originally in their capacity as conditional relatives ; and it is probable that at first KE or O.v with the relative gave the clause a combined final and conditional force, in which the conditional element gradually grew weaker as the relative particles came to be felt chiefly or only as final partiCles (312, 2). ''Iva and p.~ never take Civ or Ke in this way. 1 326 .. ('!,.) 1. ''ns Ke and w~ O.v are together much more common in Homer with the suJ::Uunctive than simple ws. 'ls Civ with the subjunctive is not uncommon in the Attic poets, and it occurs in Herodotus; but (like w> itself) it almost disappears in Attic prose. E.g. IIdeeo, ws &v p.ot np.~v p.eyaA1Jv Kat Kv8os &p'Y)at, obey, that
thou mayest gain jo1 me great honour and glory. Il. xvi. 84. Avrap o[ 1rpocpp&.w v7ro8J)U", ws KE p.O.X dU"Kl]B~s ~v 1raTp[8a yaZaJ' i:Kl]Tat. Od. V. 143. ITa[U"aTE, ws x' 6 ~EWO~ f.v[U"1r'(J oi(]"t cp[A.otU"tV. Od. viii. 251. 'AA.X i:Bt, JL~ ,L epf.8t(e, U"awnpos WS KE VE'Y)at, that tlwu mayest go the more safely. Il. i. 32. ITpoU"8eop.e8a .. U"VfL1rfJLlj;at ~p.Zv, <ils av ~f:Awp.ev EK T'lj~ xc!:JpTJS HDT. i. 36. Tovs f.p.ovs >..6yovs Bvp.(/J f3aX, ws av 'TEpp.a-1 EKp.a8vs 68ov. AESCH. Prom. 705. 'AA.X iaU"WfLEV, cp[A.ot, EKl]AOV avrov, <ils av els V1rVOV 7T"f<TYJ.
Bacch. 784. For I"~: Od. xxiv. 544; THEOG. 1307; AR. Eccl. 488. Only four undoubted examples occur in prose. 1 In the single case of d with tva, Od. xii. 156, O.A.A.' <plw eywv, tva e//16TS ij KE 8avWj.tEV, ij KEV a::\.evaj.tfVOL 8avetTOV Kett K~pet if>Df'Wj.tEV, tvet K is not used like &s Ke, etc., above, but tva is followed by a potential subjunctive with K (285). The repetition of Kl removes the case from the class under consideration. "!vet in its sense of where may have liv (see SOPH. 0. C. 405). M~, lest, may have (J.v with the optative after verbs of fearing (368).




' ' ' . 2 a ' t ' , , av S OPH. Ph85 . K() pr;aT avrov, ws <TKOTWV etpop<[ Kv~cpas. Eun. Bacch. 510. Tovri A.a(3lilv p.ov r0 <TKtttOHov ixrrpexE &vto8v, ~> liv p.~ jll>pwcrtv ol 8w. Aa. Av. 1508.

2. In Attic prose <1ls av with the subjunctive is found only in Xenophon and in one passage of Thucydides. The last is THuc. vi. 91 : (7rp.fErE) llvopa "21rapnarrw llpxovra, <1ls ~v ro1Js TE 1rap6vras ~wrd~IJ Kai rovs p.~ ()EA.ovras 7rpocravayKtt<riJ. See XEN. An. ii. 5, 16, <1ls o' ltv p.d()IJ'>, avTttKOl!<TOV. So An. vi. 3, 18. See other examples of Xenophon's peculiar use of <1ls llv with the
subjunctive in Appendix IV.

327. ("Ocppa.) ''Ocppa KE and Scf>p' flv have the subjunctive in a few final clauses in Homer. E.g. Oi'iros vvv crot 11/ fErat, Scppa KEv Eilon cro'icrw Jv2 p.eyapotcrw. Od. iii. :359. ''lop.ev, ocppa KE ()ficrcrov Jydpop.Ev 6~t1V "ApYJa. Il. ii. 440. Tov ~ei:vov ay ES 7rbAw, Scf>p' ltv EKEt()L oaZTa 1T"TWXdiJ. Od. xvii. 10. For ocppa KE and ocf>p' av with the optative, see 329, 1. 328. ("01rws.) ''01rws does not occur in Homer in pure final clauses with either KE or av. ''07rws av final with the subjunctive appears first in Aeschylus, and remains in good use in Attic poetry and prose, being almost the only final expression found in the formal language of the Attic inscriptions. One case of OKws flv occurs in Herodotus. E.g. ;r.. , , , " \ ~ " ' ' \ \ ' " ''l!ll.a<T<TE TaV OLK4} Kaii.WS1 01T"WS av apTLKOII.JI.a <Tl!/L(3 aLVIJ TaOE,
7raV 'TO oplilp.evov, 01T"WS &v elows dyyd.\ns cracpfj. SOPH. EL 40. Tour' avro vvv 08acrx', 071"WS &v JKp.aew. Id. 0. c. 575. OvK a7rt()', 07rWS &v ol AttKWVES Ka()' ~crvxav a7r[wcrw; AR. Lys. 1223. Tavra OE E7rOEE 'TWVOE EtVEKEV, OKWS av I> Kljpv~ dyye>-.:o 'AA.vtt'TTIJ. HDT. i. 22 (see 318). .6.ta TryS. crfjs xi!Jpas a~ELS, 071"WS ltv ELOWp.ev, K.T.A, XEN. Cyr. v. 2, 21. Ka cpaTE avrov 'TOLOVTOV Eivat, orrws &v cpa[vYJTat <1ls KttAAtcrros Ka2 llptcrros. PLAT. Symp. 199 A. "Av y nvas V1T"071"TEVIJ EA.eveepa cppov~p.ara ~xovras 11-~ Jrrt'Tpefew avrcp {ipxHv, (7roAep.ovs Ktvet) 01T"WS &v 'TOVTOVS p.ETU 7rpocpacrews a7roAAVIJ, that he may destroy them. Id. Rep. 567 A. Evcre(3ovp.ev Kai r~v 8tKatocrVVYJV acrKovp.ev, ovx C:va Twv d.\A.wv EAaTTov txwp.ev, dA..\' o1rws &v ws JAETa 1rA.dcrTwv dyaewv Tov (3ov 8taywp.ev. Isoc. iii. 2 (tva and o1rws d.v may here be compared in sense: see 312, 2). T~v 7r6.\w crwexew, 07rWS &v p.av yvlilp.YJV EXW(J"LV a7raVTES Ka2 11-~ Tots Jx8po'is ~oov~v 7rotwcrw. DEM. xix. 298 : so xiv. 23.
watch what goes on in the house, that these things may work harmoniously. AESCH. Cho. 579: so Prom. 824, Eum. 573, 1030, Suppl. 233. "lcr8L

''Av oT 1c in Final Clauses with Optative.

329. 1. ('ns and ll</>pa in HorneT and ws and oKws in Hemdo-

tus.) In Homer ils KE and WS av sometimes have the optative in final




clauses after both primary and secondary tenses. "Ocf>pa Ke and /Jcf>p' :J,y occur each once in Homer with the optative after past tenses. Herodotus has W> ay and OKW> :l,y with the optative after past tenses, and oKw> :l,y once after a present tense. This optative with KE or :J,y after primary tenses is certainly potential as well as final; and this analogy makes it difficult or impo<>Sible to take it in any other sense after secondary tenses, though here the potential force is less obvious. (a) After primary tenses six cases occur in the Odyssey and one in Herodotus : 'A7repplyw:n YEe<TBat &Is K' a1h0> Je8Yc!J<TaGTO BvyaTpa, they dread to go to him that he may settle (if he will) the bridal gifts of his daughter, lit. that he would settle, etc. Od. ii. 53. KYv(cfJ<Tw 8 TOG O<T<TE, WS lJ.y aHKEAWs <f>aYd1Js, I will di'l1j, your eyes, to the end that you might appear unseemly. Od. xiii. 40 l. Llvo 8ovpE KaAAt7reew, ws liv m6v<TaYTe> UolJLeBa. Od. xvi. 297. T0 Ke Taxa yyo1)> cf>t.\67-rJTa TE 7rOAAa TE Owpa J$ JJLEV, WS ay Ts <TE <TVYUYTOJLEYOS JLa Kap {( o t, so that one ~vould call you blessed. Od. xvii. 164 ( = :xv. 537, xix. 310). 'Hyd<TBw 6px'181'-o'i:o, WS dy TLS cf>a'1 yafJ-OY EfJ-fJ-EYaL EKTO<; aKovwy, let him lead off the dance, so that any one who should hear without would say there was a marriage. Od. xxiii. 134. ''I<Txe<rBe 'TrTOAEJLov, ws KEV aya{JLWT ye OtaKpwfJe'iTE TUXL<TTa. Od. xxiv. 531. KeA.evEL <re TlJ mu8[ov 8dvat, OKWS lJ.y TUXL<TTa 8wcf>8apeYf, he bids you so expose the child that he would be likely to perish most speedily. HDT. i. 110.

(b) After past tenses the following cases occur 1 : "Ye 8' apa Zei>s <TVYXES, ocf>pa K BamTOY aA7rAOa TExm BdYJ.
11. xii. 25. 'EwA7rew <re if>(j1JYOe YEe<rBat, ws (iy JLOL TiJY 7rat8a '2Kvp6BeY e~ayayo ts, i.e. I hoped fm your coming, that you 'rnight perchance bring my son away from Scyros. Il. xix. 330. Ka JLLY JLaKpo'Hpoy Ka~ 7rOO<TOYa e~KEV i8e<TeaG, WS KEV if>aL~KE<T<TL cf>A.os mfvTE(T(TL )'EvOLTO. Od. viii. 20. TvJLf3oy xdlaJL<Y, ws KEP TYJAEcf>av~s K 'TrOYTOcf>tv av8p6mY EtYf. Od. xxiv. 83. '2v 8 JL 7rpotets, /Jcf>p liv UoJL'1Y 8wpa (Bekker aYeAo[JL'?Y). lb. 333. Af.yeTaL 8tc!Jpvxa 6pV<T<TELY, OKWS ay TlJ <TTpaT07r00Y i8pvJL&YoY KaTa Yc!JT01> Aa{3ot, i.e. he is said to have dug (119) a channel, in mder that the river might flow behind the army. HDT. i. 75. TavTa o~ 7rep~
1 It must be confessed that there are some difficult questions concerning these optatives with Ke or li.v in final clauses after past tenses. It may ])erhaps be thought that the subjunctive after &s K<, oKws l!.v, etc., has been changed to the optative after a past tense retaining K or llv without effect on the verb. Compare i!ws av with the Ol)tative (613, 4; 702). Would 5Kws llv in HDT. i. 22 (quoted in 328) have changed its nature if d')'')'<IA17 had been changed to d')'')'<IA< 1 On the other hand, can we separate the optatives in HnT. i. 75 and 99 (in b) from the optative in i. 110 (in a) 1 The potential view seems, on the whole, much the more natural; but the potential force can be expressed in English only with great difficulty, owing to the ambiguity of our auxiliaries might, would, should, etc.







Ewvrov Ja-f.p.v1JE TWVbE d'vEKEv, oKw s &v p.ij 6pf.ovns o! 6p.~AtKES AV7rEo[aTo Ka1 E7rt(3ovAdJOtEv, d'AA' ~TEpo'i6s a-<j>t 8odot Eil'at p.~ 6pwa-t, in orde1 that his companions might 1wt be offended by seeing him and plot against him, but that he might appear to them to be of anothe1 nature when they did not see him. Id. i. 99. ITop<j>.Upwv Eip.a 7rEpt(3aA.6p.Evos, ws &v 7rUV8av6p.EVOL 7r AELO"TOL O"V vf. A 0 LE V. Id. i. 152. T6 vowp T6TE E7r{jKav, ws &v xapa8pw8d1) 6 xcvpos, they let in the water, in 01der that the country might be gullied. Id. vii. 1 7 6. IIEptE7rEJJ.7rOV e~w8Ev 2:Ktct8ov, WS &v p.~ 6</> d 1) a- a V 7rEpt7r A.f.ova-at Ev(3otav. Id. viii. 7. "HA.avvov Toi>s t'r.1rovs, ws &v Tdv vEKpov dvEA.o [a TO. Id. ix. 22. METaKtVEEO"I7at JMKEE T6n, WS av fJ.1J loo [a TO o! J~opp.wp.f.vovs. I b. 51.

2. '!:2, ({v with the optative in Attic prose is found chiefly in Xenophon. It is never strictly final; but ws is relative or interrogative, and the optative with dv is potential. E.g. ,'EOo~Ev ai,TtfJ ToVTo 7TOL~crat, Ws 0Tt -{jKHTTa &v 1rt.<jJ86vws <T7ravt.6s TE Ka1 <rEfLV~S </>a v d 1), he decided to do this in such a way that he 1night
appear, etc. (i.e. in the way by which). XE~. Cyr. vii. 5, 37. (Here the separation of &v f1om ws makes the potential character plainer.) '.fls o' &v Kat o! r.60ES EtEV Tifl KpctnO"TOL, El p.f.v TLS EXE pcjw a<rKYJ<TtV, K.T.A., as to means by which the horse's feet could be kept strongest. Id. Hipp. i. 16. See other examples in Appendix IV. This is the same rebtive use of ws with the potential optative which we find in DE~L vi. 3, ws p.f.v &v d1ro tTE OtKa[ovs A.6yovs ap.Etvov q,,A_[7r7r01) 7rapE<rKE.Uaa-8E. WS of. Kw A.Ua-a tT' av EKEtVOV 7rpaTTHV TavTa, dpyws f!xen, as to means by which you could malce just speeches, you are betteT equipped than Philip; but as to steps by which you could prevent him jTom doing what he does, you an wholly inact1:ve. See also DEM. vi. 37, WS o' &v E~Taa-8d1) p..aAta-T' dKpt{3ws, p.~ yf.votTo, as to any means by which the truth could be tested most thmoughly,-may this never come !

330. ''01rws c'f.v with a final potential optative occurs once in Thucydides, four times in Xenophon, and once in Aeschylus:
Tas 7rp0pas Kanf3,!pa-wa-av, orrws &v d1roAta-8avot ~ XEip irrt(3aA.A.op.f.v1), they coveTed the p1ows with hides, that the (iTon) hand when tl~Town on might be lilcely to slip off. THuc. vii. 65. ''EowKe XP'~)p.aTa 'AvTaAK[OI(-, 07TWS aJt, 7rA1)pw8f.vTOS vaVTtKOV 1nro AaKEOatp.ov[wv, Ot 'A8,7va'io~ p.aA.Aov n}s Eip1)1'1)> 1rpoa-8f.o~vTo. XE:s<. Hell. iv. 8, 16. (Here 7rA1)pw8vTos J'avnKov, if a navy should be manned, stands as protasis to 7rpO<TOEOtJITO av.) ''07rwS 8' dJ' ws JppWf1EVE<TTU"l'OV TO <TTpa"l'EVJJ.a 7rOt~<ratTo, J~ &A.A.wv m1Aewv ~pyvpoA6yH. lb. iv. 8, 30. ITaa-w eo[oov /3ovs n, 07rWS av 8.Ua-aVTES Ed'TtciJvTo, Kai EK7rc!Jp.aTa. Id. Cyr. viii. 3, 33 (one llfs. omits &v). T~v AE[av d7rE7rEJJ.tfE otaT8Ea-8at 'HpaKAE[OYJV, 07rWS dv p.ta-8os yf.vo t"l'O TOtS <rTpanc!JTats, Id, An. vii. 4, 2 (most lliss. have o1rws ')'EV1)Tat). So AESCH. Ag. 364. In these cases the final force is equally strong with the potential.




.Elliptical Constructions.
331. In colloquial Greek we often find i'va rf; that what?,where rt takes the place of a final clause, which generally appears in the answer to the question. E.g.
BA. Zva 7{; IIP. 80A.ov 70V7oyt Zva .. <!xwcnv. AR. Eccl. So Nub. 1192, Pac. 409. So DEM. xix. 257: Zva 7[; Zv' <1>s p.e7d 7rAetCT7'fJS crvyyvtJp.'fJS 7rap' -&p.wv Ka7'1Jyopw. Just before this we have 8 t i'i. 7 [; !va p.~T< Aov p.~-re crvyyv<f>f-'rj'> TVXO So PLAT .Ap. 26 c. 332. A final clause may stand without a leading verb expressed, when the omission can easily be su1)plied ; as on ~p~a, p.~ d:rro01Jf-~crw; Zvo. ye JL~ 7rpoA.o.f3wv XP~f-CL7a 7"YJS 7r6Aew<; ~ 7rp0.~H<; opaup.rp XP~U1J, because I held an office, may I not leave the counby? No: that you may not take to fligld, etc. AESOHIN. iii. 21.


333. In Attic Greek the secondary tenses of the indicative are used in final clauses with rva, sometimes with 07T'W') or ros-, to denote that the purpose is dependent upon some unaccomplished action or unfulfilled condition, and therefore is not or was not attained. The tenses of the indicative differ here as in conditional sentence~, the imperfect (the most frequent tense) referring to present time or to continued or repeated action in l)ast time, the aorist and pluperfect to past time ( 410 ). Thus Zva 70VTO (7rpanev means in order that he might be doing this (but he is not doing it), or tlwt he might have been doing this (but he was 1wt) ; Zva -rov-ro ~7rpa~ev means that he might have done this (but he did not); i'va -rovro i7re7rpaxE means that he might have done this (but he Jws not). E.g. OiJK &v Jcrx6f-YJV, l:v' ?J rv<f>A.6s TE Kd KAvwv f-YJ8v, in that case I
should not have fo7'borne (to destroy my hearing), so that I might (now) be both blind and devoid of hearing (implying that really he is not so). SoPH. 0. T. 1387. <i><v, cf>ev, 70 f?J 7<1. 7rpO.yf-a7' d.v8p<f>rrot<> ilxew cf>w~v, l:v' ~ua v f-1J8v oi omo~ A6yot, Alas 1 alas 1 that the facts have no voice for men, so that words of eloquence '!night be as nothing. EuR. Fr. (Hipp.) 442. 'Ef3ovAOJL1JV iev E-repov &v rwv ~80.8wv >..yew 7<1. f3Anu(J', Zv ha(J~f1JV ~crvxos. Au. Eccl. 151. 'Exp0v ElcrKaA.a-avras p.O.pTvpas 7roAAoils 7rapaU1Jf-~vacr8at KeAEVcTat -ri'is 8w()~Kas, Zv'' er Tt ~y[yvero avcf>tuf31JT~CTtf-OV, ~V els Td ypdf-f-aTa rav-r' E'lf'aveAee~v. DEM. xxviii. 5. (This implies that they did not have the will thus sealed, so that it is not now possible to refer to it in case of dispute.) 'E~fJTYJ<rev &v 11-e -rov 1raZoa, t'v' el Ji-~ 7rapeot8oVJI JL'fJDEv 8lKatov A.yEv ~86Kovv. DEM. xxix. 17. 'Exp0v a~-rovs 7~v 7rpo-




-r~pav t+r-quw trru/iv, ?va &.7rTJAAd.ytJ-<Ba To~ov Tov STJtJ-ayCJryov, they ought to have made the previous investigation, in order that we might have been already freed from this demagogue (but we have not been /feed jfom him). DIN. i. 10. See LYS. i. 40 and 42 ; Isoc. ix. 5, xviii. 51. 'AA.A.a u Jxpqv ~tJ-Zv uvyxwp<Zv, [va uvvovu[a ytyv<To, but you ought to give way to us, that our convenation might not be interrupted (as it is). PLAT. Prot. 335 C. T [ sqi OVK i!ppt:f' f.pavT~V T~(]'(j' d7r6 7r~Tpa>, 07r w<; TWV 7rd.V'I'WJ, 1r6vwv d7rTJAAayTJv; why did I not th1'ow myself from this rock, that I might have been freed j?-mn all my toils ? AEscH. Prom. 7 4 7 : so Cho. 195. See SOPH. EL 1] 34. OvKOVV expqv lJE IITjya(]'O'U E<v~at 7rnp6v, o1rw<; J<f>a lvov -ro!> Bw'is -rpaytKcf>npo>; AR. Pac. 135. Tt JL' ov "Aa(30v EKTnvas <VBvs, w<; i!o<t~a JL~7roT< EJLav-r6v dv8p07roww i!vBEv iJ y<ycf>s; that I might never have shown, as I have done. SoPH. 0. T. 1391. El yap tl i>1r6 yqv ~K<v, ws f'~TE B<'Os Jl-'JT< ns lL"AA.os Towo' e1r<y~ BEL, would that he had sent me under the earth, so that neither any God nor any one else should be rejoicing at these things (as they are). AEsCH. Prom. 152. "Eo<t -ra vxvpa ,-6n Aa(3<!v, ws JLTJ8' cl (3ovAETO 8vvaTO ega7raTav. XEN. An. vii. 6, 23 (the only case in Xenophon).

334. This construction is the resuJt of an assimilation, which makes more distinct the connection in thought between the two clauses. It is especially common after secondary tenses implying unfulfilled conditions and unaccomplished wishes.
335. ''Av cannot properly be added to the indicative in this construction. In the two examples in which it is found, it would seem that the construction has slipped into an apodosis, or that copyists have been misled by the resemblance to an apodosis and inserted &v. Zwvn E0L f3oTJBE'iv, 07rWS on 8tKat6TaTO<; ~~~ Ka~ or:ncf>TaTOS i! tTJ T twv Ka~ TAVT~ITUS dnJLcf>pTJTOS &v KaKWV dJLapTTJjLrlTWV ey{yvETO, in order that he might thus live while he lives, and (so that) after death he would be (as a consequence of such a life) free from punishment (?). PLAT. Leg. 959 B. T6v y< 7rpaTTOV'I'rl TL o[KaLOV ov 7rpouqKV d7rop<tV d.\,\' cl!Bvs A.ynv, 'tva'JLaAAov &ll E7rH1'TVTO v<fi i>p.wv, (possibly) that the result might be that he would be (in that case) the more trusted by you. IsAE. xi. 6. 336. The indicative can never be used in this construction, unless the final clause refers to present or past time, and unless also it is distinctly implied that the purpose is not (or was not) attained. If the purpose is future (at the time of speaking), or if it is left uncerta-in whether the object is or was attained, it must be expressed in the ordinary way by the subjunctive or optative, even though it depends on one of the class of verbs mentioned above. Both constructions may occur in the same sentence. E.g. OiJs (TWV VEWV TOVS aya8ovs) TJJL<LS &v e<f>vAaTTOtJ-EV Ell rlKpo7r6AEL, 'tva JLTJOE~S aVTovs 8 d.<f> BELpEv, d,\..\.' E7rELO~ d<f>lKoLVTO <ds -n}v ~,\,.




K[av, XP~<ntJ-OL y[yvotv-ro Tat's ?T6Acnv, we should guard (in that case) in the Acropolis, that no one rnight corrupt thern (as some now corrupt them), and that when (in the future) they should become of age they rnight . become useful to their states. PLAT. Men. 89 B. (Here it is not implied that they never become useful, this depending partly on the future.) Tai!r' &v .Y)ol) A.~yw 1rp6s ilfLas J1rxdpovv, Zv' elo~TE 1roAAov ofv a~wv OVTa TVX'iv TOV if;l)</J[ap.o,To> avrov TOVTOV[, I should (if that were so) be now undertaking to explain this to you, that you ?night (after hearing me) know that he is faT jTom deserving the honour of the proposed decree. DEM. xxiii. 7. Ka[rot xp~v <TE ~ TovTov fL~ ypa<jJHv ~ EKEtvov t\ 1\ ' I I It. JI.VELV, ovx, tvo, o f3 OVI\H <TV YEVl)Tat, ?TavTa <TVVTapac;at, I.e. you ought not to have confused everything in order that what you want rnight be done. DEM. xxiv. 44. 337. Clauses with !'-~ after verbs of fearing are never thus assimilated to a preceding indicative, as there is no reference here to the attainment of a purpose.
\ / ' t1

338. A purpose can be expressed in various forms besides that of the final clause ; as by the relative with the future indicative, or in Homer with the subjunctive (565 ; 568); by the infinitive (770) or the infinitive with W<TT or <1ls (587, 3); by the future participle (840); by w~p with the genitive of the articular infinitive (802).



339. In Attic Greek and in Herodotus, object clauses with o71"ro~ and o7rro~ P-n after verbs signifying to strive, to plan, to care jo1, to effect, regularly have the future indicative after primary tenses to express the object aimed at. The subjunctive also is used, but less frequently than the future indicative. After secondary tenses the future optative may be used, corresponding to the future indicative after primary tenses ; but generally the future indicative is retained, as the original form of the thought (319). The other tenses of the optative are sometimes used, to correspond to the same tenses of the subjunctive, or the subjunctive itself may be retained (318). E.g.
'E?TtfLEAVrat O?TWS (or

that it may (or may not) happen.

!'-~) f'EV~<TETat or y~vl)Tat, he takes care 'E,-<fG<AeZro 3mus Y'"lJ(]'<T<u, yel'l}-

<TOLTO, or y~votTo, he took care that it should happen.





(Fut.) TO pJ.- KaAWS lxov d7I"WS xpovl(ov V fJ-V'i (3ovAVTfOV, we must take counsel that what is well shall continue to be well. AEscn. Ag. 846. 6.t8o.Us 8e Tov8< cppd.{;' d7rws p.'l)8cts (3poTwv Kdvov 1I"d.pot8<v ap.cpt8v<rTU xpot SOPB:. Tr. 604. Lot 8~ p.Anv XP11 TaAX d7I"WS lf<t KaAws. EuR. I. T. 1051. Elp~v'l) 8' d"II"WS E<rrat 1I"ponp.wrl ov8v, but that there shall be peace they care not. AR. Ach. 26. LO~ p.<Aerw dKWS p.~ <r< of<rat. HDT. i. 9. ''Opa OKWS p.~ (ho<rT~<rov Tat. Id. iii. 36. Xp~ 6pav roils 'Apydovs O"II"WS <rw8~<r<Tat 1} IlcA07I"OVV'I)<TOS. Tnuc. v. 27. ''fl<r1I"<p Tov 7I" 8c'i i7rtp.<Adn8at
Owws Q"ijat TE Euo vTa t al o'lEs Kal: ,.a E7rt/r~8Eta ~ o vu t v, oVTw Kal TOv <rTpar'l)yov ~'lrfLA<w8at 8<'i d7I"WS <r<jio n oi a-rpanwrat :!<rovTat

Kal: ,.a 7rtT~0Eta E~ovcrt, Ka'i oV Evt::Ka a-TpaTE{,ovTat -roVTo ~er-rat. XEN. Mem. iii. 2, 1. 8e 8'1)p.aywy~a-ns, ~v <rKo"ll"fjs o7I"ws o1

(3A.rt(TTO fLEV TdS TLfLUS l~ov<rtv, o1 8' aAAOL fL'I)8Ev aOLK~fTOVTat. Isoc. ii. 16. <Pp6vn{;' o"ll"ws fL'IJOEv dva~wv T~> TLJLYJS ravT'IJS 1I"pa~<Ls. Ib. 3 7. T [ p.d.Ata-7 iv a7I"a<rt 8tca-7I"ovoaa-Tat To 'is VDJLOIS; d7I"WS p.~ ycv~a-ovrat o1 1I"pt aAA~AOVS cpovot. DEU. XX. 157. MKpav 1I"p6votav EX LV vp.l:v 6 8cis TOV v6p.ov 8oKt d7I"WS wvpws Ea-TaL Kat fL~T <rvyxv8~(TTUL fL~T' av fLTU7I"OrJfJ~a-Ta; Id. xxiii. 62. KaA.ov TO 1I"apaa-Kvd.(Hv d7I"WS ws {3Ana-Tat l!a-ovTa t Twv "II"OAtrwv ai fvxa. FLAT. Gorg. 503 A. Lll: <DAaf3w8at, JLcfA.ta-ra JLev 67rws fL1J yycv~<r<<r8ov, &v 8e f.yyv'l)a-8ov, d"II"WS 6n Taxwra EKHTJL~<ra-8ov. Id. Rep. 564 C. (Subj.) Xpry cpvA.d.a-a-nv Kat 1I"poKaraA.ap.f3avw 07I"WS fL'I)8' es E"ll"lvowv Tovrov rwa-t. Tnuc. iii. 46. (Tiapa<rKwa(<a-8at) o7rws a-ilv e.<ji dywvt(wp.c8a. XEN. Cyr. i. 5, 14. Ov yap 0"/I"WS 1I"Adovos a~ws YEV'I)Tat E1I"LfJ-EAtVrat, dA.X 0"/I"WS avros dTL 7I"AEtfTTa wpa'ia Kap1I"wa-<Ta t (subj. and fut. combined). Id. Symp. viii. 25. Ov cpvA.d.~c(T8' ()"II"WS p.~ 8a-7I"OT'I)V tVp'I)TEo DEU. vi. 25. ''AA.A.ov TOV E"/1"p.EA~<T ~ O"II"WS on oi 1I"oA'iTat <; FLAT. Gorg. 515 B. "Opa o7I"ws p.ry 1I"apa oo~av 6p.oA.oyifs. Id. Crit. 49 C. (Fut. Opt.) ''E('I) vrro 7I"OAAYjs i1I"LJLEA.tfas O"II"WS ws f_A.d.xta-Ta fLEV 5fotro, f.A.axta-ra 8' aKova-otTo, f.Aaxta-Ta 8' :!potTo. XEN. Oec. vii. 5. (After a primary tense this would be o7I"WS olj;Erat, aKova-aat, EP'I)Tat. But Cobet reads f.po'1).) 'E7rEr!AEtTO 0"/I"WS fLYJ aa-tro 1I"OT lia-otvTo. Id. Cyr. viii. 1, 43. See the examples under 1:30. (Fut. Indic. after past tenses.) ''E1I"paa-a-ov O"II"WS ns (3o~8Eta ~~. THUC. iii. 4. Tipo&VJL'I)8EJJTOS Jvos JKd.a-rov a"JI"WS ~ vavs 1I"po~t. Id. vi. 31. EvA.af3cZa-&at r.apKAEV(T(JE aAA~Aots, a1I"WS fLYJ A1JfTETE fnacp8apEVTS. FLAT. Gorg. 487 D. Ovo' a"JI"WS 6p8~ 11"A1!(TTa ~ 1I", TO Ka8' avr6v 0"/I"WS E"ll"t TOtS ix&poi:s Ea-Tat 1I"ap<a-Kvaa-v. DEU. xix. 250 ; so xix. 316. (Pres. or Aor. Opt.) 'Er.EJLEATO avrwv, d"II"WS d.i &.v8pa7roOa 0 taT<Aol:ev. XEN. Cyr. viii. 1, 44. 'A7reKp{vaTo, 6n avr0 JLEAO 01!"WS KaAWS EX 0 ' Id. An. i. 8, 13. 'EJLEJLEA~KH OE avroZs 07I"WS 6 t71"1raypT'I)S d8d'l) oils 8ot 1I"EfL1I"V, Id. Hell iii. 3, 9. (Subj. after past tenses.) Ppovpljcrovrl (ijga) 07I"WS Atyta-tJos ~JLaS p.~ A.d.8n. SoPH. El. 1402. So HnT. ii. 121. ''E1I"paa-a-ev o7rws




7r6AEp.os yEv'tJTat. Tnuc. i. 57. "E1rpa<J"<rov Otrws d7roo--r~a-wutv 'A(J-qva[wv -r~v 1r6A.w. Id. iii. 70. 'ilveZ-rat 1rap av-rwv 01rw~ p.~ cl1rwp.ev (v.l. li1rtp.ev) JK MaKe8ov[a~, he bribed them to effect that we should not leave Macedonia (after historic present). DEM. xviii. 32.

340. It will thus be seen that the future indicative is the most common form in these sentences, after both primary and secondary tenses ; the future optative, which is theoretically the regular form after secondary tenses, being rarely used. (See 128.)

Homeric and other early Usages.

341. In Homer, verbs signifying to plan, to consider, and to try, chiefly <j>pa(, f3ovA.evw, p.epp.'fJp(w, and 1rnpw, have o1rw~ or ~s with the subjunctive after primary tenses, and the optative (never future) and sometimes the subjunctive (318) after secondary tenses. K~ is almost always used here with ~~ and the subjunctive, less frequently with 01rws (313, 3 ).

342. The original relative and interrogative force of 01rws

is more apparent here than in the Attic construction of with the future indicative, especially after verbs of con-. sidering / though after 1rnpw the dependent clause comes nearer the later meaning. E.g. Au-rol 8~ <f>pa(Wp.8' 01TW~ ox' lipur-ra Y~V'Y}Tat, let us ourselves consider how the very best things may be done. Od. xiii. 365. if?pa(6p.e8' (imperfect) 'Apyeounv 07rW~ ox' lipurra yevotTO. Od. iii. 129. if?pa(eer8at 01T1TW~ Ke p.vYJer-r~pa~ K-rdvv~ Od. i. 295. ITept<j>pa(0p.e8a 1rav-re~ v6er-rov, o 1T w~ >.. (} Yl er t v, i.e. how he may come. Od. i. 76. if?pa(wp.er8' w~ KEV JUV 7T"1T8wfLEV. Il. ix. 112. if?paerere-rat w~ K V~'Y}Tat, J?rd 7rOAvp.~xav6~ Jernv. Od. i. 205. ''Ap.a 7rpoererw Kal fJ1T[ererw A.evereret 07rW~ ox' llpterTa YEv'IJ'Tat, i.e. he looks to see how, etc. IL iii. 110. 'Ev6'fjer 8ed. ~~ '08vcrevs eypotTO. Od. vi. 112. Ov yd.p 8~ 'TOVTOV f-1-~V if3ovA.everas v6ov avT~, ~- ~ TOt KELVOVS '08verevs d?ro-rere-rat aec!Jv; Od. v. 23. BovA.evov 07T"W~ ox' O.pterTa yho tTo. Od. ix. 420. "'HA.Oov, d nva f3ovA.1)v d1rot 01rws 'l8aK'fJV Js 1Tat7raA.6eererav iKo[p.YJV. Od. xi. 479. Mepp.1jpt(ev 01rws J.1ToA.oa-ro 1raerat v~es. Od. ix. 554. Mepp.~pt(e Ka-ra <j>peva <il~ 'Axt>..~a TtJL~erYI (or np.~erd), i.e. how he might honour Achilles. Il. ii. 3. 'AA.A: aye f-1-~TtV v<f>YJVOV Q1TWS av-rovs. Od. xiii. 386. qflpfL'Y}VEV ava Ovp.ov 01TWS ?raVer LE 1TOVOLO 8Zov 'AxtAA~a. Il. xxi. 137. MvYJerOp.8' ws x' 6 gftVOS ~)v 1TaTpi8a yaZav iK'Y}Tat, f1-YJ8~ 'Tt p.eererYJyvs ye KaKuv Kat 1r~p.a 1ra8vertv. Od. vii. 192. In Hymn. .Ap. Pyth. 148 we have ws Ke yev'Y}Ta t. lle[pa o1rws Kev 8~ ~v 1ra-rp8a yai:av i'K'fjat, i.e. try to find means by which you may go, etc. Od. iv. 545. ITnpi ws KE Tpwe~ ilrEp<j>[aAot a1T6AWVTat. Il. :xxi. 459. Tourt 8~ ?roAX E7rt'TeAA 7rtpav w~ 1T






1rl6otV &.p:vp,ova IT?JAEtwva. Il. ix. 179. In Il. xv. 164 we have <f>pa{ecr&w p,~ , o-&8~ raA.r5.crcrn fJ-fLVat (354). For a full citation of the Homeric examples with 071"W~ and w~, see .Appendix III. 3.

343. The frequent addition of Ke to ws or 01rw~ in Homer shows the relative origin of the construction (312, 2). For OKW~ av in Herodotus, see 350 ; for 071"(0~ av in this construction in .Attic writers, see 348, 349.
344. In Homer 01rw~ takes the future indicative chiefly when it is merely an indirect interrogative, with no reference to purpose, as in Il. ii. 252, r[ 71"(0 tOf1-V 07rW~ ECTTat TaO epya, we do not yet even bww certainly how these things are to be; or in Od. xiii. 3 7 6, <f>pa{Ev 01rw~ p,vrycrrijpcrw dvatoecrt XLpa~ Jcp~crH~, consider how,you will lay hands on the shameless suito1s. See Il. ix. 251; Od. xx. 38. In Q,l, xx. 28 the fntnre indicative is retained after a past tense, there ueing as yet no fntnre optative (128) j ev8a Kat Ev8a EACTCT7'0 f1-pJJ-?/P[{wv 071"71"WS 8~ p.vrycrrijpcrtv dvatOetn XE'ipa~ Jcp,]uo. ''07rws may take the future (like other tenRes) as a simple relative ; as in IL i 136, 01rw~ dvrrf.$wv lcrra t, as shall be an equivalent. 345. ''Ocppa has the subjunctive in an object clause in Il. vi. 361, ~07) yap fl-OL 8vp,'o~ E71"f(J'fTVTat, ocpp' E?rap,vvw 'l'pwecrut, and the optative in Od. iv. 463, r[~ uvp,cpprf.uuaro {3ovA.d.~ /Jcppa p,' l A.o t~; In Il. i. 523, Jp,ol 8 KE ravra JJ-EA~(J"7'Ut /Jcppa TE AE(J"(J"W, /Jcppa may mean until. 346. The single object clause of this class in Pindar is Pyth. i. 72, VEV(J"OV Ujl-Epov o<f>pa Kar' otKOV 0 <Po'ivt~ 0 Tv~avwv T dA.rf..Aaros EXYJ> grant that the Phoenician, with the Etruscan war-cry, may keep quiet at home. (See 359:) 347. As relics of the Homeric mage we fin<l tiJs with the subjunctive in sentences of this class in Eun. Med. 461, I. T. 467, PLAT, Rep. 349 C ; and with the optative in AEsCH. Prom. 203 (see 353, below). Herodotus has .1,~ with the future indicative in iii. 84, 159, vii. 161 (in the last w~ crrparryy-q(J"EL~ yAxwt), Herodotus has w~ av with the subjunctive in iii. 85, jl-?Jxavw w~ v UXWJ1-V TOVTO TO ypas, which is cited as the solitary case of w~ av in these object clauses after Homer, except in Xenophon (351 ) So also AESCH. Eum. 771, Sept. 627; Solon xiii. 38. See also SOPH . .Ant. 215 (in 281, above).



&v in Attic Greek and Herodotus. '!l~ and w~ llv in Xenophon.

348. The Attic writers sometimes use 01rw~ :lv with the subjunctive in these object clauses. This occurs chiefly in .Aristophanes, Xe11ophon, and Plato. E.g. LK071"EL 01rw~ Clv d1ro86.vwp,Ev civoptKwrara, see that 1oe die most manfully. An. Eq. 80. Llta!'-'IJXaV~(J"op,at 01rw~ Clv l(J"rov (J"U.?rpov .A6.f3vs,




I will contrive that (somehow ?) you get a rotten mast. lb. 91 7. So AR. Nub. 739, Eccl. 623, Ach. 1060, Eq. 926. MO:AAov ~ np6<T8<v <l<TVH avroils 01l"WS ltv Ka~ f.xovT~> TL o{KaO d<f>KWVTat. XEN. An. vi. 1, 17. (Here some word like ~1l"tfLEAtta is understood as the subject of El<Trfn.) Twv aAAwv E1l"LfLAc'Vrat 01l"WS ltv ()'f}pW<TLV. Id. Cyr. i. 2, 10. 'EKEA<v<T TOv -I><pavAav ~1l"LfL<A1J8~vat o1rws ltv oVTW yEl''f}Tat avpwv ~ f.~EAa(J"t!). lb. viii. 3, 6 : so v. 5, 48. See also XEN. Cyneg. vi. 23 ; Eques. iv. 3. "'H aAAov ~<f>tEfLEVOL btKd.<Tov<TW 1J TOVTOV, 01l"WS

ltv EKa<J"TOL fL1T' EXW(Tt ,.d,\,\6Tpta fL1T TWV avTWV (J"TEPWVTat; Rep. 433 E. IIdvTa 1l"OtovvTa> o1rws ltv !Jcf>{<Tt TO 1l"'fJMAwv ~1l"tTpEthJ. lb. 488 'Eav o' EABYJ, fL"1XaV1]'TEOV 01l"WS ltv 0 w<f>vyYJ Ka~ fL~ oiJ o{K'f}V 6 ~x8p6>. Id. Gorg. 481 A.


Besides the examples cited above, Weber gives :fifteen of Plato, and the following: SoPH. Tr. 618; EuR. I. A. 539; lsAE. vii. 30; DEM. xvi. 17, xix. 299. He adds HDT. i. 20, where OKWS av is certainly :final.

' \ ,.., tl '' t I ) \ 'f3 \ I 1l"tfL11t<T 8at O'il"WS av ws TaXL<TT~ a1l"OAa OtfLEV Ta XP1)fLaTa, Cod. A reads d7roAdf3wfL<V.

349. The only case of o1rws av with the optative in an object clause in Attic Greek, except in Xenophon (351 ), is PLAT. Lys. 207 E, 7rpo8vfLOVVTat 01l"WS <J.v <vOatfLOVO{'f}>, which is potential and on the Xenophontic model (see 351, 2). In DE~L xxxv. 29, EK<A<vOfLEV


350. Herodotus has OKWS llv with the potential optative four times after past tenses. E.g.
IIpo8vfLEOfLEJ ov o Ao~{Ew oKwS <J.v ")'EVOtTo, being zealous that it might (in some way) be done. i. 91. So ii. 126, iii. 44, v. 98.

351. .(Xenophon.) Although Xenophon generally follows the Attic usage in these object clauses (339), he yet violates this signally by having ~> and ~, :iv with both subjunctive and optative, and o1rws :iv with the optative; and further by having the Optative with ~S /iv and d7l"WS av after both primary and secondary tenses. He also has ~> twice with the future indicative (like 87Tws) and once with the future optative.
1. 'Q, or ~s :iv with the subjunctive, ~> with the future indicative, and oh with the optative, are used by Xenophon like o1rw> in the construction of 339. E.~. 'EmfL<AovvTat ws ~XYI oilTw>. Oec. xx. 8. (Here the regular Attic usage requires 81rws. i!~tt.) 'E7TtfL<A<'i:<T8at w> ltv 1r paX Bfi, to talce care that they shall be done. Hipp. ix. 2. 'E7rEfLEAovTo w<;; fL~ KwAvotvTo. Cyr. vi. :3, 2. 'Qs o~ KaAws i!~o Ta VfLET<pa, EfLo'i fL<Aij<Tet. lb. iii. 2, 13. IIpo<m<v ws fL'f}Otts Ktvlj<Totro fL'fJO~ dva~otTo. Hell. ii. 1, 22 (see 355). . For Xenophon's regular use of 81rws in all these constructions, see examples under 339. For his regular use of 3Trw> llv with the subjunctive, see 348.




2. When the optative follows ~s :lv or 811'ws :lv, it is always potential, and the original relative and interrogative force of ~s and 811'ws plainly appears. E.g.
'E11'tfJo11.0VTat ws av ,.-eii.TtcTTOt tV ot 7rOII.tTat trwy take care tnat ,, ( " {./'\ .. \' L ~ 1 the citizens may be best (to see how they might be best). Cyr. i. 2, 5. fls &v &mpaA~uTaTri y' l8l7jv brolovv, I took steps that (by which) I #uTa might krww most aciYUrately. lb. vi. 3, 18. l:Kom;; 01rw<; &v 8ufyoHv1 I am considering how they might live tlw easiest lives. Symp.


vii. 2. (Cf. PLAT. Lys. 207 E, quoted in 349,) For a full enumeration of all the irregular passages of this class in Xenophon, see .Appendix IV.

Negative Object Clauses.

ws in Homer (341) are negative, except that Od. vii. 192 combines ws Ke tK1JTat with p.7J8e n 1ra8ww. Negative object clauses are expressed in Homer, like most negative final clauses (315), by the simple p.~ with the subjunctive or optative, as in Il v. 411, cppa{j~J(}w p.~ Ts ol &.p.dvwv ui:o p.aX1JTat, and Il. xv. 164, xxii. 358, Od. xvii. 595, all with cppa( p.~ and the subjunctive. So p.p.{3AeTo Texos p.~ .:lavao2 7reputav, Il. xxi. 517. These examples show a common origin with clauses after verbs of fearing, but the optative in the last example indicates that the original parataxis is no longer felt. 353,, The earliest example of a negative object clause with a final particle and p.~ is .A.ESCH. Prom. 203, (]"11'Voovres (past) ws Zevs p.~7roT' llp~t:tev Oewv. In all the .Attic writers and in Herodotus the development of the negative object clause with 011'WS' p.f] and the future keeps pace with that of the negative final clause with Zva p.~, etc. 354; (M?] for o1rws p.~ in Object Clauses.) Verbs of this class (339) which imply caution, especially opw and UKo1rw, may have
352. None of the object clauses with 01rws or the simple p.~ with the indicative), even in Attic caution (365), as well as belong equally to the two subjunctive (rarely with the future prose, like ordinary verbs of fear and 011'ws p.~ with the future. Such verbs classes Band C (303). E.g. };K67ret p.?] uot 1rp6vot' Tov Oeov cpvAaKTea. SoPH, 0. C. 1180.

qOpa ~ p.~ vvv p.ev ns eilxep~s 7rapifs. Id. Ph. 519. qOpa p.~ 1rapd. yvr!Jp.7JV 7rEITTJS EuR. H. F. 594. LK<hm Tri8e, p.ry I'VV cpvy6vTes et8' dAwp.ev VITTEpov. Id. .And. 755. T1JpOV "'~ A.af3vs il11'W1I"ta, .A.R. Vesp. 1386. ''Opa p.ry p.aT1JV K6p.1ror; o..\6-yos oVTos elp7Jp.evos fj, i.e. lest this may prove to have been spoken, etc. HDT. vii. 103. ''Opa p.~ 1roA.A.wv EKauT<p ~p.wv xetpwv oe~ue t. XEN. Cyr. iv. 1, 18. 2:Ko7ret 8~ p.~TovTots ailTov J~atT?}IF1JTat Kai KaTayd.auv. DEM. xxi. 151. ''Opa ovv p.~ n Kai vvv pyau7JTat. PLAT. Symp. 213 D. So Il. xv. 164 (see 342). See the corresponding use of 01rws p.fJ for p.fJ after verbs of fearing (370).




after Verbs of Asking, Commanding, etc.

