LIBRARY ANNEX Cornell Mniversity Library BOUGHT WITH THE INCOME FROM THE SAGE ENDOWMENT FUND THE GIFT OF Henry W. Sage 1891 Virucaes Gayle NT ites bead BIN AND..... i | 9755-2 wii ARISTOPHANES AND OTHERS By THE SAME AUTHOR NOTES ON XENOPHON AND “OTHERS. 1907 [On Xenophon, especially the Opera Minora, Hercdots, Pausanias, Plu- tarch's Lives, the Erotici, &c.; also on Catullus, Propertius, Juvenal, and others.] ° Crown 8vo. Cloth. 6s. net. TWO LATIN TEXTS By A. E. Housman D. JUNII JUVENALIS SATURAE Demy 8vo. Paper boards. 45. 6d. net. M. MANILII ASTRONOMI- CON I Demy 8vo. Paper boards. 45. 6d. net. GRANT RICHARDS ARISTOPHANES AND OTHERS BY HERBERT RICHARDS, M.A. FELLOW AND TUTOR OF WADHAM COLLEGE, OXFORD Nage nal péuvac’ amioreiv.—EPICHARMUS, Those who in desperate cases conjecture with modesty have a right to be heard with indulgence.—GIBBON. LONDON GRANT RICHARDS 190 ae a A24SZA85 RicHaRp CLAY AND Sons, RiMITED, BREAD STREET HILL, H.C. AND BUNGAY, SUFFOLK. PREFACE LarcGE parts of this book have appeared from time to time in the pages of the Classical Review or Classical Quarterly. Many of the notes on passages in Aristo- phanes and the Comic Poets were printed there in 1899 -and 1901, some in 1907. The emendations of the Attic Orators will be found in the volumes 1903-1906, and those of the treatise On the Sublime, the critical writings of Dionysius, and Demetrius De Interpretatione between 1902 and 1906. Many of the Vuria appeared in the same Review in 1898, 1902, and 1907. The notes on elegiac and lyric poets were published in the Journal of Philology in 1897. I have to thank Mr. Nutt and Messrs. Macmillan for leave to reprint what was published thus in the first instance. It has, however, been carefully revised and has received considerable additions. Many suggestions on Aristophanes appear now for the first time. There is a good deal of other new material. I have added a study of the language of Aristophanes and the vi PREFACE Comic Fragments, made from one point of view only and with reference to only one question, but this (I think) an important and interesting one, the relation of their diction to that of serious poetry. It would be well if some one made a parallel study of it in its other aspect, noting minutely, that is, the reverse ways in which it falls below as well as rises above the level of ordinary serious prose language, seen perhaps best in such speeches as those of Lysias, Isaeus, and Hyperides. Besides some detached notes on a few other books or passages, I have put together here for the first time in the Silva Critica an unwieldy and, I fear, uninteresting mass of conjectural matter. The article on the use of tpaywdds and xwpwdds in Attic Greek appeared practically as it now stands in the Classical Review for 1900. Particular pains have been taken with the various parts of the Index, especially with that which directly concerns textual criticism. Perhaps I may be allowed to suggest, as I did in my former work, Notes on Aenophon and Others, that after reading an emendation in the body of the book readers might sometimes with advantage turn to the Index, as the same or very similar ideas are often applied to more passages than one, and such passages are of great value in illus- tration of one another. The causes, forms, and signs PREFACE vii of corruption recur in some cases with no small frequency and regularity. The book had already become so miscellaneous that I ventured to add to it something which is, I feel, a little incongruous, but may have some slight interest as a very small contribution to University history. At Oxford it has long been the custom for outgoing Senior Proctors to review the events of the University year in a Latin speech, and I had on one occasion —now distant—to