355. Verbs of asking, entreating, exhorting, commanding, and forbidding, which reg11larly take an object infinitive, sometimes
have an object clause with 07rws or 07rWS p.~ in nearly or quite the same sense. E.g. .6dlovs 8 r6v8e <j>pci(' 87rws p.'YJ8e2s (Jporwv Kdvov 7rclpodJev dtJ><j>L8-6o-erat xpot; i.e. tell him that no one shall put on the robe before himself. SoPH. Tr. 604: so Aj. 567. AaKeOatp.ov[wv J8ovro ru frycpto-t/ o7rws p.erao-rpacpd'YJ. AR. Aeh. 536. Kat fl alrw (3paxv, orrws go-op.a U"OL <Pav6s. Id. Eq. 1256. ''OKwS .fwvrwv yev'l]rat pyov rrapaKeAevrrap.evot, ~pyov eZxovro rrpoBv(J>6TEpov. HDT. ix. 102. To IIavaKrov JBovro Botwrovs ('I) o7rws rrapa8wo-ovo-t AaKeOat{J>ov[ots. TRue. v. 36. "07rws fJ>EV p.~ drroBavu 1Jl'rt(36An Kat tKETEvw, LYs. i. 29. IIapatTE'i'a-Bat orrw<; allrw!' p.~ Karatf'YJcp[o-'YJo-Be. ANT. i. 12. .6e~U"TaL 8' vp.wv 011'WS otK'YJV f1-1J 8lji. Ib. 23 : so alrOV(J>aL Brrws oiJ, Ibid. .6wKeAe-6ovraL 07rWS r tp.wpryo-era t 7rclvra<; rovs row-6rovs. PLAT. Rep. 549 E. IIapayyeAA.n o1rw<; p.1) o-ovrat. Ib. 415 B. "Ep.otye drr'l]y6peves o1rws p.~ rovro d7roKptvo[fh'YJV (fut. opt.) Ib. 339 A. 'A1rtp'I](J>EVOV allrr{i 01l"W s /h'YJ8ev J per (f)v ~yeZrat, when he is forbidden to say a word of what he believes. Ib. 337 E.


356. l'his is rare in Homer ; but twice in the Odyssey A[o-o-o{J>at has an object clause with orrws : . Ao-o-eo-Bat 8 fJ>LV allros o1rw> V'YJfJ>Eprea d1ru, and implore him yourself to t>peak the truth. Od. iii. 19. (Compare the regular construction, ollo o-e Ato-o-o0at fl'EVELV, Il. i. 174.) Ao-o-ero 8' alei "Hcpato-rov KAvroepry6v 01rws A.-6o-etcv "Ap'YJa, he implored him to libera.te Are1. Od. viii. 344. ' 357. A{ with Zva and the subjunctive is found in Od. iii. 3 27 : Ato-o-eo-Bat o avr6s zV a V'I](J>Eprf.s EV[(J"1f u, and implore him yourself that he 1nay speak the truth. With this we may compare DEM. xvi. 28, l'rfjAoL eo-ovrat ollx Zva 8eo-1rwt KaroLKto-Bwo-L Jl-bvov 7rOt0-6fhEVot n)v o-rrov8ryv, it will be evident that they take an interest not me1ely in having Thespiae established; in both cases the object clause falls into the construction of a pure final clause. This is very rare in classic Greek ; but it reappears in the later language, as in the New Testament: thus JvroA1]v Katv~v 8[8wfhL Vfh'iv, Zva dya1rau dAA~Aov<;, a new comnwndment I give unto you, that ye love one anothm, I oH. Evang. xiii. 34. So f.oe?}B1JV tva JK(JaA.Awo-tv, Luc. ix. 40. Compare the Latin, 1ogat ut liceat. 358. In Od. xvii. 362 we find wrpvv' ws &V rr-6pva KUTd Jl-V'YJU"Ti)pas d y d pot, she exhorted hirn that he should collect bread among the suitors. (See 329, l.) 359. The singular ease of ws with the subjunctive in Il. i. >58, rfi er' 6w ka-rav~vcrat in)rv0ov, <i>s 'AxtA.{ja TLfJ>~<TU"' dAecrvso 7riJAEas E11'~ V'YJVU"tv 'A xatwv, i.e. I believe that you promised by your nod to honour





Achilles, etc. has the appearance of indirect discourse ; but probably KaTavEDw 0s is used with the same feeling as At<r<rOJW.t 01rws in 356, promising to act here taking the same construction as entreating to act. See PrND. Py. i. 72 (in 346). 'Hs, as an adverb of manner, is here clearly on its way to its use in indirect discourse. Some read Ttfl,~<TH!> and oAE<rHs. 360. A singular use of 07rWS and the future indicative with oii <Tii in place of the regular infinitive occurs in SoPH. Aj. 556, OEf <rE 01rws od~e.s, for OEt (]" oe'i~at, and Ph. 54, T~V <l>tAOKT~TO'U <TE OEZ fvx~v 01rws A.6yot<rtv EKKAefe.s A.ywv. So Cratinus, Fr. 108, o<Z<r' 01rws Eixrx~p,ovos dAEKTpv6vos p,TJo'Ev ow <r<ts. This would be like oop,at 01rws (355) except for the object <re, which is like <re in OEt <rE TODTov, the orrws clause representing the genitive.

Object Infinitive and Indirect Questitms.

361. Some verbs which regularly take an object clause with orrwc; sometimes take an object infinitive, which may have the article Tov or T6. (See 373 and 374.) E.g. 'AE nva hrEp,eAoJ!To <rcpwv avTwv v TaZs dpxaZs dvat, they always took care that one of their own number should be in the offices (where we should expect 01rws ns ~a-Tat or ~a-otTo). TRue. vi. 54. Ovo' f.rrEp,EA+ eTJV TOV Otoaa-KaA6v p,o[ nva YEJ!e<r8at TWJI E'lrt<TTap.eVIJ)JI. XEN. Mem. iv. 2, 4. To p,ev oi'iv AEKTtKoDs y[yvEa-8at ToDs <rvv6vTas ovK ~<T7rEVDEV. Ib. iv. 3, 1. (See 793.) 362. Verbs signifying to see or look out (like <rKo1rw) may be followed by an indirect question with d, whethe1; as El ~vp.1rov~aELs Kat ~VVEp yci<rH <rK67ret, see whether you will assist me, etc. SOPH. Ant. 41. For independent clauses with orrws and orrws p.~ with the future, often explained by an ellipsis of a-K6rret or a-KorrEZn, see 271-283.

Amist Subjunctive in -(J'w and



363. When an aorist subjunctive active or middle was to be used with 01rws or orrws p,~ in any construction, the second aorist was preferred to a first aorist in -a-w or -a-wp,at, if both forms were in use. This preference arose from the great similarity in form between these sigmatic aorists and the future indicative (as between (3ovA<vo-v and (3ovA<'-va-H, (3ovAEV<TTJTat and (3ovAdJa-eTat). This made it natural also for a writer to avoid those forms of the subjunctive which were nearly identical with the future indicative where the latter could be used as well. This of course does not apply to the first aoriBt subjunctive passive, which has no resemblance to the future; and there is no reason for applying it to liquid aorists like fJ,ElJ!W and a-cp~A.w. 364:. The general rule. laid down by Dawes more than a century ago (Mise. 01it. pp. 222 and 228), the so-called Canon Davesianus,




which declared the first aorist subjunctive active and middle a solecism after 81rw!i pfJ and oil pfJ, was extended by others so aB to include 01rws (without p.fJ), and the Greek authors were thoroughly emended to conform to it. As this rule has no other foundation than the accidental circumstance just mentioned (363), it naturally fails in many cases, in some of which even emendation is impossible. In the first place, there is no reason for applying the rule to pure final clauses, in which the future indicative is exceptional (324) ; and here it is now generally abandoned in theory, though not always in practice. There is, therefore, no objection whatever to such sentences as these: (Jjy lveKa E7TTa0~vat, 07TW!i a7roAavuwp.ev Kal 07TWS yevwpeOa, XEN. Cyr. vii. 5, 82; EKKATJu[av ~vvfJyayov, 01!"W!i v7ropvfJuw Kat p.!, THUC. ii. 60; and T'ifv ayopd.v E7Tt T~V O&.A.auuav Kop.[ua, 011"W> 7rapd. Ta, vav> ap CTT07TOfjuwvTa, Kat u oAfyov Tots 'AOTJva[ots E7TXHPWCTv, THUC. vii. 39, in which the best Mss. have the subjunctive. Indeed, where the reading is doubtful, the subjunctive should be preferred in these cases. Secondly, in independent prohibitions with 01rws p.fJ, although the future is the regular form, there is less objection to the subjunctive (even the first aorist) than in positive commands with simple or.w>, since the analogy of the common p.~ 7rotfJuTJ!i TovTo, do not do this, supports 81rws p.~ 7rotfJuTJ> TovTo in the same sense (283). There is no such analogy, however, to justify such a positive command a~ 81rw> 7rotfJu1J!i TovTo, do this, and this form has much less manuscript authority to rest on. Thirdly, in the case of otl p.fJ, if both constructions (denials and prohibitions) are explained on the same principle, no reason exists for excluding the subjunctive from either ; and it cannot be denied that both the first and the second aorist subjunctive are amply supported by the manuscripts. (See 301.) Fourthly, in object clauses with 01rws there is so great a preponderance of futures over subjunctives, that the presumption in all doubtful cases is here in favour of the future, as it is in favour of the subjunctive in pure final clauses. A much stronger case, therefore, is made out by those who (like Weber and most modern editors) change all sigmatic aorist subjunctives in this construction to futures. Some cases, however, reAist emendation ; as XEN. An. v. 6, 21, K<Aevovut 1rporrTanvrrat 81rws tK11"Aevuv ~ crTpaTL<i, where we canuot read tK'IrAevu<t, as the future is EK7rA< or iK11" A< In DEM. i. 2, all Mss. except one read 7rapauKwauau()a T~v Tax[uTYJV 81rws tv0v8e f3oTJOfJuTJT< Kat p.~ 1r&.OTJT< Ta0T6v, and it seems very arbitrary to change f3oYJOfJ<rTJT< to f3oTJOfJueTf'. and leave 1r&.OTJT<. But a few cases like these weigh little against the established usage of the language, and we must perhaps leave the venerable Canon Davesianm undisturbed in the single depar1;ment of object clauses with o1rws, although we may admit an occasional exception even there. See Tra:nsactions of the A mmican Philological Association for 18 6970, pp. 4-6-55, where this question is discussed more fully.





365. Verbs and phrases which express o:r imply fear, caution, or danger take #~ lest or that, with the subjunctive if the leading verb is primary, and with the oJptative if the leading verb is secondary. The subjunctive can also follow secondary tenses to retain the mood in which the object of the fear originally occurred to the mind. M~ (like Latin ne) denotes fear that something may happen which is not desired; (ut= ne non) denotes fear that something 'ffW.,y not happen which is desi1ed. E.g.

#n ov

if>o(3ovfLat fL~ yevrrrat (vereor ne accidat), I fear that it may happen: cpof3ovfLat fL~ ov yev'fJTat (vereor ut accidat), I fear that it may not happen. b..d8w fL~ O~pEuuw lAwp Ka~ KVpfLa yEvWfL<tt. Od. v. 4 73. llE8w fL~ oil Ts Tot flTroUX'fJTat To8E lpyov. Il. x. 39. (This is the only case of fL~ ov in these sentences in Homer. The next that are found are EuR. And. 626, EL 568, Phoen. 263. See 264, above.) Ov cpof3iJ fL~ a? ., Apyos aTrOKTEtVat 8Uu. EuR. Or. 770. ITotOJI e8vos ov 8oKEt flTrEP'fJT~UEtl/ cpo(3oVfLEVOV fL~ n 7ra8u; XEN. Cyr. i. 6, 10, <I>povT(w fL~ KpanuTov iJ fLOt utyfiv. Id. Mem. iv. 2, 39. <l>vAaTTOfLEvos fL~ 86~u J-Lav8avELV n .. lb. iv. 2, 3. b..e8otKa fL~ ov8' ouwv aTrayopE1JEtl/. PLAT. Rep. 368 B. Td. 7rEp~ T~S fvx~s TroAA~v amurav 7rapexa TOtS av8pti:J7rots, fL~ ~Tra8d.v aTraAA.ayiJ TOV uti:JfLUTOS ov8afLOV ETt if, aAM 8tacp0dp'f}Tal TE Kat aTroAAll'f}'Tat. Id. Phaed. 70 A. OvKovv vvv Kat TovTo Kv8vvos, fL~ Aa(3wut 7rpouTaTas avrwv rwas TOVTWV, there is danger of this, that they may take, etc. XEN. An. vii. 7, 31. Kv8vv6s Jun, fL~ fLETa(3d.A.wvrat Kal yevwvTat fLETa Twv TroAEJ-Llwv. Isoc. xiv. 38. 'OKvw fL~ fLOt <I Avulas ra1rnv6s cpavfi. PLAT. Phaedr. 257 C. EvA.a(3ofi 8 fL:yl cpavfjs KaKOS yEy<ils. SoPH. Tr. 1129. Ov8v 8nvo2 EUOVTat fL~ {30lJ(}EWfrt ravru. HDT. vii. 235. 'YTrOTrTEVOfLEV Kat VJ-LaS fL~ OV Kowol aTro (3 ~TE. THuc. iii 53. AluxvvofLEvos fL~ cpopnKws <rKOTrWfLEV. PLAT. Theaet. 183 E. 01 fLVOot urpecpovuw ailrov T~ll fvx~v, fL~ aA'fJOEZS Giut, torment his soul with fear lest they may prove true (92). Id. Rep. 330 D. lld<ras fL~ 1rti:Js oi f.pvua [a To VEKpOv 'Axaw[. Il. v . 298. '' A(;Ero yap fL~ ::\TvKTt Oofj d'!fo0llfLtU ep8ot. 11. xiv. 261. 'Eyti> yap iifLYJII EK7rE1TAYJYfLEIIYJ cp6!3'1}, fL~ fLOL TO KaAAos aAyos f.~n)pot 7r0TE. SoPH. Tr. 24. "E8Huav oi "EAAYJI'ES fL~ 7rpoud.yotEV 1rpos TO Kepas Kat auToBs KQTUKOlfHav. XEN. An. i. 10, 9. . OvKETt E7rT[(}EIITO, 8e8otKOTES fL~ a7rOTfL'fJ(}d'fJuav. lb. iii. 4, 29. "E8Hcrav fL~ Avrra ns W<TTrEp Kvu2v ~fL'i:v EfL1r1fTti:JKot. lb. v. 7, 26. 'YTroTrTEV<ras fL~ T~v Ovya-;:epa Aeyot, -i)pETo, having suspeeted that he m~qht mention his daughte1. Id. Cyr. v. 2, 9. 'H(}{;fLYJ<rav nvEs, f.vvoolJfLVOt fL~ T<i im-;~8Ha ovK fl x o LE v 61ro0v AaJ-Lf3avotv. Id. A.n. iii. 5, 3. Ov8E2s




yap Kvovvos o6Ket il'vat P-'1 ns &vw 1ropnoJ1-f.vwv EK roil 07r!cr8u' E7rcr7r6tro. Ib. iv. I, 6. ;r.. ' ' ' ' '(3 1\ \' 0 ' ' Otc "'WKO,tfES Ta<; I'?JcrOV<; OVK E OVII.OVTO 7rWII.EEW1 VHfJ-atVOVTE<; fl-?J Ef1-7roptov yf.vwvTat. HDT. i. 165. Tcj) yap OEotf.vat fl-~ .\6yots ~cr6-ovs iJJcr t, TOAJl-'YJPW'> 1rpus Td ~pya x6Jpovv. TRue. iii. 83. llEptOE~s f'EVOJl-EVOS fl-~ f'lrt7rAEf!crWcrLv a1 V~<;. Jd. iii. 80. "EoELcra fl-~ Tpoav d8pocrn Kat ~vvotKcrn 1rriAw. EuR. Hec. 1138. 01 8Ecbf1-EVOL Jcj>of3ovvTo fl-'1 n 1ra817. XEN. Symp. ii. 11. D.ijA.os i]v 1racrtv (Klipos) on V7rEpecj>o(3EtTO fl-~ o1 6 7rri7r7T"OS d7ro ea. VYJo Id. Cyr. i. 4, 2. For the present subjunctive in these sentences denoting what rnay henajte1 p1ove to be an oLject of fear, see 92.

366. The manner in which this complex sentence expreRsing fear was developed from an independent sentence like Jl-1J v1)w; f!A.wn, rnay they not seize the ships, and a precediug verb of fearing like the two gradually becoming one sentence, has already been explained (:307). As the fear and the desire to avert the cause of fear are both implied in fl-~ with the subjunctive, it is not strange that this expl'ession can follow verbs like 6pw and oloa which do not imply fear in thewselves;


as i.~eA.86Jv ns i'ooL, fl-~ 01J crxeoov (f)(]"L KLOVTES, let some one go ont and see that they do not appToach neaT '(cf. videat 11e accedant) ; originally, let sorne one go out and look to it: rnay they not approach, Od. xxiv. 4D l. So ollof. TL tOfJ-1', fl-~ 7rWS Kat OLd J!VKTa Jl-EVOW~(]"W(]"L fl-U xecr8ru, nm do 'We know any way to pTevent their being irnpelled to fight even d~tring the night; originally, noT have 'We any knowledge: rnay they not b1' impelled to figi~t, Il. x. 100. See also PLAT. Phaed. 91 D, Tooe &o'YJA.ov 7ravTt, fl-~ 7roAAd <TWfl-aTa KaTaTptfacra ~ tfvx~ TO TEAEVTaLOV avT~ &7roAA1~?] TaL, i.e. iw one lcnows any security qgainst the soul itself finally perishing,
etc. The indirect question sometimes used in translating such a clause with p.~, as whether they may not appToach or 'Whether they. may not be irnpelled, is merely an attempt to express the hesit,ttion which the apprehemion involves, as there can be, of course, no real indirect question. See especially the cases of fl-~ with the present indicative (369, 1), which are often called interrogative. See the corresponding construction in 4 9 2.

367. (FutuTe Indicative.) Sometimes, though seldom, fL'J has the future indicative after verbs of fearing. Tht' examples arc:<Pp~v dJl-f.crcreTaL cp6f3'1:', Jl-1J 1roAts 1rv81]TaL .. . Kai Tb Kwcr!wv
0 " ' O' ' 1r011.L(]"fl- aJITWOV'lrOV I(UETat, (3 VcrcrtVOLS V EV 7rE7rii.Ot<; 7rE(J"YJ 1\.UKtS. AEscH. Pers. 115. Tavr'-oilv cpo(3ovf1-at, fl-~ 1rocrts Jl-EV 'HpaKA~s Jfl-os KaAELTat (fut.), T~<; VEWTEpas o' dv~p. SoPH. Tr. 550. 6.8otKa f1-1J ~AAov nv6s fl-E(jf.~w. XEN. Cyr. ii. :3, 6. <Pof3ovf1-aL oe ,u~ nvas ~oovos 1)oovaZs Evp~crOJl-EV JvavTas. PLAr. Phil. 13 A. 'AAA.Ii (cf>of3epov Kai crcpaAep6v) fl-~ crcpaA.eis Kdcrofl-aL. Id. Rep. 451 A.
1 1 1 \ ' ' \ ' \ '

(l'he last two examples are not given by Weber.) l<'or three cases of p.fJ with the future optative after past tenses of yerbs of fearing, representing the future indicative, se" 131.





P-~ anel the But a potential optative with &v can follow fl-~ after a verb expressing fear or anxiety, after 1oth primary and secondary tenses (168). E.g. 6.f.8otKa yap f1-1J 7rp~) Af.yo t<;; <lv Tuv 1r68ov TOV ~ Ef-ov, I fear that you might perhaps tell. SoPH. Tr. 631. Ovu 7rpo~rSoda ofJO{f-f.a (1}v) f-1J &.v 7ro7{ o1 1roAf.f1-wt E7rt7rAdJ<rHav. THUC. ii. 93. 'EK{tVo vvow fl-1J A[av liv Taxil <rwrppov~~r8d'Y)v, lest (in that case) I should be veTy soon bTmtght to my senses. XEN. An. vi. 1, 28. 6.{8~6ns f-?J KaTaAv8d'Y) av 6 8-fjf-OS. LYS. xiii. 51.

368. The particles &v and KE are never used with


369. (Present and Past Tenses of Indicative with f'~) Verbs ofearing may refer to present or past objects. (See 308.) M~ c_an therefore be used with the present and past tenses ..of .the indicative after these verbs. .. 1. M~ .with the present indicative expresses a fear that something is now going on. E.g. 6.f.8otKa fl-?J 1rAyywv of.H, I am afraid that you need blows. AR. ,Nub. 49:3. 'Opwf-{V fl-1J N~Klas oZ{Ta{ T~ -A.f.yew, let us be cautious ~est Nicias is thinking that he says something. PLA'r. Lach. 196 C. {Here oZT)Ta~ would have meant lest llicias may think, in the future.) "Opa P-0 h{Zvov KWAJJH. Ill. Charm. 163A. <Po(3{t<rl3{ /:'1J ov<rKoAw'npol' n vvv o~a.K Hfha t -1) Jv T<[l 1rpo<r&{v (3cp, you aTe Gfmid that I am now in a 1JW1'e peevish state of mind than I used to be in (where the subjunctive would have been future, lest I may hereaftm be). Id. Phaed. 84 E. 'E1r[<rX{'' <ils &.v 1rpofJ~Ep{vv~<rw O'T[(3ov, f-~ ns 1roAtTwv ev Tp(3cp <P a VT!L ( { Ta ~, Kap.ot f-EV f.A8v rpavAos WS oovA.cp tjloyos. EuR. Phoen. 92. (Here f-1J <f>ana(ETat means lest any one is now to be seen; and f-1J lest any repoTt may come hereafter.) 'AA.>..' dcrOf-{<r&a fl-~ n Kat KaTaO'x{Tov Kpv<f>?) KaAv7rTet Kap8[?- 8vf1-0Vf1-i~vYJ, 86f1-ovs 1rapa~rnxovns. SoPH. Ant. 1253. (The idea is, we le,wn the ?"esult of our anxiety lest she is concealing, etc.l) KdfJ-aVT~> 7rEpt 8f.A.w 7rv8f.~r&at, fl-1J '1rt Tots 1raAa~ KaKots 7rpO<rKdf1-ev6v n 7T~fl-a cn)v OaK vu, and I wish to inquire about myself, (in fea1) les1~ etc. Eun. Her. 481. "AJ!a~, Efl-O[ TOt, fl-~ n Kat e{1)AaTOV Tovp~!OV Too', ~ ~1!wow (3ovA{VH 1raAat. SoPH. Ant. 278. (The idea is, my m,ind has long been delibemting in anxiety lest this is the w01lc of the Gods, E<rT[v being uuderstood after ft~. 2) ''Opa, cpvA.a<r~rov, f'~ ns v crT{/3'1! f3po-



(se. f<rTtv).

EoR. I. T. 67.

In this passage and the following, if anywhere, it would seem necessary to admit the intenogative force often ascribed to Jl-fJ. But here, as elsewhere, it is plain that the dependent clause with 1'-fJ expresses the object of an apprehemion. To establish Jl-iJ as an interrogative, meaning -whethm, 1'-iJ should not only follow a verb like otoa., but also be followed by a clause expressing no object of apprehension, like d!T6J1-e!T8a Jl-TJ ol q,lA.ot !;'w!Ttv, -we shall lea.nL whethe?" our jTiends an now living; but no snch example can be found in cla~sic Greek. 'l'he nse of El, whethe1, after verbs of fearing (376) shows how the Greeks expressed an indirect question in such cases. 2 That this is the correct explanation, and that we need not emend the




2. M~ with the perfect indicative expresses a fear t:hat something has already happened. The difference between this and the :perfect subjunctive is often very slight, the latter expressing rather a fear that something may hereafter prove to have happened (I 03). E.g.
Nvv 8~ cf>o(:JovJLdla JL~ O.JL</>QTtpwv &JLa ~JLapdKaJLH, but now we fear that we have missed both at once. TRue. iii. 53. (The perfect subjunctive here would mean lest it may hereafter prove that we have missed.) b.t8otKa JL~ AA~8aJLH (~v dp~V'Y)V) i1r~ 7ToAAifi ll:yovTes, I fear that we have been 1tnconsciously enjoying peace borrowed at high interest. DEM. xix. 99. <Po(3ovJLat JL~ Myots n<Tl tj;evOE<TtV h'TETVX~ KOf'EV FLAT. Lys. 218 D.

3. M~ can be used with the imperfect or the aorist indicative, to express fear that something happened in past time.
t.d8w JL~ 8~ 1rrivTa 8e0. V'Y)JLEpTta et7rev, I feM that all that the Goddess said was true. Od. v. 300. 'A>.>.: Opa JL~ 11'at(wv E')'EY, but be careful lest he was speaking in jest. PLAT. Theaet. 145 B. 370. ("01rws JL~ for JL~ with Verbs of Fearing.) Verbs denoting fear and caution are sometimes followed .by an object clause with 01rws JL~ and the future indicative, the subjunctive, or the optative, like verbs of striving, etc. (339). It will be noticed that ihws JL~ here is exactly equivalent to JL~ in the ordinary construction, so that cf>of3ovJLat 01rws JL~ yev~<TeTat (or ytvrJTot) means I Jea? that it will happen (not I fear that it will not happen). E.g. t.8otK' 011'WS f'~ 'K T~S <TtW11'~<; T~<T8' J.vapp~~Et (Laur. &vapp~~YJ) Ku.Kri, I fear that a stmm of evil will burst forth from this silence. SOPH. 0. T. 107 4 (the earliest example). Toli 8aJLOVOS ot8otx' il1TWS JL~ T1J~OJLC1.t KaKo8a[JLovos, I fear that the luclc that I shall get uoill be bad luck. AR. Eq. 112. EiJ>.af3ovp.EVOL o-iw> JL~ OLX~< PLAT. Phaed. 91 0. L::.ootKa 01TW> JL~ O.vriyKYJ "/EY~<TETat, I fear that there may ' ~ " ' ' ' ~ be a necess~'t y. D EM. 1x. 75 . 0' 't'o(3 et 01TWS JLYJ aVO<TtoV 1rpayJLa TV')'v -~, xrfvu> 7rpriTTwv; PLAT. Euthyph. 4 E. <PvAriTT01! 011'W> IL~ d-. TollvavTtOV f.A.fJus. XEN. Mem. iii. 6, 16. 'H8ws av (fJpEy,atJLt T0v av8pa), EL JL~ cf>o(3o[JLYJV 07TW'> JL~ M alln)v JLE Tpri7I'OtTO. lb. ii. 9, 3. Tol:s 1I'pE<T/3vTepots aVTt7rapaKEAEVO/f'a JL~ KaTaurxweqvat 07TWS JL~ 86~o JLaAaKu> lvat, i.e. not to be shamed into fear lest he ma.y seem to be weak. TRue. vi. 13. Compare the corresponding use of JL~ for o1rws JL~ in ordinary object clauses, especially with opw and <TK01TW, which belong equally to both classes, B and 0. (See 354.) 371. (Indirect Discourse with .;.., or o7I'ws.) In curious contrast

passage so as to rea<l ToiJpyov T6o' (i, ~tlvvo<a [3ouA<V 1ra"Aa<, is suggested by the scholion: r, <rtlvvOLa p.o, {JouA<V<Ta< Kal of<ra< p.n Kal 8efJXaT6v < <rn .,.c, 1rpfiyp.a.. So perhaps we should read if>of3ii:<T8at p.-f, T< liatp.6vwv rO. 1rp&."yp.a.Ta JXauv (vulg. iXavv?l) in DEM.. ix. 54 (with Cod. A). But the subjunctin in both passages might be explained on the pdnciple of 92.




with the preceding construction with 61rws p.~ for p.~ (370) is that by which verbs of fearing sometimes take the construction of ordinary indirect discourse. Here ws and even 611"ws, that, may introduce the object of the fear, thus taking the place of p.~ in the common construction. This apparently occurs only when the leading verb is negatived. E.g.
M~ Beur~s ?roB' ws y~Awn 'TOV!J-OV cpatBpov ol/Jerat Kapa, do not feM' that ~he will ever see my face joyful ( = p.~ ZBy). SoPH. El. 1309: so 1426. 'Av8pos p.~ cpo(3ov ws thop~rrets &.gtov, do notjeQ/1' that you will be at a loss. XEN. Cyr. v. 2, 12. (Here the direct discourse would be d1rop~rrw, I shall be at a loss.) M~ 8drr'YJTE ~., ovx ~Bews Ka6ev8~rre-re, do not fear that you will not sleep sweetly. Id. vi. 2, 30. (Here p.~ ovx would be the ordinary expression.) Ov 7ovro BeBotKa, ws id.v <iKpoarrtle aV-rwv d?rofrJcpte'irrBe, I have no fear of this, that you will acquit them if you heQ/1' them. LYS. xxvii. 9. M~ rperrys O'II'WS rre -rts <i'll'orr'll'arret f3v-, that any one shall tear you away by force. EuR. A.. I > \ 'I I \I \I I H er. 248. M'Y\J 't'o(3 ev f-'YJTE ep.e, ws rreo 'll'etpwp.evos Aeyw Aoyov rovoe, / ,... , , \ 1 '{:. ) .-.. I f-'YJTE yvvatKa T'YJV Ef-'YJV, f-'YJ TtI rot E~ aVT'YJS yev'YJ'Tat (3' 1(3os, do not Aa fear either that I am saying this to try you (ws A.eyw), or lest any hMm shall come (P-~ yev'Y}rat). HDT. i. 9. (Here the two constructions after cpo(3ev make the principle especially clear.) In all these cases p.~ or o'l!'ws p.~ would be regular, and exactly equivalent to ws and o1rws here. In the same way, we s:ty in English he fears lest this may happen and he fears that this may happen in the same sense. In Greek we might have p.~ rperry> O'll'ws p:~ rre -rts d?rorrmfrret (3 70) in the same sense as p.~ rperrys o'll'ws rre ns <i'll'orrm:frro (above).
~ ~

372. (Infinitive.) The future infinitive may stand in indirect discourse after verbs of fearing, to represent a future indicative of the direct course. E.g.
Ov cpo(3o{;p.e8a i A.arrrrwrrerr eat, we are not afraid that we shall have the worst of it. THUC. v. 105. (Here p.~ with the subjunctive would be the regular form.) 373. The present or aorist infinitive (without p.~), not in indirect discourse, may follow verbs of fearing, to denote the direct object of the fear; as in English, I fear to go. This infinitive may l1ave the article. E.g. <Po( o~v 8 te Aeyxe tv rre, p.~ il'll'oAti(3ys, K.r.A., Jam afraid to refute you, lest you may suspect, etc. PLA.T, Gorg. 457 E. <Po(3~rrera dBtKei:v, he will be afraid to do u;rong. XEN. Cyr. viii. 7, 15. (But cpo(3~rrerat p.~ dotKfj, he will fear that he may do 'll>r0711J.) t:.e8tevat cparrKOVTWV KepKvpatwv ~XELV avrov. THUC. i. 136. Ou KareBewav ~<rA8d:v. Id. iv. llO. IT~fptl<a 'EpLVVv 'TA~<rat "Td.s Komipas, I shudder at the idea of the Fury fulfilling the curses. AESCH. Sept. 720. (But in vs. 790, rpew 1'-~ re>..Errv means I tremble lest she may fulfil




t~em.) See also XEN. An. i. 3; 17. To rhro()viJcrKELJI ovoet> <Pof3iirrat, To 8 d8tKtv <Po{JtTat. PLAT. Gorg. 522 E. 374. Verbs of caution may be followed by an infinitive (with or 'without p.~), which sometimes has the article; the infinitive or the infinitive with p.~ having the same meaning as a clause with p.~ and the subjunctive or optative. E.g. ~ ' ,, t ' , ' \ ' " ' : IT Wo OVK a,;tQJI aVTOJI ""j <PVicagacr{)at TOWVTOJI ""jJIfT () at j W hy ought he not to guard against becoming such a 'rnan hirnself? XEN. Mem. 5, 3. (Here yv~cr()at is equivalent to p.~ yivqTat.) 4>vA.aTT6p.vo> _TO AV7r~cra( -nva, taking cwe to offend no one. DEM. xviii. 258 . .PvA.a&crHv p.YJ8~va 7rpawvcr8at, to g'liard against any one's crossing over. THUC. vii. -17. 4>vAaT-r6p.<vov Kat 7rpooprflp.<vov p.~ KaTatcrxvvat '-rav1-Yjv. DEM. xxv. 11. (For p.~ in this construction see 815, 1.) ln THuc. vii. 77, we find the infinitive with wcrn after <PvA.acrcrw. . 375: Kv8vv6> E(J"Tt, the principal expression denoting danger, which ~akes J-t~ and a finite verb, is quite as regularly followed by the ili,finiti ve. E.g. Ov fT{ttKpos K[v8w6> ~crnv ~a71'aTYJ()~vat, there is no little danger ,()f their being deceived. PLA'r. Crat. 436 B. Ktvqvv<vw is regularly followed by ,the infinitive (7 4 7). 376. (Indirect Questions.) Verbs of fea1ing may be followed by an indirect 'question introduced by d, whether, or by some other interrogative. "Orrw<; as an interrogative here must not be confounded :with orrw> as a conjunction. E.g. Oi. 0!8otKa d .PU.. t71'7l'OS {jj, d,\A.' d -r~> 7r6AEWS 7'~()VYJK< TO TOVS ;d8tKovvTas p.t(J"<LJI Kat np.<op<l.crBat, I have no fear (on the question) whether Philip is alive; but I have fea1 (about this), whether our .city's habit of hating and punishing evil-doers' is dead. DEM. xix. 289. .P6(Jo> {tOt {W(J"Ljl oils Jyw Bf.A.w. EuR. Her. 791. .Pf.povcra (J"Ot liEOVS 1JKW A.6yov>, <j>6f3'1! p.f.v d ns O<crrrorwv alcr8~(]'Tat, through fear whether any one will perceive it (where {t~ at(J"B"Y)Tat would have meant lest any one shall pe1ceive it). Eua. Andr. 60. See XEN. Cyr. vi. l, 17. 4>o"f3ovvTat orrot rroTE 7rpof3~crETat TJ -rov dv8po> 8vvafttS. XEN. Hell. vi. 1, 14. (The direct question would be 7roi: rrpo(J~crerat ;) T~v B<ov 8' orrws ,\<f()w 88otKa, I am in fear (about the question) how I shall escape the Goddess. Eua. I. T. 995. (The direct question was 71'W<; A.aBw; 287.) So SoPH. Ph. 3:37. 'A7ropovvns 7rws XP~ d7rLBEZv, <Po(JovJ-t<YOL of. rrws XP~ a7r<tA.ovvn {,rraKovcrat. XEN. Cyr. iv. 5, 19. 377. (Causal on) "Verbs of fearing may be followed hy on, because, and an ordinary causal senten,;e with the indicative (713). E.g. OvK a~wv 8ta TOVTO <Po(3t(J"ea, TOlJS' 7/'0A<J-t[Ov>, OTt 7/'oA..\ot -rvyxavovcrtv OJIT>, to fear them., they happen to be many. Isoc. vi. 60. .Po(JoVfLEVYJ> T~> fl-1JTPO>, OTt TO xwp[ov E71'Vv86.v<TO VO(J'W0S lvat. Id. xix. 22. ''On o rroAA.wv apxovcrt, f-L~ <Pof3178~TE, d,\U '1l'OAV JhUAAov OLU TOVTO eappiiT<, do not be afraid because they rule many, etc..XEN. Hell. iii. 5, 10. 'E<j>o(3ei-ro, on 6<j>(J~(J"(J"eat f.p.EAA ;Ta f3acrA.Ha olKoOOJ-tELV dpx6J-tvos, he was afraid, because he was about






to be seen beginning to build the palace. Id. Cyr. iii. 1, l. if>oj3oVp.EV<)s r6 Kaw&a~ Kat ro rep.v<rr&at, 8 n dA. yetv6v, fearing them because they are painful. PLAT. Gorg. 4 79 A. So THUC. vii. 67.


Conditional Sentences.
378. A conditional sentence consists of two clauses, a dependent clause containing the condition, which usually precedes and is called the protasis, and the leading clause containing the conclusion, which is called the apodosis. The protasis is regularly introduced by the particle et, if, negatively /, f'-~ 379. Ai is a Doric.. and Aeolic form for cl, and is sometimes used in epic poetry in the forms aW< and at yap, and less frequently in at K<.l 380. The name protasis is often restricted to clauses introduced by a particle meaning if But it applies equally to all conditional relative and temporal Clauses (520), and it properly includes all clauses which naturally precede their leading clauses in the order of thought, as E7r<t iJa-8ETo rovTo, a1r~A.8cv, after .he perceived this, he departed. Such a clause may still be called a protasis, even when it follows its leading clause, provided the order of thought is not changed.

381. The adverb &v (epic IC~ or IC~v, Doric d.) is regularly joined with El in the protasis when the verb is in the subjunctive, El with &v (a) forming the compound M.v, ~v, or &v (ii). (See 200.) The simple Ei is used in the protasis with the indicative and the optative. The same adverb li.v is regularly used in the apodosis with the optative, and also with the past tenses of the indicative when non-fulfilment of the condition is implied.
382. The only Ionic contraction of El av is 'JV, which is used in Homer and Herodotus. The Attic Greek has av, ')v, and av (a); but
1 Ai for <l is usually left in Homer by editors as the Mss. give it. But Bekker (Homerische BliitteT, pp. ol, 62) '}Uotes Heyne with approval, who says that no human being can tell why we have al in one place and <i in another. Bekker cites, to illustrate this, a tO' oifrws x.o'Aov T<AeO'<t' 'A'Ya!J.E/lvwv, Il. iv. 178, and dO' &s TOt 'YovvaO' i!1rotro, iv. 313; also at K< ()e!Js t""'rat, Il. v. 129, followed immediately by &.rap d KE 'A<j>polilr1J g}..8p0'' is 'lrOA</lov. Bekker 1n his last edition of Homer (1858) gives only <l, <f0<, and <l 'Yap, without tegard to the Mss. ; and he is followed by Dolbriick.



[383 -

!v, if, was probably never used by the tragedians or by Thucydides, although the Mss. have it in a few cases.

383. The negative particle of the protasis is regularly that of the apodosis is 384. When ov is found in a protasis, it is generally closely connected with a particular word (especially the verb), with which it forms a single negative expression; so that its negative force does not. (like that of 1'-~) affect the protasis as a whole. E.g. ~~ ''"A 'A,.~''.' IT avTW> O'YJ1rOV (~ OVTWS XH, eav TE <TV Kat .'1VVTO!> OV 't''YJT eay T'



you deny it, as well as if you admit it. PLAT. Apol 25 B. El Tot> OavovTas ovK efjs ( = Kw..\vas) Od:rrntv, if you forbid burying th~ dead. SOPH. Aj. 1131. El JL~V ov iro..\..\o~ ( = 6.\.[yot) ~<rav, KaO' ;Ka<rTov &v 7rep' TOVTWV ~KoveTe, if there were only a few, etc. LYS. xiii. 62: cf. 76. Twv8e ~V ov8tv t<TOV EO"T~v, dye cl.cp' ~JLWV ye TWV Ell f'-~O"Ijl ouods OUOE71'0H lJ.pgETfH, there is no fairness in this, if (it is the plan, that) no one is eve1 to begin with us. XEN. Cyr. ii. 2 1 3. In all these cases 1'-~ could be used, even where ov seems especially proper; as in llv T' eytiJ cpw llv T JL~ cpw, whether I admit or deny it, DEll!. ui. 205. See EuR. Hipp. 995, ov8' ~)! <T(, JI-1J cpifs. The use of 1'-~ or ov was determined by the feeling of the speaker at the moment as to the scope of his negation. The following example makes the difference between oi. and JL~ particularly clear, o-& affecting merely the verb, and 1'-~ affecting the whole clause (including the ov): el }'-~ Ilpo~eJ!ov ovx i17rEOE~avTo, euwfJ'YJcrav /lv, if it had not been that t~ey did not receive Proxenus, they would have been saved, DEM. xix. 74. 385. El oi. with the indicative is sometimes found in Homer where the Attic Greek would have d }'-~; as in d 8 JLO oi.K e7rwcr' E'fft7rdUETat cl.,\...\' d..\oricrEt, I1. xv. 162. See also Il. xx. 129; Od. ii. 27 4 1 xii. 382. 386. After verbs expressing wonder, delight, and similar emotions (494), where a protasis seems to take the place of a causal sentence, d ov can be used, on the principle of 384, though here JL~ is more common. See examples of l JL~ under 494 ; and for d ov see Isoc. i. 44, JL~ 6avJL&ny,s d 7ro,\..\a TWY elpYJJLEVWV ov 7rpe7ret uot. See also

</>~Te, if

387. When two clauses introduced by f'-EV and U depeng upon a single el which precedes them both, ov is used even more frequently than JL~ ; as such clauses have their own construction independently of the El, which merely introduces each of them as a whole, not affecting the construction of particular words. E.g.
.6.nv6v &v ~z'Yf, l 0~ f"tll hdvwv gliJLJLaxot E71't OOl!Ae['l- Tfj avTWV cpepovns oflK d11'"epov<TtV, ~JI-ELS o' i1ri T<i) aflTOL <T~(,ecr()a, o0K /lpa.




lia7ra.Jrq1Top.Ev, it w01tld be a hard thing, if (it is a fact that) their allies will not refuse, etc. while we will not contn:bute. THUC. i. 121. ET,.'

' ' ' ' ' ' ' ' ~e '-~ '() ' ~ OVK atiTXpOII, Et TO JLEII 'A pyetwll 'Tri\Yf o<; OVK E't'Of3 Yf Yf TYfll A aKEOatJLOII[wv dpx~v i>fLE'i> ll~ f3r5.pf3apov </>ofJ~ITEIT()E; is it not then disgraceful, if (it is true that), while the Argive people did not fear, you are going to be afraid, etc. DEM. xv. 23. See also PLAT. Phaed. 97 A; LYs. :x::u:. 32; IsAE. vi. 2; DEM. :x::x:xviii. 18; AESCHIN. iii. 242.


388. The most obvious natural distinction is that of (a) present and past conditions and (b) future conditions. Present and past conditions (a) are divided into two classes by distinguishing (1) those which imply nothing as to the fulfilment of the condition from (2) those which imply that the condition is not or was not fulfilled. Future conditions (b) have two classes (1, 2), distinguished by the manner in which the supposition is stated. Class 1 of present and past conditions is further distinguished on the ground of the particular or general character of the supposition, as explained below in II. (394). 389. Excluding from the class (a) 1 the present and past general suppositions which have a peculiar construction (395, a and b), we have-

I. Four Forrns of Ordinary Conditions.


390. In present or past conditions, the question of fulfilment has already been decided, but we may or may not wish to imply by our form of statement how this has been decided. In Greek (as in English or Latin) we may, therefore, state such a condition in either of two ways : 1. We may simply state a present or past condition, implying nothing as to its fulfilment; as if he is (now) doing this, el ToiYro 7rpriiTITo,-ij he was doing it, ~:l e7rpaa-rn,-if he did it, el e1rpa~e, -if he has (already) done it, el 1rE7rpaxe,-if he had (already) done it (at some past time), el E7rE7rpr5.xet. The apodosis here expresses simply what is (was or will be) the result of the fulfilment of the cond1tion. Thus we may say:Et 7rpriiTITEt 'TOlrrO, KaAw<; exH, if he is doing this, it is well; el 1rpaiTITH TOVTO, ~fLapTYfKEII, if he is doing this, he has erred j el 1rpri1TrrEt TOVTo, KaAW> l~n, if he is doing this, it will be well. El E7rpae (or f.1rpaa-a-e) rovTo, Ka..\ws EXH (d'xev, f.axEv, or, if he did this, it is (was or will be) well. El 11'f1rpa.xf. roiYro, KaAws l~f.,, 'if




he has done this, it will be well. So with the other tenses of the indicative in the apodosis. (See 402.) . So in Latin : Si hoc jacit, bene est; Si hoc fecit, bene erit; etc.

2. On the other hand, we may state a present or past condition so as to imply that it is not or was not fulfilled; as if he were (now) doing thi8, El 'TOV'To :hrp<HHTE ;-if he had done this, El 'Tovro brpa~< (both implying the opposite). The apodosis here expresses what would be (or would have been) the result if the condition were (or had been) f11lfilled. The adverb av in the apodosis distinguishes these forms from otherwise similar forms under (a) 1. Thus we may say : El i!1rpa<nrE rovro, KaAw<> &v El'XEV, if he wme (now) doing this, it would be weU; or if he had been doing this, it would have been well. Ei i!1rpa~ 'Tovro, KaAw<> &v i!<TXV (or &v ElxEv), if he had done this, it would have been well (or it would now be well). On the other hand, l E7rpa~ TOVTO, KaAw<; i!<TXEV (without av) would mean if he did this, it was well. (See 410.) In Latin : Si hoc faceret, bene esset (present); Si hoc fecisset, bene juisset (past).

391. The Greek has no form implying that a condition is or was fulfilled, and it is hardly conceivable that any language should find such a form necessary or useful.

if he does this), it will be well (sometimes also cl

444 and 44 7.)

392. The question as to the fulfilment of a future condition is still undecided. We may state such a condition in Greek (as in English and Latin) in either of two ways:1. We may say if he shall do tl!is, Jav 1rpa<T<TU (or 1rpa~u) rovro (or, still more vividly, fi upd.~H TOV'To), making a distinct supposition of a future case. The apodosis expresses what will be the result if the condition shall be fulfilled. Thus we may say:'Eav 1rpd.<Trr17 (or 1rpa~u) 'Tov'To, KaAws l~EL, if he shall do this (or
1rpa~EG 'TOVTo). (See In Latin: Si hoc faciet (or si hoc fecerit), bene erit.

2. We may also say if he should do this, El 1rpa<T<Tot (or 'Tovro, still supvosing a case in the future, but less

distinctly and vividly than before. The apodosis corresponds to this in form (with the addition of av), and expresses what would be the result if the condition should be fulfilled. Thus we may say:El 1rpa<T<Tot (or 1rpd.~tE) rov,-o, wAws &v ~xot, if he should do this, it would be well. (See 455.) In Latin : Si hoc faciat, bene sit.




393. The Latin commonly employs the future indicative, si hoc faciet (correspow1ing strictly to 1 TovTo 7rpagt., if he shall do this), or the future perfect, si hoc fecerit, to express the form of protasis which
the Greek expresses by Mv and the subjunctive (J{w TovTo 7rpacruv or 7rpag'[]) ; and it uses the form si hoc faciat to represent the Greek .d TovTo 7rpacruoL, if he should do this.

II. Present and Past General Suppositions.

394. The supposition contained in a protasis may be either particular or genemJ. A particular supposition refers to a definite act or to several
had had the power, he would have helped me, if he shall receh;e it (or if he receives it), he will give it; if he should 1eceive it, he would give it. So if he always acts justly (or if he never c01nmits injustice), I h01wu1 him; if he acted justly on all these occasions, he will be rewarded.
A general supposition refers indefinitely to any act or acts of a given claRs which may be supposed to occur or to have occurred at any time; as if ever he receives anything, he (always) definite acts, supposed to occur at some definite time (or times); as if he (now) has this, he will give it; if he had it, he gave it, if he

gives it, if ever he received anything, he (always) gave it, if he had (on any occasion) had the power, he umtld (always) have helped 111e; if ever any one shall (or should) wish to go, he will (or would) always be pe1mitted. So if he ever acts justly, I (always) honour him; ij he ever acted justly, he was (always) rewarded.

395. Although this distinction is seen in all classes of conditions, present, past, and future (as the examples show), it is only in present and past conditions which do not imply nonfulfilment (i.e. in those of 390, 1) that the Greek distinguishes general from particular suppositions in construction. Here, however, we have two classes of conditions which contain only general suppositions. (a) \Vhen the apodosis has a verb of present time expressing a customary or repeated action, the protasis may refer (in a general way) to any act or acts of a given class which may be supposed to occur at any time within the period represented in English as present. Thus we m(ty say:'Eav ns KAE'Tf'T'[], KoAaCTat, if (ever) any one steals, he is (in all such cases) punished; iav ns 7rpauuv (or 7rpagv) TowvT6v n, xa"AE7ravoJLEV ailT0, if (ever) any one does such a thing, we are (always) angry . with him; lav ns To-6Tov 1f'T/, &.7roBvjuKn, if any one (ever) drinks of
this, he dies.
(See 462.)




(b) When the apodosis has a verb of past time expressing a customary or repeated action, the protasis may refer (in a general way) to any act or acts of a given class which may be supposed to have occurred at any time in the past. Thus we may say : Ei' n> KA~7rTot, iKoAcf.{eTo, if (ever) any one stole, he was (in all such cases) punished; ei' Tt> 7rpcf.fTIFOt (or 7rpcf.~ete) 'TOWVT6v n, ixaAe7ravop.ev a'llT<e, if (ever) any one did such a thing, we were (always) angry with him j t. TL> 'TOVTOV 7rot, a7rt8vlJfTKEY, if any one (ever) drank of this, he died. (See 46 2.) 396. Although the Latin sometimes agrees with the Greek in distinguishing general conditions from ordinary present and past conditions, using si faciat and si faceret in a general sense, like iav 1rpcf.fTIFlJ and el 1rpafTfTOt above, it yet commonly agrees with the English in not recognising the distinction, and uses the indicative alike in both classes. Even the Greek sometimes (especially in poetry) neglects the distinction, and uses the indicative in these general conditions (467). 397. In external form the general present condition coincides with the more vivid future condition, 392, 1, as both are expressed by Mv and the subjunctive, the form of the apodosis alone distinguishing them. But in sense there is a much closer connexion between the general present condition and the ordinary present condition expressed by el and the present indicative, 390, 1, with which in most languages (and sometimes even in Greek) it coincides also in form (see 396). On the other hand, Uv with the subjunctive in a future condition agrees substantially in sense with el and the future indicative (447), and is never interchangeable with el and the present indicative.




398. It is impossible to discuss intelligently the origin of the conditional sentence until the etymology and original meaning of the particles el, al, av, and KE are determined. On these questions we have as yet little or no real knowledge. The theory of et or al which identifies it with the pronominal stem sva (fTfe), Oscan svai, and Latin si, is perhaps the most common. By this the original meaning of el, or rather of one of its remote ancestors in some primitive language, would be at a certain time (or place), in a certain way.l But, even on this theory, we can hardly imagine any form of el as existing in the fffeek language until the word had passed at least into the relative stage, with the force of at which time (or place), in which way, under which circumstances. It cannot be denied that the strong analogy
I See, Oonj. u. Opt., pp. 70, 71, who terms this a "wahrschein. liche positive Vermuthung."




between conditional and relative sentences and the identity of most of their forms give great support to any theory by which the conditional sentence is explained as an outgrowth of the relative, so that the conditional relative sentence is made the original conditional construction. Thus cl ~A.lhv might at some time have meant in the case in which he 1vent, and d f.)..Bv, in the case in which he shall go (or in 1 case he shall go), etc. But here we are on purely theoretical ground ; and we must content ourselves practically with the fact, that in the earliest Greek known to us ,d was fully i:lStablished in its conditional sense, like our if and Latin si. 399. The regular types of the conditional sentence, which are given above (390-395) as they appear in Attic prose, have been mainly sifted from a rich variety of forms which are found in earlier Greek. In Homer we have all tenses of the indicative used as in Attic Greek, except that the imperfect has not yet come to express an unreal present condition, but is still confined to the past. The future indicative sometimes has K& in protasis, and the future with KE or av can stand in apodosis. The subjunctive in protasis can have Ef K (even E av), lJV1 Or El alone j and it can stand in a future apodosis either alone or with av or KE (like the optative). The optative sometimes has d Ke in protasis, and occasionally stands in apodosis without (;.v or KE. Once we find d KE with the aorist indicative (Il. xxiii. 526). Thus, while we have in Attic prose two stereotyped forms of future qonditional sentences, Hw (~v, :iv) 8f, EAOVfJ-L and El oo{7J, EAO[JLYJV av, we have in Homer ~V of, d KE ocf, d o</l, and El OOtYJ, t K oo!YJ, in protasis ; and JA.ov}Lat, EAOVJLa[ K, V . wJLaL, EAWJLa K, and EAOtjL'f}V K (or av), rarely EAO[JLYJV aloni', in apodosis; with every variety of combination of these. (For the details and examples, see 450-454 and 460.)

400. There is a tendency in Homer to restrict the subjunctive with simple El (without I(E or av) to general conditions (468), and a similar but less decided tendency to restrict the subjunctive with conditional relatives without Ke or iiv to the generic relative construction (538). But the general condition with el appears in Homer in a primitive stage, compared with the corresponding relative construction, which is fully developed. Both subjunctive and optative are freely used in general relative conditions in Homer, as in Attic Greek; while in general conditions with ~;l the subjunctive occurs only nineteen times and the optative only once (468). On the supposition that the clause with El is derived from the relative clause, this would appear as the ordinary process of development.




401. It is perhaps the most natural view of the various conditional expressions, El, .r Kt, l iiv, etc. to suppose that at some early stage the Greek had two perfectly analogous forms iu future conditions, one with two subjunctives, and one with two optatives, e.g. ei B.p Towo, v. wpa~ and tl oo['l} TOVTO, V..ofp?JV. The particle KE would then begin to be allowed in both of these conditions aqd conclusions, giving to each more distinctly its force as a protasis or an apodosis. 1 It would thus be allowed to say .r K o<tJ TOVTO, V...wp.a[ KE and EZ KE oo1] TOVTO, V...o[p1JV KE, both of which forms actually occur in Homer. Gradually the tendencies of the language restricted the use of K more and more to the subjunctive in protasis and the optative in apodosis, although for a time the usage was not strict. This state of transition appears in Homer, who preserves even a case of an otherwise extinct use of er K< with the aorist indicative. Shortly befote this stage, however, a new tendency was making itself felt, to distinguish the present general condition from the partic.mlar in form, the way being already marked out by the conditional relative sentence. As this new expression was to be distinguished from both the really present condition El {3ovA<Tat and the future d K< (3ovA.rwu, the half-way form El {3ovAYJTa~ (which had nearly given place to Er K< {3ovA?JTa~ in future conditions) came into use in the sense if he ever wishes. 2 This would soon. develop a corresponding form for use after past tenses, el {3ovA.o~To, if he ever wished, of which we see only the first step in Homer,.Il. xxiv. 768. (See 468.) It would hardly he possible to keep the two uses of d with the subjunctive distinct in form, and in time the form with KE (or av) was established in both (381 ). But we see this process too in transition in Homer, where et K Or SOme form Of El aV is USed in all future COnditionS except nine, and has intruded itself into five of the nineteen general conditions. We must suppose a corresponding process in regard to KE or ({v in conditional relative clauses to have gone on before the Homeric period, with more complete results. 3 In Attic Greek, except iu a few poetic passages, the usage is firmly
1 As I do not profess to have any oistinct theory of tl1e origin or the original meaning of either KE or av, I have not attempted to deJiue their force, except so Jar as they emphasise what we see by usage may be implied by the sentence without their aid. 2 Monro (Hom. Gr. p. 263) tlinks "the primary use of l1v or KEv is to show that the speaker is thinking of pa,tic>tla? iustames or occasious." If this is so, we should expect these particles to be first used iu future couditions, while the later general conditions would first take th.e simple .Z, as is here supposed. 3 See A1n. Jo>tr. Phil. iii. pp. 441, 442, where Gildersleeve refers to the use of el, iirE, etc. with the optative in oratio obliqua, representing Uv, lirav, etc. with the subjunctive in the direct form, as evidence of an old use of fl, lire, etc. with the subjunctive.




established by which the subjuncti1e in protasis requires av in both particular and general conditions.



1. Sirnple Suppositions (chiefly Particular).

402. When the protasis sirnply states a present or past particular supposition, implying nothing as to the fulfilment of the condition, it takes a present or past tense of the indicative with d. Any form of the verb may stand iu the apodosis to express the result if the condition is or was fulfilled. E.g. El i(3p6v-rrJIT<, Kal {j~T-rpatfv, if it thundered, it also lightened.
(This implies no opinion of the speaker as to the reality of the thunder.) El o' OVTW TOVT E<rTlv, Ef:J-Ol Jl-fAA ,PLA.ov lvat. 11. i. 564. El TOT Kovpo> ea, vvv av-r f- y~pa> (hra(Et. Il. iv. 321.

El pAA.a Kapnp6s E<r<rt, (h6> 1rov <roi -r6 y eowK<V. Il. i. 178. El 8~ X p-I] Kal 7rdp U'ocpuv dvn,Pepga.t, pf.w, but if I must match myselj against the wise one, I will speak. Pmn. Py. ix. 54. El e.o TL 8pw<rtv ai<rxp'Dv, oBK d (T l V ew, if Gods do aught tliat is base, they a?e not Gods. EuR. Bell. Fr. 294. El yw if.>at:Opov dyvow, Kat EfkavTov E'lrtA.EA.'f/<' &A..\a yd.p o-&Bf.-repa E<rn -rov-rwv, if I do 1wt know Phaedrus, I have forgotten myself; but neither of these is the case. PLAT. Phaedr. 228 A. El Jk~V (' AU'KAT)'IrtbS) ewu ~v, OVK i]v aiU'xpoK<p'ln)>' .l 8' al<rxpo!po;)>, OllK ;)v ewv. Icl. Rep. 408 C. El 8 iKeivo<; aU'8VEITTEpO> ;)v, Emmf.i -rov 1ra8ov> afnov ~yf;<ra-ro. DEM. xxiii. 54. 403. The imperative, the subjunctive in exhortations or pro-

hibitions, the opkttive in wishes, the potential optative or indicative with av, or the infinitive may stand in the apodosis. E.g.
'AA.X <looK<< U'ot, U'-rdx, if thou art resolved, go. SoPH. Ant. 98. (Here av 8oK'fi would refer to the future, while El ooKtZ is ~trictly present in its time. Cf. Ant. 76.) 'A).X el ooKEi, 1r A. WJJ-EV, &pJJ-aU'e w TUX1J'), Icl. Ph. 526. El JJ-EV r<TTE jt TOLOVTov, . f:J-?)OE q>wv;)v aJ!a<rXTJ<T8t. DE~f. xviii. l 0. 'AA.A' 'lrOV 'lrTWXWV y< 8w2 Kat Jpwu., ei<r2v, 'Av-rlvoov 1rpu yaJJ-oto TEAo> Bava-row KtXd'f/. Od. xvii. 475. 'A.\A' El 8oK<t U'ot -rave', ,',rra ns dp(3vA.a<; A..)ot -raxos, b-u.t ijth~:s pleases you, let so'!M one qu.icklyloose my shoes. AESCH. Ag. 944. KaKw-r' hoA.ojt'l}V, !Eiav8iav el Jt?J .j>tA.w. AR. Ran. 579. Ilo.\.A.;) yap &:v EVDa~JLov[a Err; Trtpi ToiJs v(ovs, d ~rs fk'i.v p..Ovos a-DToVs Ota<-pOEipt:i. 0~ 8' aAAOL d:,,P<AOV<TLV. PLAT. Ap. 25 B. See also Il. vi. 128, d . elA.;)A.ovBa>, oBK &v p. a X o tJk 'fJ v. Tuv 'Y7r<pd8rJ1', d1rep





d.A:l]O~ JLOV vvv Karrrrop~ J-taAAov liv lK6rws ~ r6v8' i8twKv,

if he

is now bringing true charges against me, he would have prosecuted Rypereides
with much more reason than he does this man. DEM. xviii. 223. (See 479, 2; 503.) 404. This form of condition may be used even when the supposition is notoriously contrary to fact, if the speaker does not wish to imply this by the construction; as in DEM. xviii. 12, rwv JLEVTOL Karrrroptwv, L7rp 1]o-av a:Ar/Nis, oilK Evt rii 'IT6At 8{KTJV :Aa.j3'iv, but if the charges were true ( = emnt, not essent), the state cannot obtain adequate satiifadion. So in English, we can say if three times six are twenty as well as if three times six we1e twenty, or if all men are liars as well as if all men were liars,-from different points of view.

405. A present or past general supposition is sometimes expressed by the indicative: see examples in 467. Here the Greek neglects the distinction which it regularly makes between general and particular suppositions of this class. 406. Pindar uses these simple conditions with l and the indicative
more than all other forms. 1 But among his forty-eight cases are many general conditions (467), which most writers would have expressed by the subjunctive.

407. (Future Indicative in P1esent Suppositions.) Even the future indicative with d may be used in a present condition, if it expresses merely a present intention or necessity that something shall be done hereafter; as when d rowo rrwqo-ot means if he is (now) about to do this, and not (as it does in an ordinary future c,ondition) if he shalt do this (hereafter). E.g. Atp 11'A~Krpov, d fka.x<Z, 1aise your spur, if you are going to fight. AR. Av. 759. (El /ka.X1: in protasis commonly means if you shall fight, like eav Jl-UXYJ) "'H vvv ey<i> }-t~V oilK rlv~p, a:VTYJ o' av~p, d Ta.vr dva.rt r1]8e Keto-Ta.t Kprin], i.e. if this is to pass unpunished. SoPH. Ant. 484. T Ota<f>povcn TWV E~ avayKYjS Ka.Korra.8ovvrwv, d ')'E 'ITHv~o-ovo-t Kat OLt{~rrovr:rt Kat ptyr!>o-ovrrt Ka~ aypv'lTV~CTOlJCTt Kat rd.A.A.a. 'lTavm J-tox81]{]'ovrrtv eK6vr<s; how do they diJfer, etc., if they are to suffer h~m/!m, thint, etc. 1 XEN. M em. ii. l, l 7. So d rr6J.p.o<; T 8afki Kat AoLfkOS 'Axa.tovs, if both war and pestilence are to lay the Achaeans low, Il. i 61 ; and El 8ta.f3AYJ0~rJ'Oj-taL, if I arn to be slandered, EuR. Hec. 863. In 11. v. 715, 1) P' aA.tov TOV p/v8ov V11'EO"'TYJfkV MEvAa<p, d OVTW j-ta.lvo-8a.L eao-op.Ev oi'i:Aov 'Ap~a., vain is the word we pledged, if we are to 1Jennit, etc., the verb of the apodosis is past, showing that the condition is not futl1!'e. 408. It is important to notice that a future indicative of this kind could not be changed to a subjunctive with Mv without an entire change of sense and time. It must therefore be distinguished from the future in future conditions, where it is generally interchangeable with

See Am.

Jour. Phil. iii. p. 438.




the subjunctive (447). Here it is nearly equivalent to the periphrastic future expressed by p..EAA(l) and the infinitive (73), in which the tense of Jl-EAAw (as in d J1-AAo1XF TOVTO 7rodv = l 'TOV'TO 'lrOt~uovuw) shows that the condition is re\t].ly present and not future. So with the Latin periphrastic future, si hoc facturus est. 409. A present condition may be expressed by a potential optative in the protasis, and a present or past condition by a potential indicative; as ft7rp aAAcp Ttp &.vBpt!J7rwv 7T8owYJV &v, Kal uol?rd8o}J-a, if (it is true that) I would trust any one of mankind, I trust you, PLAT. Prot. 329 B; El 'TOVTO luxvp?w 'ljv &v TO~Ttp TKp.~pOv, Kdp.o1 yEvEuBw TEKJ.~pOv, K.T.A., if (it is true that) this would have been a strong proof for him (if he had used it), so let it be a proof for me, that, etc., DEM. xlix. 58. (See 458, and other examples in 506.)


2. With Supposition contrary to Fact.

410. When the protasis states a present or past supposition, implying that the condition is not or was not fulfilled, and the apodosis expresses what would be (or would have been) the result if that condition were (or had been) fulfilled, the past tenses of the indicative are used in both protasis and apodosis, and the apodosis contains the adverb liv. The imperfect here, in either protasis or apodosis, refers to present time or to an act as going on or repeated in past time, the aorist to a simple occurrence in past time, and the (rare) pluperfect to an act completed in past or present time. E.g.
El Tovro 1rpauuE, KaAws &v EtXEV, if he were (now) doing this, This may also mean if he had been doing this, it would have been well (implying that he was not doing it). The context must decide, in each case, to which time the imperfect refers. El TOVTO E7rpa~E, &v ~<TXEv, if he had done this, it would have been well (implying that he did not do it). El TovTo ~7T'E7rpttxH, KaAws &v ElXEv, if he had finished doing this (rww or at any past time), it would be well (implying either he has not or he had not finished it). (Impf of P1esent Time.) El OE p.' ~8' dd A.6yovs E~~p X Es, ovK &v .fjuBa A.vmjpii KAvnv, if you always began your talk to me in this way, you would not be offensive to listen to (as you are). SoPH. El. 556. So El. 992, 1331, 0. T. 1511; and AESCH. Sept. 662, Ag. 1395. Kat vvv l cpo(3ep6v T JvwpwfuV, 1rav av <TO 7rpoEcppa(op.EV, if we saw any cause of alarm, we should tell it all to you. HnT. i. 120. TavTa otJJ< UV EOVVaVTO 1T"OGV, l f.L~ Ka1 0aTll fETp{'f EXPWVTO, they WOUld rwt be able to do this, if they did not lead an abstemious life. XEN. Cyr. i. 2,

it would be well (implying that he is not doing it).




16. Ev L!J'8' on.;: n Ef'-OV EK~Oov, ol,0110) &11 OV'TW f'- d7rO!J''T(ptV icpvA.arrov W$ a~u:)f'-a'TO<; Kat 'Ttf'-~'>, if you ca?ed for me at all, you would take precaution, etc. Ib. v. 5, 34. IIoAv &11 8avf'-a!J'rbc -rt:poll ~11, L ETLf'-WVTo, it 'would be much more wondmful, if they we1e honoured. PLAT. Rep. 489 B. Ayov!J't mi11-ra fj f.xn Karot d f'-1J .1n1yxall11 av-rol:s E7r(J'TfJf'-YJ El'Ovua, ol,K llll oTo -r' ~(J'CI.II TOVTO 1rot~a-nv, they tell everything as it is: and yet if knowledge did not chance to be in thmn, they could not do this. Id. Phaed. 73 A. Oi'X OVTW o' av 7rpo8Vf'-WS E7rt 7'0117r6Af'-0l' vpJis 7rapKaAoVII, Lf'-1J T~V .:lpfJvYJV wpw11 (J.L(J'xpall EU"Of'-EIIYJV, I should not exhort you, did I not see (as I do), etc. Isoc. vi. 87. (Impf. of Past Time.) Kat -ravr' av ODK ~7rpCI.(J'(J'OII, L P-0 f'-0 1ftKpcw a0r0 r' &.pas 1)paro, and this I should neve,. have done, had he not invoked bitter curses on myself. SoPH. 0. C. 951. 0DK a11 v1)!J'wv eKpa-rn, d f'-17 n Kat 11avnKov t:ix<v, he would not have been master of islands, if he had not had aklo some naval force (implying 11avnKov lX<II ancl v~!J'wv f.Kpart:t, he luul a na.vy, for he was mastm of islands). THUC. i. 9. (Tavra) OVK av 7rpoU.. y11, d 1'-~ e7rfTTV11 dA.?]e.V<T11, he would not have declared these things (referring to several), had he not been confident that he should speak the truth. XEN. Mem. i. 1, 5. El ~a- a 11 &vop<s &ya8ot, ws uil <P?)>, ovK llv 7rOT< -rav-ra l1rau X o 11, if they had been good men, as you say, they would never have sujfeTed these things (referring to several cases). PLAT. Gorg. 516 E. (Aorist of Past Time.) Eli'-'J opKOLS iJp81')v, oDK U117ror' li<TXOII P-0 ov -ra8' ~n1re'iv 1rarp, had I not been bound by oaths, I should never have nifmined, etc. EUH. Hipp. 657. Ka Zo-w> av 8ta rav-r' a7r8avov, .:1 tL0 ~ &px~l Oid raxiwv KCI.TEA,;e'l'}. PLAT. Ap. 32 D. T 1foT' li.v E7ra8<;v V1f' avTwv, EL 7rAdw xp6vov E1f'Tp07rVe1')1'; 1 KTEA.ecp&YJVf'-EV EViavuws, ~~ E7'1) 8 7rp'o<T1f'Tpo7rd&1]11 V7r' avrwv, ol,o' av TQ f'-tKpa TO.VTa 7rp' avrwv d7rEAa(3ov. DEM. xxvii. 63. El -rovvv 6 <{>[Ai7r1f'OS 7'07' TaVTYfJ! E(J' X -r0v YJ!Wf'-YJV, OVOEJI av 6lv vvvl 7r1T'O[YJKEV if1rpa~E11 1 ol,8 TO(J"O.VTYJV (K-rf]!J'a-ro OVVO.f'-V. Id. iv. 5. (DijfeTent tenses in Protasis and Ap01losis.) El f'-1J vpt:l:<; 1)A.8en, E7ropev6P-t:&a &v E7rt (3aa-tAa, if you, had not come, we should (now) be on O'Ur way to the King. XEN. An. ii.1, 4. ''0 t:l d7rKp[J!W, iKaJIW<; av l}SYJ 1rapcl O'ov r>)v oa-tbT?)Ta Ef'-Ef'-a8fJKYJ, if you had given this answm, I should have alTeady leamed, etc. PLAT. Euthyph. 14 C. Aotm)v 8' av ?j 11 ~jpZv E'T 7rEpt -ri)> 7rbAEW$ 8w.A..x81JIIaL -rijs 0f'-ETEpa>, el f'-1) 7rpo-repa TWV aAAwv T~V <ip1)V1')V E7r1f0 [?)TO. (This implies aAAa T?JV dpf]vY)v 1rporpa 7r<1T'o[1')rat.) Isoc. v. 56. El yap l:K -rov 7rapeA"Y)Av86ros xpovov 7'd OEOVTa OVTO uvvt:(3ovAV<Tall, oiJo'Ev UJI vp.l2s vvv f!oet (3ovA.eveu8at, if they had given the necessary advice in time past, the1e would now be no need of your deliberatiny. DEM. iv. 1. Twv d8tK'Y}I'-a-rwv Ull Efl-E/L111]7'0 TWV avrov, eZ TL 7rpt Ef'-OV y' eypcpv. Id. xviii. 79. These examples show the fully developed construction, as it appears in the Attic writers and in Heroclotus. For the more primitive Homeric usage, see 435 and 438.




411. This construction is equivalent to that of the Latin imIJerfect and pluperfect subjunctive in protasis and apodosis. With regard to the tenses, the L~tin imperfect subjunctive represents the . Greek imperfect indicative referring to present time, and rarely that referring to past time; while the Latin plupPrfect subjunctive represents the Greek aorist and pluperfect indicative, and also most cases of the Greek imperfect referring to past time.

412. L It will be seen that, when this construction is used, it is usually implied not merely that the condition of the protasis is not (or was not) fulfilled lmt also that the action of the apodosis does not (or did not) take place ; thus d TOVTo i7roJI, E7f[<rfJYJ aJI, if I had said this, he would have been tJersuaded, generally implies not merely that I dirl not say this hut also that he was not persuaded. But this denial of the apodosis is not an essential character of the construction, as we eau see if we change the apodosis to o-GK &J1 E7r[<rfJYJ, he wmdd not have been persuaded, when it is not implied that he ?'eally was persuaded. We have seen that there is notbing in the nature of the potential indicative which makes a denial of it,; action necessary (244); and when this form is made the apodosis of an unrenl condition, it simply states that something would happen (or would lune happened) in a case which di<l not arise. Denial of the apodosis can follow as a logical inference from denial of tbe protasis only in the rare cases in which the unreal condition is the only one under which the action of the apodosis could have taken place, as when v.-e say if the moon had entend the euTth's .shadow, she woulrl have been eclipsed, where the denial of either clause canies with it by necessity the denial of the other. But if we say if it had rained, the g1'mmd ~could be wet, the denial of the protasis cuts off only one of many conditions under which the ground might be wet. Such sentences as this are, however, very common, though they are not used to prove the opposite of the apodosis (ti1at the ground is not wet) ; but they are arguments in whkh the apoclosis is assumed to be false (on the ground of observation or experience), and from this it is aTgued tllat the assumption of the protasis is false ; tl1at is, since the g1ound is not wet (as we can see), it cannot have mined, which is a good argument. This is the case in THUC. i. 9, and PLA'l'. Gorg. 516 E (quoted in 410, al1ove); where it is argued that Agamemnon had rt na.vy because this was a necessary condition of his ruling islands, aud that ce1'iain persons were not good m.en because they suft'ereu what they did, the facts of ruling islands and of suffering Leing assumed in the argument as establishetl on independent evidence. In other cases, where it is stated that the apodosis would follow as a consequence from the fulfilment of the condition, as in SOPH. Aj. 45, Kllv E~E7rpa~a.T' El KUT1Jil-~A1Jrr' ytil, he would even have accomplished it, if I had been canless, whatever negation of the apoclosis is im11lied (here ovK ~~E7rpa~a-ro) comeB from a feeling that when the only condition under which it is stated that an action would have taken ])lace fails, there is no reason for believing it to have taken place at all. We may. doubt whether ally




negation of the apodosis is implied in the form of expression in such cases. Certainly, in many cases in which the apodosis states a consequence which would follow from the action of an unreal protasis, this negation is assumed as already known apart from the construction; thus in SoPH. El. 556 (quoted in 410) the apodosis means you would not then be offensive to listen to, and the only ground on which we mentally add as you now are is our knowledge of Clytaemnestra's feeling towards Electra. If the sentence were if all men began their speeches politely, they would not be offensive, we should not think of supplying as they now are without some knowledge of the facts. 2. When the sentence merely affirms or denies that one act, if it had occurred, would be accompanied by another act, and there is no neces~ary relation between the two acts as cause and effect, and there is no argument drawn from the admitted unreality of the conclusion to prove the opposite of the condition, no denial of the apodosis is implied in the expression, although we may know from the context or in some other way that the action of the apodosis does not (or did not) occur. Thus in PLAT. Ap. 17 D, l -rcj) Svn ~vos f:-r{;yxavov <Ov, ~vvytyvw UKn o07!"ov aJI p.ot d f:v lKlV?J -rfj t:{>wvfj EAyov, etc., if I were 1"ea.lly <!> foreigner, you would swely pardon me if I spoke in my own dialect, etc., it is not ilU})lied that now you do not. pa~don me. We should rather say that nothing at all is implied beyond the statement you would pardon me in that case. If the apodosis were you would not be angry with me, the impossibility of understanding but now you are angry would make this plainer. Again, in XEN. An. vi. I, 32, ovo' &v eywy<{ov <l &.\..\.ov ei:A.w8, neither shQuld I (auymore than Xenophon) be quarrelsome if you had chosen another man, nothing like uracrui'w is implied ; . on the other hand, any such implication as ov (J"Tauta{w must come from the circumstances' of the case, not from the form of expression. In SOPH. 0. T. 220, ov ylip &v p.aKpdv tXVVOV avr6s, if the protasis is d Zxvwov avr6s, if I were undertaking the search by myself (alone), the apodosis I should not be very far on the track does not imply p.aKpliv lxvvw, or anything more than the sentence states. (See 511.) Again, in SoPH. Tr. 896, l 7!"apovua 71".\.Y}uta EAV(J"(J"S oi' eopaO"E, Kapr' av 4\KnU"aS, the statement does not imply ODK ~KTtfTas, although this may be true. 3. Further, in concessive sentences introduced by Kai d or l, even if or although, or ovPf d, not even ij, where it is stated that something would be true even in a supposed case (which does not arise), we have what amounts to a statement that the thing in question would be true in any case. Here, therefore, the actiou of the apodosis is distinctly affirmed; as in Isoc. xxi. ll, NtK[as p)w, el Kat T0v a.\.\.ov xpovov dthu-ro fTVKot:{>avTtv, r6r' av E11"aVU"a"TO" Ev8vvovs oe, Kat l f-'YJOE 71"W71"0T OtevofJ8YJ d8tKtV, r6r' av i71"~ p 81], i.e. N. would then have slopped, while E. would have been urged on, in any case. So DEM. xxx. avrO. a.v E11"pag Kat 11"pWTYJ 14, and xl 23. See PLAT. Rep. 620 D, AaxovO"a (=Kat El 'Trpw-r"l ~Aa)(6Y), it would have done the same even if it had drawn the first choice.





413. In the unreal conditional sentence, therefore, the unreality of the supposition is alwAys implied, and that of the apodosis is geneTally either assumed or implied. The implied opposite of an imperfect is always a present or imperfect, that of an aorist is an aorist, and that of a pluperfect is usually a perfect or pluperfect. Thus El ;7rparr(TE, when it means if he were doing, implies &.>..>.: o-& 1rp&o-(Tn, but really he is not doing j when it means if he had been doing, it implies &.>..>.: o-&K ;7rpa(T(TE, but really he was not doing: El p.~ ;7rpa~Ev, if he had not done, implies &.>..>.: e1rpa~EV, but really he did do : El ~7rE7ro~KE TovTo, if he had already done this, implies either &.>..>.: o-& 1fE7ro1JKEv, but really he has not done it, or &.>..>.: o-&K ~7fE7fo~KE, but really he had not done it, according to the context. The aorist, however, is very often used here, as elsewhere, where the pluperfect would express the time intended more exactly; as in the sentence quoted in 410 from DEM. iv. 5, olloev <iv illv vVIIt 1fE7ro[1JKEv ;?rpa~Ev, where the perfect 7rE7ro1JKEV shows that the pluperfect might have been used for e1rpa~Ev (see 58). 414. Sometimes an aorist not referring to past time is found in the apodosis, after a protasis in the imperfect referring to the present. This occurs chiefly in Plato, and generally with Et1rov iiv, d7rEKpw&p.1Jv iiv, or a similar verb, meaning I should at once reply. The aorist excludes the idea of duration which the imperfect would express, and for the same reason it cannot be strictly present ; in effect it does not differ much from an aorist optative with iiv, the apodosis really being the result (in the case supposed) would be (~v iiv) that I should reply (d1rop. av), etc. E.g. El p.ev ovv (T{; JLE ~pwTa<; T nov vvv 8~, El1rov <iv, K.T.A., if then you were asking me any one of the questions before us, I should (at once) say, etc. PLAT. Euthyph. 12 D. "il(T1fP ltv El tT-6yxavv t:Jv V1f001Jp.&Twv 01Jp.wvpyo<>, d7rEKpvaTo liv o~ 1ro-6 (TO (TKVToT6fl-o'>, as, if he chanced to be a maker of shoes, he would answer that he ll!as a cobbler. Id. Gorg. 447 D. See also PLAT. Symp. 199 D, Men. 72 B, Theag. 123 B; .ANT. Tetr. .A. (3. 13. In PLAT. Prot. 311 B, C, we have f T[<; (TE ~PETo, T[ ltv d7rEKpvw; with the answer El1rov <iv ~>, K.T.A., twice, referring to present time; but in D, El ovv T> ~f-a> epoTo (future), followed by Tt ltv avT!f d7roKptva{fl-f{ja; An example of this is found in SOPH .Ant. 7 55 : El /1-~ 1ra-r~p ~(T8', El7rov av (T) OVK EV cppovE'iv, if you We1e not my father, I should say you were not right in mind. See EuR. Ale. 125, ~A8Ev av, ie. (the result would be that) she would return. So .Ale. 360.



av.____:'Eon, XP~II,


415. A peculiar form of potential indicative without dv consists of an infinitive depending on the imperfect of a verb of




obligation, propriety, or possibility, like on, XP~v or EXP1JV, elK os 1jv, or r.pocr~Kev. This expression refers to past or present time, and generally implies a denial of the action of the infinitive. Thus EOE -rovrov &rro&av<i:v in this iuiomatic use means he ought to have perished (but did not) ; ~8~t 1JJ-Lfis -rouro r.oteL'v means we ought to be doing this (but we are not) or we ought to have done this (but we did not do it). This combination contains in other words what might have been expressed substantially by a past indicative with llv of the verb of the infinitive, qualified by an adverb or other expression denoting obligation, propriety, or possibility: thus 8et rovrov d:rro8av~i:v is (as a construction) eqni valent to oi'ITos OcKa[ws (or d~[ws) &v d:rrEBavev, he would justly have pmished, and <lKd> ?JV erE 'TOVTO r.aB~L'v is equivalent to TOVTO elJ<6rwr; av,, you would properly have suffered this (implying otiK f.7ra.8e,). Strictly, the expression involves also an unreal protasis, as (in the last case) el Td EiK6s E7ra8es, which with the apodosis -rovro r.a&e, ilv appears substantially in elK6s o}v erE rovro 1ra6~i:v. (See 511.) \Vhen the present infinitive is nsed, the expression is present or past ; with the aorist infinitive it is always past. 416. The following imperfects may take the infinitive in this sense: ~o~t, XP'lv or Jxp~v, EiKos 1)v, 1rpocr~K~v, (v~v, -~~v, >jv (or v1r~PXEV 1 ),d was possible, one might, the impersonal 1}v with adjectives or nouns expressing obligation, propriety, possibility, and similar ideas, as 8Kawv ?JV, a~wv ~v, KaA6v (KuAAtov, 2 Kpe'i:Trov,3 Kpan4 5 CTTOV ) ?)v, alcrxp'Ov ~v, 1rpo~Kov 1}v, o~ 1)v, d<.r<j>uAecrTEpov ijv,6 i:a-ov 1}v/ e1!Aoyov ijv,8 a-uyy\'w<.rTdV 1jv, oT6r; r' 1jv, ipyov >}v, 1)v with the verbal in -reo>,-also e71"pE7rcl', cruvcpep~l', 9 EAvcrtrEAEt}0 with other verhs of the same nature. To these mnst be added the expressions specially mentioned below in 424-431. 417. The~e are all originally expressions of past necessity, obligation, etc., involving no reference to any condition (mlfulfilled or otherwise); and in this sense they may always be used, as in DEliL xix. 124, ~/Sa JJ-Emv, he was obliged to stay (and did stay), and HDT. i. 8, xp~v /'UP KavoavA.v /'Evecr6at KaKwr;, for C. was doomed to fall into trouble. It is only by idiomatic usage that the denial of the action of the infinitive comes to be implied in them, and that a past tense comes to express present time, both of which characteristics are found in Greek, Latin, and English ; as ~8~t <.re avrov cpLAAv, debebas eum colere, 'lJ016 ought to love him (but
ANT. v. 13. DEM. lix. 112. The imperfecta not included in these references will be found among the examples in 419-422. The above list could doubtless be greatly extended.

1 See Isoc. v. 34. 3 DEM. xx. 23.

2 IsAE. ii. 15; A:arsTo-r: Etl1. x. 9, 18 (p. 1181 a, 4). 6 LYs. vii. 24. Isoc. xx. 14. 5 DE~L xviii. 248. s AnrsTOT. Eth. x. 9, 19 (p. 1181 a, 6). 9 LYs. xiii. 28.






you do not), ought being the past of owe. The infinitive is felt to be negatived, even when the negative belongs to the leading verb.

418. Like the potential indicative, this form of expression can either (1) be used alone, with no external protasis expressed or distinctly implied, as in xpfjv (T i..\e,zv, you ought to have gone; or (2) stand as apodosis to an unreal protasis, as in d iKeAwa-, xp~v a- i..\8,zv, if he had commanded it, you ought to have gone.

419. I. When these expressions are thus used alone (418), the denial of the action of the infinitive is always implied. E.g. T015a-OE yap JL Yj ( 0v oo, fm these ought not to [,e living (as they
are). SOPH. Ph. 418. "E 0 t JLEV TOll<; ..\yovTus a:;ravTa<; JL1JTE 7rp'Us x8pav 'TfOtEt(r()at ..\6yov JL>JOEJ'a JL>JTE trp'Us xaptv, i.e. the speakers ought not to say a wmd out of regard either to enmity or to favour (and yet they do so). DE;f. viii. 1. '2:ty>ja-a<; 1)vi/ oo .AyEtv, keeping silence when he 01ght to speak. Id. xviii. 1b9: cf. xviii. 191. Xpfjv yup (TE JL'IJT avT6v 'TfOT' EL<; Tpo[av jLOAtJI, ~JLOS T' d7rdp{'Etv, for you ought ?JOUrse/j nerer to have gone to 'f'Toy, aud you ought (now) to keep me away f?'om it. Sorr-r. Ph. 136:3. See AE<;CH. Ag. 879, Oho. 930; :SoPH. El. 1505. 6aVEZv, eavLV a-, 7rpeu(3v, XP~V trapo<; rKvw;;. EuH. And. 1208. T ixp~v JL' 7f0ttl'; JL1J 7rpoa-ayEw ypafat (Tovs 7rpeo-(3H>); what ought I to have done (u;hich 1 did not do)? Ought I not to have proposed (as I did) to invite the arnbassadors? DH:M. xviii. 28. 'Exp~v JLEl' ovv Kal otKa.wv 1JV TOV<;; TOV UTE<j>avov


KaK<u> AEy<tv E7ft01) oE TOVTo 7rapVT'> EKHvo 'Tfotovutv, K.T.A., J.e. those who think they 01ight to receive the crown ought to show that they dese?ve it themselves, and not ve auus'ing me; but since now they have neglected the fanner nn<i do the latter, etc. Id . .li. 3. L;q)<pv 8', iJi TcKv', oils p.f.v ELK<h 1)v 7fOVtV TUOE, those of you who ought to be bearing these lauoun. SoPH. 0. C. 342. Tip~> T015Tov<; Tov dywl'U ICU.TUUT?Jl'U.t, ov<; ELKO<; 1ll' np JLEl' n8VWTt TtJLW[JOD<; YEVE(J" eat T<'; 8' E7f~t01'Tt f3oYJ8015<;, ?rho properly should have come fonvard to avenge the dead and to help the prosecutor. A;sT. i. 2. El il'TfO Twv 7fOAJL[wv JLEl' Juwel)JLEV, ov<; dKb<; ?JV OtaKWADtV JL?J (T(gEueai, i.e. who would natumlly have tried to Jwevent us frorn being sa'ved. LYs. xx. :36. See DEM. xl. :30. Kal JLaAuna ElK~> 1)v VJLUS trpoopaa-fJa,, avTa Kal JL1J JLUAaKw<;, WU'TfE(l vvv, ~VJLJLUXEtl'. THUC. vi. 78. (The orator adds, d..\..\ ov8' VjLEt<;; VVI' y 1rW oM' ol a..\.Aot E1rl TaVTa wpp-YJue.) l\1v<tv yap E~'fJI' r<{i l<amnopovvn TWV fi..\..\wv, he might have stood his ground (but really he ran away). DEM. iii. 17 : cf. xviii. 14, xxvii. 58; LYS. xii, 31. 1\)v Sw()~K'fJV 1)<j>uvtKan, ~ ?J> i) v l0vat 7rEpl 'Tfavrwv T~v &..\~8<tav, yo'u have concealed the will, j?-orn which we (now) rnight know the truth about the whole matter. DEAL xxviii. 10. T?)> 1JJLTepa> ;xepa<> ~JLfiS J<j>' ~JLWV aVTWV o[Katov 1JV T'Vv E~ETU<:TJLOV 7f0tt(T8at, i.e. we should justly settle up om quarrel by ouTSelves. Id. xviii. 16 : cf. 13,


HV' a~TOV~.-. a~LOV~ f.7rLU,l.~VVl'at. "'TOV'TOV,



fl'>} ~JLE





where oKatov ~v is understood with XP~u8at, he would justly have used them. IIp&:rrwv ?ToAv (3..\nov ~ u~ 7rpou~Kov ~v (se. 7rprirrnv), being much better off than you deserve to be. Id. xlv. 69. Ka2 fl-~V U~tov ?jv aKovuat, indeed, it was worth your while to hear them (J.ttws av fJKoDuan). PLAT. Euthyd. 304 D. The person addressed had just said ovK oios i' ij KaraKm5Hv. ''A..\..\qJ ~7rp7rV ..\yctv Cl. A~')'EtS, another would have becomingly said what you say (0...\..\os &v ;>,)'< 7rp<7rovrws). PLAT. Rep. 474 D. T6 Bvux<peurarov rwv ovoprirwv, 'TWV .p8ovoVV'TWV epyov ?jv AeyHv, J.)._)..' OV 'TWV 7rp0UTWTWV 'T~S rotaVTTJS 7rat8VuWS, the most disagreeable of names (Sophist), which the envious ought to use rather than those who stand at the head of the business in question. Isoc. xiii 19.

420. II. When this form is made the apodosis of an unreal condition (expressed or distinctly implied), it states that what the infinitive denotes would necessarily, properly, or possibly be done (or have been done) if the case supposed were a real one. The chief force of the apodosis here always lies in the infinitive, while the leading verb acts as an auxiliary (which we can generally express by ought, rnight, or could, or by an adverb), modifying the idea of the infinitive more or less in different cases. But when the chief stress is laid on the necessity, propriety, or possibility of the act, and not on the act itself, so that the real apodosis is in the leading verb, this takes av, like any other imperfect in such an apodosis (423). In some cases, however, even when no Civ is added, the force o the infinitive is so modified by the idea of the leading verb that the opposite of the apodosis (which is generally inferred) cannot be expressed without including both ideas (see examples in 422, 1). 421. In the following examples the infinitive represents the real apodosis, and its action is denied as when no protasis is added (419):El i7r' ~peas povvovs iuTparTJAaTa 6 IIpuTJ>, xp~v mlr6v 71'rivTwv 'TWV aAAWV a7rXOfi-VOV Uvat ollno J7rl 'T~V ~fl-ETEPTJV" Kat &v f.B~A.ov 7raut ws i7rl '.Sd8as f.A.avvH, if the Pe?"Sian were making his expedition against us alone, he should leave all others and be marching directly into our count1y ; then he would show eve1ybody that he was marching against Scythians. HDT. iv. 118. .6. e fv ( = Efi<t) 8, et7r<p 'ljv ovvaT6v, lfv<u TWV a.A.Awv avT6 ..\y<u 8av vvv OE ao-6varov. PLAT. Theaet. 202 A. Xp~v u', &rep 1ju8a p~ KaK6s, 71'e[uavrri Jl- yap<Zv yripov TOv8', d..\A.<i p~ utyfJ <f>[.Awv, i.e. if you we1e not base, you slwuld make this marriage with my consent, and not (as you do) in secret from your friends. EuR. Med. 586. Ei' TtVa (7rpOZKa) i88ov, dKO<; 'lj" Kal 'T~V oo8etuav V7r6 TWV 7rapayevf.u8at <f>auKoVTWV paprvpe'iu8at, ie. if he had given any dowry, it would naturally have been attested by witnesses. lsAE. iii. 28. See Id. iv. 18. 'Ef<~ d f<~V Jv lift.Aats rtuiv -Yjpf.pat> ~OtKTJUE Tt TOV'TWV lotWTr}V OV'T<I., lOE'f ~ai o{KTJV 7rpoufjK1> O.~Til 8toovat,






i.e. in that case he would' properly have given satisfaction by a privat~ 1rpocrqdvTws lorc- OlKYJV O:v J08ov). DEM. xxi 33; see xxxiii. 25 and 38. Ov yd.p Jv.f}v p.~ 7rapaKpov(J'8evTwv vp.wv ( = d P-~ 7rapeKp0V(J'8YjTE) p.etvat <PtAt7r7r(jl, for Philip could not have remained (as he did) unless you had been deceived. Id. xix. 123. El ~(J'aV avopes, W(J"lrEp cpa(J'tV, dya&o, 0(J'tp aAYJ1rT6repot 1](J'aV TOts 7rEAas, TO(J'cjjoe cpavepwTepav J~.f}v avrow T~V apeT~V OEtKVVVat, i.e. in that case they 'might all the more plainly manifest their virtue (which they do not do). THUC. i. 37. El J{3ovAETO 0Kaws elvat, J~.f}v avTciJ p.t(J'Ow(J'at r6v olKov, ?] y.f}v 1rpufp.evos K TWV 7rpO(J't6vTwv TOVS 1ral:8as Tpe<f>Etv, i.e. he might have let the house, or have bought land and supported the child-ren from the income. LYS. xxxii. 23. 'Ev avTii TU 8[KYJ J~~v (J'Ot cpvy~s np.~<ra(J'Oat, l {3ovAov. PLAT. Crit. 52 C. (See Isoc. xvii. 29.) lloAAol:s 86~w, ws oi6s i lf>v <re a-c{lCEtv el i)OeAov &vaA<rKtV xpqp.aTa, &p.eA~<rat, many will think that, whereas I might have saved you if I had been willing to spend money, I neglected it. lb. 44 B.
~tUit (as if he had said


422. 1. In the following examples the idea of the infinitive is so modified by that of the leading verb, that the real apodosis (the opposite of which is implied) includes both ideas; but the chief force still remains in the infinitive, so that no av is added.
El yap v1r6 686vTos Tot il:1re reAevT0(J'etV p.e, XP?JV 80 a-e 1rotetv ra 7rOtEEtS' vvv 8 V'lrO alxp..f}s, for if the dream had said that I was to be killeli!by a tooth, then you would pToperly do what you now do; but it really said I was to be killed by a speaT. HDT. i. 39. (Here the real apodosis is not in 7roteetv alone, which is affirmed in Ta 7roteets, but in the combined idea you would do with prop1iety; and it is the opposite of this which is implied. Xp.f}v av, which might have been used, would throw the main force on the xp.fJv, with the meaning it would be your duty to do.) El p.ev ovv arravTES cilp..oAoyovp.ev <PAt7r7rOV TV 7r6Aet 7rOAep.el:v, ov8v aAAo 8et TUV 7rapt6vTa Aeyo V Kat a-vp.(.1 \ 1 " 1 > ../.. \ I l \ > I :.f h ,..,ov~~.evetv TJ 07rWS aa-..t'aA.E<TTara avTov ap.vvovp.e a, I.e. ~J t en we wme all agreed that Philip is at war with us, the speaker ou.ght to say nothing else and to give no other advice than this, etc. (but it is added that, as there is a difference of OIJiuion, it is necessary, dvayKYJ a-T[v, to speak on another subject also). DEM. ix. 6. (This imiJlies not he does speak, etc., but he is bound to speak, etc. "Eoet av would merely have thrown the balance of force upon the necessity, whereas now it falls on the speaking and advising.) El yap 1rap' f.p.ol h8YJ Td ypap.p.are'iov, v.f]v alna(J'a(J'Oat 'A1raTovpltp f.yw ?}cpavtKa Ta> a-vv8?)Kas, joT if the account-book had been given me to keep, A. might possibly have cha1ged rne with putting the contract out of the way (implying that, as it was, he could not charge me with this). Id. xxxiii. 37. El p.ev J~pa. p.erap.eAov Tfj 7r6AEi TWV 1rE7rpayp.vwv, OVK a~tov ~V eavp.&Cetv ai>Tov, if J,e !.ad seen that the state repented of her acts, we should have no good reason for beiny surprised at him (imvlying we nlYW have good reason for suTprise, dgtws Oavp.&Cop.ev). Isoc. xviii. 21. The preceding examples confirm the reading of the best Mss. in

e '


>I " a1ra.~<; En1


' "' ~~ O"Vy')'VWO"'TOV TJV a-o~ 'TOW if you had remained still childless, you might pa1donably have become enamoured of this new 'll'Utrriage, the apodosis being equivalent to 1]paaB1j> av with an adverb meaning pardonably (if you had done what would have been pa1donable). This implies not you were not enamou1ed, but you were not pa1donably enamoured. If no protasis had been added, <T1!')')'VW{]''TOV 'l}v pafJ'6'l]vru (in its potential sense) must have meant you might pardonably have become enamoured (but you did not), and then av would have been required to give the sense it would have been pardonable (b1tt is not so). The other reading, <Tvyyvwrrr' &v 'l}v, would make the same change in the balance of force that x.p'l]v av, iioEt &v, ev'l]v &v, and li~wv &v ~~~ would make in the preceding examples.