conform to the custom. I am under great obligations to Messrs. Clay and to their reader for the trouble they have taken in the processes of printing and revising. I hope that this volume will be followed before long by a third, for which I have large materials, in part already published, and which will be concerned mainly with the text of Plato. HERBERT RICHARDS. Oxrorp, May 5, 1909. CONTENTS ARISTOPHANES . Tue Greex Comic FRAGMENTS ; Tue Diction or ARISTOPHANES . Tue Diction oF THE Comic FRAGMENTS Notes oN THE ATTIC ORATORS—- Demosthenes Antiphon Andocides . Lysias Isocrates Tsaeus Lycurgus Aeschines . Hyperides Dinarchus Rhetorica ad Alexandrum CriticaL Writines oF Dionystus oF HALICARNASSUS Tue TREATISE zep! tous DEMETRIUS epi Epunveias THE PHILOsTRATI WESTERMANN’s Biographi HIPPOCRATES mep! d&pxalns inrpicys PAGE 59 116 160 173 204 217 218 224 229 230 235 237 238 240 242 248 260 263 272 274 x CONTENTS HIPPOCRATES ep) dépwy H.7.A. PassaGEs FRoM GREEK Exreci1ac anp Lyric Ports Varia . Silva Critica ON THE USE OF THE WORDS rpaypdds AND rwppdds APPENDIX: Oratio Procuratorin . Inpex I. 5 Il. 3 LIL, IV. 2 PAGE 277 278 291 309 334 367 373 386 391 395 ARISTOPHANES AND OTHERS ARISTOPHANES ACHARNIANS 1 6oa 89 dédnypot TH euavrod Kapdiar, joOnv 8 Bad, révy St Bard, rérrapa, «.7.AX. The first lines of the play present difficulties that have not at present been cleared up. Why the emphasis of Thy éuavrot xapdiay instead of tiv kapSiay simply? Sedyypac must be passive, like é87y6nv in 18 and jo6nv in 2, though some one has tried to make it middle. Why again such poetical words as Baia and still more xéap, not to mention aduvyOnv (8, 9), the odd yarpnddvos (4), and éyavaOyv (7) One would suppose the whole passage must be a parody of something, and the reason for ryv éuavrod may have lain in that. Bad, révy 5& Bard has been suspected, but cf. Aesch. Ag. 1456 pia tas modAds, ras mdvy mwoAAds | yoxas éXécaca. Before noticing the parallel I had thought of puixpd, mavu dé Bad. Tapa 68 kat dir’ érpuxdpec ba. Suk a } Kaiorpiwy mediwy ddotrAavodvres. mapa mediwv being nonsense and 8 zediwv open to the objection that the plural ze&éa, used of one plain, is not suited to very ordinary comic iambics, we have to look for something different from either. wzapa K. zediov (Dindorf) hardly makes sense, for why should they skirt the plain? and rapa K. worapdy (Blaydes from Herod. 5. 100 ropevdpevor rapa worapov K.) is a rather bold change. B 2 ACHARNIANS Perhaps we may conjecture wept K. reSiov If it be asked why they should have gone round and about, not straight along, one may point in answer to é8ourAavotvres, which in itself almost suggests zepi. 196 abrat pev dLovo’ dpBpocias kal véxrapos kal way “mernpely our’ HyuepGv Tpidy kav 7G oTdpatt Aéyovor ‘Bair’ dry Gédes’. Difficulty has been felt about the infinitive éurypeiv, and the line sometimes therefore put after 198, with a possible change to émirjpe. But probably it is right. Besides poetical uses like Alcest. 879 ri yap dvdpt Kaxdv peilov dpapreiv (i.e. tov dpapreivy) moris dAdyou; Agam. 788 76 Soxely eivas (1.€. Tov elvat) tporiovar (so it seems best taken) and Herodotus’ occasional use of dvri with an infinitive, where Cobet would always add an article, e.g. 1. 210 dyri 88 &pxeoOa1, we find a close parallel in Apol. 39 c GAAG xpypdrov kat dedéa0ar (Tiyujowpat) ; first substantive, then infinitive. Whether émirnpeiv is in sense the right word is a further question ; it seems strange, but I am not sure it may not stand, 313 GAN’ éyd éywv Gdt TOAN’ dy aropyvayp’ éxelvous 200’ &~-Kadixovpévors. modAd. has been called in question, and its separation from é6’ & is certainly remarkable, due no doubt to metrical convenience. The suggested dAAa makes only imperfect sense, while peydAa and faora are very unlike it in form. I have thought of éAcya, as éAcyou and woAAof seem occasionally to get confused, and éAéya might go with déyov, in a few words ; but on the whole I think voAAd is probably right. Cf. Xen. Hell. 6. 