.epaa-8~vaL AExovs, which Jnay be translated, for

' 'laE UR. Med. 4901 Et' yap .J: ()'

2. In concessive sentences introduced by Kai El, even if, ovo' El, not even if, or <l, although, containing unreal conditions, where the action of the apodosis is not denied but affirmed (see 412, 3), the real apodosis may be represented by an infinitive and a leading verb like ;sa, J~rjv, etc. combined. E.g. OvK e~'l]v aimji OtKa(Ea-6at 7rEpt 'TWV 'T6TE YEYEV'Y}fLEVWV, ov8' El 1rdvra ravr' ~v 7rE7rOLYJK<il> 3. 4''1JfJ'LV oi'!ro>, he could not maintain a suit
about what was then done, even if I had really done whut he says I did (implying OVK f!gwnv avr<p 0LKa(w6aL, but with the chief force on 8tKa(Ea-0at). Isoc. xviii. 19. Ov8' Ei yvfJa-wL ~a-av Ela-7raLYJTO~ 8, w> OVTOL ;cpaa-av, ov8' OVTW 'TI'poff''l]KEV avrovs EvK'TryfLOVOS" dvat, not even if. they were genuine sons and were afterwards adopted into anothm family, would they now prope1ly belong to E.'s house (implying they do not p1operly belong there). lsAE. vi. 44. See also Hnr. vii. 56 ; Dn:11. xviii. 199, xx.iii. 107. Ovo' El yap ?]v TO 7rpay/La fL'J BEfJAarov, aKa6aprov {JfLfis dd., ?]v OVTW'> eav, for even if the duty wen not IM"!Jtd upon you by a God, you ought not to leave the guilt unpurged as you do. SoPH. 0. T. 255. (Here the apodosis as a whole is affirmed, although the infinitive itself, not to leave, is denied. So in the two following examples.) KaA.ov 8' ~v, El Ka~ 1JfLapTaVofLEv, TOL(J'OE d~aL Tfj ~fLETEPif 6pyfj, ~fL'i:v 8' ai<Txp"Ov (se. i)v) (3Laa-afJ'Bat -n)v fLETpt6n]Ta, if we had eve11 been in the 1/Yrong, they rnight jctiTl:y have yielded to o1tr w-rath, while we could not have done violence to thei1 moderation without disgTace. THuc. i. 38. ''A~ tov ~ v, El K( 'fL']il'Ev avTUl:s 1rp6npov {J1rrjpxw d.yaBov, (Tavras) Tijs fLEY[<TT'Y}S owpd1s 7rapa TWIJ 'EA.A.l]vwv T1!XtV, i.e. these cities, even ~f they had had no other meTit to ul!J on, deserTed to nceive (ought to hcve nceived) the greatest nwa?d j1orn the (}1eelzs (which, it is said, they did not receive). Isoc. xii. 71.

423. ("EilEL av, etc.) The examples in 421, 1 and 2, show that the common rule for distinguishing (on etc. with the infinitive (without &v) from oH l:lv etc. with the infinitive,-that the former is nsed when the action of the infinitive is denied, the


"EoEt /iv




latter when the obligation, propriety, or possibility is denied, -often cannot be applied, though as a working rule it can be used in the great majority of cases. While there are many sentences in which either form would express the required sense, the essential distinction is, that the form without av is used when the chief force of the apoJosis falls on the infinitive, the leading verb being an auxiliary (see 420); but the leading verb takes ({v when the chief force falls on the necessity, propriety, or possibility of the act, rather than on the act itself. The following examples will illustrate the form with El Jl-EV yap ey<il En EV ovvdfl-E! ~V 'TOV pr;.o[ws 7r0p15nr8a! 7rpds 'Td auTv, o~OEV a V (]" EOH Oevpo leva! &A...\' ~JLEtS liv 7rapa U"E Vfl-EV"

av :-

l'VV OE (]"{ XP~ 7rVKV6Tepov oevpo leva!, i.e. in that case there would be no need (as there now is) of your earning hither. FLAT. Rep. 328 C. T<IJ fJ-EV 1ra-rpt a~'TTJS, El 1rai:oes appeves fl-~ (yevov-ro, O~K liv e~Tjv Ci.vw 'TUVTY)'> O!a(}eu(}a!, her fnther, 'if he had had no nwle children, WMdd not have been allowed to leave her out of his will (implying &A.A.' e~Tjv). lsAE. x. 13. El ovv 7rapeKaAovf1-EV &A.A.qA.ovs e1rt Ta olKoOOJl-!Ka, 1r6Tepov E/lo &v ~Jl-as U"Kefau8a! ~JLas a~rovs Kai e~eramu el E7r,uniJ1-eea T1JV TEXVYJV j EO Et av ~ oll j i.e. in that case would it be needful or not to examine ourselves and inquire whether we understand the art? PLAT. Gorg. 514 A. See also DEM. iv. l, quoted in 410. A comparison of these examples with those in 422, 1, will show the distinction between the two forms and also the narrowness of the line which sometimes separates them. For a discussion of xv)jv and XPTJV av in DEM. xviii. 195, and of XP'JV and expT}v av in LYS. xii. 32 and 48, and for other remarks on these constructions, see Appendix V.

424. 1. The imperfect w<f>eA.Aov or &<f>eA.A.ov of d<f>eA.Aw (Epic of d<f>dA.w), owe, debeo, and the aorist /J<f>eA.ov or &<f>eA.ov are sometimes used with the infinitive in Homer like xpryv, EOH, etc. in the later construction ( 415 ). E.g. 'l'LfJ-~V 7rEp Jl-OL 5<jleAAEV '0A.1~Jl-7rLO<; eyyvaA.[~aL Zevs -Dtj;Lf3pef1-ETYJS. VV11 o' O~OE fl'E T1>r86v ETtU"ev, i.e. Zeus ought to hnve sec1Lred me honour ; but now he hM not honoured rne even a little. ll. i. 3 5 :3.
:Xvv 5<f>eA.ev Kara 1rdv-ras dpLu-rryas 7roveu8aL ALuu6Jl-evos, now ought he to be labouring arnong nll the nobles, beseeching them. Il. x. 11 7. 'AAA' iJq>EAEV d(}avaTOLU"tv evxeu(}u.L, but he ought to have pmyed to the (;ads. Il. xxiii. 546. For the reference to present time in ll. x. 117, see 246 and 734.

2. From this comes the common use of this form in expressions of a wish, in Homer and in Attic Greek; as /J<f>e A.e Kvpos (rjv, would that Gyrus were living (lit. Gyrus ought to be living),

An. ii. 1, 4.

(See 734.)

425. Similar to this is the occasional use. of ef3ovA6Jl-'f}V (wit4-




out av) and the infinitive, to express what some one wishes were now true (but which is not true). E.g. 'E tJOVII.OfL'r)V JMV OVV K<U 'T'r)V tJOVII.T)V Kat 'T(t') EKKII.'f)rTta<; .op (}~ ~ \ ~ W<;
I ' .. \ \ \ ' ' ' ' \ I '

8totKe'i:rr9at Kai rovs v6p.ovs lrrxvHv, I would that both the Senate and the assemblies were 1-ightly managed, and that the laws were in force (implying the opposite of op9ws 8totKE'irr9at and lo-x,vew). This is analogous to /J<f>eA.ev elvat, would that it were, and ~OH elvat, it ought to be (but is not). AEsCHIN. iii. 2. 'Ef3ovA.6p.1]V fl-~V ovK Jp!Cetv v(M.8e, I would that I were 1wt contending here (as I am), or I would not be contending here. AR. Ran. 866. 'Ef3ovA6f1-1]V T~v ovvafl-tV roil A..iyetv Et itrov p.ot Ka 8 E<TT aVat Tfj o-vp.<f>op~. I would that powm of speech equal to my rnisfo1tune were granted me. ANT. v. i. 'Ef3ovA6p.YJV Kay~ TaA.YJ(}~ 7rp6s ei11'eZv 8vvl]8~vat, I would that I had found the power to tell you the truth. IsAE. X. 1. 'Ef3ovA6p.YJV JLl]O' v<f>' Jv6<; a8tKe'io-8at TWV 71'0AtTWV, I would I had not be~n W1'0nged by'a single one of the citi~ens. Id. Frag. 4 (Scheibe): see Frag. 22. 426. 'Ef3ovA.6p.l]v &v, vellem, I should wish or I should have liked, can always be used as a potential indicative, like eoet av etc. (423):
see AR. Eccl. 151; AESCHIN. iii. 115. (See 246.)

427. (a) The aorist of Ktv8vvevw is used with the infinitive, as a periphrasis for the verb of the infinitive with av. E.g. 'H 7r6At<; JKtvovvevrre 1!'ao-a 8ta<f>8ap~vat El avep.os J'TI"ey~vero, the city ran the risk of being utterly destroyed if a wind had, a?isen. TRue. iii. 74. El p.~ J~e<f>vyop.ev els 6.eA.<f>ovs JKtvovvevifap.ev d. 11' o Arr (}a t, we 1an the rislc of perishing had we not fled to Delphi, i.e. we should ve1y probably hme perished if we had not fled. AESCHIN. iii. 123. For JKwDVvevo-a av see (b) below. So with Kfvovvos ~v: as in AND. ii. 12, el r6re 'T'tt E'TI"tT~OHa p.~ elo-~x8YJ, ov 'TI"Ept TOV (TWo-at Ta> 'A8~va<; 6 Kfvovvos ~V avroZs p.fiA.A.ov ~. K.T.A., i.e. they mn a risk, in case the supplies had not then been broLtght in, not so much about saving Athens, as, etc. (b) When the chief force of the apoilosis lies in EKtvovvev(Ta, even
though the meaning is not much affected by the distinction in form,

av is use<l (as with EOEt etc. in 423). So in XEN. An. iv. I, 11, el 11'Adov> o-vvd.eyl](J'av, htv8vvev<TV llv 7roAv 8w<f>9ap~vat Toil cnpaTvp.aros, if more had been collected, there would have been danger of much of the army being destroyed.

428. (a) The imperfect of p./..Aw with the infinitive may express a past intention or expertation which was uot realisPd, and so take the pla.ce of the Yerh of the infinitive with :iv. E.g. "H p.O.A.a o~ 'Ayap.ep.vovos <f>8<Teo-9at Ku.Kuv olrov ;p.eA.A.ov, el
p.~ . . . let11'E<;, i.e. l should ha-re pe?ished Ulce A. (lit. I 1vas to have perished), ~f thou hadst not spoken. Od. xiii. 383. M~AAev p.ev 'lron: o'lKoS oo' d.1wn0<;; Kat d.p.vp.wv EfJ-JLEVU.t' vilv 8' hipw<; f.[3oAOVTO 8w[, this house was to have bem 'rich and glorious; but now the Gods have willed it otherwise. Od. i. 232. 0~ v-vrrrpaTEV(J'EtV lp.eAA.ov, they were





not going to join him, or they would not have joined him (in that case). DEM. xix. 159; see xviii. 172. "'HTTov TO J))[K1JJLO. 1T'OAAwv ovcrwv ;JLeA...\e Sq>..ov ;cTfT(Jat, the offence would have been less plain when there were many (olive tnes). LYs. vii. 24. See Tau c. v. 38, JLEAAo VTE~ 7rpoT<pov, el Tavra E7rtfJ'av, 1T'Hpci(J'Hreat. Compare the Latin: Hoc facturi erant, nisi venisset, they were to have done this (would have done this), had he not come,l (b) A single case. of liv with EJLEAAev occurs in AND. i. 21 : d Ka2 ' '{3 1\ ( I \ ,.J,.I\ > ' '11'aTYJp E OVII.1'0 V71'0f.tEVEtv, TOVS 't'!II.OVS O.V OWT E E7rt1'pE7rEtV avr</), d...\>..' OVK ll.v 7rapatTEW8at Kat Oew8at dmevat 01T'OV Cl. V I" EA A V uw8~fJ'efJ'8at; i.e. tn depart to a JJlace whe1e he would have bee'll l-ihly to be safe. Most critics repudiate this liv; but it seems perfectly ana-logous to liv with ou, XPqll, etc. (423). ( 429. Similar is the use of ;</>YJV in Od. iv. 171 : Kaf. JLtV e<f>?JV ~A.86vro. <{>tA-IjfTEJLEV e~oxov O.A.A.wv, ei VWtV VOfTTOV i!ilwKEV (Zevs), i.e. I intended to love him (and should have done so) had Zeus gmnted us a return. 430. An analogous case is LYs. xii. 60: d7roA(J'o.t 7raperrKV&.CovTo T~v 7rOAtv cl I"~ ot' livopas dyo.8ovs, they were p1eparing to dest?oy the city (and would have dest1oyed it) had it not been for good men.
t\ ''

ToVT!f 8' Elf.t~ U)J.tOADyovv a OVTO<; f.{JovA<To,'{ tYJJLl'{ <!vo xos ?j v, but if they had not acknowledged to him what he wanted, he was liable

431. A few expressions which have no dependent infinitive are practically equivalent to a potential indicative with :I.v, and so can stand as the apodosis of an unreal condition. E.g.

to no cha?ge (i.e. he could not have been accused). LYS. vii. 37. '.12s, d pv TO :71 avTo<f>Wp!fJL~ 7rpo(J'eyypo.7rTo, voxo~ wv (=i)v) Tfj a7raywyi), assunoing that, if the ords f..,f aiJTo</>Wp!f had not been added, he rnight pToperly have been t1ied by d1ro.ywy~. Id. xiii. 85. ITtfTnvovro~ "/UP EJLOV EJLO'I. eloeva.t a, A.eyw, Ka.Aw~ elx EV ?), if I tmsted ( = el i:rdfTTEvov) to any knowledge of rny own about what I a1n saying, the consolation which you off~?' wo1.Zd encoumge rne (lit. your consolation was good on that supposition). PLAT. Rep. 450 D. (We might have had KaAov 1jv fTE 7raprtJLv8e'i.cr8cu in the same sense.) El TO KWAvfTo.t T?)v Twv 'EAA1)vwv Kowwv[a.v E1T'E7rpaKetv iyw hM1l'7r!f, fJ'o! TO f.t?J crt')'qfTa.t Aot7rOv 1]v, in- that case it rmnained fo1 you not to lceep silent (i.e. you should not have lcept silent). DE~f. xviii. 23. (The mticle with fTt')'qcro.t only slightly distinguishes this from the examples under

432. The same explanation applies to other cases in which a rhetorical omission of Ul' in apodosis is commonly assumed; as in
1 This use of #p.<'A'Aov with the infinitive corresponds precisely to the Sanskrit use of the past jutu1e tense in the sense of the Greek aorist indicative with llv. Thus" if he had said (avak~yat) this, he would have slain (ahanisyat) Indra" (Cat. BTahrn. i. 6, 310 ), where the two verbs are augmented past futures, meaning literally he was going to say and he was going to slay. See Whitney's Sanskrit Grammar, 950.




EuR. Hec. 1113, cl 8~ f1J <Ppvywv 1r6pyovs 1rT<)vras iJ<rfEV 'EAA~vwv OOp~, cp6(3ov 1rape<rXV o~ fE<rWS o8e KT67roS, but if we ha,d not known

that the Phrygwn. towers had fallen, this noise gave us cause fm terror in earnest (i.e. would ea,sily have tm'ified us).

8wJLa, for it had been a greate1 gain if Ulysses had rehwned (for Kepowv -i}Ev '08vcra vocrT~uaL). Od. :x:x. 331. Compare MATTH. Ev. :x:xvi. 24, KaAov fiv a~T~, <l ovJC y<vv0(Jry o dv&pw1ro<; eJCEtvos, it had been good for that man if he had not been bom (for KaAov -i}v a~nji 1;,1) yeVF1]BFjvat). El 8~ a7ro<f<6~eTat, Kp<fTTWV i)F o dy<~ll' f1l ye)'El'17fJ,EVOS (=cl P.0 yeyf.v1JTO ), but if he is acquitted, it we1e better that the iT1:al had neveT taken place (for Kpd:rrol' 1]v Tov dyw11a p.1) yEyev~<rB<u). AE~CHI;><. i. 192. This occasional suLstitution of a protasis does not indicate that the infinitive in Kpe'iTTov 1)1' a~T<fl lA.fJeiF, he had bette1 have gone, was felt as a protasis. We could suLstitute for this English it were better if he had gone, but only by a change of construction.

433. Occasionally a protasis takes the place of the infinitive in the construction of 419. E.g. 'E1rEt T60E Kepowv ~ev, d l'b<rTqrl 'Oou<T<~s Kat inr6Tpo1ros tK<To

434. In Homer the construction of the unreal conditional sentence is not completely developed. It is not improbable that in the primitive language the optative could express in a rough way both present and past unreal conditions, and in Homer the present unreal condition is still expressed only by the present optative ( 438). . 435. The aorist indicative in Homer, both in protasis and in or KE, is used as in Attic Greek j but the imapodosis with perfect is always past, never present. 1 E.g.


KE oi] ~t<fEE<T<T' a~Touxeo(w o~ra(OJ!TO, et Jl-1l K'!JpvKS they would have wounded each otheT, had not heralds come. ll. Yii. 273. "Ev&a KE Aotyus E>7P Kat d01/xal'a lpya yf.povro, cl JL>) llp' o~u J!01)<TE 7raT1)p aFOpWF TE B<<'vl' TE, tlum the1e would have been, etc. ll. viii. 130. So viii. 366. Kat v6 KE 01J 7rpcrpw er'> )'El'ET' dJL<Ponpot<rlJ!, cl p.0 'Ax,i\A.evc; a~TOS dvturaro Ka! KaTepvK.v. Il. :x::x:iii. 490. See Il. xi. 504; Od. :x:vi. 221, :x:xiv. 51.

'l A &o




1 Mr. Monro (Ham. G1. p. 236) doubts this statement, and refers to Od. iv. 178, Ka.l Ke Oaf.'' (p()ai5' <oPTes '~-'"''Y6J.'e()', ooM KEP f};da.s <L\t.i> otiKp<e, as a case in whicl1 ''the imperfect EJ.'L<Y"fOJ.'e()r;. takes in the present time, we should (from that time till llow) have been meeting." It seems to me tbat, according to the Homeri.c usage, we can find no more in Oa}-'a. fJ.'UT"fDp.E8a Ke than we should have hadjTequent meetings, and the rest comes front the context. In any case, this use is far removed from the Attic hropwowea /J.p ,,.4 fJacnAea, we should (no-w) be on our way to the King (410). A nearer approach to the later use perbaps appears in Il. xxiv. 220, El -yap r p.' af.f.os heA<VV, ij any otheT (had .1) commanded 11w. But see Il. ii. 80.




Ka[ v~ Jl fln 1rA.ova> AvK[wv Krav< ow<; 'Oova-a-<v>, El !J-~ lip' &~iJ fJ-EJ'OS Kopv8ao/...os ''EKrwp, i.e. Ulysses would have killed still more, had 1wt Hector peneived him. IL v. 679. Ka v~ KEV ~ta 7ravTa Kar1>8tro Kat fJ-EPI dPopwv, El ~J-fJ rs !J-E 8Ewv &A.o<f>~paro Ka[ ~J-' a-dwa-EJ!. Od. iv. 363. But {},<f>EAov with the present infinitive may be present, even in Homer, both as a potential expression (424) and in wishes (734).

436. We find the imperfect referring to present time in Theognis : see vs. 905, El fJ-~V yap KanoEw j3torov rA.os ~v, clKos &v ~v. See PINn. Nem. iv. 13. 437. In Il. xxiii. 526, Et K is found with tile aorist indicative in protasis, KE apparently adding nothing to the sense : Ei o K' i!n 7rporpw f'EVETO opofJ-OS d~J-cf>orepota-w, T!f KEV !J-<V 7rapEA.aa-a-' ovo' dfJ-cpljpWTOV f181)KEV. 438. (Optative in present unreal Conditions.) In Homer a present unfulfilled condition is regularly expressed by the present optative with El, alld its apodosis (if present) by the present optative with I<E or /lv.
The only instance of this form in both protasis and apodosis is Il. xxiii. 274, El fL~v vvv e1rt d.AA.<J:> d8AE~OtfLV 'Axawt, ~i r' <i.v eyw ra 1rpwra Aaj3ivv KAta-[,)vO </>Epo[fL')v, if we were now contending in honouT of any otheT (than Pat1oclus), I should take the jint pTize and bea1 it to my tent. Twice we have the optative with dv in apodosis with the regular imperfect or aorist indicative (past) in the protasis : Il. ii. 80, El fJ-EJ! ns TOV 8vap0ll uAAos EV0"7rV, tfEVOO<; K11 cpai:JLEV Kat voa-cf>tCot(LE8a fLUAAov, if any other had told the dream, we should call it a lie and mtheT tum away f1om it; and the same apoclosis after El: ,.[, 0' d.A.Aos JKEAl>EJ!, in Il. xxiv. 222. In Od. ii. 184, OVK &v roa-a-a 8w7rp07rEWJ! ayopEVES, OUOE KE TryAEJLaxov KxoAW!J-EVOV wo' &vH1)>, we have first the imperfect with dv as a past apodosis, (in that case) you would not have rnade this speech with all its divination; and then the present optative with KE as present, no1 would you be urging Telemachus on, as you now aTe; both referring to an unfulfilled past conditiou, if you had perished, suggested by Karacp(Ja-8at {},<f>EA> in vs. 183. 439. See the corresponding use of the present optative in Homer to express an unaccomplished present wish (739). In both wishes and conditional sentences, it must be remembered, the use of the optative in its ordinary future sense is completely established in Homer. See exa~nples in 455 and 722. 440. (Optative in past unTeal Apodosis.) Homer has four cases of the optative with KE (three aorist and one present) in the apodosis referring to the past, with the regular indicative in the protasis expressing a past unfulfilled condition. These areKat vv KEV f.v8' <h6A.otTO avag avOpWll Alvcias, d fL'J tip' &~v vo'l)a-< 6.tos 8vyarw 'AcppooT'lJ, Aeneas would have peTuhed, had 1Wt Aphrodite quickly peTceived him. Il. v. 311. Ka vv KEV Jv8' d1r6Aotro




HAp'YJS d'T'OS 7ro'Ap.oto, El /k~ 'H<pt(3o[a 'Epf-k~'{/ ~~~)")'ELAEV. Il v. 388. Ov KE Oav6vn 1r<p t':>o' dKaxo[f-kTJV, d f-k<Ta o'l:s hapoun'YJ Tpwwv vi O~f-k<p, I should not have felt so grieved if he had perished, Oav6vn ( = d EOavEv) being further explained by d ... 80./kTJ Od. i. 236. "EvOa KE pEta cppot KAvra ret)xm, El""~ oE ayauuaro <Poif3o> 'A?roAA.wv, he would easily have borne away the famous armour had not Phoebus Apollo g1udged him. Il. xvii. 70. Here dm.SAEro, dKaXOf-k'JV, and cpep would be the regular forms even in Homer, corresponding to the regular protases.

441. In the transitional state of the Homeric language we see that the past tense; of the indicative had fully established then1selves in the protasis of past unreal conditions, but not so thoroughly in the apodosis, where the optative occasionally occurs. In present unreal conditions, the optative alone is used in both protasis and apodosis.
442. Besides the full conditional sentences above quoted, we find in Homer many potential optatives with K~ or t1v which seem to belong to the borderland between past and future conclusions, and are not definitely fixed in the past (like the apodoses in 440) by a past tense in tl1e protasis. Such are especially cpaJTjs KE, as in Il. iii. 220, xv. 697, and of!o KE cpa[TJ'>, as in Il. iv. 429, xvii. 366, Od. iii. 124, I1. iii. 392. In the first four cases it seems most natural to translate them as past, you would have said, ?W1' wo1tld you have said; but in the last two cases it is more natural to translate nor would you say (future), and so with cpa[TJv Kev, Il. vi. 285. But in the fluid state of the language which allowed both d?ru)AEro Ke and ti?T6Aom5 KE to mean he would have perished, and cppot KE to mean both he would carry (fut.) and he would have carried, according to the protasis which was used with them, i.t is easy to understand how cpa[TJS Ke (without a protasis) might have a vague potential force, you might perchance say, which could be felt as either past or future as the context demanded. We must, therefore, hold that the optative with KE in such cases expresses merely what could happen, without any limitations of time except such as are imposed by the context; and according to the limitations thus imposed we translate such OIJtatives (with more exactness than they really possess) either as past or as future. In one case the feeling of past time is seen in the dependent verb: Il. v. 85, Tv8d'81Jv 8' ovK &v "Y v o [ 7J s 1rorpo ten f-k E TE [ 1), you would not have known to what side he belonged. (This occurs in the same book of the Iliad with both the examples of dmSA.ot76 KE for d1rt.sA.er6 Ke.) Other examples are the fol1owing : OvK &v ETrELr' 'OoviJ' y' f.pi(J'IJ'eHv (3por'Os O:A.Ao>, no other mortal could then vie with Ulysses (after a past verb). Il. iii. 223. ''Ev8' ovK O.v f3p{ovra roots 'Ayaf-kEf-kVOVa owv. Il. iv. 223. "Ev8' oll KEV pea t1!'17'os i11'(3al1J, 7r(Cot o f-kEvo[vEov El T<Afovuw (the connection with f-kfiVO[Vf.OV gives ~1]'(3a{'t) a past direction). Il. xii. 58. "Ev/1a K' etrELTO. KO.L a8avar6s 7rEp E7rEA86Jv BTj~O"UtTO loinv Kat rapcpfJdTJ cppeutv




y<nv. Od. v. 73. 'Ds o~K &v ~A1row v{rrpov dvna<Tav-ra, i.e. as you would not expect (?) a younger person to do. Od. vii. 293. 0~8 KfV ~P1J~ KlpKOS op.apr~<THfV. Od. xiii. 86. Further, compare Od. ix. 241 with 11. i. 271 and v. 303.

443 . (a) Herodotus has a few cases of the potential optative with the same vague reference to time which has been noticed in Homer (442), and we may sometimes translate these, like those in Homer, by past expressions. E.g. Taxa 8 &v Kat o1 d1ro86p.vot A.yo tEV dmK6~vot Js "21r&pr1Jv ws d7ratp81)<Tav {nro 'Lap.lwv, and perhaps those who sold it (the cup) might come to Sparta and tell that they had been robbed of it. HDT. i. 70 (see Stein's note). All that the optative itself seems to express is that this would be a natural story for them to telL In vii. 214, d8d'YJ p.'Ev ilp llv Kat f.wv p.ry M1JA.tevs -ra{;T'YJY rryv d.-rpa1rov 'Ov~T1JS, d rij xc!JpTJ 1ro.A.AO. op.tA1)KWS E~1J, for Onetes, even if he was not a Malian, might know this path, supposing him to have had much acqu,aintance with the country, the optative in protasis (expressing no condition contrary to fact) shows that el8e[1) O.v is not felt to be past. See also vii. 180, r&xa 8' &.v Tt E7ra{;potTO; viii. 136, r&x' &v 7rpo.Ayot, might perhaps warn him; ix. 71, ravra llv d7rotev, they 1night say this. For f~'I}<Tav 8' llv O~TOt Kp~rfs, HDT. i. 2, and similar expressions, see 238. (b) In EuR. Med. 568, o~8' &v <TtJ <jlat1)S er (]"E P.TJ KvCot A.xos, the condition seems to be present and contrary to fact, like d JLTJ ~Kvt(ev. See also PLAT. Menex. 240 D, f.v ro{;r<p 80 &v ns yev6p.vos yvo [1) oTot &pa f.-r{;yxavov ovTEs, K.r . .A. Such examples are extremely rare in Attic Greek.

1. Subfunctive or


Indicative in Protasis with a future A podosis.

444. When a supposed future case is stated distinctly and vividly (as if I shall go or if I go in English) the protasis generally takes the subjunctive with Uv, ~v, or llv (ii) (Epic r " or a~ 1a). The apodosis takes the future indicative or some other form expressing future time, to denote what will be the result if the condition of the protasis is fulfilled. E.g.
'E&v n .Aa(3w, 8w<Tw <Tot, if I (shall) receive anything, I will give it to you. 'E&v ,., .Aaf3n>, 86s p.ot, if you receive anything, give it to me. El 0 KEV ws ~P~TJ> Ka rot 1reifJwvrat 'Axawt, yvw<TTJ t1retfJ' os (J' ~yp.6vwv KaKos os r~ vv .Aawv, but if you shall do thus and the Achaeans




obey you, you will tlwn leam both which of the leaders and whi1h of the soldiers is bad. Il. ii. 364. Af K' avruv yv~w J!r'jfJ-Epr~a 7ravr' EJ!~7rOl'Ta, f(TITW fJ-W x.AaL'v&v TE xmvv& n, EtfJ-UTa KaAa. Od. xvii. 549. So aZ KE ow1n, Il. i. 128. Et fJ-EV K<V Mcvc.Aa.ov 'A.Af.~avopo> Kara7rE</>l''[l, avrus l7rH8' 'EAEV?]V X Er w Kat KT~fJ-aTa 7ravra, fJtJ-ELS o' EJ! l!1JECTITL VE~fJ-E ea 7rOJ!T07rbpounv El o K' 'A.Afavopov KT vv ~av86s McvEAaos, Tpwas ~7rtt8' 'E.Af.v?)v Kat Kr~)p.aTa 1r&v-i d1rooovvat. Il. iii. 281. Here f.xeTw, V<WtJ-eec~ (subj. in exhortation), and a1rooovvat (infin. for im})erative) are in the apodosis. AEKn r~vo> AT/ K<pa.uv Tpayol', alya Ti> .Aaffi. THEOC. i. 4. "Av of. TL<; Ul'BtcrT~Tat, cri>v vp.'iv 7rHpaO"bp.8a xnpowBat, if any one shall stand OJ)poscd to us, with your help we will try to overcome him. XEN. An. vii. 3, 11. K&v p.1] vvv f:Bf.AwtJ-EV EKE'i: 7rOAfoELl' avr0, v8ao' t(TWS dvayKarrB7]<rbp.()a rovro 7rOtEfv, and if we shall not now be willing to .fiyht him the1e, we shall pe1haps be jmced to do so here. DEM. iv. 50. (Here vvv refers to time immediately following t1Je present: if we a1e not now willing would bed P-0 vvv U)f.Aop.Ev.) ''Hv yap TaVra KaAW<; 6ptrrwp.cBa, /lp.nvov /3ov.Aevrr6p.d)a Kat 7rep2 Twv O..A.Awv. Isoc. viii. 18. ''Hv OE TlJV tlp~VYJV 1T'Ot?]O"~P,.E8a, Kai TOWVTOV<; '1tJ-a<; avrovs 7rap&crxwfl'EV, tJ-<ra 7roll)..~s dcr<f>aA.das r1]v 1r6Aw olK1J<TOfJoEJI. Id. viii. 20. 'Eav o:Ov fus vvv, 1r6re- gcr<' ofKot; XEN. Oyr. v.. 3, 27. Kat xpw allrol'>, f:av of.v TL, and use them, if there shall be any need. lb. v. 4, 30. "Hv p.~v 7r6AEtJ-OV a1p~cr8, fl-'JKETt 'lKET Ocvpo a!'V 37rAwv, d crw<PpO!IfVrE' ~V 8~ clp-f)ll'fJ> OOK~T< 8wBat, llvEV 07rAWV ?JKETE' ,t, o KaAws f~H Ta vp.f.npa, ~~~ <f>[A.o~ YEY1)<r(h, f.p.o2 p.tA?JfJ'EL. Ih iii. 2, 13. 'Ectv y&p T[ (Tf <f>avw KaKOY 7rE7rOL?]KloS, op.oAoyw UOtKElV' eav pf.vrot p.YJ8f:v <f>al' KaKuV 7rEtrOtl]I<WS fl-1JOE (3o1>A1J()<2s, ov Kat fTV ail 6p.o.Aoy~(TLS fl-Y)OEV v7r' f.p.ov dotKEwBa~; Ib. v. 5, 13. (Hel'e op.oA.oyw, I am ready to confess, refers to the future.) 'Ea1' fl-'l 1) o! </>tll.6uo<f>o~ (3autAcuuwcrtv ?) o1 (3arr~A.~s <PtA.orrocpl)trwcrtv, ovK ~rrn KaKwv 1ravA.a TnZs 7rbA<<r~v, unless either the ]Jhilosophen ~holl become kings or the kings philoso]Jhers, theTe is no escape from tloubles joT states. PLAT. Rep. 473 D. 6.[8wd ~KltlV KTfvLv avr6v, -i)v rao 1/;<urrBiJ A.yw1', he of!en himself willingly to suffer death in case he shall be proved false in this that he says. SoPH. Ph. 1342. M'JXO.l'YJTEov, f.av T xpvcr[ov TJp7ra.KUJ<; 7rOAV, 00 d7ro0</) TOVro, Eriv TE Bavarov /l~ta ?JOt KLK<os iJ, orrws P-0 U7ro8avE'iTaL, if he shall (p1ove to) have stolen much gold, we must contrive that he shall not 1estoTe it ; and if he sltall have committed crimes deserving death, that he shall not die. PLA'l'. Go1g. 481 A (for the perfects see 10:3). "Hv ue Tov AoL7rOV 7rOT d<f>A.wtJ-a.' xp6vov, KUKLfJ'T' a7rOAOlfMJV, i.e. may I pmish, ij I ever take them away. AR. Ran. 586. (See 181.)


445. It will be seen that the apodosis here (444) may consist of any future expression,-the future indicative, the imperative, the subjunctive in exlwrtations and prohibitions, the infinitive in any future . sense, the potential optative with or the optative in a wish. It may also contain a present indicative including a reference to the future, like XP~ or oer or the verbal in Tos, or the present





used emphatically for the future, like 6fLoAoyw above quoted (444) from XEN. Cyr. v. 5, 13, or 'lTav'Aa Jtrn in PLAT. Rep. 473 D.

446. The English, especially the colloquial language, seldom expresses the distinction between this form of the future condition and the present condition (402). Thus modern custolll allows us to use the inexact expre&sion if he wishes, not merely for cl f3ovA<Tat, if he now wishes, but also for Jav f3oi1AYJT<H, if he shall wish. The sense, however, generally makes the distinction in time clear. It is worth noting that the Authorised Version of the English New Testament never uses forms like if he does, if he. s, in either future or present conditions, even when tl1e Greek has ft1e present indicative with El; hut it has either the subjunctive or the future indicative in future conditions, and tl1e subjunctive in present conditions. The Revised Version, on the other hand, admits the present indicative (as if he is) in present conditions, but not consistently. See Luc. xxiii. 35, d oilT6~ trnv 6 XpttrT6>, A. V. if he be Christ, R. V. if this is the Christ; but in MAT1'H. vi. 23, cl oi'iv TO </Jw> TO v O"ol O"K6Tos JtrT[v, both versions have if therefo?e the light that is in thee be darkness. See also Cor. ii. v. 17.

447. (Future Indicative in Protasis.) The future indicative with Ei is often used in the protasis to express a future condition. This is a still stronger form of expression than the subjunctive, though it sometimes alternates with it in the same sentence. Both, however, correspond to the English if I shall do this, if I do this, etc. The future, as an emphatic form, is especially common when the condition contains a strong appeal to the feelings or a threat or warning. 1 It is thus a favourite construction with the tragedians. E.g. El yilp 'AxtAA<tJ~ olo~ t'lTt TpwEtrtrt fLax'iTat, oi8~ 0vvve' l~ovtrt 'lTOOWKm llry'AEtwvcL, if Achilles shall fight alone against the TTojans, not even a little while will they keep baclc the swift son of Peleus. Il. xx. 26. El 8~ trv y' (~ 7r6AfWV 'lTWA~trcat, ~ TE tr' Ow pty~trHv 'lT6A<fL6v y<, Kat cf x' ETEpwet 'lTlJeqat, if you shall mingle in the battle, ve?ily do I believe you 1oill shudder at the very name of battle, even if you hear it elsewhere (away from the war). Il. v. 350. El 8 (LOt Ttrovtrt f3owv E'lT,HKE' d{Lot(31]v, 8vtro{Lat <1> 'A8ao Kat iv vcdcO"U't <jJa<[vw, but if


In "minatory and monitory conditions": see Gildersleeve in Tmns. qf This article contains an enumeration of all the cases of iav with the subjunctive in future conditions and of <l with the future indicative in the three tragedians. It appears that in Aeschylus there are 22 cases of the future and only 8 of the snbjnnctive; in Sophocles 67 futures and 55 subjunctives; in Euripides 131 futmes and 177 subjunctives. If we omit the futures which are equivalent to fJ-EAAw with an infinitive for which the subjunctive could not be substituted (see 407), we have in Aes~hy lus 15 futures in future conditions and 8 subjunctives; iu Sophocles 46 and 55; in Euripides 98 and 177. In Attic prose Thucvdides and Lysias have the largest proportion of futures; but in prose, as in Aristophanes; the subjunctives always preponderate.
Am. Phil. Assoc. for 1B76, p. 13.




they do not pay rJW a proper requital for my cattle, I (the Sun) will descerul, to Hades arul, shine among the dead. Od. xii. 382. Et 8~ 7rpds ro1houn 'I \ I \ {3' EV, OVTO<; EKE WO<; TOY CJ'V ~'Y)TE!<; 011. lOS '1' 't' f. " Jl' \{3 ET! TE II.EVT'Y)CJ'H TOJI tOJI KEKA'ijcr&at &gu5<; ~ern, and if besides he shall still end his life well, he is that happy man you are seeking. HnT. i. 32. 'AA.A.' d ere p.apfH frjcpos, /J.A,\.' JpEI:s rd.xa, but if the judgrJWnt shall lay hold of you, you will soon tell another story. .A.ESCH. Eum. 597. See Prom. 311, Sept. 196, Suppl. 472, 474, 924, Cho. 683. El ravra A.~o>, ix8ape'i: p.~v ~ Jp.ov. SoPH. .Ant. 93. See .Ant. 229, 324, 0. T. 843, 846, 0. C. 628, Ph. 75, El. 465, 834, 1004. El r.;Jo' dpdcret>, KaKd<; cpavL, if you aid this man, you will appear base. Eun. Hec. 1233. M~ (c{r']v, El p.~ cpacryavov cr1rd.crw. Id. Or. 114 7. See Hec. 802, Or. 157, 272, 1212, Med. 346, 352, 381. El p.,) Ka8~~ets yA.wcr(Tav, ecrrat crot KaKcf.. EuR. Fr. 5. El 8~ p.~ rovr' ~'lf'todtt, 1rws XP~ ravrv rfi 1rpoKA~crL 1rpocrexw i!p.a<; r6v vovv. DEM. xxvii. 52. El 8' vp.E'i> aA.A.o n yvr!Jcrwfh, 8 f'-IJ yvwro, rva. ornr8E aVr~V fvx~v :JgHv; but if you shall give any other judgment, etc. Id. xxviii. 21. (Referring to the same thing, xxvii. 67, Demosthenes had said Jav yap ar.ocp-6yy fJ-E of!ros, iJ P.,~ yevotro, T~Y Jr.w(3<Aav ocpA~(J'W.) "Hv J8Awp..EV a1I'orivyjcrKELJ! vr.ep 'T'WJI OtKa{wv, EiJOOKLP.,f](J'op.ev el OE cpof3"1(!'6p.-8a 'Tovs Ktvovvov<;, Els 'll'oAAas Tapaxas Karacrr~crop..v a-6r01!s. Isoc. vi. 107. Here what is feared is expressed by the emphatic future as a warning, while the alternative that is preferred has the subjunctive. See also DEM. xviii. 176, where .l 7rpoatp~f3 , , , , (J'OfJ-E "Jf'-EL'i, EL Tt OVfTKOII.OV 'lrE7rpaKTa 8 1] atOtS 7rpv'> ?Jp..a<;, 'T'OVTOtt pp.v~cr8at, if we sha.ll prefer to remember every unpleasant thing the Thebans have ever done to us, is vividly stated by the future, as this is the cour'se which the orator spec~ally fears and wishes to warn the people against; wl1ile he pnts his own proposition into the milder subjunctive form, ~~~ p..~vrot 7r<t(J'8fjr' ip.o~ 1<at 7rp0s T<fi (J'I<o7rel:v &A.A.d. fl--'J cptA.ovELI<Etv y~vYJcr8E. See also Isoc. xv. 130. In other cases it is difficult to detect any distinction, as in DEM. xxvii. 67 and xxviii. 21 (above), and in Hnr. i. 71; cf. Il. i. 135 and 137. 448. The future in protasis is also appropriately used when a future apodosis is implied in a pa.~t tense ; as in SOPH. 0. T. 843, L AE~El 'T't)JI aBrov apt8p..ov, OVI< iyw 'Krd.vov, if he shall tell the same number (it will follow that) I did not kill him. So EuR. Med. 1249. 449. This use of the future must be distinguished from its use in present conditions (407), where it is equivalent to pEAAw and the infinitive and cannot be interchanged with the subjunctive.
.! ,.., \ \

e' , , ,

, ,


In the Homeric language the following peculiarities appear in this construction:. 450. By far the most common Homeric form with the sub-




junctive in future conditions is ei K, often El p,f.v KE, El of. KE, etc. (218). "Hv also is frequent, being the only Homeric contraction of El av. El o' av occurs in Il. iii. 288, and Ei 7rEp av in Il. V. 224 and 232. "Hv 1rep yap ,( MJEA.wcnv is found in Od. xviii. 318.

451. El KE or al KE is sometimes found even with the future indicative in Homer. E.g.
A't KEV &vw Jp}.f}Ev 'L\.[ou 7rE</>LO~CTE7'UL ovo' crat, icrTw Tovro. 11. xv. 213. (See 196.)


452. The subjunctive with KE is sometimes used in the apodosis instead of the future indicative, thus making the apodosis correspond in form to the protasis. E.g. El OE K JL1J owvcrw, eyw Be KEV avTdS Uwp,at, and if he do not give her up, I will talce her rnyself. Il. i. 324 (compare i. 137). This
gives a form with two subjunctives analogous to that which has the optative in both protasis and apodosis (460). See 399. (For o in apodosis see 512.) For the Epic use of the future indicative with Ke or &v in apodosis, see 196.

86. El o' av TLS palvcrt e.wv fvt ol'V07TL 1l'OVTitJ, TA~IJOjLaL lv cr~8EO'CTLV exwv'TaAa7TEV8~a 8vp,ov. Od. V. 221. SoIl. i. 341, V. 258, xii. 223, 245; Od. i. 204, i. 188, xii. 348. Only these nine cases occur, and the more co:mmon use of the simvle d with the subjunctive in Homer is in general suppositions (see 468).

453. The simple El (without KE or civ) is sometimes used with the subjunctive in future conditions in Homer, apparently in the same sense as Et KE or ~v. E.g. El 7rEp yap O'E KO.TO.KTavv, ov cr' ET' eyti! YE KAaVcrop,at. Il. xxii.

r!lp' iyw, qov crTEpYJ8w. Id. o. c. 1443. So Ant. 887. El p.-~ q' h cp d. y w K r~cr8< r~> y~>, ov8i71'on (3twcrop,at. An. Eq. 698. So AEscH. Pers. 791; Eun. Or. 1534, I. A. 1240, d 1mcrOs (Mss.); all in dialogue. In Sappho 118, 1 we have a/: n> p'YJrat. 2. In Attic prose, this construction is extremely rare and always doubtful. The M~s., .however, have it in a few passages, as THuc. vi. 21: Ov vavnKf}S IJTpanas JLOVOV 0(~ aA.A.a Kai 'lrECov 7rnAtJV tvJL7rAi:v, /IA.)cws TE Kat El ~viJTwcnv ai 'IT'Ot\.t'LS <f>of37J8iuraL. (Here a few inferior Mss. read 1jv.)

454. l. Thib Homeric use of the simple El with the subjunctive in future conditions "as allowerl by poetic licensA in a few passages of the Attic drama, chiefly in tragedy, even in the dialogue. E.g. El yap Bavv> Ka2 TAE-v:r~cras cl<f>vs. SoPH. Aj. 496. ~vcna'Aawa





2. Optative in Protasis and Apodosis.

455. When a supposed future case is stated less dis tinctly and vividly than the subjunctive would state it (as if I should go in English), the protasis takes the optative with l. The a1)odosis takes the optative with &v to denote what would be the result if the condition of the protasis should be fulfilled. E.g.
El V..Bot, 1rrivr' &v Zoot, if he should gu, he would see all. EZ <I ovrws ~BeAot <j;tAe<tv Kf]ootr6 n 8vp,rp, rip KEV ns KE[vwv I'' Kal iKA<A<iBotro yrijJ-OW, if she should be willing thus to love you, etc., then some of them would cease even to think of nut?'?'iage. Od. iii. 223. "H K]1 '}'']()~(]'at llplafi-O'> ITptd.fLOt6 T 7raL8Es, aAAOL TE TpwES fLE'}'a KEV Kexapo aro BvfL(j}, el (]'<Pww rd.oe 7rrivra 71'VBo [aro fLapvafi-EVOtW. Il. i. 255. 'AA.A' d Ji-O Tt 1r8ow, r6 KEV 71'oAil Kepowv d17. Il. vii. 28. EZYJ> </JopYJTOS ovK &v, et 71'pri(J'(J'Ots KaA(;Js, you would not be bearable if you should evm be in 1J1'0SJle1ity. AEBCH. Prom. 979. OtKo<; 8' avros, El <P8oyy~v A.af3o, (]'a</J~crrar' &v A.etHEV. Id. Ag. 37. Ov8~ yap &v M~ooK6s fi-E (3acrtAEils J71'atvoYJ, El itEA.a{Jvotfi-' roils <VEpyras. XEN. An. yii. 7, 11. Ovo' El 1ravns U Bo tEJI ITp(]'at, 7rA~8et y< ovx il71'<pf3aA.oJJ-EB' b'.v roils 7rOAEJJ-ov>. Id. Cyr. ii. 1, 8. Ov 7rOAA~ av rlA.oyta tl'], El <Pof3o'iro TOV BJ.varov 0 rowvros; PLAT. Phaed. 68 B. El 8 ns roils Kparovvras rov 71'Af]Bovs M rlpeT1JV 7rporpElfLEV, &.jJ-<f;orEpov<; av ovf]crEL<. Isoc. ii. 8. EZ TLS TWV IJOL crvv6vrwv J7rap8d'ij 71'0LELV (]'V rvyxavas d>.oywJI, 7fW~ oiJK <'tv aBA.ufY'raros er']; Id. xi. 47. IIr;ls of>v OVK av olwrp6rara 71"UVTWJ1 EYW 7rE7rov80s dYJv, d JfL lf'Y)</J[cratvro dvat ~vov; how then should I not have suffered (lit. be hereafter in the condition of having suffered) the 1nost pitiable of all things, if they should vote rne a fo1eigner ? DEM. lvii. 44. (See 103 for other examples of the perfect optative.)

456. This form of the conditional sentence in its fully develo1)ed use, as it a1)pears in Attic Greek, must be carefully distinguished from that of 410 ; the more so, as we often translate both d'YJ av and 1)v by the same English expression, it wo1dd be; although the latter implies that the suppoBition of the protasis is a false one, while the former implies no opinion of the speaker as to the truth of the supposition. We have seen (438-440) that the more primitive Homeric language had not yet fully separated these two constructionR, and still used the 01}tative in the apodosis of present, and sometimes of past, unreal conditions. On the other hand, the distinction between this form and that of 444 is less marked, and it is sometimes of slight importance which of the two is used. As it is often nearly indifferent in English whether we say if we shall go (or if we go) it will be well, or if we should go it wou'ld be 'Well, so may it be in Greek whether we say Jd.v tA.Bwfi-EV





KaAws e~H or d EAOotp.EII KaA(os &11 gxot. In writing Greek, this distinction can generally be made by first observing the form of the apodosis in English; if that is expressed by should or would, it is to be translated by the Greek optative with fl11; if it is exlJl'essed by shall or will, by the future indicative. Other forms of the apodosis, as the imperative, will present no difficulty. The form to be used in the protasis will then ap1)ear from the principles of the dependence of moods (170-178); the optative will require anothe.optative with El in the dependent protasis, while the future indicative or any other primary form will require a subjunctive with M.11 or a future indicative with El. 457. In indirect discourse after past tenses we often find an optative in protasis, which merely represents the same tense of the subjunctive or indicative in the direct discourse. See 667, 1; 689 ; 694. For the occasional omission of &11 in an apodosis of this kind, see

458. The potential optative with &11 may stand in the protasis with El ; as in El EAOotp.t &11, supposing that I would go, easily distinguished from El f.AOotp.t, SU1JPosing that I should go. Such an expression does not belong here, hut is really a present condition. (See
409; 506.)

459. The future optative cannot he used in protasis or apodosis, except in indirect discourse to represent a future indicative of the direct discourse. (See 128 and 203.)

460. Er KE with the optative is sometimes found in Homer, and d 7rEp ;{11 occurs once.l This is a mark of the unsettled usage of the e[\rlier language, in which KE or :iv was not yet required with the subjunctive in protasis, and was still allowed with the optative or indicative (401). It is difficult to see any essential difference between these protases with d KE and those with the simple El and the optative. E.g. El &~ KEV "Apyos iKo[p.Ee' 'AxauK6v, o1;0ap dpo-6prys, yap.f3p6s Kf.v p.ot f.ot, and if we should-ever come to Achaean Argos, then he would (shall) be my son-in-law. Il. ix. 141 ; cf. ix. 283, and Od. xii. 345, xix. 589. ITws &v Jyw en ootp.t JLET' dOavaToun Ow'irnv, El KEV "Aprys OLXOLTO xpf.os JCaL &wpov dAv~as. Od. viii. 352. Twv KEJI TOt xap[a-aLTO 7raTTjp d:;n:pE[<rt', ti KEV Efl~ (wOv 1rE7rV()otT!J E1rt VYJV(]'LV 'Axatwv. Il. vi. 49. The distinction between these cases and those of 458 is obvious. In Il. i. 60, ,;: KEV with the optative forms a subordinate protasis,

ii. 597) ; besides 11. v. 273 ( = viii. 196) and Od. xvii. 223, mentioned in the text (461).

twenty-six cases of .t t with the optative in Homer, and one of

See the examples in Lange, Pa?-tikel El, pp. 185, 186.

er 7rp liv

There are




with a remoter and less emphatic supposition than the main protasis El oap.i (future) ; JIVII ap.p.E 7rd.Aw 7rAayx8f:wras Mw f!.lj; d.7rOVO<TT~<Tetv, d KfJI e&vaTOV YE cp-6yotp.ev, el 0~ op.ov 7r6Aep.6s TE oap.~ Ko.l Aotp.ds 'Axo.w-6s, now I think we shall be driven back and shall return home again-that is, supposing us to escape death---if both war and pestilence are at the swme time to destroy the Achaeans. In 11. ii. 597 we have 7rEp Clv O.VTal Mov<Tat ddooHv. These constructions are never negative.


461. In the strange protasis, d To-6Tw KE >..af3otp.Ev1 Il v. 273 and viii. 196, the separation of el from KE might compel us to recognise a potential force, if we could (possibly) secure these; but the difference between this and the Attic examples of el with the potential optative and ilv (458; 506), and the difficulty of seeing any difference between this and El ToflTw >..af3otp.ev, if we should secure these, induced Bekker to read el To-6Tw ye >..a{3otp.Ev here, and also TOV y' d p.ot oo[YJS (for TOll K' El) in Od. xvii. 223. The Homeric use of the optative in present and past unreal conditional sentences has been discussed (438).


462. In present or past general suppositions, the apodosis expresses a eustornary or repeated action or a general t1uth_ in present or past time, and the protasis refers in a general way to any act or acts of a given class. Here the protasis has the subjunctive with after present tenses, and the optative with el after past tenses. The apodosis has the present or imperfect indicative, or some other form which implies repetition. E.g. ''Hv Jyy:Us v..ev eavaTos, ovods ,13-:>vAETat Ovrf<TKEW, if (or when)


death comes near, no one is ( e>er) willing to die. EuR. Ale. 671. "Hv f'-~V aOYJ Kat V~cpov<Tt, x_piwvrat a:&rc{) ~V OE fk~ aOYJ, f'-ETtEUJ't. HDT. i. 133. 6wuAel: p.urwv, atiK ~v r{s n av,-uv &otKfj, &A.>..' Uv nva V7r07rTe>)<TYJ f3e>..,-tova ~o.vToil d'vat, he continues to hate, not if any one wrongs him, but if he ever suspects that any one is betteT than himself XEN. Cyr. v. 4, 35. "A1ras >..6yos, &v &1ri} ,.d, 1rpayp.aTa1 p.rf.Tatov Tt cpalveTat Kat Kev6v, all speech, if deeds are wanting, appears mere emptiness and vanity. DElf. ii. 12. 'Ed.v o o6~u Ta o[Kata E['Ka>..efv Kat l >.. Y/ TdJI oeopaKOTa TOV cpovov, ovo' OVTW dpw; y[yveTat Tov d..\.ov-ros-. Id. xxiii. 69 (so 74, 75, 76). 'A..\.X et 7't p.~ <f>tpotp.ev, b\,-pvvH cpepew, but if we ever stopped bringing him food, he always urged us to bring it. EuR. Ale. 7 55. El TtS d.vTE7rot, eti8:Us TE8v~KH, if any one objected, he was a tkad man at once (52). THUC. vi:ii. 66. Et' TtVO. 7r1!V8dvotTO vf3p(ovTa, Toifrov EOtKa[ev. HDT. i. 100. El o~ nvas 8opvf3ovp.~vovs aZcr8ot-




\ "' ' ,.., Q I ' \ .) "' To, TO ULTtOV TOVTOV I"TK07rWV kUTUI"T{-'VVVVU T'YJV Tapax1JV E7r tpaTo, whenever he saw any making a disturbance, he always t?ied, etc. XEN. Cyr. v. 3, 55. OvK a1r A.d11"To i!n avTov, d p.~ n &.vayKa!ov d'YJ, he never left him, unles.~ there was some necessity fo1 it. Id. Mem. iv.J, 40. "H V TOtS fLElJ ocf;BaA.p.oZs f7rKOVp1]fLU r'ljs xt6ros, d TLS p.~J..uv Tt i!xwv 7rp0 TWV ocpBaAp.wv 7r0pVOTO, TWV OE 7rOOWV d TLS Ktvoi:To. Id. An. iv. 5, 13. 'E7rt.Sh) 8 tF.toov avrov TaxurTa, rrvA.A.a{36vns ayovaw aVTLKpvs 6!s U1l"OKTVOVVTES, OV7rP Kat TOtJS aAAous aTre(J'cpaTrov d nva ArJ(J'T~v 1) KaKovpyov (J'VAAa{3ov, i.e. where they had been in the habit of killing any others whom they took. LYS. xiii. 78. 463. This optathe referring to past time must be especially distinguished from the optative in ordinary protasis referring to the future (455). El and Uv in this construction are often almost equivalent to OT or ora.v (which are the more common expressions), and the protasis has precisely the same construction as the relative sentences of 532. 464. The present and aorist subjunctive and optative here do not differ except as explained in 87. The future optative of course is never used here (128).

465. The examples in 462 exhibit the ordinary Attic usage. In Homer we find this construction in a partially developed state: see 468. 466. The gnomic aorist ( 15 4) and other gnomic and iterative expressions (162; 163) may be used in the apodosis of these general conditions. The gnomic aorist, as a primary tense, is followed by the subjunctive. E.g.
''Hv (J'cpaAW(J'LV, J.vTEA7r[rravTES aAAa.~7rA~pW(J'UV T~V XP[av, they fail, they always supply the deficiency, etc. THUC. i. 70. "Hv of. TLS TOVTWV T 1rapa{3avyJ, (YJp.[av avTots ~Tr~BHrav, they (always) impose a penalty upon evmy one who transgresses. XEN. Cyr. i. 2, 2. EZ TWS LOOLEV 7rrJ TOVS (J'cj;Tepovs ~7rKpaTOVJ!Tas, UV8ap(J'Yj(J'UJI av, VJhenever any saw their friends in any way victorious, they would be encoumged (i.e. they were m~couraged in all such cases). THUC. vii. 71. See XEN. Mem. iv. 6, 13, quoted in 162.


467. (Indicative.) The indicative is sometimes found in the place of the subjunctive or optative in these general conditions, that is, these follow the construction of ordinary present and past suppositions, as in Latin and English. Here the speaker refers to one of the cases in which the event may occur, as if it were the only one,-that is, he states the general supposition as if it were particular. E.g.
MoLpo:.t s> dcpcrravT}, r TtS Ex8pa 7r~Af.t OfLo'Ybvot.s, alOW KaAI.Jif;at, the Fates stand aloof to hide their shame, if there is enmity among kindred. PrND. Py. iv. 145 ; cf. Ol. i. 64. (See 406.) EZ ns ovo 1] kat TrAeov>




n> iJfh~pas Aoyl{;erat, f-U1'at6<; EO'rw, if any one ever connts upon two or even more days, he is a fool. SoPH. Tr. 944.. 'EA<v0pws ?ToAtnDofJ-<V, 0~ 6py~s TOV ?TEAas, cl Ka8' ?]oov1v 7't opfi, xovns, i.e. not (having a habit of) being angry with O!t1' neighbour if he ever ads as he pleases. THUC. ii. 37. (Here the indicative oplj, is used as if some particular act of one neighbour, and not any act of any neighbour, were in mind.) Et y&.p ns v BYJJ-1-0Kpctr[?- 'TETtfJ:rwvo<; 'TOAfhi fJoYJ8E'iv roi>;; 1rap6.vo/ha ]pa<f>ovuw, Ka'TaA1~t 'T~V ?ToAt'TE[av v<f>' i]s 'TETlfJ-YJ'Tat. AESCHIN. iii. 196. EZ Tls Tt E?TYJPWTa, d?TeKplvovro, if any one asked anything, they replierl (to all such). THuc. vii. l 0. 'Efh[O'tt O~K d 'TLS KaKWS mfuxwv ~f-VVETO, d'AX d Tt<; EVEPYETODfJ-EVOS axaptO''TOS </>a{vo t'TO. XEN. Ag. xi. 3. Here, without any apparent reason, the writer changes from the indicative to the optative. (See 534.)


468. In Homer the subjunctive appears in protasis in general suppositions (462) only nineteen times, and the optative only once. Here the subjunctive generally (in fourteen cases) has the simple El (without K~ or ltv ). E.g.
Er 1Tp yap TE x6A.ov YE Kat avTqp.ap KaTa?TEfYJ, dA.'Aa ')' /<0.2 jJ,E'T6?Tt0'8ev fXH K6Tov, o<f>pa TEAE(J'O'YJ, for even if he swallows his wrath for the day, still he keeps his anger he1eajter, until he accomplishes its object. IL i. 81. T wv ov Tt fJ-ETa'TpE?TOfJ-' ov8' dA.Eyl(w, d 7', E?Tt 8e~[' iwO't, d T' -;( dpurTEpa, I do not heed them nor care for them, whether they go to the right or to the left. Il. xii. 238. SoIl. iv. 262, x. 225, xi. ll6, xvi. 263, xxi. 576, xxii. 191 (the last four in similes); Od. i. 167, vii. 204, xfi. 96, xiv. 373, xvi. 98 (r 116). "Hv ?TOTE 8aufJ-bS ZKYJTa t, O'Ot Tb y~pas ?ToAV f-E'i(;ov, if ever a division comes, your prize is always much g1eater. Il. i. 1G6. So Od. xi. 159, qv P-0 ns EX'[) Besides these two cases of ~v, Homer has two of fi KE, Il. Xi. 391, Xii. 302 ; and one of et 1Tp dV, Il. iii. 25 (ftve in

The single case of El with the optative in a past general condition in Homer is Il. xxiv. 768: d'AA' Ei Tls fh Kat O:'A'Aos v?T'Tot, &A.A.d. O'V Tov '}'E Ka'Tf:pvKei>, but if any other upbmided me, you (always) restmined him.

469. Pindar has only eight cases of the subjunctive in protasis. These all have general suppositions and all have the simple el ; 1 as ?ToAA.oi o fJ-EfJ-VavTat, KaAOv et 'Tt ?TO va Bfj, but rnany remember it if a noble work is done, 01. vi. 11.

470. The other lyric and elegiac poets show no preference for the simple d. The following cases may be cited: CALL. i. 13 el V (but ijv


profess to be complete.

2G6, 27:~ (perhaps also 263) ; Nem. vii. 11, ix. 46 J lsth. iii. 58, iv. 12; Frag. 171 (Bock,h), 5. The references to the otl1er poets in 470 and 471 do not

Jour. Phil. iii. p. 443.

The examples are 01. vi. 11 ; Pyth. iv.




in 1 i) ; TYnT. xii. 35 d ,P{;y'[] (hnt ?JV xi. 16); SoL. iv. 30 d iJ 1 (hut i}v xii. 1, xiii. 29); THEOG. 121, 122 El AE>-..~817 .. iix17, and 321 d ?nraO"<r?J (hut ~v 93, 186, 379, 697, 929, 932, 1355, 1356, 1385); all (Uoth tl and ?jv) in general conditions. See Snf. AMORG. vii. 15,
69, 97 (1jv).

471. In the Attic poets we find a few cases of the siiDj)le El in general conditions. E.g. 'AAX avop(L, K Er 'TL" 1J <rorpos, TO fW.V8avf.JI m5A.X a.iO"xpcJl! oDOEV Kat TO fL~ TE[vnv IJ:yav. SoPH. Ant. 710. So Aj. 521; 0. T.198, 874; 0. C. 50[); AEscu. Supp. 91, Emn. 234. For the simple d in future conditions, see 453 ; 454. For the prolxtble relation of El to d KE, ijv, Edv, etc., see 40 I.

and Ellipsis in Frotas'is.-1'1otasis without a Ve1b.

472. Often the protasis is not expressed in itR regular form with El or Uv, but is contained in a participle, or implied in an adverb like oihws or OtKa[ws, in a preposition with its case, or in some other form of expression. \\"hen a participle represents the protasis (841 ), its tense is always that in which the verb itself would have stood in the indicative, subjunctive, or optative. The present (as usual) includes the imperfect, and the perfect includes the pluperfect. E.g. TovTo 1rowvFns ED 1rpa~ovO"tF (i.e. Jal' ;;-o l<VO" tJ!), if they (shall) do
this, they u-ill prospe1. TovTo 7rot,jO"aJ!TES ED 1rpa~ovO"tV (i.e. <iav 1ro (]"WO"l]!). TovTo 7rOWVJ!TES ED dJ! 7rpaTTOlEV (i.e. el 7rOlOLEV), if they should do this, they would prospeT. TovTo 7rot~O"a.vns ED aJ! 1rpaTTOtEV (i.e. tl7iot'l}a-attv). 'To.UTo 7rOtoVvTES' tD Uv g1rpa.:TTOV (i.e. El f.7ro{olll'), if they were doing this (or if they had been doing this), they would be in prospMity. To.VTo 7r0l1JO"<XJ!TES ED av E7rpaTTOJ! (i.e. El E7r0 c.,,(J'a J!), if they had done this, they wo-uld be in p1ospe1ity. Ilws 8-fjTa' 8Dcqs o1J0"1)S 0 ZEvs OVK dm5AwAEV TOV 7raTEp avTOV o,jO"as; i.e. how is it that Zeus has not been dest1oyed, if Justice exists? AR. Nub. 904. (Here 8KTJS OlkT'']S represents El o{K1) JO"T[v.) 'AAA.' El(]"(JfLE0"8a OOfLovs 1rapaO"TE {X o VTES (av 7rapaO"TdxwfLEJ!), but we shall know, ifweenterthehouse, SoPH.Ant.1255. 2:v o KA<iwJ! EiO"El Taxa (EdJ! KAV1J>), but yon n:ill soon know, if you, listen. AR. Av. 1390. So fh1J fLae..:)]!, unless I leam, for av fL'J fLaew, Nub. i92. Ka[ J<EJ! TovT' Wi>..o'fL' tu6s ye otooT'Tos dpO"Ba, (6,os OoovTos=El ZEvs otoo[lJ), and this I shonld like to obtain, if Zeus would only give it. Od. i. 390. TowvTa TaJ! yVJ!at~i O"vJ!va[wv f.xots (El O"Vl'v<dot>), such things 1could ym su.f}'er, if you should live with wom-en. AESOH. Sept. 195. 003' av t:rlW7r~O"UlfLl T1JJ! aT1)JI opwv O"TdxovO"aV aO"TOLS (i.e. El op<l)v). SOPH. Ant. 185. 'ABT}vaJwv OE Tb aDTo TOVTO 7ra8ovTwv, omA.a,o-{av a.v





T~v Mvap,w dKa&w-Bat (o'lp,at), but if the Athenians should ever suffer this (1ra&ovTwv =El 1ra8otev), I think it would be injmed that their powm wa;; twice as great. THUO. i. 10. (Here only the context shows that 1ra86vrwv does not represent El ~1ra&ov, if they had ever suffered.) llptv yv~cr&at ~1rlcrT7JCTEV ltv ns dKovcras (i.e. el ~Ko1xHv), before it happened, any one would have disbeliewd su~h a thing if he had heard it. Tauo. vii. 28. OD yap &v P,Ta1rd&w ilp,os JNra P-'J rotaVT1JS oiJu17s T~s 1nrapxovcr1]s -inroA.~tj;ews, for he would not be seeking to change your minds, if such were not the prevailing o-pinion (i.e. el p,1) row.VTYJ ~v). DEM. xviii. 228. ''Eo-rw ovv 07TWS TO.VT 1 <iv, JKEtVa 7rpoetp1]KWS, 0 aDros dv~p p,~ (hacp&apet<;; -roAfk1)<J'EV d1rel:v; is it J'o,ssible then that the same man, after saying that, would have dared to say this unless he had been corrupted (d fk'l otecp&ap1J)? Id. xix. 308. M~ KO.T'l]yop~uavros Alux{vov fi-1]0~V ~~w T~S ypacp~s oDD' <iv Jyw A.oyov ovo~va E1f'OW1~f1-YJV upov (el 11-~ KO.TrJyoprJrrEv). Id. xviii. 34. Ta avra <iv E1rpa~E Kat 1rpwrYJ r\axovrra (i.e. el 7rpWTYJ EA.axev), it (the soul) would have done the same, even if it had had the forst choice by the lot. PLAT. Rep. 620 D. Mafkp.av o' <iv alT~uavros i)Kov UOt cp~pwv <iv ilprov, and if you ever asked for something to eat, I used to corne bringing you b1ead. AR. Nub. 1383. (Here alr0rravros represents El ah~o-<ta'> in a general supposition, 462. For ~KOV av see 162.) Olire rr&Covrrt 1rAdw ~ ovvavrat cp~pnv, 8wppaye'i:JI yap &v oiJr' &p,cptEVVVVTat 1rAE[w ~ OYVO.VTO.t </>~.pew, a1r01rVtyei:ev yd.p &v, they do not eat more than they can bear, for (if they should) they would b1trst, etc. XEN. Cyr. viii. 2, 21. Avro~ av J7ropE-68YJUO.V 0~ &A.A.ot Ta o' -inrot..)yta ovK 1jv fiAATJ ~ TaVTTJ ~Kj3~vat, they would have gone thernselves wheTe the others went ; but the anirnals could not go otherwise than as they did. Id. An. iv. 2, 10. So ~ yd.p av A.wf30<J'ato, Il. i. 232. . 'HfktV 8' J~ 7TOAA~s <iv n-eptovrras vewv fkOAtS TOVTO il1f'~PX~ Ka'i fk0 dvayKa(op.vots, wrr1rep vvv, 1rao-ats cpvr\arrO'etv, but we sho1ld ha1dly have this advantage if we had a g1eat supmiority in nurnber of sh1:ps (=El 7roAA0v 1repwvuav etxop.Ev) and if we we1e not compelled (el fk0 1)va.yKa.(ofk<Ba), as we a1e, to use our whole fleet in guarding. THuc. vii. 13. 1 To p,f.v E7r EK<lv'l! 7rOAAaKtS av OtA-68Yjrra.v, if it had depended on him, they often would have been disbanded. Isoc. iv. 142. 6uf y< aD roils '1raAat llv a1roA~AetTE, if it had depended on yourselves, you would long ago have been T1tined. DEM. xviii. 49. (So sometimes Ka&' vp.a.<;.) yd.p <iv ~v<Ka ye 1{1]</>t<J'fkarwv Joeo~K<< 0tK1Jv, for, if dec1ees were of any avail, he would long ago havesujfe1ed punishment. Id. iii. 14. {Here the protasi,s is implied in v<Ka tf1JcpLO'fkaTwv.) Oihw yap OVKET rou r\ot1rov 7rr5.o-xotfkEV <iv KaKws, for in that case we should no longer suffer. Id. iv. 15. So WS OVTW 7TptyVOf1-VO<; av, XEN. An. i. 1, 10. 0-.lo' av O,tKdws ES KaKbV 1rEUOtpJ n. SoPH. Ant. 240. In such cases the form of the apodosis generally shows what form of protasis is implied. When the apodosis is itself expressed by an infinitive or participle (479), as in THuc. i. 10 (above), the form of the protasis is shown only by_the general sense of the passage.






473. The future participle is not used to represent the future indicative in future co!l<litions (447); it may, however, represent the future in present conditions (407), where it is equivalent to JLEAAw and the infinitive; as in DEM. xxiv. 189, JL~ 7rEpt To-6,.wv DJLWV ol0"6vTwv T~v !ffi1>ov, ,.[ 81: TaVTa AeyovTa vox AEtv fU vvvt; if yov are not to give your vote about this, 1'-~ ol0"6vrwv representing El JL~ o!O", n =El 1'-~ JLEAAETE q,epEw. The present and 1wrist participles, when they represent the present and aorist subjunctive, express future conditions, thus making the future participle unnecessary. The aorist participle in protasis can always represent an aorist subjunctive in the sense explaineu in 90. 474. The verb of the protasis is suppressed in the Homeric El8' aye, come now I This is commonly explained by an ellipsis of, if you will, come now I But it is probable that no
definite verb was in the speaker's mind in such expressions, even when we find it necessary to supply one. E.g. El 8' ayE, TOt KEq,a> KaTaVEVO"OjLat, come now! I will nod my assent to thee. Il. i. 524. El 8' ayE fl~V, 7rdp'Y)O"at, rva yvwWO"t Kat ot8E, well! co'llJ;e now, try it. Il. i. 302. El JLEV 8~ 8E6s EO"O"' 8wi6 TE ~KAvE<; aiJ8~s, l 8' ayE JLOt Kat KEtVOV ot&vp?Jv KaTcJ.Ae~OV (the apodosis being introduced by El 8' aye, corne now, tell me). Od. iv. 831.

475. ('.!:h d.) There is a probably unconscious suppression of the verb of the protasis when dJs El or dJs d TE is used in comparisons (especially in Homer) with a noun or adjective or with a participle. E.g. Twv veEs cbKEuu ws El 7rTEp6v 1) v6YJJLa, their ships are swift as (if) a wing or thought. Od. vii. 36. '.Qs I riO"v1>YJAov ~pE~Ev 'Arpd8YJS w<; Ei' nv' riT[JLYJTOV JLETavciO"T?]V,for the son of Atreus insulted me like (i.e. as if he were insulting) some despised wanderer. Il. ix. 648. 'E1r AEOJLEV Bopeu riVEJL'f PYJL8ws ws 'TE KaTa p6ov, we sailed on with the northeast wind easily, as if (we were sailing) down shearn. Od. xiv. 253. In all these cases no definite verb was in mind after El, but the addition of El to ws shows that a conditional force was felt (at least originally) in addition to the comparison ; and this is the only difference between these examples and those with the simple ws or ws TE, as ~O"T~KEW &s ,.[, TE Aewv, he stood like a lion. 1 In .Attic poetry we find p,6.n7p WO" t ns 7rtO"Ttt, like sonM faithful mother, SoPH. EL 234 ; and 7rT1JO"a> WO" l T 8vO"jLEvij, spu1ning her ltS an enemy, .Ant. 653. With Od. vii. 36 compare Hymn . .Ap. Py. 8, 1rp6s "OAvp,1rov WO"TE v6YJJLa dO",, and 2 7 O, E7rt v~a v 617 I ~ s iLho 7rE'T0"8at.


1 See Lange, Partikel El, p. 234. Lange is at great pains to show that there is no ellipsis here, or indeed in any cases of <i without a verb like <! 1rep dva-yK?), if necessary. By" ellipsis" we often mean merely what one language finds it necessary to supply to translate an idiom of another. There are few ellipses of which a speaker is really conscious when he uses them. In this sense, it seems to me that, whenever we use if without a verb, there is at least a supp1ession (if not an ellipsis) of a verb.




'O>..o<PvpJJLEVOt ~. d eava:rovOE KLOVTa, bewailing him as if going to his death (in full as if they were bewailing him going), for which we say (changing the construction) as if he were going. Il. xxiv. 328. See also . rl,' ~' ' ' 't ' ~ ' ' ' Il , XVl, 1 9 2, V, 3 74. 'AJl-'t'L O K(J,7TVO<; ')'L')IVETat ' UVTTJ'> W<; Ey17TVpOS al&oJLf:vow, i,e. the smoke rises from it (the fountain) as if (it rose) from a blazing fire. Il. xxii. 150. So Od. xix. 39. What seems like a more natural construction with ws El or ws Et TE is that of the optative with the apodosis suppressed (485). In all these cases there is also a suppression of the verb of the apodosis (see 485). For the participle in such expressions se.e 867-869.

476. (El JL~-) El JL~ is used without a verb in various expressions to introduce an exception. 1. With nouns and adjectives. E.g. Tts y&p TOt '.Axatwv aAAOS 6JLOWS, d "'~ ILfTpOKAO<;; who is like to you, except (unless it be) Patroclv..s? Il. xvii. 4 75. See Il. xviii. 192,
xxiii. 792 ; Od. xii. 325, xvii. 383. Such expressions are like the tlJLewov, if this is better, Il. i. 116; d n6v 7TEp, xiv. simple el T6 125; et 7Tep dvayKTJ, xxiv. 667.

2. With participles.


El JL~ KPEJLriU'as TO v6'1JJLa, i.e. I could never have done it, except by suspending thought. .AR. Nub. 229. So ovoev 7TOT' el JL~ ~l!v&avOVJLEV'fJV, AESCH. Ag. 1139 ; el JL~ KaTaot)(]'aVTES, THuc. vii. 38 ; Jd.v JL~ T~s doda<; oo8E[U''IJS, DEM. xxiv. 46. 3. In the expression el JL~ otd. TOVrO (or TOVTov). E.g. K(Lt d JL~ Otd TbV 7TpvTavw, EVE7TEU'V av, and, had it not been for the Prytanis, he would have been thrown in. PLAT. Gorg. 516 E. (Compare ota ye VJLa>, DEM. xviii. 49, quoted in 472.) Ov yd.p ws d JL~ otd. AaKEOULJLOv[ov<;, ovo' ws cl JL~ llpo~EVOV ovx V7T0E~avTo, ovo' El JL~ Oi 'H ')'~U'7T7TOV, ovo' ws el JL~ Otd. TO Kat TO, EU't!J(J'IJO'UY &v o1 <PwKEtS, ovx OVTW T6TE U7T~')'')'ELAV, for he did not then repmt that if it had not been for the Lacedaemonians, or if they had not refused to receive Proxenus, OT if it had not been for Hegesipp1LB, or if it had not been jo1' this and that, the Phocians would have been saved. DEu. xix. 7 4. 4. In the rare expression d JL~ except if, except in case that.



'0 XP'IJJLUTLU'TLK6s T?)v Tov nJLaU'8at ~oov~v ry n)v Tov JLav&avew ovoev6s d~lav <P~U'H Elvat, El JL0 d TL UVTWV dpyvpwv 7TOL~ the money-maker will say that the pleasure of 1eceiving honour or that of learning is not worth anything, unless (it is worth something) in case either of thern produces money. PLAT. Rep. 581 D. In Prot. 351 0, yw yap AEyw, KafJ' 6 ~Of.a EuTLv, dpa Ko.Td. ToVro oVK J:ya80., fL~ Et 'Tt &,,r UUTWV &rrof3~U'Tat aAAO ;-fo'l' r aslc this: 80 fa1' as they a?'e pleasant, are they not just so jar good, without taking into account any other result (i.e. other. tban their pleasantness) which may come from them ?-JLfJ is not a mistake for El JL~, but it seems to imply a conditional participle like




1xrroA.oyt(oJMVO<> (though no precise word can be supplied), very much as P.0 and P-0 01rwS imply a verb of saying (707). The meaning clearly is, Are not things good just so far as they are pleasant, if we take no account of any other (i.e. unpleasant) element in them ? This sense would hardly be found in the emended reading el p.~ n. In Tauc. i. 17 the Cod. Vat. reads el p.~ n, although el P-0 et TL can be under~tooa as in PLAT. Rep. 581 D (above).


477. Equivalent to <l P.0 el ( 4 7 6, 4) is 1rA el, except if or

1rA~v represents the apodosis. E.g. Ta ov6p.ara ol'6v TE O.VTWV elo~vat, 7r A~ V er Tt> KWJl-lf>OLO'TrOLdS rvyxcfv<t ~v, it is not possible to know even their names, except in case one happens to be a comedian. PLAT. .Ap. 18 C.

unless, in which

478. In alternatives, el o~ p.~, otherwise, regularly introduces the latter clause, even when the former clause is negative. Et o~ p.~ is much more common here than Jd.v o p.fJ, even when Jd.v p.~v with the subjunctive precedes. The formula el o~ p.~ was fixed in the sense of otherwise, in the other case, and no definite form of the verb was in mind. IIpils ravra P.0 dJ7rT d o~ p.0, <TaVT6v 1ror' alnd<T<t, therefore do not beat me; but if you do, you will have yourself to blame for it. AR. Nub. 1433. El p.~ eavovp.a[ -;. d 0~ p.~, OV Ae{tf;w 'TrOT~, if I do not die (I will leave the place) ; otherwise (if I die) I shall never leave it. EuR. And. 254. See SoPH. Tr. 587. II6Aep.ov ovK <twv 7rotel:v d 0~ p.~, KO.C O.VTOC dvayKa(l'e~ue<Tfiat ecpacrav cp{Aov> 1r0Lft<T8at oils ov f3o{,A.ovrat, they said that otherwise (El o~ p.~) they sho1dd be obliged, etc. Tau c. i. 28. El'1rov (ITav<Tav[\l) -rov K~pvKo> p.~ Ae7rerr8at el o~ p.~, 7rOAep.ov avr~ 2:7rapn6.ros 7rpoayope{,nv, they OTdered him not to be left behind by the herald: and if he should be (el o~ p.~), (they told him) that the Spartans declared war against him. Id. i. 131. M~ 'lrOL~<TV> 'TCLVTO.. d 0~ p.~, alr[av g~ELS. XEN. An. vii. 1, 8. 'Ed.v p.~v 'TL vp.l:v OOKW dAYJ8ES A~yELv, ~11VOJl-OAoy~<TCLTE. d 0~ p.~, 7rO.VTt A.6yp avnT<[vere. PLAT. Phaed. 91 C. So av p.Ev 7re[crYJTE, .. el o~ p.~, K.r.A., DEM. ix. 71. El o alone is sometimes used for el o( JL~ ; as in PLAT. Symp. 212 c, el p.~v f3o{,, . . . el o. So el o' ovv (se. p.-IJ), SOPH. Ant. 722, EuR. Hipp. 508. The potential optative and indicative with av, so far as they are apodoses, might be classed here; bnt these have higher claims to be treated as independent sentences. See Chapter IV., Section I.
S1tbstitution and Ellipsis in Apodosis.
479. The apodosis, in any of its forms, may be expressed by an infiniti1re or participle, if the structure of the sentence requires it.




l. It may be expressed by the infinitive or participle in indirect discourse, each tense representing its own tenses of the indicative or optative, the present including the imperfect, and the perfect the pluperfect. If the finite verb in the apodosis would have taken av, this particle is used with the infinitive or participle. E.g.
'H yovftat, El TOvTo 7rotei:Te, miv-ra i<a..\ws ~ xo v, I believe thut, if you a:re doing this, all is well. 'H yovftat, av -rovro 7rot~Te, 1nfvTa KaAws E!~uv, I believe that, if you (shall) do this, all will be well. 'H yovt-tat, El Tov-ro 7rotoZTe, 1rav-ra KaAws &v ii X et v, I believe that, if you should do this, all would be well. 'H yovftat, el TOVTO E1f'Ot~o-a.n, 1rav-ra KaAw> &v l X EL v, I believe that, if you had done this, all would now be (or 1vould have been) well. Oloa vt-tiis, av -rovro 7rotqTe, 1r pagovras, I lmww that, if you do this, you will prospe1. IIws yO.p oteo-th ovo-xepws aKo-6etv '0..\vvetovr;, El Tls Tl Aeyot KaTa 'PtA[7r1f'OV KaT JKeivovs To-ils xp6vovs; how unwillingly do you thin!c the 0. heard it, if any one said anything against Philip in those times ? DEM. vi. 20. (Here dKovew represents the imperfect 'lJKovov, and el .\~yot is a general supposition, 462.) For examples of each tense of the infinitive and participle, see 689. For the use of each tense of the infinitive or 1)articiple with av and examples, see 204-208; 213-216.


2. It may be expressed by the infinitive in any of its various

out of indirect discourse, especially by one depending on a verb of wishing, commanding, advising, pTeparing, etc., from w_hich the infinitive receives a future meaning. Such an infinitive is a common form ef future apodosis with a protasis in the subjunctive or indicative. E.g.
BovAe-rat JA()e'iv Jav -rovro yv']Tat, he wishes to go if this shall be done. IIapacrKeva(6fJ-eea a7rd.()e'iv ~V OVVdJftEea, we an preparing to depa1t 1j we shall be able. KeAevH ere a7re.\()eZv el (3ovAet, he bids you depaTt if you please. (See 403 and 445.)


d7roeavel:v, whereas he might easily have been acquitted (dcpeUJ'l} av), if he had done any of these things even in a moderate degree, he chose to die. XEN. Mem. iv. 4, 4. LKEftftaTa TW]J pq.olw; a7rOKTlVVVVTWV Kat dva.{3tW<TKOftEVWV y' O.v, el oio -re 1)crav, considemtions fm those who readily put men to death, and who would bring thern to life again too if they could. PLAT. Crit. 48 C. (' Ava{3tw<rKOftEVWV tl.v = dv<j3u~o-KovTo tl.v.) '!ls ol6s T 4v o-e <ri{>Cetv el ijfJe..\ov dvaMcrKel!l XP~ftaTa, whereas I might have saved you if I had been willing to 8Jlend money. lb. 44 B.

3. The apodosis may be expressed in an attributive or circumstantial participle. E.g. 'P?-Mws av dcpe()ds El Kat ftETpws n TOV7'WV J7rof1)<TE, 7rpoeAeTO

480. A verbal noun may take the place of an apodosis. E.g. '!h OVT &va<TTTJT~pa KaOftE{wv xeovos el ft~ eewv TlS f-jl1f'OOWV




~CTTYJ 8op[, as one who would 'have laid waste ( = avECTTl](]'<V &v) the Cadrneans' land, if some one of the Gods had not stood in the way of his spear. AESOH. Sept. 10 15.

481. Other forms in which an apodosis may appear, as a final clause, need no discussion. (See 445.) In indirect discourse, after past tenses, an optative in the apodosis often represents an original indicative or subjunctive. (See 15 and

482. The apodosis is sometimes omitted, when some such it is well or it will be done can be supplied, or when some other apodosis is at once suggested by the context. E.g. 'A'AX d p,~v 86J(]'OV(]' ypas p,eya8vp,ot 'Axawl, ap(]'aVTtS KaTd
~xpression as

8vp,dv, 07rWS dvragwv ECTTa1 - el 8E K Jk~ 86Jw(]'V1 Jyw OE KV EA1upat, if they give me a prize,-well; but if they do not, I shall take one for myself. Il. i. 135. (Here we must understand something like e\3 ~et, it will be well, after ecrrat.) EZ 7r<p yap K' EB>-.:ycrw 'O'Avp,1rws dCTTep07r1Jn)s ~ 8wv crrv4>eM~at -6 yap 7roAi> 4>prar6s icrnv. Il. i. 580. (Here we must understand he can do it after the protasis. The following yap refers to this suppressed apodosis.) El p~v f.yw ilp,as tKavws 8toacrKw olovs 8EI: 7rp'Os dAA~Aovs Elvat -l 8~ p~, Kat 1rap?i rwv 7rpoyeyv1Jpevwv p,av8avere. XEN. Cyr. viii. 7, 24. ~e'ivot 7rarp6Jtot evx6p,dJ' i{vat, l 7rEp T yepoyr dpl]at E1rA8wv AaepTlJV, we boast that we are friends by inheritance, (as you may know) if you go and ask Laertes. Od. i. 187. ITpO(]'l]yopv81JS ~ Llt'Os KAnv~ 8ap,ap JLEAAovd E(]'(T8', t TWVO 7rpocrcravn ere T .AESOH. Prom. 834.

483. Sometimes the adverb C:,v, without a verb expressed, represents an apodosis in the indicative or optative, when the verb can easily be supplied. E.g. Ot olKeTat peyKov(]'tv dAA' ovK &v 7rp'O rov (se. ovrws tppyKov), but they would not have been sno1ing at this late hour in old times. .AR. Nub. 5. (See 227.) So 7rWS y?ip av; (se. dlJ), how could it be?
484. there is apodosis wcr1rep.
In 6J(J'7r<p &v d with a noun, as 6Jcr7rp av el 7raZs, like a child, originally a suppression of the verbs of both protasis and (227 ; 485) ; but in use the expression hardly differs from (See 868-870.)

485. ('52s d and wcr7rep d.) There is an unconscious suppression of the verb of the apodosis when el, EZ TE, and 6xr7rep el are used in similes and comparisons. E.g. Aaol E7rov8', ws d re fLETd KTDwv E(]'7!'To p:ljAa, the hosts followed as if sheep followed a ram. Il. xiii. 492. (No definite verb is understood here, either with ws in Greek or with as in English, but the origin of the expression is the same in both.) <i>1aAav ws eZ ns 8wp+ crerat. PrND. 01. vii. 1. Ka p< <j>[A'Y)d ws <Z T< 7rarTJP 8v 1rai:/la 4>tA~crv. Il. ix. 481. Oi 8' ap' i:crav ws eZ TE 7rvpl x&wv m'icra V~JLOGTO, i.e. their march was as if the whole land should flame with fore (originally






as it would be if, etc.) Il. ii. 780. B-lj o' tftEV, 7raVTo<T xerp 6p~ywv ws cl nwxos d1J, holding out his hand as if he had long been a beggar (438). Od. xvii. 366. For other optatives with ws cl, see Il. xi. 467, xxii. 410; OJ. ix. 314, x. 416, 420. "f!rnrep 1d 7rape~rTa,-w;, as if you, had dwelt near by. AESCH .Ag. 1201. ''Ottota W!T7rep et ns 1roAAa J~r&iwv ft1JOE7ron Jft7rL'Tf'Aat-ro, jnst as if one should eat much and neve1 be filled. XEN. Symp. iv. 37. There is the same suppression of tl1e apodosis in the examples in 4 7 5, where the protasis also is wanting with cl and similar expressions.


.Apodosis C()ntained in the Protasis.

486. A protasis may depend on a verb which is not its apodosis, the real apodosis being so distinctly implied in the form of expression that it need not be stated separately. 487. 1. This is found especially in Homer, where et Ke (ai: Ke) or ijv (without an expressed apodosis) often seems to have the force of in the hope that/ as in 1rarpos ttov KAo<; tterpxottat, ~v 1rov dKo.Ua-w, I am going to seek tidings of my father, if I shall chance to hear of him, i.e. that I 1nay hear of him if perchance I shall, or in the hope that I shall hear of him (Od. iii. 83). Here the protasis carries with it its own apodosis, which consists of an implied idea of purpose. 1 The whole sentence (both protasis and .apodosis) is thus condensed into the protasis; but the apodosis is always felt in the implied idea of purpose or desire which is inherent in the idiom. As we have seen (312, 2) that final clauses with /{v or ~~ and the subjunctive originally included both a conditional relative clause and a final sentence, so here we have both a conditional and a final force included under a single conditional form ; and this double force is felt also in the English translation, if haply, in the hope that, in case that, etc. E.g. Aurap a-ol ?rVKwws im-oe~a-ott<&', aE KE 1rl&1Jat, but we will make
you a wise suggestion, jm you to obey it if you will. Il. xxi. 293. (Here the protasis aE K 'Tf'[(J'f)at with its implied apodosis seems like 7rd8oi av d 7rBo,', you can obey if yov. please, AESOH . .Ag. 1049, and xapotf aJI El xapotr', Ib. 1394.) So Il. i. 207, 420, xi. 791, xxiii. 82; Od. i. 279. IT~fi-fW 8' ES L7rapT1]V v6(T'TOV 7rV(T6ftVOV mnp'Os <f>>..ov, ~v 1rov rlKo.Ua-v, 1}8' ha ruv 1<Aeos ~xva-w, I will send him, to Sparta, to aslc about his fatheT's nt11.rn, in hope that. he may hear of it, and in order that glOTy may possess him. Od. i. 93. (Here the
1 The English translation of certain conditional clauses in the New Testament wl1ich have this peculiar construction preserves the sense of purpose or desire with the original form of protasis. Thus, that they should seek the Lord, if haply they might feel afim him and .find him, Acts xvii. 27 ; and he came (to the fig tree), if haply he might find anything iheTeon, MARK xi. 13.




added final clause shows th; distinction between this and the protasis ~V 7rOV aKOVCTV.) So Od. i. 281, ii. 216, 360, iii. sa. El7rE pm, at KE 1ro8 yvww TowvTov MvTa, if haply I may recognise him. Od. xiv. 118. Ba.AA' ovTws, at KEv n cp6ws Llvaoun yJ!'Y]a, if haply you may became (i.e. in hope that you may becorne), etc. IL viii. 282. So Il. xi. 797, 799, xiii. 236, xiv. 78, xvi. 39, 41 (cf. 84), xvii. 121, 692, xviii. 199. Ka of v7ro<rx(T8a ovoKa[oeKa {Jovs t(pevCTEJLEI', a[ K' EAE~<T1) fiuTv, aZ KEV Tvo&os vluv dm5a-xu 'IA[ov Zp~s, let lie? p1omise to sacr~fice twelve oxen (to Athena), in hope that she may pity the city, . . . if haply she nwy lceep the son of 1'ydeus f1om sacnd Ilios, etc. Il. vi. 93. (For at: Kev dr.6axu Aristarchus read ws KEJ'.) Evcf>YJp/fjcrai TE KEAEcr8e, 6cppa [l,i Kpovloy ap1)crop.e8'' ai t<' Ae~a-y, in order that 1ve ?nay pray to Zeus tp pity us if he will M haply he shall pity us). Il. ix. 171. So Il. vi. 281, 309, xvii. 245, xxii. 419, xxiv. 116, 301, 35 7 j Od. xiii. 182. See also et: KEV 7rWS (3ovAETl AOl')'bV apvva, Il. i. 66. llaTp6KA<p ecj>E7rE Kpanpwv1>xas l7r7rOV,, at KEV 7iW<; pw Ayr>, owv of. TO eiixos 'Ar.6.\.\wv. Il. xvi. 724. So Il. xv. 297; Od. xxii. A ~ ' ' 'c I / ' r' 76. Ll<Vp' KOJL 8' , a,, KI '710 El!'> Es0'11<TW 'lrtp 7rUVCFrJ 0~VOS. Od . iv. 34. So Od. i. 379, ii. 144, xii. 215, xvii. 51, 60, xxii. 252. 1 ' EKTopos opa-wpev KpaTep~v fhEV0<; 1 ~V TWa 7rOV Llavawv 7rpoKaAecr<TTa. Il. vii. 39. 'Yf6cr o' avyYj ylyveTa a[m:rovcra 7rEpKn6ve<T<r locr8at, at KEV 'lrW> <Tt'V VYJWiv dp~s aAKT~p<s l'KWVTU. Il. xviii. 211. El o K' i!n 7rpOTEpW 'lrfJpav~~OfhCJ., i)v 71'011 cpe1ipw 1/~6vas, odow, K.T.A., but if I shall swim, on still jct1ther, to find a shore if haply I ?nay, I fear, etc. Od. v. 417. (Here ~v rrov Jcp<vpw depends on an ordinary protasis, which, however, is not its apodosis.) 'AA.X &yeT', at Kiv 1rws- 8wp1]tOJl,EV via> 'Axatvv, i.e. let us Mm tlwm if we can. IL ii. 72 (so 83). "2Ke1l"TEo vvv, a.E K< i8rya (wov / 'AJ!Tl.\oxov, if haply you may see. xvii. 652. 2:0 'oEKcp Swpov 7rOT0E')'jJ-EliO<;, aE K 7r6pTJITV, expecting a gift, if haply he shall give one (i.e. in hope that he will give one). Od. ii. 186. So Od. xv. 312. 'AAX ov yap a-' UJ.\w (3a.\f.ew ToovTov J6vm .\d&py dr.7l"cVrTO.S, d.\X apcpa86v, er KE TVXWJLl, if haply I may hit you. Il. vii. 242. NvJ! o.DT yxd17 7rELp~a-opat, ar K TVXW!J-L, I will t?y with my s2Jea1, if haply I may hit you. IL v. 2i9. '!Js OTE ns Tpoxov KEPUfhE:Vs 7rHp~<rETa, ar K (:)f.?JCTV, i.e. tries a wheel, in case it will mn (i.e. to let it run if it will). 11. xviii. 600. (The analogy of the two l)receding examples shows that there is no indirect question.) llapCw Kai .\af3 yovvwv, aK KEv 'lrws WAy<Ttv ~1rl Tpwecra-v ap~~a, i.e. clasp his knees in the hope that he will aid the 1'1ojans (that he may aid them ih case he will). Il. i. 407. So Il. vii. 394, x. 55, xiii. 743, xviii. 457; Od. iii. 92, iv. 322. For these last examples, see 490, 2. For al KE in the common text of Homer, here as elsewhere, Bekker and Delbriick write er K (see footnote to 379).

e z'


2. In alternatives with two opposite suppositions, this construction implies that the subject is ready for either result, though the former is hoped for or expected. E.g.




'II:Ii>s cj:>~peratJLEVct, i]v Ttva 7recf:>vTJ dv8pwv ~ ailT0s cj:>l:i{ETat 1rp<fmp Jv OJLOvp, i.e. he (a lion) rushes on, ready to slay or to pe1ish. Il. xx. 17 2. In Od. xxiv. 216, the common text has mtTpb> 7r!p~U'O}Lat, at KE (or Et K<) / hrtyvwv . . . ~ KEV dyvotfiU't, I will try my father (ready for either result), in case he shall recognise rne or shall not know me (where Kev alone in the second clause is very strange). But La Roche reads i] Ke / J7rtyvwu, as an indirect question, one ll1s. lutving 1] K: see also Orl. xviii. 265. 'Emyvwu is Hermann's conjecture for J7r,yvo{YJ or yvo[YJ. 488. The optative with el (rarely d K<) is sometimes used in Homer like the subjunctive after primary tenses in sentences of this class. It is also very common after past tenses, representing a subjunctive of the original form, though occasionally the subjunctive is retained in indirect discourse (696). E.g. 'AA.A.' en TOv 8-6U'T1JVOv dtoJLat, d 7raBEv f.A.Bwv dv8pwv JLV1JO"T~pwv U'Ke8aU'w KaTa 8wJLaTa !:Id?), but I am still expecting the poor man, if haply he should come and scatte1 the suitors. Od. xx. 224. So Od. ii. 351. 'A,\,\a ns d1J 'AyaJLEJLVavt, el 7rAdovas 1rapa vavcj:>tv J7rOTpvvH< ve<U'I:Ia,, let some one go to A., in hope that he 1nay exhort, etc. Od. xiv. 496. See also 491, below. BovA.evov 07rW!) ox' aptU''Ta yevotTo, d nv' ~TapotU'V l:lavaTov AVU'tV EDpo[JL1)v, i.e. if haply I might find some escape. Od. ix. 420. 'A,\,\' Jyw oil mi:IOJLYJV, ocj:>p aV'TOV 'T t8o'JL' Kat El: JLO' 8o'1}, but I disobeyed them, in order that I might see him (the Cyclops) and in hope that he would show rne hospitality. Od. ix. 228. (The final clause and the protasis are here again clearly distinguished: see Od. i. 93 under 487, 1.) Ilo,\Aii 8 T ayK f.mjA.()E JLeT dvepo> t:xvt Jpevvwv, .r 7ro8w E~<vpot. Il. xviii. 321. n.,p~()YJ 8~ .:o aBTOV EV <ivTEU't, .l ot cj:>apJLOU'U'H Kal EV'TPEXO' a yA.aa yvi:a, i.e. he tried himself in his armour, eager for it to .fit him and joT his limbs to pluy fTeely in it (if haply it should fit hirn, etc.). Il. xix. 384. (See the cases of the subjunctive after 7rHPWJLaL in 487, l. Here there is no indirect question, for Achilles can have no real doubt about the fit.) 'Ev 8~ 7r{()o, oZvow fU'TaU'av, et 7rOT' '08vU'U'EV> otKaoe voU'T~U'H<, i.e. the caslcs of wine weTe waiting joT the return of Ulysses. Od. ii. 340. "'HU'TO Ka'TW op6wv, 7rOTt8EyJL<VO!) EL Tf JL'V dm),, i.e. he sat looking down, waiting for Penelope to spealc. Od. xxiii. 91. To8' 1]vwy., El7retv E7ro>, .Z K' 8AYJTE 7ravU'aU'ea, 7rOAEJLov, he bade me say this word, if hwply you might be willing to stop the war. Il. vii. 394. (This appears in vs. 387 as d KE yevotTo, and the direct form of the command in vs. 375 is Et K' f.()A.wU'tv.) In Il. xiv. 163165 We have d 7rW!) [JLdpatTO . , , Tt{; 8~ X<V1J after a past tense. N~XE 1rap~~' d 1rov cj:>d;pot 1JtOvas. Od. v. 439. (Compare vs. 417, d 8 KE 7rapav~~o}Lat, ~v 1rov cj:>Evpw, under 487, 1.) See also Il. ii. 97, iii. 450, iv. 88, x. 19, xii. 122, 333, xiii. 807, xx. 464, xxiii. 40; Od. i. 115, iv. 317, ix. 267, 317, 418, x. 147, xi. 479, ~28, xii. 334, xiv. 460, xxii. 91, 381. 489. This construction (487; 488) with both subjunctive and




optative is found also in Attic Greek and in Herodotus, but with less variety of expression, and at the same time with some extension of the usage. Especially to be noticed are the protases depending on verbs like and OEA.w in Herodotus. E.g.
e~f3a~ ~p.fis 7rEp.fov, Uv 7rWS 8taKwA:6o-wp.ev l6vTa <j>6vov 1'0t(]"V ~p.aip.ots, send us to Thebes, to prevent, if haply we 'ln!J,y, etc. SOPH. 0.


Ti]s ep.i]s "(VWflXJS aKOl/<TOv, 1JV -r O'Ot BoKW Aeyetv, hear my


judgment, in the hope that you may thi7UC the1e is something in what I say. EuR. H. F. 2 78. 'E8ovTO 'TO V 'Apunay6pew, KWS a,)TOUT 7raplf<rxot 86vap.v Ttva 1<ai l<aT~AIJouv e<; T~V wvnov, they besought A., if in any way he might supply them with an armed force and tkey might be restored to tkei1 own land (to do this). HDT. v. 30. iPpov~<TaVT<; eZ Kws ~v yvotTo TO 'E..\A.1]vtK6v, having it at heart that, if it were in any way possible, the Hellenic race should be 'ln!J,de one. Id. v:ii. 145. BovA.ofJ-El'YJV er KWS dp.cp6npot yevoiaTO f3aO'tAES, i.e. wishing that both might be made kings, if in any way this could be done. Id. vi. 52. 'EfJovA.evero OEA.wv d KWS 'TOlJTOV~ 7rpwTovs i!A.ot. Id. ix. 14. IIp61Jvp.ot ~o-av mXEPEE!V (se. Tfjcn V>]Vcrt), eZ KWS i!A.otev aimf.s. Id. viii. 6. IIp.favns 7rap' 'Afhjva[ovs 7rpEO'f3w;, 7f(J)~ 1l'eio-etav JL~ O'cf>wv 7rEpt VEW'Tep(etv P-~78f.v, to persuade them if they n~ight, etc. TRue. i. 58. IIopwop.evot is ' ' O'!UV 'f3 ' ' Et ' ''.r.ll .lb/3'' T'JV A ' WS UO'tii..W.," 7rWS 7rEtO'Eta.V f.t1JTOV1 JO OWe~o~ J OVII..OfJ-EVO 1rei:o-at aiYTov, el ll6vatvTo, O'TpaTevO"at, in nea1ly the same sense. Id. ii 67. ITvvOav60evot TOtJ~ 'A8YJvaovs ES -r~v Kap.cfptvav 7rpCrf3eVor8at, 1/'W~ 7rpOO'ayayotVTD aiJToD<;1 that they Went on an embassy to 0., to bring the town over if they could. Id. vi. 75. (Compare J.s 'AKpaya.vTa. LtKa.vov &rrcrTEtAav, !hws 1nra.yayotTo T~V 7r6Aw el 06vatTo1 vii. 46. This might have been ei: 7rWS fJ7rayayotTo TTJI' 7r6Atv, and in vi. 75 we might have had omJls 7rpD<TayayowTo aiYToi>s el 8vva.wTo, with nearly the same force, but with different constructions.) See v. 4, e~ 7rw~ . . Stao-wO"etav. '1KETat 7rp'Os cr~ 8evp dcpyp.e0a, d nva 7r6A.w <j>p6.crnas ~p.rv e-!Jepov, we have come hither to you as suppliants, in the hope that you might tell m of some city soft as a fleece (to have you tell us, if perchance you might do so). An. Av. 120. "AKovO'ov Ka~ f.p.ov, M.v crot TavTct lloKfj, listen to me too, in the hope that you 'ln!J,y think the same (in case the same shall seem. t1ue to you). PLAT. Rep. 3 58 B : so 434 A. ''Opa ovv Kat 7rpo0vp.ov KG.Tt8efv, eav 7rWS rrp6TEpo~ f.p.ov i:llns 1<at J.p.ot cf>paO"n>, i.e. for the chance that you 'ln!J,y see it first and tell vw. Ib. 432 C: ~o 618 C, Theaet. 192 C, Soph. 226 C. See also XEN. An. ii. 1, 8, av T o6vwvTat, and AR. Nub. 535. On this l)rinciple we must explain An. Ran. 339, ovKovv &rpjl e~ns, ~v n Kat xopo~~ A.af3ns, will you not lceep q1tiet then, in the hope of getting some sausage too (i.e. to have some sausage 1j you chance to get any) ?




490. 1. The apodosis may, further, be suggested by the context, even by the protasis itself, without implying that the protasis expresses a purpose or desire of the leading subject. This gives rise to a variety of constructions. E.g. KTavefv EJ.Lot vtv ~8oO"av, ei:TE JL?J KTavci!v OD.. otp.' /J.y~O"IJat :._




'Apye[av x06va, they gave her (Helen) to rne to slay, oT, in case I should prefer not to slay her but to ca?'TY her back to the land qf A rgos (for 'l'iw to do this). EuR. Tro. 87 4. ''Hv (r~v ~vpp. axtav) y< ovK 1rt TOL~ cpA.ots E7T'Ot~IJUIJ0e, TWV 8~ x0pwv ~V TtS cp' .,)faS ru, i.e. you nwrle it (to use) in case any of your enernies should come against you. TRue. vi. 79. llpos r0v, .t E1ft{3oryBouv, Jxr!Jpovv, they marched towa?'ds the city, (to be 1eady) in case the citizens should ?'ush out. Id. vi. 100. Ta.AA.a, ~v ;n vav0axdv o 'AfJTJva'iot roAp:quwtJt, 1rapetJKevaCovro, they rnade othe?' p1eparations, (to be ready) in case the Atheni1ms should venture on fuTtheT sea-fights. Id. vii. 59. Ko)pvyfLa 7!"oLOi!vnu .. -rwv VYJtJtWTwv d ns (3ovA<Tat br' A.evBepf!- ws O"cpas d7l"tevat, they maJw proclamation, in case wny of the islanders wishes to come over to them with promise of f1eedom (joT him to do so). Id. vii. 82. OvoefL[a (3A.af3TJ -rwv 1rpos -rtl.s 71"6AHS Ota7rOf7rWV !is n KaratJKo7r0v Kat i)v Tt &A.A.o cpavryrat mr~oewv, theTe is no harm in the envoys whom we have sent to the various cities, paTtly for inquiry, and also in case any othe1 advantage may appear (to secu1e this), i.e. to secure any other advantage that may, appear. Id. vi. 41. So Kat d nva 7rpiis aAAOV 8ot, Id. v. 37. 'Aptl.s 7rotovvrat, d Tt~ EmK']pvKeverat IT~ptJats, they invoke curses, if any one (i.e. to fall on any one who) sends hemlds to tlw Persians. Isoc. iv. 157. if>tAoTLfLEWBat !J.']O' evt Jcp' lf.;\AI(l i) E1ft XP'lfLaTWV KT~O"H Kat Uv 'Tt lf.A..Ao els rovro cpf.pu, i.e. for anything else that may lead to this. PLAT. Rep. 553 D. See ARISTOT. Eth. x. 9, 2 : EXELV (-r0v Jpm}v) Kat xpf)uBat 7r<tparov, ~ d 7l"WS li.A.;\w; dytBot yw6fLeBa., 1ve must tTy to possess and employ vir6ue, o1 if then is any other means of becoming virtuous (to use this).
2. In the Homeric examples in which the protasis consists of an infinitive depending on Wf.A.w (487; 1, end), the apodosis is sugge~ted by the infinitive rather tl1an by Jef."Aw. This shows that a.l K lfJEA.utJt in itself has no final force. See also Od. xxii. 381, 7rd1T'T'}Vev 8' 'OovO'EVS Kar' 0v 06fLoV, TtS ;r' UI!Opwv (wos -&7roKA07rEOtTO aAVIJKWV K{jpa p.f.Aawav, he peeTed through his house, in case any man might still be alive and hiding himself (i.e. to find any such man), where no desire or hope is implied, and the construction is like that of THUC. vi. lOO (above). In PLAT. Rep. 327 O, oBKovv !in JA.A.e7rETUt Tb ~v trdO"WfLEV -&jLaS WS XP0 ~fLU~ Jcpel:vat; the subject of v. A.d7CETUt is a protasis introduced by r6, into which the apodosis has been wholly absorbed. The construction is, is there not still left the SU]J1Josition of our 1Jenuading you that you nw,st let us go ? But the meaning is, is it not left for us to pmsuade you that you must let us go, if we can (i.e. 7rELO'at ?)1' 7rELO"WfLEl') 1 This is an important example for explaining this whole class of sentences (486-490). The cases in 490 malce it plain that the final force often ascribed to ei or i)v comes from the suppression of an apodosis containing the idea of purpose or desire, since the same form of protasis which is sometimes called final has no final force when a slightly 'different apodosis is implied (as in Tnuc. vi. 79, 100, vii. 59).





491. Sometimes a clause with d K or ~v (rarely ei) and the subjunctive, or with Er KE or El and the optative, in Homer is the object of o'loa, i8ov, or a verb of saying, expressing in a conditional form a result which is hoped for or desired. These clauses have the appearance of indirect questions; but the analogy of the preceding examples (487-490) shows that all are based on the same idiom,-a protasis which involves its own apodosis so that it would be useless to express the latter separately. The examples are these:Ts o'lO' d KE Kat a1h6s iwv KoO'YJ' J?Tl VYJ'Os -r~AE <j;[A.wv d?TOAYJ Tat, who knows the clwnces that he too rnay perish, etc. '! or wlw k1wws the chances of his peTishing, etc., if ha11ly he 'rnay? Od. ii. 332. (We may
translate colloquially: who lcnows ? supposing he too slwll peTish ?) Ts o'lo' d K 'AxLAEVS <jJB~v EJL0 {nro 8ovpt 1'V'11"t> d?To evJLOV OAerrrrat; who knows the chances that Achilles may fint be struck (the cha%ces of his , being jiTSt sbuelc, if haply he shall be)? ll. xvi. 860. (We should naturally express this by a different construction, whetheT he 1nay not be fwst struck.) Tt, oi8' d KEV o1 ITVV 8a[f.LOVL 8vf'OV op vw 7rapH7rWV; who knows the chances of my 1ousing his spiTit by persuasion, if haply I shall do so ? IJ. xv. 403. In Il. xi. 792 we have Nestor's advice to Patroclus, r> olo' Et1 KEV 0~ <TVV oa[JLOVt (Jvf'OV op[vaL<; 7rapet7T0v j who know> the chances that you could ?'Ouse his SJYi1it by persuasion ? (op[vats K being potential), Ov JL1JV o'lo' El a:OrE K~Koppa<f;YJ> aA<YHV~> ?Tp0rYJ 1ravp1)aL Ka.{ rr< 7TAlJn)rrtv !JLa<T<T(tJ, I mn not sure of the chances of your being the first to enjoy youT own device, etc., i.e. I am not so sure tlwt you may not be the fint to enjoy it, ~f it shall so chance. Il. XV. 16. ZVS ydp 7T'OV r6 Y" olo<: Ka1 d8dvaTOL Owt aAAoL, Et Ke f'LV dyy<{AatJLL l8wv hrl ? 8' d>..~81)v, Zeus and the other i?nmoTtals (alone) lcnow this, the chance of my bringing news of lvim, if haply I have seen him and so might do this. Od. xiv. 119. El o' <fy< 8~ }LOt TOVTO, 8a, VYjf'PT~> vmrt>, d 7TW<; T~V oAo~v JL~V {J?T K7T po <jJvyo LJLL Xapv{38tv, n]v o K opvva{f'YJV OT JLOL <TvotTQ y' hapovs, i.e. tell me this without fault, the chance of 1ny escaping Cha1ybdis if haply I should do this, and of my then lceeping Scylla off if I could (lit. tell me this, supposing I should escape ChaTybdis and could then lceep Scylla off). Od. xii. 112 (this translation supposes K' to be potential, affecting only df'vvaf'1JV). "H JLEV1' Tpwas <TX<8ov f.A8JLV, o<jJpa Z8YJ-r' aZ K' 15JLJLLV V7rEP<TX?J xdpa Kpovwv; are you waiting f07' the 1'rojans to come 1teaT, that you vwy see the chances of the son of Cronos holding his hand over you ?-or that you 1nay see hirn hold his hand ovm you, if haply he may do this ? Il. iv. 247. (We might say, is it that you 1nay see it,-supposing the son of Cronos to hold his hand oveT you?) Twv <T1 a:Ons JLV?j<Tw, Zv' &:7roAA.?J~vs J:;rarawv, O<jJp' fo17s ~V 1'0. xpa {<T }L'{/ <jJLA6T1JS 1' Kat eiJv~, i.e. that you rnay see the chances of youT device availing you, or that you may see it if pe1chance your device shall avail you. Il. XV. 31. See also Il. XX. 435, dAA' ~ TOt f'~V ravra 8WV iv yovvaU"t KEtTat,




ai: KE ; xnpoTpos 7rp Jwv &7r0 evJLOV E>.. wJLa!, i.e. this rests qoith the
Gods, for me to take your life away, wealce1 though I arn, if perchance I may. The conditional construction is more obvious here than in Il.

iv. 24 7 and xv. 31 ; but in all three we naturally fall into an indirect question when we attempt to express the thought in English. 492. A comparison of these peculiar conditional constructions (491) expressing hope or desire with clauses with JLfJ expressing anxiety and desire to prevent a result, both depending on ol8a or el8ov, is suggestive. With Od. ii. 332 and Il. xvi. 860 (in 491) compare Il. x. 100, oii8E T! tOJLeV, JLfJ 7rWS Kal VVKTa JL<VO!v'ljcrwcn xecr8a!, nor do we ktWw any way to prevent thei1 being impelled, etc., and PLAT. Phaed. 91 D (quoted in 366); and with Il. iv. 247 and xv. 31 (491) compare Od. xxiv. 491, tOO! fL'l 01J crxeoov i!Jcr! K!OVTES (366). This comparison shows that el8eva! (or 18e'i:v) d KE -rovTo yevYJTat means to know (or see) the chances of gaining this (object of desire); while d8vat (or l8efv) JL0 TOVTO yEVYJTat means to lc1ww (or see) some way to p1event this (object of fear). The idea of desire or anxiety belongs to the dependent clause, and not at all to the leading verb. 493. These Homeric expressions (491), in which nearly all the force is in the ]Jrotasis, so that the apodosis is not only suppressed but hardly felt at all, helps to show how the ]Jarticle el came to be an indirect interrogative, in the sense of whethe1.- But in Attic Greek, where the interrogative use is fully established, only the simple el (never ~V or eav) can mean whether, even when the verb j~ subjunctive (6~0).





494. After many expressions of wonder, delight, contentment, indignation, disappointment, pity, and similar emotions, a protasis with el may be used to express the object of the emotion. When the supposition of the protasis is present or past, a causal sentence would generally seem more natural. Such expressions are especially ()avfi-at;w, alax6VOf1-at, arya7rw, aryava/CTW, and 0VOV anv. E.g.
ealfMi{;w 8' eywye l JLYJOEtS VJLWV JL0T ~vevw'Vrat p:q! 6py{;eTat, opwv, K.-r.A., I wonder that no one of you is either concer1wd or angry, when he sees, etc. (lit. if no one is either concerned or angry, I wonder). DEM. iv. 43. , AAA' he'i:vo eavJLO.[w, El AaKEOa!JLOVlOtS JLEV '!rOTE &vT0paTe, vvvt 8' 6Kve'i:TE J~teva! Kat JLEAAeT elcrcpepew, but I wonder at
this, that you once opposed the Lacedaernonians, but now are unwilling, etc. Id. ii. 24. (The literal meaning is, if (it is true that) you once opposed, etc~ then I wonder.) OiK &ya-;r~ e l JL0 0KYJV OEOWKEV, &AA' l f'-1 K<fl xpvO'cj) O'T<<{>av'f! O'TE<f>avwfh)crETU.L ayavaKTEL, he is not content if he has not been punished; but if lw is not also to be crowned with a





Kat 6ls riATJ8ws riya.vaKTw, e l ovTwU"t il vow JL0 oi6s T, dJLt el-Irei:v, I arn indignant that (or if) I am not able, etc. PLAT. Lach. 194 A. Ov 01) 8avfLaU"T6v EU"Ttv, el U"Tpanv6JLevos Kat 7Tovwv EKel:vos a1hos {>JLWV p.eAA6vTwv Kat frJ~t(op.f.vwv Kat 7Tvv8avop.f.vwv 7T<ptyyv<Tat, it is no wonder that he gets the advantage of you, etc. DJM. ii. 23. MT}o~ p.f.J,TOL TOVTO p,e'tov ~'t. ,, ) ( K llpEWL 7Tp0U"OEV U"VV 1)p.LV TO.TTOfLEVOL l'VV O.'I',. I fl ' t ,., ' " ',.J,. OOt;TJT EXELV 1 L OL U"T1KaU"LJ11 i.e. do not be discontented, if (or that) the Oyraeans have now Alvw U"e, d Knvei:s OUJLapTa U"~v. withdrawn. XEN. An. iii. 2, 17. EuR. Tro. 890. Llewuv &v EtT} 7Tp-qyp.a, d LaKaS JLEV oovAovs ~XOftEV, ''EAATJI'O.S of. oD TLJLWPTJU"6f'e8a. HDT. vii. 9. AlU"xp6v EO'nv, El yti> ftEV Ta pya V7TEftHva, Vftei:s 8~ ftTJOE Tovs A.6yovs dvf.~eU"8e. DEM. xviii. 160. Llewov av EtT], el oi JLEV EKdJIWV ~{,p.ftaxot o-&K d7TepoVO'LV, 1JJLELS o oDK &.pa 8a7TaV~O'op.ev. Trruc. i. 121. Tpa> Af.yn>, el .. oDK ll11 MvatJ!To Aa8el:v. PLAT. Men. 91 D (see 506). Llewov d oi aDTot ftd.pTvpes TOVTOLS fLEJI llv p.apTvpoVJITES 7TLO'TOt ?)<:rav, Jp.ol o p.apTvpovvTES a7TL<TTOL ;<:rOJ!Tat, it is hard that the same witnesses testifying for them would have been trustworthy, and testifying for me an to be unt?ustworthy. ANT. vi. 29. See AESCHIN. i. 85. In all the preceding examples the protasis belongs under 402, the futures expressing present suppositions (407). For d oD see 386 and 387, with examples.

golden crown, he is indignant. AESCHIN. iii. 14 7.

495. The same construction is sometimes used when the leading verb is past. E.g.
KaTEfLEfL~ETO awdv Kat TOVS <:rtJJ! avTtj), el o1 &A.A.ot riKp,ct(ELJ! p.fiAA.ov avTwv 86 KovJ!. XEN. Cyr. iv. 3, 3. But generally such sentences are affected by the principle of indirect discourse, and have either the optative or the form of the direct discourse : see XE;N. Cyr. ii. 2, 3, ~x8rr&YJv El: n p,etoJ! ooKo tEJ! gXHV (where OoKov<:rt might have been used). See Eun. l\1ed. 931, d<:rqA.B p.' olKTos, d yeJ!~<TETat, and XEN. An. i 4, 7' <fKTELpOll El aAW<TOLVTO. For such sentences see 697.

496. These expressions may also be followed by 8n and a causal sentellce, as in PLAT. Theaet. 142 A, f.Bavfta(oJ! on ovx oT6s T, ~ eypi:J!. The construction with el gives a milder or more polite form 6f expression, putting the object of the wonde?' etc. into the form of a supposition, instead of stating it as a fact as we should do in English. They may also be followed by protases expressing ordinary conditions, which have nothing peculiar: see Isoo. xv. 17, dya1rTJTOv (se. E<:rT[J!) ~J! KA.af3etJ! 8vvTJ Bw<:r t UKawv, they must be content 1j they a1e able (cf. xix. 20); and PLAT. Prot. 315 E, DEM. ii. 23 (d 7TcptqftE11).


497. This construction must not be mistaken for that in which el is used in the sense of whethe1, to introduce an indirect question; as /jpWTwv el1]A.8ev, I asked whether he had come.




498. The forms of protasis and apodosis which are contained in the classification above (388-397) include by far the greater number of the examples found in the classic authors. Many cases remain, however, in which the protasis and apodosis do not belong to the same form. Especially, the weat wealth of conditional expressions which the Homeric language exhibits in both protasis and apodosis (399) allowed great variety of combination ; and the early poets used much greater freedom in these sentences than suited the more exact style of Attic prose.

I. Optative in P1otasis, with Futu?'e or Present Indicati1'e or an equivalent form in Apodosis.

499. (a) In the earlier language a protasis with the optative is not infrequently followed by an apodosis with the future indicative or impera.tive or (in Homer) with the subjunctive. The subjunctive or future indicative in Homer may also take KE or lfv (452). E.g.
Eg 'r[~ fWt O.vljp &Jl ho tTo Kat lL\A.os, fhaAAov BaA.1rwp1j 11.~ Bapa-aAH!Jnpov iia-Tat, if any other man should follow with me, there will be mare corrifo'rf and gnater couraqe. Il. x. 222. (The want of symmetry in the Greek is here preeisely what it is in the English; and d'Y} fiv is no more required in the apodosis than would be is, though both are 'the conventional forms.) , See Il. ix. 388, and xxiii. 893, 1rOPWfhEV, d UJEAot<;. Tov er 7rWS a-v o.UvatD AOX'YJO'dJLEVOS AeAa{3e(]'Bat, or; K~V TO cb1rTJ(]"t o86v, he will tell you, etc. Od. iv. 388. See Il. xi. 386, d 7rHp'YJ8E['YJS, OVK fiv TOt xpu{rJ'fl'TJ(]"L {3t6s j alld Il. ii. 488, XX. lOO, Od. xvii. 539. Elo~ oa[fh<iJV ep7rot, 62 Tovl 'Evva;\{itl T' EKO<!Ja-ofhEV 1rpWnTtv. PIND. Ol. xiii. 105.1 So in an old curse, cl ns TUO 7rapuf3avot, eva-y~s ~U"TW, AESCHIN. iii. 110. See SOPH. 0. T. 851, d n KaKTpe7rotTo, ovTo< TOV y< Aut'ov cpovov cpa Vet OtKalw<; op86v.

500. (b) A present indicative in the apodosis with an optative in the protasis is sometimes merely an emphatic future expression. E.g. IlavT' :ix<ts, er (]" TOllTOJI/ f.LOip' f.q>f.I<OtTO Ka.A(tll', Y01 have the
whole, shonld a share of these ulmies fa.ll to your lot. PIND. Isth. iv. (v.) 14. So Katpuv d cp8y~aw, fhdwv E7rETo.t fhWfhO'> Ul'()p07rwJ!, i.e. should you spealc seasonably, you are sure to be followed by less censure of men, Py. i. 81. In THee. ii. 39 we have Ka{Tot el paBvfht\t JLaAAoJ! ~ 7TOJ!WV fhEAhTJ (jf.J..otfhV KtVOVVVHV1 7rcpt-y{yveTa.L ~}LtV, K.T.A.,
1 For the cases iii. p. 444. ill

Pindar here and in 500 and 501, see Am. Jour. Phil.




and now supposing that we should choose to meet dangers with a light he(lfft rather than with laborious tmining, we secure the advantage, etc. This sentence is loosely jointed, like the others which have this combination; the condition is stated as a remotely supposed case, in the vague future form, but the apodosis, we at once gain this advantage, etc., is adapted to a present supposition. 'fhe optative is generally emended to UJD...oJMiV, although it is one of the best attested words in Thucydides, being in the hest Mss. and also being quoted by Dion. Hal. as a faulty expression. The criticism of Dionysius (de Thuc. Idiom. 12, 1) is instructive: ev"Tav8a yap 'TO JL~v J8AotfLEV p~fLa Tov JLEAAov"T6s E(J'T xp6vov 01JAWTtKOv, TO OE 1T'<ptyyve'Tat 'TOV 1T'ap6v'TOS' aK6A.ov8ov o' ltv ?JV el (J'VVe(ev~e 'T<{i e8eAOtfLV 'TO 1T'EptE(J''Tat, i.e. the future expression el J8AotJL<V should have a future form like 1T'epd(J''Tat to correspond to it. In DEM. xviii. 21, <l yap it:va n ooKo[1J -ra JLJ)-..tu'Ta Jv TOV'Tots ao[K1JJLU, OVOEv f(]''T 0~1T'OV 1T'p0S f.JLE, the apodosis refers to the real protasis if theTe is any apparent fault.

501. (c) In most cases, however, the present indicative in the apodosis precedes, containing a general statement, and the optative adds a remote future condition where we should expect a general present supposition. E.g. Ov JLOL 8JL''> ~(J''T', ovo' d KUK[wv (J'Eeev U8ot, ~ELVOV UTLJL~ (J'at, it is not 1ight for me-even supposing a more wretched man than you should come-to dishonour a stmngm. Od. xiv. 56. 8ap(J'aAeos yap dv~p ev 1T'a(J'tv aJL<[vwv ~pyot(J'tV HAE8H, d Ka[ 1T'08<v ClA.A.o8<v U8ot. Od. vii. 51. So v. 484, viii. 138; Il. ix. 318. Ov-r' ovv ayy< ~n 1T'd8oJLat, .r 1T'o8<v A.8ot, ovT< 8<o1T'po1T'[1J> EJL1T'tl(oJLat, ~v nva JL~T1JP E~<PE1J'Tat, neitheT do 1 any longeT put t1ust in reportsshould any one come-noT do I 1ega1d any divination which my mother may 1 ask. Od. i. 414. (Here the remoteness of the supposition in .l (>..8ot is contrasted V.ith the greater vividness of that expressed in e~<pE1J'Tat). 1 1 ' ' ' A '~' r/ f ( '(3 '1 ''t_ 't L.J.HVOV 'T, tit K E'f' UJLUt;UV V1T'tip WV ax O<; O.Hpa.<; t;OVa KUVKO.t;;L') "Ta o <j>opn' aJLavpw8d?J, it is hard, . . . supposing you should break your axle and your load should pe1ish. EEs. Op. 692. lGpoos <j>A.-ra"Tov, EKOV'TOS Tt'> EK o6JLWV <j>epo!., it is the dearest gain, if one should bring it from the house of a willing givm:. Pnm. Py. viii. 13. See Isth. ii. 33. So SoPH. Ant. l 032. In most of these examples a general supposition with the subjunctive (or present indicative) in the protasis would have agreed more closely with the thought. If the protasis had preceded, so as to determine the character of the sentence, the apodosis would naturally haw had the optative with K~ or Clv, or some future form (as in the cases under a).

e ''



502. (d) The optative in protasis sometimes depends on the present of a verb of obligation, propriety, or possibility with an infinitive, the two forming an expression that is nearly equivalent in sense to an optative with :l.v. E.g.




El yd.p d1J(J"O.V ovo nv~s ~va.wdot v6pm, ovK dJL<f>o-r~pots ~vt lh)7ro11 1f'fJ<f>tra.tr8a.t, for if thme should be two laws opposed to each other, you could not surely vote fm both. DEM. xxiv. 35. This is analogous to the apodosis formed by i!oct, xpijv, EV?JV, etc., with the infinitive (415). There, for example, vijv a-lri} VdhZv, he could ha've gone, is nearly equivalent to 1jA8cV av, and here ~VEITTW <J.VTi) JA8e'iv, he could go, is nearly equivalent to i!>c8ot av. This use of the optative is more common in the corresponding relative conditional sentences (555),

II. Indicative



Subfunctive in Protasis, with Potential 01' InAiicative in Apodosis.

503. (a) A present or past tense of the indicative in .the protasis with a potential optative or indicative (with &v) in the apodosis is a perfectly natural combination, each clause having its proper force. E.g. El 8~ ns JOavarwv ye Ka-r' ovpavov ELA~Aov8as, OVK av iiywyc
OwZtrtV E7rovpavfottrt 0a X o JL?) v, but if thou art one of the immmtals come f10711 heaven, I would not fight against the Gods of heaven. Il. vi. 128. lloAA~ yd.p av EVOatp,ov[a et1'} 7rEpt -roD<; VEOVS, el eis fJ-EV JLbVor; Vrous 8 t a 1> E [pet' ol o' aAAOt w EA 0 VIT t 11' for there would (naturally) 1> be gnat happiness, etc. PLAT. Ap. 25 B. Er n 8tr<j>arov 1ra.-rpl XP'fJ' (J"JLOUJLV ( K V E 1:8'' WITTE 7rp'o<; 7ro.owv 8aveZv, 'ITWS av OtKa{ws -rovr' 6 V EL 0 0 t S EfJ-Ot i if a divine decree canoe fo my father through Oracles that he was to die by his sonrJ hands, how can you j~;stly 1eproach me with this ? SOPH. 0. c. 9 6 9 ; so 9 7 4-9 7 7. ,, iltr-r' er p,ot K<J.t JLEITW<; ~yov JLEVOt JLfi).. >cov E-rpwv 7rpotrdvat a-&ra 7rOAEJLEtV E7rtltr8qn, OVK av elK6rws vvv -roil ye dotKdv al-rtav <f>po[JL1JV, if you we1e persuaded to make war by thinking, etc., I should not now justly be cha?ged with injustice. Tnuc. ii. 60. El yd.p ov-rot 6p8w<; d7r~trT1'}trav, VJLeZ> llv ov XPEWv /1. p X o LTE, for if these had a right to secede, it uo,uld follow that your dominion is unjust. Id. iii. 40: see vi. 92, and DEM. xxi. 37. Ka[rot TOTE TOV 'Y7repe8'1Jv, d7I'p J>c'fj&ij JLOV vvv Ka-rqyopE~ JLUAAOV av elJ<o-rws ~ -r6vo' f.OwKev, and yet, if he is now nwJcing tYUe chMges against me, he would then have p?osecuted Hypereides with ?JLuch more reason than (he now has for prosec~dinq) this man. DEM. xviii. 223. El yd.p yvvaLKE<; tls r68' ?J~OVITtV 8patrovs, . .. 1I'ap' OVOEV av-raZs ?JV <iv 6.\>cvvat 7rbtrHs, for if women are to come to this height of (tudacity (407) it would be as nothing for them to slay thei1 husbands. Eun. Or. 566. Tov-ro, El Kat ra>cA.a 1rav-ra &7roa'-rEpovtrtv, &1ro8ovva.t 1rpotrijKev, even if they steal all the 1est, they ought to have 1esto1ed this ( 415 ). DEM. xxvii. 37.

504. (b) An unreal condition in the indicative followed by a potential optative seldom occurs and is not a strictly logical combination. E.g.
El 'ToVr:l 1rexeipovv AEytv, oVK
~a-8' Ocrrtr; oVK &v lK0Tws




E7rt-np.~o-H~ p.ot, if I were undertaking to say this, (the result would be that) eveTy one would censuTe me with reason. DEAL xviii. 206. (Here

many Mss. and Dion. Hal. p. 1054 read E7rerJLYJ0"1 the ordinary form in such an apodosis.) See [LYs.] xv. 8.

505. (c) When a subjunctive or a future indicative in protasis has a potential optative in the apodosis, there is sometimes a distinct potential force in the apodosis (as in 503), and sometimes the optative with Clv is merely a softened expression for the future indicative (235). E.g.
El fJ-EV KEV 'lraTpO<:; (3oTOV Kat VOO"TOV aKOVO"W, ~ T &v TPVXDfJ-VO<; 7rEp en TA a[?] V EVWVTOv, if I hear of rny father's Z.ife and 1"et1<Tn, wasted as I am, I can still enduTe it fo1 a year. Od. ii. 218. (See the next verses, 220-223, El OE K n8vWTO<; aKOVO"W, with future forms in the apodosis. See also the corresponding verses, Od. i. 287-292.) 'AAA: (n fJ-EV K Kat /},c; KaKa 7rp 7rUO"XOVT<; tKOto-8, ar K' f.eEAl)<; O"bV 8vp.ilv JpvKaKew, but still even so, though suffeTing evils, you 1nay come home, if you will curb youT passion. Od. xi. l 04 ; so xi. 110 and xii. 137. See Il. xxi. 556. El 8 KEV of' ap6o-v<;, TOOE KEV TO~ <foapp.aKOV dry, but if you plough late, this may be your Temedy. REs. Op. 485; so 665. 'AAA' i)v <fo1/> pol, .. A.talp.' &v opOws, i.e. I would fain speak. SoPH. El. 554. So 0. T. 216, Phil. 1259 ; Eun. Hel. 1085. 0{,8~ yap &v 1roAA.a ycj>vpa~ ilio-tv, f.xotp.ev &v o7rot <jovyovnc; ~p.EZs o-wOwp<v, for not even if thme an (shall be) rnany bridges, could we (in the case supposed) find a place to fly to and be safe. XEN. An. ii. 4, 19. El yap Tt AE~Et<; (/i xoAc!Jo-Tat O"TpaTo<;, &v Ta<jJE[YJ 11"0-tS oo' OtKTOV T1)xot, fo1 if you say anything by which the aTmy shall be made angry, this child cannot be buried or find pity. EuR. 'rro. 7 30 ; see i.J.>povpwv d 'lrOt~O"Ol'Ta~, r~s p."Ev Y~' Suplll. 603, Oycl. 474. f3 AU7rTO tev Ci1' Tt [Lf.pos, ov fJ-El'TOt L<m,ov YE ~o-Ta t KbJAVEW 1)pJis, K.r.A., if they (shall) build a fort, they might peThaps injuie some paTt of ou1 land; bu.t it will n.ot l!e s11ficient to prevent us, etc. Tauc. i. 142. In the following examples the optative with d.v seems to form a future apodosis to the future protasis ; though in some of them it may be tlwught to be potential : El of. KEV EV7rAOt'Jv oc!Jv KAVTb<; dvoo-[yaw<;, ~fLaT KEJI Tpmh'{' i.J.>8ryv p(3wAov 1Kop.?Jv, i.e. on the thi1d day I shall arrive. Il. ix. 362. (The reference to this in PLAT. Crit. 44 B shows that tKo{p.,)V dv is a mere future.) See Il. xiii. 377, xvii. 38 ; Ocl. xxi. 114. 'AotKO[?JfJ-EV &vEl fl-'J a7rOOW<TW, I should be guilty o..f wrong, should I (shall I) not restme her. Eun. Hel. 1010. See Imi. :~74, Suppl. 520, I. A. 1189, Oycl. 198. ''Hv ovv 17 , fLOt rov u.o~Kov roi!Tov A6yov, oBK &v U1rOOo[ryv ov8' &v o(3oA6v OVOv[, if you (shall) learn this cheating reason fo1 me, I will not (or I would not) pay even an obol to any one. AR. Nub. 116. Kal oVTwc; &v OHJIDTa.Ta mivTwv 7ra8otEl',



Tat. LYS. xiii. 94.

l oVTot Ofk6i.frr:f>ot KaT' JKElvwv rWv dv8pWv rot's rpt.d.Kovra y<Cv1jcrov(Here we should expect El yf.vo~VTO.) Twv aT07r<u




TChwv 1dvl &v tt'l), el, & vvv &votav 6cf>A.trrKavwv iKAaAd, Tavra 8vv'IJB<'is p.~ 1rpa~<t. DEM. i 26.

III. Potential Optative


Indicative (wm;, Cfv) in the P1otasis.

506. A potential optative (with av) in the protasis may express a present condition, and a potential indicative (with :i.v) a present or past condition. E.g.
El p.ryo~ OOVAOV dKparij OE~atp.d)' av, 7rWS O~K ll~wv a~TOV ')'E cf>v,\6.~a<T8at TOWVTOV yEvea-8at; if we would not tuke even a slcwe who was internperate, how can it be other than fitting to guanl onf,self against becoming so? XEN. Mem. i. 5, 3. Kal y!iJ, drrep /l,\,\tp rtp J.vBpwrrwv rrtt8ofkrJJ! av, /(Ut <TOt, and I, ~f I would trust any 'l1um, tr1ut you. PLAT. Prot. 329 B. OvTOt 7r((,J!TAW>, d ft1J 7r0L~(TO.LT' a ]I TOVTO ws E')'W')'E <I>'Jfl-L OEtJ!, UKQTa<j>poJ!1)TOJI Ea-TLJ!, this (preparation) is not wholly to be despised, even if you would 11ot do this as I say you ought. DEM. iv. 18. Notice the difference between this supposition that you would not do this if you could (i.e. o~K &v rrot~a-atre TOVTO) and the ordinary et Jl-0 rrodwatTE Tov1"o, supposing you not to do this. El Tovvv TovTo l<Txvpov i]v &v To.Ur\u TEKf<ryptov, Kopo'i y<J!e<TBw TEKf<~pwv, K.T.A., if then this would have been a st1ong proof for him (se. had he had it), so let it be also a p1ooj for me, etc. DEM. xlix. 58. Ei , , , , 1\ 'c,, , , , , f<') Via TO TOVTOV<; f3 OVA(T at <TWo-at, Eq;WA')S 0.1f"OA0Lft'JV KQL 7rpOWNI}S > \ (3 I > ,, ' / / \' \ / (3 EV<Ta, H 7rp01TA<l WV Y a V apyvptOV 7r((,JIV 7r0AV ftETa TOVTWJI 7rpE<T had {t not been jot rny wish to save these (captives), nwy I pmish uttmly and befme my day if I would have gone on an embassy with these men eve1~ for very high pay. DE:~iL xix:. 172. (Here the protasis to which the apodosis d7roAo[p.1)V refers is really the whole expression ei E7rpa-{3v<Ta O.v el f<~ . <rw<Ta.L, if I would have gone except to save these, Jrrpf.<T(3ev<Ta, av in the protasis being itself the apodosis to et f<~ <TW<Tai.) In DEM. xviii. 101, K((,t T[s OU/( &v arrEKTELJIE ft OtKa,{ws, d Tt Twv ~7ra,px6v-rwJ! T1l 7roAn KaAwv ,\6yp f<6vov KQTata-x.UJ!HJ! E7r EX E [ PYJ a-' a V;-if we retain the final U]l (strongly supported by Mss.), we must translate if (it is hue that) I would (under any circumstances) have undmtalcen, etc., and not simply if I had unde1talcen (tl E7rXdPYJ<Ta,). (See 557.) 507. It is obvious that such forms (506) express simple present or past conditions, the real protasis always being if it IS (or WAS) the case that sornething would now be (or wo'uld have been), or if it IS the case that something would hereafter be undeT ceTtain einurnstances. (See 409.)




IV. I1ngula1 Oornbinations.-Present or Past with Future in one P1otasis.

508. In a

few irregular constructions, which are only cases




of anaeoluthon, the speaker adapts his apodosis to a form of protasis different from that which he has actually used. E.g. 'Eyw fLf.v liv, tl <Ixo~fL~, w> 'TaxrTa <hrAa J7row6fLYJV 'll"acr~ IHpcrat>. XEN. Cyr. ii. 1, 9. (Here f.7roW{,fL'JV &v is used as if d elxov, if I wme able, had preceded. We should expect 1rowfL'7V &v, which is found in one J\Is.) El oDv d8EI:ev O'Tt fJEaTaL allToiJs, i'V'T"O av
f_7ri ToiJs 1rovovs . Kai KaTEpya(otvTo Civ a'lh,)v, if then they knew that she (virtue) sees them, they would rush into labours and would secure he1. XEN. Cyn. xii. 22. Ei JLEV yap els yvva'iKa crwcppovecrTf:pav ~[cpos fLE8EtfLEV, ovcrK)w]s Civ ~V cpovo<;. EuR. Or. 1132. (Here we

should expect d1J.) 509. The same }}l'Otasis may have one verb in the indicative referring to present or past time, and another in the optative referring to the future. E.g. 'Eyw oilv onva Civ r'Y)V Elpya(TfLEVO>, el, OT fLEV ;u o iipxovus
ETO.T'TOV, TOT fLEV ffJ-EVO v, TOV of. 8wv TaTTOVTO<; _,\[1J"OLfLl T1JV 'T<L~~v, I should thenjo1e (pTove to) have behaved outmgeously, if when the state auth01ities stationed me I stood my g1ound, b1;t (if) 11ow when God stations me I should deseTt my post. PLAT. Ap. 28 E. (Here the snp}}Osed combination of the two n,cts is the future condition to which the future apodosis refers.) 'E1re{,xofLa' 1racrt Tof,Ttns, d a/..;78~ 1rpos BfLaS td1fOLftL Kai El1J"OV Ko.i ToT' 0evs EJ! 'TrfJ Mwif, EVTvxav pm oovvat, i.e. if I should speak the t?uth and if I did speak it then, etc. DEM. xviiL 141. El OE fL1JT EITTt fL~TE ~V fLYJT av Ei1fELV i!xo fL1J0eLS fLYJOE7rW Kat rs)fLepov, T[ 'TOV tTvfLJ3ov/..ov Jxp~v 7TOte'iv; but if there neither is nor ~oas (any such thing), and if no man yet even at this day could possibly tell of any, what ought the statesrnan to have done ? Ib. 190.

V. Seve?al P?otases in one Sentence.

510. Two or more protases, not co-ordinate, may belong to the same sentence; but one always contains the leading condition, to which the rest of the sentence (including the other conditions) is the conclusion. Here several protases may belong to one apodosis ; or the leading condition may be followed by two subordinate conditions, each with its own apodosis. E.g.
Kat yap ilv oDTO> TL 1J"a8y, Taxew<; VfLEtS npoll .P/..t7T7rOJI 'll"<n+ uen, ilv1rep ovTw 7rpocrxrJT -roi:s 7rpayfLa(Tt TOv vovl', for if anything shall lwppen to this Philip, you will soon CTeate anothm if this is your way of attending to the business. DmM. iv. ll. So xviii. 195, 217 (two cases in each). El o' ~ fkV J!EOL ois Kai ypoJ!TES, Et Tt<; f~'Y} fLrlpTave, Ot'li"AOV j3ov A.ax6vTE> ~wp8o{,fLE8' aJJ, if 'We we1e twice young and twice old, in case any one of 1ts was in fault we should seC?re a double life and set ou1selves right. Eun. Supp. 1084. See AR. Ran. 1449. El ~EVO'i f7{,yxavol' ~v, ~vveytyv<VITKET ory7TOV &.v fLO~ el EV EKELVTJ Tij cpwvij Te Kai 'Tii Tp07T'f} EA.eyo v v oi<J7rep !n8p.iJLfL'fJV, i.e.




if I were a fo7eignm, you would pardon me if I spoke in my oum dialect, etc. PLAT. Ap. 17 D. Et -rs u dvEpo tTo ToVTo, rl uTt a-x~P-a j
d7r<; OTl t:npoyyvAoTYJS, d (J'O d7rV a1f'P iyliJ, l1f'<; av OTl crx~JLcl Tt. Id. Men. 74 B. El JLEV 7rcpt Katvov nvos 7rpayJLaTos 7rpovT8To A.~ycw, i1f'tCJ'XdJV av EW<; o1 1f'AfLO'TOt TWV clw8oTWV YVWJLYJV a7rcp~vavTo, d JLEV i)pcrd T JLOL TWV inro TOVTWV pY)O~vTwv, 1)crvxav llv ~yov, cl 8 JL~, To! av dTO> i7rctpWflTJV YLYVWCJ'KW A.~y<tv, i.e. if the subject of debate were

d avT(/)


mw, I should have W(tited for othen to speak; and then, if I liked any-

thing that was said, I shonld lcee2J qniet, and if not, I should try to say something m,yself. DEM. i v. 1 ; see also xxxiii. 25.

511. It will be noticed that when the leading condition is unreal (as in Erm. Supp. 1084, PLAT. Ap. 17 D, and DEM. iv. 1, above), this makes all subordinate past or present conditions also unreal, so far as the supposed case is concerned, without regard to their own nature. Thus, in DEM. iv. 1 and xxxiii. 25 we have two directly opposite suppositions both stated as contrary to fact, which could not be unless the leading supposition had made the whole state of things supposed in the sentence unreal like itself. It is obvious, therefore, that such a subordinate condition may refer to a case which is not in itself unreal, although it is part of a supposition which as a whole is unreaL This can be seen more easily in English. '\Ve can say, if he had been an Athenian, he wonld have been langhed at if he had talked as he did; hut we Hr8 far ftom implying that the latter supposition (the subordinate one} is contrmy to fact, although it would be expressed in Greek by cl EAcyEv. Still it is ]Jart of a supposed unreal state of things. This explain~ an apparent incousistcncy in respect to sentences like dKo<; ~v er TovTo 7ra8E'i:v, yott ought J1l'O]JeTly to have wffend this, when the opposite of the infinitive is implied (415), the expression being practically equivalent (as a conditional form) to TOVTO E1f'a8e:; av cl TO clKO<; !hra&cs. As TovTo and To clKo<; are here identical, tl1e apodosis is denied in the denial of the protasis. But if a new unreal protasis is added, the opposite of the infinitiYe is not necessarily implied (see 422, 1); and if we add a concessive protasis an<l say Ka1 cl JLYJO~v ~8wrJcra<;, ElK6<; i}v CJ' TOVTO 7ra8c'i:v, even if yot had done nothing unjttst, you ought (still) to have suffered this, TOVTO generally represents wl1at actually took place (see 422, 2). Here a new chief protasis has come in and changed the whole relation of the apoclosis to the sentence. This offers a satisfactory ex1Jlanation of the a1)parent anomaly in SoPH. 0. T. 221, oD yap aJ! JLUKpav rxvvov avT6s, JL1J OV/( i!xwv Tt 0'1~JLf3oAol', where p,~ oDK exwv is ohviously equivalent to the condition l JL1J clxov, while there is yet no such opposite implied as but I have a clue. The chief condition lies in the emphatic avTo>, which is especially forcible after ~evo<; JLEV and ~evo<; 8, and involves cl JLDVO> Zxvvov. The meaning is, for I shonld not be very fa1' on the track, if I weTe atternpting to trace it alone without a clue. Thus without a clne becomes part of the unreal supposition without being itself contrary to fact, while JL~ in fl~ ovK


ll., :\:\a,



xwv shows that xwv is conilitional, and not merely descriptive (as if it were o~K xwv). For p.~ o~ with the participlc, see 818.

ll., :\Xa,


512. The apodosis is sometimes introduced by ~. dcf, or a-imf.p, but, as if the apodosis were co-ordinate with the protasis, and were not the leading sentence. This occurs when the apodosis is to be emphatically opposed to the protasis. It is especially common in Homer and Herodotus. E.g. El ~ KE p.~ ~w, ~yw KEV all-Tos, but if they do not give it to me, (then) I ~oill tc::e one myself. IL i. 137. El 'lf'Ep yap T' <'ioL YE 'lf'EptKTEWWp.EOa 'lf'aVTES V)/vulv ~Tl 'ApyElwv, vol 8' o~ os VT' d'lf'ovOat. IL xii. 245. El 'lf'EP Ka.Ta'lf'frl, da TE Ka.l p.ETO'If'tvOEv lxn KoTov. Il. i 81. El OavovTwv 'lf'EP KaTa.-?Oov.f Elv 'Alao, a~Tilp yw Kal KE0t tf>[ov p.Ep.v{pop ralpov. Il xxii. 389. El '11p.v VTt TovTo p.~ uvaTov 'lf'ot1j<Ta.&, 'p.Eis Tt Ka.l vvv K Tov p.vov '?p.:v lCEvOE. HDT. ,iii. 22. 'A'K El p.YJ8'1. ToVTo f3ovu d'lf'oKptvavOa&, a-V 81. To~VTEvBEv ye. XEN. Cyr. v. 5, 21. 513. This apodotic cannat be expressed in English ; as our averbs tlu:n, yet, still, etc., neccssarily fail to give the force of the Greek , which is always a conjunction. The expression d. viiv, now at least, is elliptical for fi p.~ 'lf'poTEpov d vvv (with apodotie d<f); as v TO lKawv d vv UHYJTE 8pv, if even now (thoztgh not befme) you 1oill do what is riyht, .AR. .Av. 1598. Sec DEY. i. 33. Sometimcs da alonc seems to imply El p.>/ TL <'io; as in Aa. Nub. 1364, Kwv' a~ov d p.vpptv'Y/v a{3oVTa TWV Alvx6ov ~at -rl p.ot, I bade kin~ at least (if nothtg more) take a myrlle ranch and give me a bit of Aeschylus. So 1369. In PLAT. Rep. 509 C, El p.'? Tt d . . 8LEgLwv, if fm notMng (else), that yoz' 'TIULY at least dcseribe, ete., a iutroduces an apodosis after El P.'? Tt (sc. <'illo~ For 8 use in the samc way to introduce the sentence UIJon which a relative clause depends, see 564.


Relative and Temporal Sentences.

514. Relative sentences may be introduced by relative pronouns and pronominal adjectives, or by relative adverbs of time, place, or manner. They . includc thereforc aU temporal sentences.




Clauses introduced by [w;, 7rptv, and other particles meaning until have many peculiarities, and are treated separately (611-661).
515. Relative sentences may be divided into two classes:First, those in which the antecedent of the relative is dejinite; that is, in which the relative pronouns refer to definite persans or things, and the relative adverbs to definite points of time, place, etc. Secondly, those in which the antecedent is indejinite ; that is, in which no such definite persans, things, times, or places are referred to. 516. Both the definite and the indefinite antecedent may be either expressed or understood. E.g.
(Definite.) Tailra xw ~P~>, you see these things which I have; or <xw ~p~>. "On ~f3ov<ro >]J...B.v, (once) ~chen he wished, he came. (Indefinite. ) llavra v f3ovwvrat l~otxnv, they will have everything which they may want ; or v f3ovwvrat l~ovcnv, they will have whatever they may want. "Orav By, ron rovro 1rpa~w, when he shall come (or when he cames), then I will do this. "On f3ovotro, roilro rrpacnnv, whenever he wished, he (always) did this. 'ils v .rrw, ovrw> rrotwp.v, as I shall direct, so let us act.

517. The relative may be used to express a purpose (565), or in a causal sense (580). The antecedent may then be either definite or indefini te.
518. When the antecedent is indefinite, the negative of the relative clause is p.'lj ; when it is definite, o is used unless the general construction requires p.i, (as in prohibitions, wishes, final expressions, etc.).

519. A relative with a dejinite antecedent has no effect upon the mood of the following verb; and it therefore may take the indicative (with o for its negative) or any other construction that can occur in an independent sentence. E.g.

Ayw oloa, I say what I lcnow. Ayw ~KoV<Ta. Agw K~Koa. ~KOVCTaV. llavra yEL y<v~CTETUL. llpaCTCTOV<TLV .

{3ovovrat (or 6Js {3ovovrat), they are doing what (or as) they please. (On the other hand, rrpagovCTLV v {3ovWVTat, or W<; v {3ovwvrat, they will do what they please, or as they please; the antecedent being




iiHldinite.) Ayw a OlJK dyvow, I am saying that of which I a;m not ignorant. 'AA. X on 8~ p' ~K Toto ovu)oed,-"7 y v, ,-' ~~s, Krti ,..5,.. 8~ 1rpos ~'OA.vp.1rov Zcrav 8wi alEv 6vns, but when now the twelfth day from that came, etc. Il. i. 493. Ts cr8' 0 xwpos O~T', EV r[i (3e(3~KafLEV, SoPH. 0. 0. 52. ''Ews ECTTt Katpos, &vnA.d(3ecr8e TWV 7rpayfLdTwv, i.e. now, while there is an opportunity, etc. DEM. i. 20. (If the exhortation were future, l1e woulJ say i!ws av Y1 Katp6s, so long as there shall 0\ l (~ \ <I ' (.! I 9 > \ ~ be an O)JJ101"tUnt"t y. ) '0 UE ava{Ja<;, EWS fLEV {JaCTtfLU 'f) v, E'TfL TOV L1f7rOV ?)yev hrd OE li(JaTa 'q Jl, KaTaAt7rWJI TOV l7r7rOJI ECT7rEVOE 1re(f7. XEN, An. iii. 4, 49. So Il. i. 19:3, Etos wpfLaLVE. 0l7rEp OE KO) TWV &7ro(3atJ16vTWV TO 1rAoJ1 '"rfJs a.l-rias EgofLEV, oVrot J(at KaB' ~a-vxfav Ti avnoJI 7rpotllwfLEl', we who we to bear, the greatm J'art of the blame, etc. THuc. i. 83. ''08eF S' o-Dv pfi<rnt fLafh}<re<r8e 1repl aVTwv, Jvnv8ev VfLaS Kai Jyw 1rpwToJ1 7retpd<rofLat 8tllacrKetv. IlEM. xxvii. 3. (Here EFnv8ev refers to the point at which he intends to begin.) "H ll0 Ao[yw <py', OTE fl <x8ob07r~<raL f_<j>~CTELS ''HpYJ, of &v fl pf.81)CTLJI .dFHOe[ots E7reecrrrtv, sunly theTe will be sad work, when you shall impel me, etc. Il. i. 518. (Here on refers to some time conceived as definite ; Whereas OT' av EpE()1)CTLV, 'tdten (if eve1") she shall p1avoJce me, is indefinite ; see 530.) Ni'~ 8' i!rrmt oTE 01) rrTvyep'Os yafLos dvn(JoA.,]rro ovA.o(J-EV'f)'> fLWEF, n)s TE Zevs IJA.(Jov d7rYJ1)pa. Od. xviii. 272. (The time is conceived as Jefinite.) 'l'')VLKavTa, OTE ovo' 0 Tt XP'J 'TfOLEi:JI fi~ETe, then, when you will not even be able to do what you ought. DEM. xix. 262. ''Ap~OfLaL 8' EJITEV8EJI o8w Kat {,"'''' tkicrf a]l fL68otTE Kayw 'TaxLrrT' av OLOa~aLfLL. DE~f. xxix. 5. (With the potential optative compare the future indicative in DEM. xxvii. 3, above.) Nvv 8~ TOVTo ovK E7ro,]rrEJI, Jv iS TC)]I 07)JhoV eTfh')CTEJ' tlv, but he did not do this, in which he might hd'Ce honoured the people. Id. xxi. 69: Els KCLAC)]I -DfLZv 'Avmo> oOE 7rapeKa8(eTo, ip fLETaOWfLEV T7JS ('l)T~rrews. PLAT. Men. 89 E (sulJjunctive in exhortation). OvKoVJI li~tov Toi:s TWJ' Ka.T')"/6pwv A6yots 'TftCTTEVCTat fLUAAoJI i) TOt<; lpyoL<; Kat T0 xp6v'{l, /)]1 l'fL''' CTacpErrTaToJI A.eyxoJ' Tov &A.'78ovs JIOfLlcraTE. LYs. xix. 61. (Here the imperative vofL[rran is used in a sort of exclamation after oF, .where ordinarily llei: JIOfL[rrat would be used. See 253.) "AF yap &1rocpvyv /U OVTOS, 0 !"0 yvo LTO, n)JI E7rW{JEAlav o<j>A.~rrw. DE)f. xxvii. 67 (optative in wish). So in fLEfLFYJfWt on and similar expressions. E.g. Ov fLEfLl'17 OTE ,.' eKpEfLW -Dfo8ev; do you not 1emem.ber (the time) when you hung aloft? Il. XV. 18. El fLEfLII')CTat OT' eyw IJOL d'TfEKptvap-')V. PLAT. Men. 79 D. Oirr()' OTE Jcpav'). EuR. Hec. 112. (See




520. A relative with an indefin-ite antecedent gives a




conditional force to the clause in which it stands, and is called a conditional 1elative. The conditional relative clause stands in the relation of a protasis to the antecedent clause, which is its apodosis (380). The negative particle is JL~ Thus, when we say & vopJ(ct rawa A.yn, he is saying what he (actually) thinkS, Or a ev6p.t(E TUVTU V'-EJ1'1 he WaS Saying What h~ thought, the actions of vop.(E and f.v6tu(e are stated as actual facts, occurring at definite times ; but when we say & Clv vop.[(y (Tawa) A.yn, he (always) says whatever he thinks, or & vop.l(ot (ravTa) ~A.cyev, he (always) said whatever he happened to be thinking, vop.(-g and vop.(ot do not state any such definite facts, but rather what some one rnay think (or rnay have thov,ght) on any occasion on which he may speak or may have spoken. So, when we say & vop.(a ravra A.e~EL, he will say what he (now) thinks, vop.(et denotes a fact; but when we say & av vop.t(v A.~et, he will say whatever he happens to be (then) thinking, vofkt(YJ denotes a supposed future case. Again,-to take the case in which the distinction is most liable to be overlooked,-when we say & 01lK oioa ovK oi' lovat, what I do not know, I do not think that I know, ovK oioa, as before, denotes a simple fact, and its object a has a definite antecedent; but when Socrates says & p.~ oioa ovo~ el8vat, tbe meaning is whatever I do not know (i.e. if there is anything which I do not know), I do not even think that I know it. In sentences like this, unless a negative is used (518), it is often difficult to decide whether the antecedent is definite or indefinite: thus & oloa. oZOfkUL eio&vat may mean either what I (actually) know, I think that I know, or whatever I know (if there is anything which I know), I think that T know it.
521. The analogy of these indefinite relative clauses to conditional sentences will be seen at once. The following examples will make this clearer : ''0 n f3ovA<Tat 80r:rw, I will give hirn whatever he (now) wishes. EZ n f3 oDAETai, owr:rw, if he wishes anything, I will give it. (402.) "0 n if3o-6AeTo ~8wKa liv, I should ha1'e given hirn whateve1 he had wished. ''0 n 1'-YJ EJEVETO OVK av el7rov, I should not have told what had not happened. Et n f.(JovA.ero, ~owKa. U.v, if he had wished wn'Jthin[t, I should have given it. Er Tt 1'-'J lyev1"0, OVK av l7rov, I should not have told anything if it had not happened. (41 0.) "0 Tt Clv (JovAr]Tat, 130r:rw, I will give hirn whateve1 he shall wish. 'Eav n (JovA'fJTat, 80r:rw, if he shall wish anything, I will give it.

''0 n f3ovll.otTo 8ofYJV liv, I should give hirn whatever he rrvight wish. 'Ei' n f3ovAotro, oo[r]V av, if he should wish anything, I should give it. (45 5.)




"0 n &v (3o-6A'fJ'Tat 88wp.t, I (always) give him whatever he wishes. ''0 n {3o1)Aot'TO e88ovv, 1 always gave him whatever he wished. 'E&v n (3o-6A-r]Tat, 08wp.t, if he ever wishes anything, I (always) give it. Et n (3o-6A.ot'To, iotoovv, if he ever wished anything, I (always) gave it.

522. The particle (Epic Id) is regularly joined with all relative words when they are followed by the subjunctive.
With ore, d7rO'TE, E7rEL, and E7ret81, fiv forms 8rav, 01rOTav, E7rd)J or (Ionic E7re&.v), and E7retO&.v. In Homer, where KE is generally Ke), also o'T' &v, where used for &v, we have O'TE KE, E7rEL KE, etc. (like in Attic we have orav, o1r6rav, E7ret8&v. 'E1r~v, however, occurs often, and 1rd &v once, in Homer. Both e1r~v and e1rav are rare in Attic.



523. The classification of common conditional sentences, with four classes of ordinary conditions and two of general conditions, given in 388-395, applies equally to conditional relative sentences.

524. The conditional relative sentence has jou1 forms, two of present and past (52 5 and 52 8) and two of fnture conditions (529 and 531), which correspond to the four forms of ordinary protasis.

525. When the relative clause simply states a preseut or past supposition, implying nothing as to the fulfilment of the condition, the verb is in one of the present or past tenses of the indicative. The antecedent clause can have any form of the verb, like an ordinary apodosis. (See 4 0 2 .)

"A p.i) oloa, ovo oZop.aL elo~vaL (like er 'TLVa p.?] oioa). PLA'l'. Ap. 21 D. (See 520.) Xp+r&wv on {3ovAovraL, let them deal with me as they please (i.e. T (3ovA.ovraL). AR. Nub. 439. 'E7r< opav 8' Oft p.<, Ko:Ux opav <lp.1J 7r p 7r EL' I know how to see anything which I ought to see, and not to see anything which I ought not. EuR. Ino, Fr. 417. ("A 8ei: is nearly equivalent to eZ nva oef, and <lp.1) 1I'pE7rE to d nva p.ij 7rpE7rEt.) To0s 7rAEl<rTOV<; gvea7r<p g1rE(J'OV EKci(J'TOV> geafav oils o !'-~ evpt(J'Kov, KEvonfcpwv aV'TOL> E7rO['fJO'av, i.e. they raised a. cenotaph for any of them whom they did 'f!Ot find (like d nvas p.0 <vpLU'Kov). XEN. An. vi. 4, 9. T yap; oU'ns oa1ravrypos ~v p.1) a-DrapKYJS e 2v, <lA.>.: (J'T &et TWV 1rAIJU'[ov Oft'Ta t, Kat A.ap.f3avwv p.ij o-6va'TaL U1r0000VaL, p.ij





AafLf3ri.vuJV 8 TG!J Jl-1J OLODVTO. punt:, OOKEl U'OL Kat OUTOS xaAE7rbS cp[Aos eCvaL; (i.e. SU]Jposing a case, <t ns , . . p1) aimfpKYJS JiTrlv, K.r.>...). Id. Mem. ii. 6, 2. So 1)ns fl-'18apov ~vf-paxet:, 'l'Huc. i .. 35. "A yap ns fL~ 7rpoU'E00K'Y)IJV, ovo~ cpvAa~arrfJaL eyxwpe, for there is no opportunity even to guard against what we did not expect (like d nva fL~ 7rpOIJEOOKT)IJ~ ns). ANT. v. 19. Eis rd.. 1rAoia rolls 1' drrOevovvros Jvef3f:3rumv KoJ TWV (TKfVWJI ocra fL~ dvayK'Y} l}v ~XEtV (like d nva TW!J CTKm;;v }"~ dvriyK'Y} 1)v ;XEw), i.e. any which they did not need. XEN. An. V. 3, l. 'Av0pw7rOVS odcpfJHpEV (0 ea>..arrcra) OCTOL JL1J io1lva J!TO cpOrjvaL 1rp6s rd.. JLErwpa dvaopafLOvnr;, i.e. if any weu unable to escape soon e1wugh to the high land, so many the sea destToyed. 'l'HUC. iii. 8 9. Ois f-EY oXptrrLs y<yevryra' rll>..>..a. drvxovrr, 7fOAA1J lfvoa 1T"OAEfL~cra EL o' dvayKatOJ! ijv, K.T.A., for any who have had the choice given them, while they are p1osperous in other respects, it is gnat folly to go to wa1 (i.e. Et TLrTiV atpHns yeyvqra). Id. ii. 61. IIdvTES rrr}"EV Xaf3ptav oVrE T117fT011TU. oVO' O.p7r&(o!ITa rOv a-rf<jJa.vov oiJ8' OAwr; 7rpoCYt6v6' f!nrot f-'J 1rporrqK<V a1lT<(J, nor going anywhen at allwhen it was not lawful for hirn (i.e. d 1fOi P-0 1rpocn}<ev). DEM. xxi. 64. IIws oDF oi dyaOoc 'Toi<; dyaOoZs cJ>O...o< EiTOVTa<, oi' fl-~T< d1r6vns 1rofJuJ1ot dA.A)Ao<<; I"''JT 7rap6vn<; xpdav airwv (xovrrL; (i.e. El fL1J iixovU'i). PLA'r. Lys. 215. 13. ~ <K{>YJ ll' n 1racnv {,yZJJ JJ-EAA o-wo[a-etJ' (i.e. n fLaA<), may any plan prevail which will ben~fit you all. DEM. iv. 51. 526. Care must be taken here (as in conditional sentences) not to include in this class general suppositions which require the subjunctive or optative (532). On the other hand, the examples falling under 534, in which the incticative is allowed, might properly be placed here, as they state a general supposition as if it were a particular one.



7rap~iTra, each rnan felt that all progress was at an end in any affai1 in which he was not persona.lly to take !?ad. THC'O. ii. 8. The direct form was v rourtp KEKWAvrai (51 ; 122) ~ 1~1) 1rapf.iTo}"aL. Oi'> 8~ aA1Jeda<; ns dTVX1)a-Et, 1rOTE T01;TOV J1fLCTT1JfJ-WV EiTTUL; b7tt if one is to miss the tntth of anything, 1vill he ever understand it? PL.AT. 'l'heaet. 186 C. So probably XEN. Cyr. i. 5, 13, o n yap fl-'l rowvrov a1fof3~(TET(~b 7rap' VfLwV, <ls ~}"E TiJ V .. Amov g(TTaL, i.e. if there is to be any failure on your paTt to come ~LP to my expectations, the loss will jaU on me. This is the only form of conditional relative sentence that regularly takes the future indicative. (See 530.)

527. A conditional relativQ clause (like a clause with El, 407) may take the future indicative to express a present intention or necessity. E.g. 'Ev TOVT!p KEKWAviJea, JooKEi EKclO""T!p rd.. 7rpay{LaTa .p 0~ TLS O.VTO<;

528. When a relative clause expresses a present or past

condition, implying that it is not or was not fulfilled (like a protasis of the form 41 0), the verb is in a tense of

the indicative.




The antecedent clause generally has a past tense of the indicative with &v ; but it may have a past tense of the indicative in an unreal condition, in an unaccomplished wish, or in a final clause. E.g.
"A fLfJ Jf3ovAETO 8ovvo.L, olJK al' ~0WKV, he would not have given what he had 1wt wished to give (i.e. d Two, fh'l /JovA.ei-o 8ovva, ovK &v (8wK<V). '01r6npov TOVTWV J'Tro{Yjrr<v, oiJOEVOS &v ljTTOV 'ABY)va{wv 1rAovrrw 1]crav, whicheve1 of these he had done (he did neither), they would be as ?"'ich as any of the Athenians. LYS. xxxii. 23. OvTE -yap &v avTot E'Tr<X<povfLEV 1rpd.Trnv fh'l 1J7rLcrTrip.,<Ba, ovT< ToZs &A.IIms E1rETp7rOfLEV WV i)pXOfJEJ! dAAO TL 7rpaTTELV i) 0 TL 7rprf.TTOVTES 6p{)ws (p.,EAAOJ! 7rpa~ELV' TOVTO 8' 1JV av ov E7T'Lrrn)p.?J1! Eixov, joT (if that wme so) we should not /,e undeTtaking (as we a1e) to do things which we did not undentand, nor should we pennit atny othe1s ~oho1n We Were 1uling to do anything else than what they wete lilcely to do propeTly j and this would be whatever they had knowledge of. PLAT. Charm. l 71 E. (Here il fL~ 'l7rLCTnip.,E8a = d TLVa JLij >}7ra:rnip.,<8a, if there were any things wlu:ch we did not know,-iliv ijpXOfLEV = Z TLVWV >JPXOfJ-EV,-0 'TL Jp.,EAADV = ot Tt f.p.EAAov,-and oi5 f.mcrT~fl-YJV <lxov = d nvos dxov. It is implied that none of the cases here supposed ever actually arose. "!Jcr1r<p TOLJIVV 0:.\A.wv TtvWv TET-rO.pwv, ti Ev Tt f.(YJTOVjl-EJJ a?YrWv 1' &rq.>oVv, 01r6T 1tpWTov lKdvo g-yvwp.,H, lKa.vws a1' EiXEV ~p.Zv, El 8 TU Tp[a 7rpoTEpov -yvwpcrap.,Ev, aVT\0 all ro!JT'f:' f.-yv<f>puno TO C?JTOi!w:vov. PLA1'. Rep. 428 A. (Here the antithesiH of ,l1f6n 1rp.wrov EK<Zvo 1!-yvwp.,w, in (whatever) case we had ncoguised this first, and E i -ra Tp[a 1rp6np01' f.-yvwp[rrap.,v, if we had recognised the thTee sooner, makes the force of the relative eBpecially dear.) 'E(3acrd.Fo(ov av fLEXP' ov aVTOts E86 KE ,, they would have questioned thern (undeT to1tun) so long as they pleased. DEM. liii. 25. El 8 otKo Elxov f/((L(J'TDt TUS 8Ka<;, TD!JTOV> av a1rwAAV(J'O.l1 o!'nll<; </>LAOL p.,aALCTTa ~rraJI 'AB17vawv T0 o,)p.,'{J, if each had theiT t?ials at home, they would ~uin any who wme especially friendly, etc. XE:<:. Rep. Ath. i. 16. (Here otTLV<<; ~(J'av, = EL TWES ~crav, forms a second protasis to the apodosis d'TrWAAvcrav tiv. See 511.) Kd 67r>)V[Ka f.<j>aiv<TO Tavnt 7r1T'OL?JKC;;s, WfLOAo-y<Z-r;' al' ~ K<lT7J"iOP[a TOt<; p-yoos aVTOV, and if he ever appeaTed to have done this, his form of accusation would agne with his acts. DEM. xviii. 14. El ~vas hv-yxavov <l'w, t1)Jieytyv<brrKETE 80rrov dv p.,ot el El' JKdl!TJ <j>wvfj TE Kat T<{i Tp<\7f'{J iiAE-yov Ell oTcr7rEp ETeBpaJLJLYJV, if I happened to be a foreigner, you would surely ]Jardon me, if I '!l'M'e (now) addnssing you in both the language and the man1wr in which I hc!d been bro'!tght up. PLAT. Ap. 17 D. 'fls So) -yw -y' o<j>EAOJI p.,aKapo> v.U TEV ffJ-fJ-EVUL Vto> dvipo~, 3v KTEancrcrLv o!c; br, -y~pas ETETfJ-H, 0 that I weTe the son of some fortunate man, whom old age had found upon his own estate (i.e. if old age had found any snch man, would that I had been his son). Od. i. 217. SoIL vi. 348 and35l. So when tl1e relative sentence depends on a past indicative in a iinal clause (:333); as in DEM. xxiii. 48, TavTa -y< 307rov rrpocr~K





ypafat, rva. orcr 1ror~ rol!pyov pri x 01), ro-Drcr ra eK rwv vofl-wv il1rijpx 0Kata, he ought to have 11Yritten it in this wa.y, in order that any one by whom the deed had been done might have his rights according to the laws. (This implies that the law was not so written, so that the case supposed in OT!f:' enpaxOYJ never arose.) So DEM. liii. 24, tv' aKo-6(]"0.VTES h rovrwv ffYJ<f>la-a.a-Oe a1rot6v rt VfJ-tV J86KH, tlw.t you might have voted whatever seemed good to you. All examples of this form fall equally well under the general rule for assimilation (559).




529. (SubJunctive.) When the relative clause expresses a future condition of the more vivid form (like a protasis of the form 444), and the verb of the antecedent clause also refers to the future, the relative is joined with /Iv (or KE) and takes the subjunctive. .E.g.
TU:wv ~V K UJEA.wfkt <f>lArtV 7r0t~O"Ofl-' tl.KotTtV (like KE nva UN~ AWfJ-t), whomsoeve1 of these I may wish I shall make my wife. Il. ix. 397. 'E yap '0p<O"T(J.O 'TtO"tS EO"O"ETO.L :>ATpowao, 07r7rOT ay 'fJ f3 'f)O"'{J T K(U ' ,f.~ ' ' ~K :fjs lfkdpeTat a.Zl)>, i.e. vengeance will come from Orestes, when he shall grow up, etc. (like Uv 1roTe 1}f3~a-v). Od. i. 40. Ton B' a.On p.axr)a-E'rat, o7r7ron Kf.v evJLo> evt a-r~Oea-a-w dvwyv Kat Beos /Jpa-17. Il. ix. 702. 'AAA' llyeB', ws av ywv d'1rw, 7retBwJL<Ba 1ravns, let us obey as I may direct, i.e. if I give any direction (av 7rWS d1rw), let us obey it. Il. ii. 139. 'Hp.ei'> avi dMxous T <f>[ Kat v~ma TEKVa tl.~OJLEV Jv v~ea-a-w, E7r~v 7rToA[d)pQV lAwJLEV, U'he7~ we shall have taken the city. Il. iv. 238. So VT av 1ri7rTWO"tY, Il. i. 242. OvKovv, orav 01) p.~ a-Bvw,, therefore, when I shall have no more strength, I will cease. SoPH. Ant. 91. 'l'avra, E'lretOd.v 1rept Tov y~vovs d1rw, TOT epw, I will spealc oF this, when I shall have spoken about my birth. DEM. lvii. 16. (See 90.) 'E7rELOdv Ota7rpd~Wfk<Lt (l OEOfJ-O.t, ~~w. XEN. An. ii. 3, 29. T!va oiw-&e avT1JV !fvx~v :tgew, OTO.V Efk~ rov -rwv 1rarp<{lwv d7rea--rEpYJJLEVOV; what feelings do you th~1tk she will have, when (or if at any ti1lle) she shall see me, etc. 1 DEAf. xxviii. 21. Tovrwv 8~ 'AB'f)valovs </>1)JL2 Oetv el'vat 7rWTUKOO"[ovs, E~ ~ .. av nvos VJLfV ~AtK[a., Ka.Aws lxew 8o Kij, from. whatever age it shall seem good to you to take them. (i.e. if from any particulctr age, etc.) Id. iv. 21. Twv 7rpayf1-drwv , ' / ~ ~ ~ ~ TOVS f3 01JAUOfi'EV018 Yj')IELO" at Oet, tVa a ay EKELJIOLS UOK'{J TO.VTa 7rp<iTT1)Tat, in mder that whatever shall seem. good to them shall be done. Ib. 39. Oil fl-Ot q)6f3ov p.A.a0pov EA7rtS JL7ra"rtV, EWS av a [ Ov 7rvp <j>' Ja-T[as p.'ljs Afyta-Bos, so long as Aegisthus sha.ll kindle ji1e upon my hearth. AESCH .Ag. 1434.






') " , " , /

530. The future indicative is very rarely used in conditional relative clauses, as it is in common protasis (44 7), in the place of the subjunctive; as it would generally be ambiguous, appearing as if the ante-




cedent were definite. Some cases of oO"o<; with the future, as f3ovA~U"ovro.t, THuc. i. 22, are perhaps exceptions. (See 527.)

531. (Optative.) When the relative clause expresses a future condition of the less vivid form (like a protasis of the form 455), and the antecedent clause contains an optative referring to the future, the relative takes the optative (without av). The optative in the antecedent clause may be in an apodosis with Clv, in a protasis, in an expression of a wi<Jh, or in a :fiual clause. E.g.
MaAo. KEV Opa.(]"l!Kapows Ei~l, Bs TOT 'YTJ(}1)(]"tEV lowv 'll"OVOV o1l3' dKa x o tro (i.e. d ns ;rle~(]"o<, pAA.a Kev Bpa(]"vKapow> dTJ), any one who shonld then Tejoice wonld be veTy stout-heaTted. Il. xiii. 343. BovAo[p.qv K' f7rapovpo> Hvv eT)TEV~fLEV dA.A.t/) 0 fl'J f3oTO'; 'lrOAVS e !.'"I, I should wish to be a smf attached to the soil, smving another man 1uho had not much to live on. Od. xi. 489. z'lJ''"'' OVK &v ([(]"(]"01! iKOLJLrJV, dTE fi'J avTo<; -ye KEAoiot, unless he should himself bid 1ne. 11. xiv. 247. SoIl. vi. 329 and 521; and O(]"Tt<; K a AEO' te, AR. Nub. 1250. 01lK &v 001 Bpf.lj;at<; avop(t, ocrn<; f_(}A.ot T Ka.l OVJIO.LTO (I'OV a7rEpVKHJI TOV<; inxetpovJ!Tas d8"K<cv (1'; wo1dd you not suppOTt any 1nan who sho1tld be both 1oilling and able, etc. I XE:'l. l\1em. ii. 9, 2. ITcwwv q>ayo" &1' 07roTe f3o1!Aotro, 1vhen he is hungry, he would eat wheneveT he might 1vish (like Er 7roTE f3ovA.o"To). lb. ii. l, 18. So i. 5, 4; i. 7, 3; iv. 2, 20. llw> of>v aJ! doELYj> 7rEpi TOlJrOV TOV 7rpa-yj.LO.TO> oil 'lrO.VTa'ffO.U"LV a'lrEL{JOS E ( T} S; how then COUld you know abou.t that thing of Which ymt, had no expe?ience at all? PLA'l'. Men. 92 0. .,.Ap' (lp ~-yoL'o TO.VTO. era ElJJat, (]'Ol E E [ 'Y) Kai d7roOocr8a.t J<ai 8ovva.t Ka.i ever at OT<p f3 0 vA 0 L0 BED)]'; ~ Id. Eutllyd. 302 A. T av 'ffaBdtJ (8vvatTo), 8 fL1J Kai vf avrov 7ra8o L; what could he suffeT, unless he should suffer it also j1om himself? (i.e. El /L'l 7ra8ot). Id. Lys. 214 E. ''0 0~ fL'l dya.'ff0TJ, ovo' &v </> tAOL (i.e El Tl fL'J d-ya.7r0rl, ov/3' dv <:/JtAoL' 'TOUTO.). lb. 215 B. 'Iol.av EKO.(]"TOS &v KO.T(.HTKV0v KO.TQ,(]'KEVa{;otro, ;;ns EKUCf'TOV apE(I'J<Ot. Id. R.e]J. 557 B. "OO'<p ol 7rp<cr/3{JTEpo<; y fyvo LTO, fLUAAov ad acr7ra{;o lTO aJ' (XP~fhara), the oldeT he shonld !JTOv;, the more he wonld always cling to it (i.e. d n 7rpecrf3unpos y[yvotTO, TOCf'OVT<tJ fl-UAAOJ! dcr7ra{;otTO av), Ib. 549 B. So 412 D. .P,)uofl-EV jL?)OE'ffOTE fLYJO~v 8v fLEZ(,ov fL'l}O~ EAO.TTOV -y<vcr8at, i!ws tcroJ! EL?/ ai!r6 avT<j), so long as it should Ternain to itself. Id. Theaet. 155 A. El8 f3ovAoto TWJI ,pA.wtl 'TlVa1rporp~fa.cr8at 01l'OTE a7rOO'ijfLO[T)S E7rtfLEAEr:(l'ea, TWI! crwv, TL &v 1l'Ot0LT)S; XEN. Mem. ii. 3, 12. ElK6rws v Kat 7rap0. BEWv trpaKTtKWTEpos Ei 1], Oa-TI.S fJ-1) Orr0TE Ev d7r6pots r1} TOTE KOAaKdiot, aAA OTE Ta aptO'TO. 7rpaTTO TOTE fLrlAt(I'TO. rillv 8EWI! fLEfl-ViJTo. Id. Cyr. i. 6, 3. '[1., a7rOAOtTO Kai G.A.A.os:, 0 TL> Totavnf y< p~ {; o ', 0 that any othe, rnan might likewise perish who shonld do the like (i.e. Ei: ns Towvru f)f.{;ot). Od. i. 47. El yrip JL<I! 8av6.Tota




O'T p,w p,opos atvos iKdvot. Il. :s:viii. 464. 6.wpa 8wv IIxoL, oTTL ~t8oi:v, rnay he have gifts of the Gods, whateve1 they rnay give. Od. xviii. 142. 'Eyiyvwa-K<: DEtV TOVS D1r7JpETOS TOVTO U(]"KfLV, ws 'lrcfvTa vop,i(oLEV 1rpi7rELV avTo'is 1rprlTTEW o<ra 6 llpxwv 1rpo<r-rcfTTOt. XEN. Cyr. ii. 1, 31. For KE or liv l.n these 1elative sentences in Homer, see 542. All these examples fall also under the general rule for assimilation (558).

~VtT1Jxios iDBe Bvva [p,'Y) V votr4>tv b:;roKp{;fat,

532. A conditional relative sentence may express a gene?al supposition, when the verb of the antecedent clause denotes a customary or repeated action or a general truth, while the relative clause refers in a general way to any act or acts of a given class. Here the subjunctive with Sc; fhav, etc., follows primary tenses, and the optative (without &v) follows secondary tenses. (See 462.) E.g.


'Ex8p6s ycfp JlOL KEtVOS 6p,tJs 'Atoao m)A1)<TW, os x' ETEpov p,~v K{;()17 e11i <f>pE<rtJI, liJI.A.o 8~ d7r1J, joT that vwn (i.e. any man) is hated by ?ne like the vmy gates of Hades, who conceals one thing in his mind and speaks anotheT. Il. ix. 312. N EJLEtTtTWJLaC ')'f JLEV ovbEJI KAa[Ew os K uvv <T L f3poTWV Kat 7rOTfLOl' EJl t<r7rJ), I am nevm at all indignant at weeping for any 'nwrtal who nwy die, etc. Od. iv. 195. Oivos-, os n Kat liAAavs j3Acf7rnt, os lill f-tLV xaJIOOJI H 17 fLYJ0 ai'a-tjLa 1r [ VJ). Od. xxi. 293. Kd yap trVJLJLa)(E'iv TDVTots JBI:Aov<rw &1ruvns, oils a11 6pw<rt 1ra(JE<TKVa<TjLEJIOvs, for all men ate (always) willing to be allies to those whom they see prepand. DEIM. iv. I\.a1rp rwv dFBpw7rwv, 11 l{i f'~l' av 'lrO AE fLW<T t, TOV 7rap6vm (11'0AJLOV) dd fLE')'L<TTOV !((JLV6vTWJI, although men always conside? the present wa1 the g1eatest, so long as they aTe engaged in it. Tauc. i. 21. IlopE{;OVTU.[ T yd.p at V av avTOS ev8{;Jiw(J"[,V ob vopErs, v~p.owra.t T xwp{a Ecp' 07rofa &v aVTds E<j:aW(JtV, > / '<' ,\ ) \ ) ' \ "' /" a1rEXOJITULI TE WV aJ' a1nas a7rupyw<TL" Kat TOtS Kap7roLS fW<TL 'TOVS l'OJLEUS XP~<r&at OVTWS 07rWS' QJI aVTOt j3o0AWVTUL' av8pw11'0! 0~ e7r' ov8vas fLUAAov <Tvv<rTavrru 1) f7rt TOVTOVS ovs a]! aZ<r8wvTaL ap)(LV aDTwJI mxtpovFras. XEN. Cyr. i. l, 2. NoJLCw 1rpoa-TaTov lpyo11 El'v(u o?ov Oef, Os Civ OpWv ToVs cjJ[Aovs E~a-n-aTwp..fvovr; f-1-?1 1rtT p f1r1], i.e. such as one onght alwa,ys to be, wlw, etc. Id. Hell. ii. 3, 51. KaTa<f>pov-IJ<TLS o~ (yyyF<To.t), os a11 Kat ')'VWJLTJ 7r wTdlJ TWJI f.vavT[wv 1rpo~xw, ~fLi:v D1rapxEL. THuc. ii. 62. (Here the ll refers to all that precedes, as a definite an teced en t.) Ov JL~v yd.p JL<i:Cov KAos dvepos, Jcppa K' ;TJ<Ttv, i) on 7ro<r<riv T< p~~17 Kat )(Ep<rv. Od. viii. 147. ("O<f>pa K' EJ)<TLV, so long as he lives.) (8w-D>) 7rapaTpW7rW<T' avBpw7rOL AL<T<TOfLEJIOL, OT KEV TLS V7rpf3~17 Kat; a{LcfpTTJ. Il. ix. 500. ''HJLL<TV yO.p T U(JfTfjs J:rroaivvTaL <{,p{;o7ra z.i>, J,,.~pos, eii-f av fLLV Ka'ra oovAwv ~fLap i!A[)<TLV. Od. xvii. 322. .Pt'An 8 Kws 7rpo<r7JJ"a{vHv, dh-' <lv JLgAYJ fL<yaA.a. KaKa i} 11'oAt i}







8vei" EO"E0"8at. HDT. vi. 27. .Pevyo110"t yap TOt xol 8pa.O"El<;, OTO.V 71"EAO.<; TOJI "AtOrJV elO"opwO"t TOV (3ov. SOPH. Ant. 580. 'Hv[K'_ av o' ofKot yevwvTat, opwaw OVK dvaO"XETa. AR. Pac. 1179. 'E7rHOav 0~ ~ hq)opd. fl, Aapva.Ka<; ayo110"W &jka~at. THUC. ii. 34. 'E7rlOav 0~ KpvfwO"t y{j, dv1]p UPrJJkEVO<; inro T1j<; m5A.ew<;, o<; llv yvw!k17 T OoKfj }k'} d~1jJ1ETO<; elvat, A.f.yet i'll a.ljTo'is E7rO.WOl' Tov 7rpE7rOVTa. Ibid. "Ew<; &v 0"0trJTO.t TO O"K!fcpo<;, T6n XP~ 7rpo8vp.ovs elva.t i1retoaJI o~ ~ 8aA.aTTa imepO"xYJ, JkaTaws 1) O"'Tro11o,). DEM. ix. 69. So EO"-r' O:v OeLO"WO"tv, XE:{. Mem. iii. 5, 6. "Ov o' av 01JfL01! avopa tOOt (36wvTa T' Jcpevpot, TOJI O"K~7rTp<p VviO"amv, whatever man of the people he saw and found brawling, he dTove hirn with his scepbe. Il. ii. 198; see ii. 188. Ov Two. yap T[eO"Kov E-nx8ovwv dv8pw7rWJI, ov KO.KOV ovo~ jk~V E0"8A.<w, 0 T[<; O"cpeas eLO"!LcpKo t To, i.e. they were never in the habit of honouring any one who came to them. Od. xxii. 414. ''On Jk~V O"KtpT</iev, . . . 8eoJI. ll. xx. 226; so 228. See Od. xx. 138. Kai ovs jkEJI tOot EVTrLKTw<; Kai O"tw7rfj l6vra.s-, 7rpoU"EAaVvwv aVroZ~ r[l,Er;. TE ElEv '~fpWra, Ka'i E7rt'i 7rV8o t.TO E7rrJl'H. XEN. Cyr. v. 3, 55. (Here 1]pwTa and hrz)vet denote the habit of Uyrus.) Ka.i To 'is jkEJI 'A8rJvaot> rJV~eTo TO JlavnKOJI d1ro T1js Oa7raVrJS -i)v EKElJIOL ~11jkcpEpotev, avToi OE, omhe d7rOO"TaZev, &:rrapaO"K110! Kai a11"etpot E<; TUV 11"6AEjkOV Ka80"TO.JITO, and the Athenian navy continued to inc;ease jTom the money which these contributed (pres.), and they, wheneve?" they 1evolted (aor.), al;ways found themselves unprepared and inexperienced for war. THuc. i. 99. 'E7ri Mo[pws (3wnA.os, oKw> EA.8o t o 7roTaJkO> 7r' OKT:iJ 1r~xws, apOEO"K Aty1111"T0l' T1JV vep8e Mjkcpws, i.e. whenever the river rose. HDT. ii. 13. Tov o xovv Tbv EKcpope6}kevoJI, OKws yf.votTo v1,~, Js -.(w T[ypw J~ecp6peov, i.e. they carried it away every night. Id. ii. 150. Oi o (KO:pes), OKW<; M[Pw<; oeotTo, E'71"A~pow ol Ta<; JIEa<;. Id. i. 171. 'E7rH01J o dvotxfJdrJ, elO"iJHJkEv 7rapa Tov L.wKpaTrJ, i.e. each moming, when the prison was opened, etc. PLA~'. Phaed. 59 D. "On ~w TOV OHVOV yevoLVTO, 71"0AAot avTOll d71"EAH71"0l', rnany used to leave him when they were o'ut of dange1. XEN. An. ii. 6, 12. (If yvovTo had been used, the whole sentence would refer to a particular case.)

533. The gnomic aorist and the other gnomic and iterative tenses (154-164) can be used in the antecedent clause of these general propositions. The gnomic aorist, as usual, is a primary tense, and is followed by the subjunctive (171 ). E.g. ''Os /( 8eaZs E11"L11"E8rJTO.t, jkrLAa T' EKAVOJI avrov, whoeveT obeys the Gods, to him, they an ready to listen (EKAVOJI is aoristic). Il. i. 218. "OTaJI TL<; W0"11"Ep oVTO<; 10" X VO"YJ, 1) 7rpWTrJ 7rp6<j:>aO"L<; &1ravTa d veX a[. rtO"E Kai OL~AvO"ev. DEM.ii. 9. '011"6Te 7rpo0"{3A~IfELE T!Jia<; TWV iJI TaZs TrL~EO"L, el7rEJI av, 6) avopes, K.T.A., i.e. he used to say, etc. XEN. Cyr. vii. 1, 10. Ovl aAAOTE 71"W71"0TE 7rpos xaptv dA.6jk'f}JI A.eyHv, 0 Tl. Clv fJ-11 Ka2 uvvo{rrl..JI 7(7rEUTft~VOS m, I have neveT on other occa..c;ions prefernd to say anything to please which I have not been convinced would also be joT your !tdvantage. DEM. iv. 51. (Here eiA.6p]v has a sense




approaching that of the gnomic aorist, and is followed by a subjunctive. See 156.) Homeric examples of relatives with K~ or and the subjunctive in general conditions are here included with the others, because this construction is fixed in the Homeric usage. In the greater number of general relative conditions which have the subjunctive, however, Homer uses the relative without KE or as he prefers the simple el in the corresponding conditional sentences (468). See examples in 538.



534. (Indicati?Je.) The indicative is sometimes used instead of the subjunctive and optative in relative sentences of this class. (See 467.) Here one of the cnses in which the event may occur is referred to as if it were the only one. This use of the indicative occurs especially after the indefinite relative ocnts; as the idea of indefiniteness, which is usually expressed by the subjunctive or optative, is here sufficiently expressed by the relative itself. E.g.
'Ex8po<; y<fp fLOt KEtFos dJlJ;J, 'A8c.w 1n)A.yyrw ryvTctL, O<; 1fEV[YJ ELKi!)V rho.Tf)Ata (Jri(o. Od. xiv. 156. Compare this with I!. ix. 312, the first example under 532. 'EI',o~ yd.p oa-n<; 1raa-av d8uvwv 1fOAW M~ rwv dp[rrrwF d7rTETat (JovA<VfLUTwF, 'AA.A' f.K <f>o(Jov rov yA.wmrav iyKA.da-as EXEt, KriKurTo<; dFat vvv TE Kat 8oKEZ Kal fLd(ov' oa-ns dvrl r'lj<; O.VTOV 1fchpas NA.ov VOfLi~EL, TOVTov ov8o.fLOV .Af.yw. SoPH. Ant. 178. (Here we migllt have had ;)<; av . . fL'l d1rT1JTILl, rlA.X . . . EXT/ and Ss &v vofLt(IJ, without auy essential difference in meaning.) OZTLvEs 1rpo> rd.s ~vfL<f>opd.s yv<JfLYJ fLEV 'JKlrrra Almovvrat, ~PY'' 8 fLd.AW"Ta d vr X ovrr LV, o-DTot Ka~ 'lfOAEwv Ko.t latwrwv KpaTLrrTo[ eirrtv. THuc. ii. 64. So in the same chapter, orrrts Aa.fLfJ ri VEL. ''OrrTL> ~, ,~ ~ ~ , f3 Q.(l"l/\.EW$ 1rpos Q.VTOV, 1fQVTO.S OVTW Ui!LTL Et<; ', , ,, , ., " U IL'f'tKVElTO TfLlJ! 1rapa d7rE71"EfL7rETo, whocve1 came to him, he alwctys sent away, etc. XEN. An. i. l, 5. ''07roll 3 XLAO<; rr7rd.JitO'; 1fUV1J er,], ILVTb<; 8' 8uva,To 1rapafJKE11UfJO.(J"8at, 8ta7rfp7rwv EKEAEVE rovs <f>[A.ovs t'lr'lrots ip(Jri'AA.Etv rovrov.


Ib. i. 9, 27. (In the last two examples there is some Ms. authority for the more regular rl<fnKvo"Vro and 8vva.tro.) 535, This use of the :indicative (534) is rare in tempoml sentences. See, however, the following : IIEpt ,-wv &A.A.wv rwv dotK011v,-wv, orE a.Ka~ovra, aE;; ,-wv 4WT1Jy6pwv 1rv8f.a8at. LYs. xxii. 22. Eixov fLa.xaptov, rp rr<f>aTTov
G>v KpaTEZv 8Uvawra, Kat a1fOTEfLVOVTES av TdS KE<f> ~XOJITE<; E1f0(lEUOJITO, O'lfOTE o[ 7r0Afptot o:vrov<; otj;Errea.L EfLE A.A.ov. XEN. An. iv. 7, 16. So 01fO'T d<f> [rrraro, ii. 6, 27 .

.All these examples fall under the first class of conditional relative eenten_ces (5 25 ). 536. The Greek generally uses the indicative in relative clauses




depending on general negative sentences, where in Latin a subjunctive is more common. A general negation is really particular. E.g. Ilap' f.p.ot OE ov8ds JJ-LCT8ocpopel:, oa-ns Jl-1J lKav6s EO"TLV Zero 7TOIJtV JJJ-o, i.e. no one who is not able (no one unless he is able), nemo qui non possit. XEN. Hell. vi. 1, 5. Ovods yap OV0J!t wpy(ero OCTTLS JJ-1J ,PE TO d7ToAZcr(hu, for no one was angTy with any one who did not think that he was abont to peTish (i.e. d Jl-1J 0To). lb. vii. 4, 37. Ovoafl-OV 1rc!J1To8', /hot 7rpcr(3VT'f}S E7TEfJ-cp 81] V vcp' V(J-&'w f.yliJ, 'lTT1Jed, d7TijA8ov T&'Jl! 1rapa cp,,\.{7T7TOV 7rpcr(3wv, nowhere, whither I was sent as arnbassador, did 1 ever corne ojj' WO'l'Sted by Philip's ambassadors. DEM. xviii. 244. Here the leading sentence is particular, on no single occasion was I wo?'Sted, so that E7TEfl</;BfJv is regular; if tbe ncal'ly equivalent universal affirmative on every occasion I pToved supeTior had been intended, we should have l1a<l 7TEflcpBdryv. See xviii. 45, 7rpovAEyoJ Ka.i OLEfLapTvpbfJ-YJV Kat 1rap' VfLtV dEt gat o1rot 7TEfL<j;Be,Jv; and the following in 244, iv ois KPD-T'}BdEv ol 7rpcrf3Hs avTov T0 A6y<tJ, TavTa Tots o7rAots E7rt~JV KaTEcrTpE<j;ETO. Notice the imperfects in tl1e two affirmative examples, and the aorist in the preceding negative example.

537. 1. The indicative is generally useu in Greek (as in Latin) in parenthetical relative clauses, like o Tl 7TOT, rnv, whatever it is (quidquid est), ocrns m)/ ECTT[v (or E(TTat), etc. E.g. Zevs, ocrns 7TOT' E(TT t V) El TbO' a-f,T0 <j;[Aov KEKA1JfLEV<tJ, TOVTb VLV 7rpocrevv~7Tw, Zens, whoever he may be, etc. AESCH. Ag. 160. 6otAnlofLEV Owi<;, 0 Tl 7TOT' d(J"tV ew. EuR. Or. 418. 'Hfll'J' ye KPE(T(TOJo . . . 3ov"Ary1]v V7TOfLEZvat ?)ns ECTTa t, b1d it is bette1 jo1 us to submit to slavery, whatever it may be. HDT. vi. 12. Soon 8~ KOTE Ea-Tt, vii. 16. 2. But ocrns in such expressions can have the construction of an ordinary conditional -relative, so that in future and general conditions it may take the subjunctive. E.g. . 'AA. A' 0 7rpocralj;Jpevos a.VTWV, OCTTlS av if, A.oyov 7rapxH, but each one uho has to do with thern, whoever he may be, gives his own account of them. AESCHIN. i. 127. 'AA.>.: v<j) VflWV EOH KEXHPOTOV1JfLEJ!OJI ELva.t TovTov, OU'TL<; &v J, but this officer ought always to be elected by you, whoever he may be. DEM. iv. 27. See THEOG. 964.

Horneric and othe1 Poetic Peenlia1ities in Conditional Relative Sentences.





538. In general conditions which take the subjunctive, Homer commonly uses the relatives without KE or tfv. This corresponds to his preference for the simple el in general conditions ( 468); but relative clauses of this class are much more frequent with him than the clauses with El. E.g.





"OTTt p.a..\' ov 01Jvat'Os os &8av&roun p.d.xrrrat. Il. v. 407. 'Av8pcfnrov<; l<f>opi, Kat rvvTat O<; Tt<; ap..<fpTTJ Od. xiii. 214.. Zd.><; ~' ' ' ' "' R ' ' ' AOt<; '' " <J <I.VTO> l'f1-Et OlltJOV '0' ' ll1!f1-7rtO<; av {) pW7rOUrtV, ECT {)' " 1)U KO.KOLCTLV, 07rWS UJE>-.. ncr t V' eKaCTTlp. Od. vi. 188. Ov p.iJv (TO{ TrOT t(TOV lx(j) ypas, 61r7ror' 'Axawt Tpwwv lK7rEp(Tw(T' <vvatop.<vov 7rToA[E8pov. I!. i. 163. So also Il. i. 554, iii. l 09, xiv. 81; 0<1. viii. 546, xviii. 1:34. Here the meaning is essentially the same as when KE or llv is adde<l, as in the examples under 532. The greater development of the general relative condition in Homer, especially in the use of the optative, compared with the less developed general condition with El, has already been uoticel (1 7; 400; 468).

r~p.a(T()' os ne; llpt(TTOS dv,jp Krtl 7rAt(TTa 7rOP'[I(TtJI, (tell her) to marry whoever may be the /,est man and may offer the most. Ocl. xx. 335. But in vs. 342, refening to the same thing, we have y~p.a(T8' 0 K' Wf.A.1J, to 'I'IJ;a?'?'Y whom she may please. ITd8eo o' ws . . . EV </>p(TL w. Il. xvi. 83; so Od. vi. 189. Ov p.;)v yap 1ror </>'f}(Tt KaKov 7rd(Te(T8a~ 07rt(T(TW, o<f>p' dpeTfJV Tra.pexW(Tt 8wt Kat yovvar' f>ptiJp'[l, he says he shall never su;{fer evil hereafter, so long as the Gods >hall supply valoUJ, etc. Od. xviii. 132. SoIl. xiii. 234.

53;). The relative (like El) is sometimes found in Homer without KE or av in future conditions. E.g.

e. [

540. "Av may GOmetimes be omitted in relative conditions with the subjunctive in lyric, elegiac, and Jramatic poetry, as in HoD;, chiefly,in general conuitions. A few examples occur in Herodotus; and even in Attic prose exceptional cases are occasiopally found in the manuscripts. (See 469-471.) E.g. Mya. :.ot KAeos a.iei, ~nvt (T'ovypas t(TTr'f}T' dy..\aov, great alway8 is his glory, whom thy illustrious honour (Olympia) follows. PrND. 01. viii. 10. So 0!. iii. 11, N cm. ix. 44. ITavras l7ra.v1wt Kat <jnAew EK<iJv O(J'TtS ~p01J fl-l)OEV a1crxpov. SD10:-l. V. 20 (but Ss av p.r) KaKO<; ii in the same ode). , See TYR1'. xii. 34 ; SoL. xiii. 9 and 55, xxvii. 3 ;
Snwx. lviii. 5, lxxxv. 7 (o<f>pa ~X'[!, but orav fJ in vs. 10). repovra o' op8ovv <f>Aavpov, os JIEOS 7rEa''[l. SoPH. 0. C. 395. Twv o 7f''f}fJ-OVWV p.d.A.una Av7rovo-' ai' cpavw(T' av6aperot. Icl. 0. T. 1231. SoAESCH. Sept. 257, Eum. 211, 661, and probably 618 (o p.1j K<AEW'[I, for :Mss. KEAEV(TEL, after <t11"ov denoting a habit). Tot'<n yap p.'qrE 'i' '<' )/ a(TT0. J1.1)7'E TE<XEU IJ EKT~<J'fi-EVa, . , KWS OVK av EL>]CTO.V OVTO ap.axotj HDT. iv. 46. So i. 216, ii. 85, iv. 66. 'E1rtxwpwv 13v ' ov p.v {3paxds dpKW(Tt p.fJ 7roAAoZs XPYi(T8at, it being our national habit not to use many wmds where few suffice. THuc. iv. 17. (Here p.v 7roAAo~s make five feet of an iambic trimeter, and the words are probably quoted from some poet. See Classen's note. Tl1e sentence continues, 7rAdo(T~ o ~JI 0 &v Katpo> V> K.T.A.) See also PLAT. Leg. 737 B, oi> Vand OCTOL'> fl-('1"0, In SorH. El. 225, orj>pa lx:o is particular. 541. In the lyric and elegiac poets, as in Homer, the form with or K~ was in good use in these sentences. See PIND. Py. i. 100 (os
3/ ' /






av av




tyKvprro), v. 65 (oTs &v UJf.A.u); MrMx. ii, 9, iii. (i1r~v 7ro.pap,E[lf'Ero.t); 75 ; THEOGN. 405, 406 (8. p,~v i'i KaKd, .. & 8' llv fi (For ordinary protasis see 469 and 470.) In the dmmatists the relative with l1v is completely established with the subjunctive as the regular form (like M.v, etc.) in both general and particular conditions.
SoL. xiii. XP~IYLJ.w).

(See 4 71.)





542. In Homer the conditional relative (like El) sometimes takes Ki Or QV With the Optative, the particle apparently UOt affecting the sense. E.g. 'H oe K' t1TetTa yq{'at8' os KEV 1rA.d:O"'Ta 1ropot Kai p.6prrtfOS A Bot,
and she then would marry whoever might give the most gifts, etc. Od. xxi. 161. ''its KE 13olYJ ~ K' J8A.ot, that he might give hm to whomsoever he pleased. Od. ii. 54. In these two cases lls- 1r6pot and ~ JBf.Aot would be the common expressions. In Od. iv. 600, however, owpov 8' O'T'Tt KE p.ot 00 LYJS', KHJJ-'IJAWV EfTTW, whatever g?jt you might choose to give me, etc., may be potential. Nvv yap x' ''EKrop' i!A.ots, E7rd av p.aAa 'TOL (jxf.l3ov UOot. Il. ix. 304. "Os 'Td KaTaf3p6gELEV E7r~ V KPYJT~PL fLYELYJ, o-15 KEV i<f>YJp.Eptos YE f3aAOL KO.'Td OrtKpv 7raptWv, whoever sho1tld drink this when it was mingled in the bowl, uould let no tear fall down his cheeks on that day. Od. iv. 222. So f.m]v . . . d1JV> Il. xxiv. 227. One ca.~e occurs of OTE Ke with the optative in a general relative sentence of past time : E7r1J0op,E8a . . . OH KEV im{a<j>eAos XOAO> 1Kot, Il. ix. 525.





543. In Homer similes and co!llparisons may be expressed by the subjunctive with <ts oTe (rarely <ils otron), as when, sometimes
TE, as. Except in a few cases of ti!s o-! by ti!s or nor Ke is found i11 these expressions.


av, neither av

544. With <ils ore or <ils D7r07e the subjunctive clearly expresses a general condition, and the meaning is as ha:ppens when, etc. E.g.

oTE KLV~fT'(J Ze<J>vpos f3afFv A.r)tov i."A&wv, \ 'f3 If ' ~ I ' 1 Aa pos etraL')'L~wv, e1rt/ T '7fhVEL afTTaxvefT<TLl', .:Is TWV trafT' ayop~ KwfJBYJ, and as (happens) when the west wind comes and move. l~ deep grain field, and it bows with its ea1'S, so was their whole assembly ?noved. Il. ii. 14 7. 'its 8' or 01rWPLVOS BopEYJS <f>opE7JfTLV aKaJJfias ilf 1rEOLov, 1TVKLVat 0~ 7rpos aAA?]Arww EX 0 vra L, J}, 'T~V ap. 7rEAayos avep.ot <J>f.pov ev&a Kat EJ!ea. 'Od. v. 328. See IJ. v. 597, vi. 506, viii. 338; Od. ix. 391, xix. 518; for ws (mon, Od. iv. 335, xvii. 126.
'its 8'

210 'lls




a-r~B<a-a-tv dv<a-r<vaxtt' 'Ayaf'-EJI-vwv.

Bi &v &a--r pli7rTTJ 7r01TtS "Hpl]S Il. x. 5.




So Il. xi. 269, xv.l70;

Od. v. 394, xxii. 468.

545. With ws or ws n the conditional force of the subjunctive is not so obvious, especially as it depends directly on the verb of the antecedent clause, which is always particular and generally past. Here we should expect the present indicative, whicl1 sometimes occurs (548). We may suppose that the analogy of the far more frequent (544) 1 caused the same construction to be used clauses with ws also in these, in which the meaning is clearly the same. E.g. 'g, o yvv~ KAa [TI(T t <f>lA.ov 7r01Ttl' dfL<f>t1r1TOV(Ia, os -re ~~s 1rporr8<v 1r6A.w<; A.awv n 1rEr:TYJITtv, ws 'Oour:reDs A.mvbv {,7f r3<f>pvr:rt oaKpvoJJ df3eJJ, i.e. Ulysses wept as a wife weeps, etc. OJ. viii. 523. tn <;:'<\ \ 1 f3 OVr:Tt opWJJ O.VXEVO. O.TJ 7ropnoc; 7)E f3' 'e \ :!.c_ ) r )/f. I )\ oS OE 11.WV EV OOS, ws -roils dJL<f>orpovs J~ t1r1rwv T voeos v16s f3~r:r<, and as a lion leaps (tmong the cattle anrl breaks the neclc of a heifer or wn ox, so did the son of 'l1Jd eus dismount them both frorn thei1 chariot. Il. v. 16 L So Il. ix. 3 2 3, X. 183, 485; Od. v. 368.


546. In all the cases of 0s n the pronominal article oi or -rovs precedes, referring to the subject Ol' object of tl1e antecedent clause. E.g. Oi o', WS r' dJL'lJT~PE'> dA.A.~A.otr:rw (}YJLOI' EAO.VJ)(J)(Ttl', (},,. Tpwe>; Kat' Axawi f.-i dAA'qAotr:rt 8op6vre;; o1)ovv, and they,-as reapers each othe1 drive theiT swaths,-so did T1ojans and Achaeans leap upon each other and destroy. Il. xi. 67. So Il. xii. 167, xv. 323; Od. xxii. 30~.

547. When a simile has been introduced by the subjunctive with ws or ws ore, it may be continued by verbs in the present indicative, which seem to be independent of the original construction. Even the aorist indicative may be used to add vividness to the description. E.g. 'lls B' aTE r> 7 f.Ae</>aJJra. yvv') <f>oJJtKt {Lt~v17 MTiovlc; ,j( KaHpa., 1rap~wJJ EJLJLEVat Zmrtp
Ke'i:rat TOto v 8aA.ap.~l, 7r0AE',; TE JI-W ~ p~(Ta wro h1r~es <f>op~ew (Ja.r:rtA~t o Ke'i:rat ll.yaAJLa.'


n, .S' or' &.<fl ilfYJA~> Kopv<f>qs opws {1-<yaA.ow

KtV~r:TYJ 1rVKtli~JJ ve<f>~A'lJV f1'TEf01rlJY<p~ra.

MEVDca.e, fLMiv8'Yjv a.'tJLaTt JL'Yjpo. I1. iv. 141.


EK r' :! cp a.,, e v 1ra1Ta.t ITKo7rta.2 Kat 7rprflopes ll.Kpot Kat Frf7rw 01lnavo8eJJ o' iJ1reppay'l) ar:T7rETOS a.i81jp &s L}al'O'Ol vt1WJJ fLEV d1rw1Ta{LeVOt oijtal' 1rvp TVT8bl' d.vi7rvevr:rav 1/'0AEfLOV 8' ov yfyv<r' epwf]. I!. xvi. 296.
1 Delbriick, G01ij. u. Opt. pp. 161, 162, cites 63 cases of this construction {49 in the Iliad, 14 in the Odyssey), of which 35 have ws IYr, 10 ws ai i!.v, 3 Ws 01r6-re, 8 Ws, and 7 Cbs 'Tf,




'ils 8' bTE Ka:rrv6s lti>v els o-&pavov ei>p:Vv tK'f}Tat aunos ai&oJLEVOto, 8EWV OE J JL0VtS av~KEV, traut o' 1Ke mivov, tro>..>..orm 8~ Kf/M <f>~Kev, ws 'AxtA.evs Tpwnnn trovov Kat w!JU ;e1JKEJ!. Il. xxi. 522.


548. Sometimes the first clause of the simile has the present or aorist indicative. E.g. ~> > I (./ 0 ' r" " I H> o aJ!ajLatjLaft fJa E ayKea eutrwaes trup, ws o ye traVT'YJ Bvve. Il. XX. 490. ns 8' otr6TE 7rA1}Bwv 1rOTafLOS tr0ov0E Ka'TEtUtv, 1rOAAOS 3~ opv> JuqJepeTat, i),s f<{lEtrEJI. IJ. xi. 492. '1 HpttrE c' liJs b'TE n> opvs ij pmev, and he fell as when ttn oak fttlls (once fell). 11. xiii. 389. 'ns 3' b'TE 'T{s 'TE 3paKOV'TU lowv traA.lvopcro<; atreCTT1). Il. iii. 33 : so ws n A.wv Jxti.p1J, iii. 23. 549. Another form of Homeric simile consists of ws with a noun, followed by a relative with the subjunctive, which may be followed by an indicative as in 547. E.g. '0 o' El' KOJ!L2)CTL xafLat 'll"EU"EV, ai'yetpos lils, 1) pd. -r' Jv EtaJLEVfj i! JLEycfA.oto trE rpvKTJ A.e1], &rcfp -r o1 B{ot J-ff &Kpo-rcf-r17 1rerpvacrtv T~JI p.f.v (J' app.aT01r1JYO> dv~p aWwvt crt3~p<p thaJ.L', Brppa l-rvv KUJ.Llf17 trEptKaA.A.t O<j>p<p' 1J J.LEV T d{OJ.LEVYJ KElTaL 1rOTaJ.LOlO 1rap oxBa.s ~ " ~. 'c t 'TOtOV ap"AV(} EfLW'f}V ,, ~LfLOEtUtOV Et;EVaptc;EV AZac; Otoyev~<;. Il. iv. 482. For w> el or ws d TE with the optative in Homeric similes, see 485.



"0 n


and 81Tov p,?] without a Verb.

550. "0 n ,.~ and oa-ov fL~, like el 0~ (476), are used in the sense of except, unless, with no verb expressed. E.g. ~o TL yap JL;1 'AB~vat, 1JV OVO~JI aA.A.o 1rOALCTfLO A.oytfLOV, for except Athens (what was not Athens) there was no (Ionic) city of ttny ttccount. HDT. i. 143. So i. 18, ovOaJ.LO' (', n JL~) X;;ot JLOVVOl. Ov ylip ,')v Kp-!Jv~), on 0~ J.L[a iv a-&-riJ TU dKpo11"6Aet, for them was no S]JTing, except one on the veTy citttdel. TRue. iv. 26 : so iv. 94, vii. 42. 0{)-r' Etrt 0wpav it0A.Oes b 'Tl ,~) atrat d<; 'IcrB,.<w, OVT aAAocrE OVOO.Jl-6()" et. Jl-'1 1ra. <npaTevu6,.evos. PLAT. Crit. 52 B. So Phaed. 67 A, Rep.
"lcre, ylip OOKWV J,.ot Ka' ~VJ.L</>VTEVCTat Tovpyat' elpyacrea, 8', bUO'JI J.L~J xepa-~ Kavwv, i.e. and to have done the deed too, except so far as you did not slay with ymu own hands. SOPR. 0. T. 346. 551. Homer once has o n fL~ or o-rE 0~ in the same sense : oiJ T~ Te<p cr1revSecrK< 8ewv {) TL f-1'~ tra-rpi, i.e. except to Zeus (o n Jk~ = El JL~), Il. xvi. 2 2 7. Here Lange (p. 161) reads JL~






Special .Fonns of Antecedent Clause.

552. A conditional relative clause (like a protasis with El) may depend on an infinitive or participle (with or without ~v), on a final clause, on a protasis, or on a verbal noun representing the antecedent clause (or apodosis ). E.g.
See D1<:M. xxi. 64 (quoted in 525); PLAT. Ap. 17 D, DEM. xxiii. 48 (quoted in 528); AESCH. Ag. 1434, DEM. iv. 21 and 39, xxviii. 21 (quoted in 529) ; PLAT. Euthyd. 302 A, Theaet. 155 A, XEN. Mem. ii. 3, 12, Cyr. i. 6, 3, ii. 1, 31 (quoted in 531). 'Opw (TOL -roDTwv oefj(To V OT<J.V E1rL81'JL~(jr;s <fnArav 7rpos nvas 7fOtfL(T8at. XEN. M em. ii. 6, 29. Kai Jp. 8Et: &m1 >..>..axBat KaTii .. <rvvB~Ka>, 7roo~ ..0 1!'epi Tov ITpwTayopov Myov TEAo<; (TXO[YJ, i.e. I ought to be released accotding to what we agned to do when the discussion of the doctrine of Protagoras shou.ld come to an end. PLAT. Theaet. 183 C. '


553. After past verbs of w,aiting or expecting in Homer d'Tl'OTE with the optative sometimes has the meaning of until, like ws. E.g. Oi 8' laT' . . . 7ron8~yJLEVOt crtr7roT' &p' f.A.Bot 'I8a2os, and they sat waiting until (for the time when) Idaeus should corne. Il. vii. 414. So iv. 334, ix. 191, xviii. 524. (See 698.)

.klixed Conditional Constructions.

554. The relative with the ~ptative sometimes depends on a present or future tense. This occurs chiefly in Homer, and arises from the slight distinction between the subjunctive and optative in such sentences. E.g. Alrrv oi JovEcTat v~as vt7rp'lj(Tat, 8n JL?J aVTos YE Kpovwv JJLf3aAot
af()6jLVOJI oaAu!' l11JE<T(Tt, it 1JJill be a M?'d task for him to ji1e the ships, ?tnless the son of K1o1ws should himself hurl a flaming bmnd upon the ships. Il. xiii. 317. (I{egularly on KE JL0 ~JLf3aA.17, unless he shall huTl.) So Od. xix. 510. Ka~ S' (}),A.v VEJLHTW 1} ns Totav-ra YE p~Cot, and I am a.ngry with any other woman who says (should say) the like. Od. vi. 286. (This resembles the loosely jointed examples in 500.)

TowvT\u oE: iiotKas, E'Tl'EL Aova-atTo cpayot


<vop.<vat JLaAaKws,

and you. seem like such a man as would sleep cornjmtably (like one likely to sleep corrifmtably) aftm he had washed and eaten. Od. xxiv. 254. (This resembles the examples in 555.) The optative regularly follows an optative in a wish (177).

555. In Attic Greek an optative in the relative clause sometimes depends on a verb of obligation, propriety, possibility, etc., with an infinitive, the two forming an expression nearly equivalent




to an optative with :lv, which would be expected in their place. (See 502.) E.g. 'AAA: ov 'll'OAIS (J"T~O'HE, rov8e xrn] KAvew, we should obey any one
whom the state might apJJoint (if the state shonld appoint any one, we onght to obey him). SoP H. Ant. 666. (Xp~ KA-6nv is followed by the optative from its nearness to 8tKa.[ws &1:' KAvot ns.) 'AAAa rov p..ev avTov A-yELv f'J <Ta.<j>Cis d8e,7 <j>e[8e<T&<u 8ei:, i.e. 1Ve ought to abstain, etc. ; like <j>e[8otTO TLS. XEN. Cyr. i. 6, 19. Oils o 'll'Ot'q<TaO'()a{ TLS (3 o (; Ao tro <Tvvepyovs 7rpo&0p.o?JS, ToioTov> '1l'U.J1Ta7raaw 'fp..otye ooKE'i d.y<t&oi:s &Y)paTEOV dvcu (&1Jparov Elvat= &J]pav oEZv). lb. ii. 4, 10. 'Y7repopav ov 8vvaTOV vp.Civ 0.v8pt os etoe[J] Kvp[ovs ~vTaS o Tt (Jo,)Aw-(JE avr<iJ XPTf<T6at. Id. Hell. vii. 3, 7. So Ib. iii. 4, 18. '2wfp6vwv EO'T' fl-')OE el p..tKpa Ta 8w<j>f:povra f f'l) 'll'OAep..ov avatpitcreat. lb. vi. 3, 5. So after 7roAv pfj6v (f.a-rt), lb. vi. 5; 52. '2wfp6vwv EO'TLV, el fh'YJ aOtKOLJITO, ~O'vxa(nv, i.e. it is ]Jroper for prudent men, etc. THUG. i. 120. 'A7ro8oTEOJI o-&8' 67rwa-novv TOn, 6JToTE ns fL'l <Twcpp6Fws a1latTOL. PLAT. Rep. 332 A.


556. An indicative or subjunctive in the relative clause may depend on a potential optative (with ~v ), sometimes when the potential force is felt in the apodosis, and sometimes when tho optative with &v is treated as a primary tense from its nearness to the future indicative. E.g. OVKoVv Ka2 Tb {ryut[ve~v Ko.i TO vocreZv, Oral' &:ya8oV T~vos- air~a
-ytyJI'I)TU.t, dya.8u, av E?-IJ, thmejO?e, both health and disease, when they prove to be the causes of any good, wonld natumlly be goad things. XEN. Mem. iv. 2, 32; so ii. 2, 3. "OTaJI 8 TLS 8EWV (3 A U'll'TTJ, 8uva.tT' aJI ovo' &JI i.<Txvwv cpvy<i:JI, when one of the Gods does rnischiej, not even a st?ong man could escape. SoPH. EL 696. "fl<TT a'll'o<j>vyot<; av ~VTL!I' av (3 o (;A r; OfKYJV, so that you can (could) get off in any S'<Lit you please, AR. NuL. 1151. o;:TLJIES roi:s f'-oEJI LO'OLS fh'J dKoVa-t, roi:s 8 Kpe[O'O'OU'I. IW.ACic; 'll'(lOO'q)epovTat, 7rpds 8 Tot>s ~<.TO'OVS p..hpw[ d<Tt, 'll'AEL<TT' av op0o2VTO. THUC. Y. 111. "() 8 f'):Sv KU.KOV 7!'0 i z, ovS' &.}I 'TLVOS el'ry KaKov ai'-nol'; and what does no harm could not be the canse of any harm at all, conld it? PLAT. Rep. 379 B. 'Eyw o TaDTY)J! f'-EV 'T~V elpqvqv, EWS av El> , A(Jl]Vatwv Ad'll''I)'Tat, OVOE'll'OT av O'llfhfJovAEVO'U.Lfhi 'll'Ot+ <Ta<r8at TlJ m5AEt, I wonld never advise the city to make this peace, so long as a single Athenian shall be left. DEM. xix. 14. (Here i!ws AEi'll'otro, so long as one shou.ld be left, would be more regular.) "0TaJI 8' acpaF[<Tas -r~~ TdKpl.{3'fs A6f'<t> f.a7raTfLv 1rttpCirat, 1rW~ &v 3tK.alwr; 7Tt<rre-6otTO ). Id. xxxiii. 36. (See 178.)

557. A conditional relative clause may contain a potential optative or indicative (with dv), which has its proper meaning. E.g.
~Eg Wv iiv Tl.S' 15 AEf'Wll 8ta(3d.AAo~J, eK To-Drwv aVTo.Os 7uda-U"Bat (ifcprJ), he said that they wonld form their opinion upon any slanders which any good speaker might chance to utte1-. THUO. vii. 48. "Ovnv' av vp..dc;



Jl .


, '\ 'f / 1' .... , 'i' " TC1V1'T)I! Tl)il Ta,;LV KaTHTT1)CTC1TE, OVTOS TWV !CTWV atTto'> 1)11 av KaKu)v oCTwv1rep Kat oVTo>, any one soeve1 whom yot! rnight have appointed to this post wo tdd have been the cause of aB great calarnities as this rnan has ueen. DE,1. xix. 29. (Without av, SvTLJIC1 KC17'1TT~UC17'E would be equintlent to Ef TLJ!Ct flAAOV KC1T1TT0CTaT1 if YOU had appointed any One else (which you did not do). With av, it is a potential indicative.) See 506, and for the optative with KE in conditional relative sentence~ in Homer (probably not potential), see 542.


Assim,ilation in Conditional Relative Clauses.

558. When a conditional relative clause referring to the future depends on a subjunctive or optative referring to the future, it regularly takes by assimilation the same mood with its Jeadiijg verb. The leading verb may be in a protasis or apodosis, in another conditional relative clause, in an expression of a wish, or in a final clause. E.g. 'Eav nve<; oi' &v OVJIWVTa rovro 7T'OWCT, KaAws iifEL, if any who shall be able do this, it will be well. Et TWES ol' 8vvatvTo TOVTO 'l!'owiev, &v l!xot, if any who should be able should do this, it would be well. EZBe r.aJITes oi' 8v vat vTo TovTo 1rowZev, 0 that all who rnay be able would do this. (Here the 1)rinciple of assimilation makes oi ovvatvTo after an optative preferable to ol' llJI BvvwvTat, which would exprfiSS the same idea.) T eBJia[rJV Sn pm fH]KET TaVTa fL A o t, may I die when these n1e no longe1 rny delight. MrMN. i. 2. So in Latin : Si absurde ~anat is qui se haberi velit musicum, turpior sit.-Sic iujurias fortunae quas ferre nequeas defugiendo relinquas. For examples see 529 and 531. 559. When a conditional relative clause depends on a past tense of the indicative implying the non-fulfilment of a condition, it regularly takes a past tense of the indicative by assimilation. The leading verb may be in a. protasis or apodosis, in another conditional relative clause, in an expression of a wish, or in a final clause. E.g. EZ TLJIES oi' UhhavTo To1ho ~7rpa~av, Ko.Aws llv eCTxev, if any who had been able had done this, it would have been well. EZBe 7ravn<; ol' ~8vvavTo ToVTo 7rpa~aJ', 0 that all who had been able had done thio. So in Latin : Nam si solos eos diceres miseros quibus moriendum esset, neminem tu quidem eorum qui viverent exciperes. For examples see 528.
560. It will be seen tl1at this principle of assimilation accounts for the unreal indicative and the optative in conditional relative sentences, which have been already explained by the. analogy of the. forms of protasis. (See 528 and 531.) In fact, wherever this assimilation occurs, the relative clause stands as a protasis to its antecedent clause.




Occasionally this principle is disregarded, so that a subjunctive depends on an optative (1 7 8). For the influence of assimilation in determining the mood of a dependent sentence, see 176.

561. The indicative in the construction of 525, referring simply to the present or past, cannot be affected by assimilation, as this would change its time. E.g. 'Yp.-ets S' VwurfJe 0 'TG Ka~ 'l"ii 7r6AH Kat d:rra(rt O"Vvo{o-Hv vp.rv p. A A et, and ~nay you choose what is likely to benefit the state and all
of you. DEJvi. iii. 36. Compare this with DEll!. ix. 76, 3 n 8' {;1-'l.v 136~eH (so L: originally), TovT', iJJ 1rd.vTes Owi, o-vv<veyKo~, whatever you rnay decide, rnay this be fm our good. In SOPH. Ant. 37 3, os ,.&.8' epOH would belong here ; but 8s ,.6,8' i!pl3ot (Laur.), =d n<; Tall' i!p8ot, falls under 558.

562. The principle of 558 and 559 applies only to conditional relative clauses. If the relative refers to a definite antecedent, there can be no assimilation, and the indicative or any other construction required by the sense is used. E.g. El TWV 1l"OA!.TWV oTo-~ vvv 7r to-T do p.<v, TOv'To~s ri'll"teTT~<<v, ol<; o' OV xpwp. TOV'TOtO"~ XPTJO"Q{j-terJ'fJ', rU'WS O"WfJ<'ip.EV &.v. AR. Ran. 1446. EW' 1)(]'fJa. ovva.Tos Spav d<Tov 1rp68vp.os <l, 0 that thou couldst do as much as thou art eager to do. EuR. Her. 731. (With ?JU'Oa for <i the meaning would be as much as thou wert (or mightest be) eager to do.)


563. Conditional relative clauses depending on a subjunctive or optative in a general supposition (462; 532) are generally assimilated to the subjunctive or optative; but sometimes they take the indicative (534). E.g. Ovo', e1rnoO.v wv &v ;.p[YJ.,.at Kvpws yvTJTat, ,.0 1rpo06TT1 o-vp.f3ovA.<p '7l'ep~ 'TWV AOt1l"lOV gn xpryrat. DEM. xviii. 4 7. See PLAT. Rep. 508 C and D (reading illv o 1) KaTaAaj-t7rEt); Charm. 164 B. '0 8~ r6T j-tUA~o-Ta i!xa.tpev, orr6n Taxwra Tvx6vTas wv OEOtVTO
a1l"01l"EJ-'1t'Ot. XEN. Ag. ix. 2. Alr[a f'EV yap o-nv, omv ns l{tA.{ XPYJ<rd.p.evos A6y<p p.i'J 7rap6.<TXYJTO.t 1r<rnv wv A.y, eAyxM 8, 3Ta.v 0v &v ei'1r77 ns Ka.~ nlA.YJOf.s 6j-tov OE [~YI DEM. xxii. 22. (Here 0v A.y and tilv &v d1r'[/ are nearly equivalent.) 'EKaAEt of. Kat ETLj-ta 071'61' nvd.s root TOWVTOV 1l"Ot~<ravTas 3 1ravras f3ovAE'To 1l"O~EL'v. XEN. Cyr. ii 1, 30. (Here f3ovA.ot'To for f3ovAe'To would correspond to 8owTo in Ag. i:x. 2, above.)

in the Antecedent Clause.

564. The conjunction Be sometimes introduces the clause on which a relative depends. Its force here is the same as in apodosis (512). E.g. 0 ?YJ 1rep cpvAA.wv YEVE~, To{YJ of. Kat &vopwv. Il. vi. 146. 'Erret




n 7rOA<p..o> KaTECTTYJ, 8 cpatv<Tat Kat ev T01JT<p 1rpoyvoi>s T0v Mwap..w; and when the war broke out, (then) he appears, etc. THuc. ii. 65. / ' 'l' c/ 'l' ' ' 1\1 EXPt p..<v ovv ot' TO,oTat HXOV TE Ta (3' \ avTO!<; Kat owt TE ?)CTav 1\.Yj XP~cr8<u, o1 8 dvTEtxov, so long a8 thei1 aTche1s both had thei1 a?Tows and weTe able to ttse them, they held out. Id. iii. 98. 'E7rHD?J o acptKOfJ-E!IQt p..axu eKpanwav . . cpa[vovTat 8 ollo' evTav8a 7/'acru Tfj XPYJCTUfJ-EVOt. Id. i. ll. "flcr7r<p oi ,hAtTat, OVT(d o Kat oi 7rEATaCTTa[. XEN. Oyr. viii. 5, 12.
0 ' ' "'


In Attic Greek a relative with the future indicative often expresses a purpose, Iik~ a final clause. It~ negative is fM). E.g. ITp<cr(3dav o 7I'EfJ-7rHV, i)ns TavT' ep< 'i. Kat 7I'apf.crTa t TOtS 1rpa yp..acrtv, and to send an embassy to say these things, and to be pTesent at the tmnsaction. DEM. i. 2. PYJp..t 01) OEiv ?Jf'J1s 7I'pos 8<TmAovs 7rp<crf3dav 7I'Efh7rEtV, ?J TOVS fhEV &06.~1 TO.VTa, TOVS o 7!'apo~VVL. Id. ii. 11. "Eoo~E TcjJ O~Jh<p TptaKOVTa avopas f.Acr8o.t, oi' T01',S 7rUTp[ovs Pop..ous ~vyypatfov<n, Ka8' oils 7roAtT1Jcrovcrt, the people voted to choose thi1ty men, to compile the ancestml laws by 1vhich they were to govern. XEN. Hell. ii. :3, 2. EZcrw OE 7rEp..tfat (EKEAE11<T) nvas, OtT!VES aimj! Ta v8ov 18ovns d7rayyAovcrtJI. XEN. Oyr. v. 2, 3. ~avnKov 7rap<ctK<va(ov o n 1rf. fh tj;o vert 11 es T?JV', Kat vo.1!apxov 7rpocrf.Taav 'A,\K[oav, 8s EJ-h<AA<v em7rA<vcrEcr8ac Tauc. iii. 16. See DEjf. xxi: 109. Ov yap ern p..ot XP0P..aTa, o1r68v eKT[crw, joT I have no nwney to pay the fine with. FLAT. Ap. 37 0. 'Pl:tj;6v f'-< yrj, EK T{jcr8<, o1rov 8vYJTWv cp a 11 o {j fh <H f'-')O<vos 7rpocr'Jyopos. SoPH. 0. T. 1437 ; so 1412. Mf.AA.ov<Tt yap et' eJ!mvea r.ifhfnv, :!J!ea fh'J 1ro8' ~Atov cpf.yyos 1rpocr6tj;o, (wcra 8' VfhJI~cros JWt<a, they are to send yo1t wheTe you shall neveT behold the sun's light (to some place, that there you may 1wve1 behold, etc.). Id. :El. 379. So Aj. 659; Tr. 800. 566. The antecedent of the relative in this construction may be either definite or indefinite ; but the negative is always 1'-~ hecause of the final force. ']'he future indicative is regularly retained after past tenses, as in object clauses with o1rws (340) ; but see iJ7:3 aud 57 4. 567. A past purpose may be expressed JJy the imperfect of f'-EAAw. See 76; and Tauo. iii. 16, quoted in 565. 568. (Subjunctive and Optative in HorneT.) In Homer these final relative clauses have the subjunctive (generally with K) after primary tenses, and the present or aorist optative (without K~) after secondary tenses. Eg.;1 ~Y<fh6Jl JcreAilv &1ra<T<Tov, os K fhE KEur dyJ,yu, and also send a good guide, who shall lead me thithe1 (to lead me thither). Od. xv. 310.

565. (Ftd'lt1'e Indicative.)




Avro<; J!VJ! ovo/ e{ipw, 0 'Tt KE Ofjat 7rat8o<; 7rat8l q,fA<p, find a name to give the child. 0<1. xix. 403. Teov oi'wofLa el1r~, Zva TOL 8w gevwv. <{i KE <TV xafplJ'> Orl. ix. 355. AvTfKa fLrlV'Tt<; EAEl~fTE'Tat, o<; KEJ! 'TOt d7rlJ<TLJ! o8r!v. Od. X. 538. ''EAKO<; 8' r~Jn}p E7rLfLrl<T<TETaL, 1)8' E7rt01)<Tet q,ripfLax', & Kev 1ravo-no-t fLEAruvawv 68w6wv. Il. iv. 191. 'AAA ayeTE, KA7JTOV<; 6TpllVOfLEV, oZ KE nixLo-Ta ~A0w<T' E<; KALo-[~]V TilJA:rwio<w 'AxtA~o>. Il. ix. 165. ''EKooTE, Kai np,1)v drronvEfLEV 1)v nv' EOtKev, 1) 'TE Kat f.o-o-OfLEVOL<Tt fLET dvOpr/moLo-t 7rEA'f)Tat. Il. iii. 459 : this verse (al~o in iii. 287) and Od. xviii. 336 are probably the only cases of the subjunctive without KE in these sentences. "AyyeAov ~}Kav, o<; dyydAoE yvvatK[, they sent a messenger to tell the ~voman. Od. XY. 458. ITarrT>JVEV 8' dva rr1~pyov 'Ax<uw,, et nv' r8ot'TO 1JyEf10Jwv, rfs oi dp1]v hapot<TW Uf1VVat, Il. xii. 333. This optative is rare.


569. Tl1e earlier Greek here agrees with the Latin in using the subjunctive and optative, while the Attic adopts a new construction with the future indicative.

570. The futme indicative occurs in Od. xiv. 333, WfLOCTE vi}oc KaTEtpvo-Oat Kat f.rrapTEa<; Ef1f1EV f.m[pov<;, 84 7rE fL fo VfT t <j>[A'f)V e<; 7raTp[8a yaiav. The potential optative with Ke may take the place of a fnttJre form; as oi>8f. oi aAAot err/, oZ KEV KaTa 01JfLOV aAaAKOttV KaKoTJ)'Ta, Od. iv. 166. So TWV K' emf3a'f)v, Il. v. 192 (cf. xxii. 348). In noue of the Homeric examples of this comtruction is the relative clause uegative.


571. A final force is seen in a few Homeric temporal clauses with OTE (or' av, OTE KE) or 07r0TE with the subjunctive, which are chiefly , expressions of emphatic predidion ; ''E<T<TE'TO.t ?JfLUp or' &v 'TrOT 6A0Au "lAw<; ipry, Zevs OE o-<fotv aVTd> E7r t o-<Te [ lJ<T t v EfJEfLV1JV aly[oa 1rao-tv, a day shall come v:hen sac1ed Ilios shall fall (ie. a day for the fa.ll of Ilios) and when Ze?cS shall shaJce his tmiule aegis b'fore them all. ll iv. 1 G4 ; so vi. 4<18. See Il. viii. 373, xxi. 111. See IIIomo, Horn. G1. p. 209.
572. 1. In Attic Greek the subjunctive is not used in final relative sentences as it is in Homer (568). A few expressions like EXEL rt trrv, he has something to say, follow the analogy of ovK EXEt o rt drrv, he knows not what to say, which contains an indirect question (G77). E.q. TowvTov i!Oos 1rap8oo-av, wo-re EKa.Tipov<; i!xew l<jo' ois <j>tAortf1'78wo-LV, that both may have things in uhich they may glmy. Isoc. iv. 44. (Here there is really no indirect question, for the meaning is not that they may know in what they are to glory.) Q,',8v ETt ow{o-EL avr\o, Jav f1DI'Ol' ~Xl/ OT'f OtaAey'f)Tat, if only he shall ha~:e smne one to talk with. PLAT. Symp. 194 D. ToZs fLfAAovo-tv i!~etv el<T4>pwo-tv. XEN. Oec, vii. 20. Compare drropEt<; n Ayvs and n!rropeZs n A.yvs in the same sentence, PLAT. Ion. 536 B. 2. The subjunctive and optative may be used with a deliberative force, even when the relative has an antecedent, provided the leading clause expresses doubt or perplexity. E.g.





0?! yd.p &Uov oi8' OT<:? >..f.yw. SoPH. Ph. 938. ODK EXW u6cptup.' OT<[l 1f"fJf1-0vYJ'> a'/l"aAAayw. AESCH. Pr. 470. 0?!8/.va ixov OUTL'> E'/l"LUTOAas 7rEJLlfH. EuR. I. T. 588. So iKavovs ot<; ilw, XEN. An. i. 7, 7 (cf. 677). See Soru. Ph. 281. 573. The present or aorist optative occurs rarely in Attic with a final sense, where there is no deliberative force. E.g. Kpvifau' avr~v (vBa JL~ TL'> du{ilot, (Jpvxaro. SorH. Tr. 903. So oun<; Aci.Kot, AR. Ran. 97 .. See PLAT. Rep. 398 Band 578 E. For the constructions of 572 and 573 see Appendix VI (p. 411).

574. The future optative also occasionally o