3.11 kat iptv 8% eywye bpd bia 7a dyvuydves mpaxévra eorw Ste ToAdG avrituma. yryvopeva, and Plato Laws 754 B otk é&yvody Ore woAAai Trav KarorKirbet- civ (médewv) Siddopor tails Karoiiodoats modAdkis eye yeydvacn 318 xdy ye phy Aé~w Bicoua. wyd? TE TAYE Sod, brép émijvov "Oedjow rhv epadiy exwv déyev. Editors are pretty well agreed now that the last line cannot be right. The dactyl in rhv xepadijv is very ACHARNIANS 3 questionable. We need not lay stress on the illogical expression of the offer (zf I don’t convince the people, I will speak with my head on a block), for it is only an abbreviation, as Van Leeuwen says, and probably many readers never notice it. vy and «i are indeed often used in this way on the chance that, but the clause containing them does not usually come first. But the awkwardness of déyew after ph Aw is considerable. All sorts of substitutes for kedadyv have been proposed (Sépyv, odayyv, Adpvyy’), or other alterations of the line suggested. Porson strangely thought tov Képadov might be read. Keeping in view the two defects indicated and believing Aéyew to be wrong as well as xepaAyv, I am inclined to put forward the somewhat bold suggestion that riv xepadtv éywv A€yew is nothing more than a gloss or explanation which has taken the place of some more striking expression. The poet might for instance very well write brép émEqvou OeAjow Tov wept Wryis Spapeiv, using dep é. as a brief expression. He has the phrase in Wasps 375, roujow Saxety riv kapodiav Kal Tov Tept Wuyns Spduov Spayetv. I add some other passages which illustrate it, and which illustrate also in some cases the ellipse of ,Spdjuos or dywv: Il. 22. 161 zepi Wuxijs gov "Exropos tarodduowo: Herod. 7. 57. 2 rept éwvrod tpéxwv: 8. 74. 1 wept rod mavrés 78n Spdpov Oéovres : 1b. 102. 4 oAAovs aydvas Spapéovrar wept odewy airay and 140. 5 Oéew aie rept tucwv atrdv: 9. 37.3 tpexwv wept THs Yryis: Eur. Phoen. 1330 80x6 8 dyava tov wept yvyis, Kpéov, 45 mwempaxOa, and I.A. 1456: Xen. Mem. 3. 12. 1 6 wept ris Wuxis mpos Tovs oAculous dydv.: B.L. 10. 2 xupious rod rept THs Wuxis aydvos: de Re Hq. 7. 11 a horse rév abroduy (1.€. Spdmov or zpdxov) Sdtarpoxdlwv: Plat. Theaet. 172 & mept wuxijs 6 Spduos: Polyb. 18. 49. 1 édv, 7d 89 Acydpevor, tpexwot THY eoxdrnv and so 1. 87. 3: Zenob. 4. 85 Aayas Tov mept Tov Kpedy Tpexer: Plut. Mor, 1087 B ri zepi tov kpeav éemdéer: Dion. Hal, Isocr. 12 rév qept- puxis tpéxovre kivdevoy év Sixacrais and Ant. Rom. 15. 3. 9 rév éoyatov wept Woxis Tpéxover Spduov: Alciphr. 3. 72. 1 rév wept Woxijs dyiva tméuevay. Cf. further the constant omission of dpdpos or dywv with such words as 6 8éAtxos, 6 drdirys, etc., B2 4 ACHARNIANS eg. Laws 833 a, B 6 7d orddiov dépidAnodpevos . . 5 Sedrepos 8 6 rév SfavAov Kal tpiros 6 rév éfirmov Kat dy Kal rérapros 6 rév 8éAxxov : and in Aristophanes such an ellipse in other phrases as rynay ry paxpdy, Sixdoot piav, varpaxjoa piav, érépay éyxeov, yewpyety ToAAHv. 410 AIK. EipuriSy, EYP. ri AéAaxas; AIK. évaBadyv Touets, efov xaraBddyv ; obK érds xwdovs motets. drip ri ra pax éx tpaywdias Exets, cob’ édeewiy ; odk érds wTwXOdS ToLEls. The idea that dévaBd8yv here and in 399 means upstairs ought to be dismissed once for all. (1) The sense with the legs up is established beyond any reasonable doubt by the occurrence of évaBddyv xabjobat in Plut. Mor. 336 c, uv. Wyttenbach: Dio Chrys. 62 (45). 6: Athen. 528 F (all apparently taken from Ctesias): by Pollux’ recognition of the phrases dvaBaSyv xabypevos (3. 90) and dvaBddyv xabilew (6. 175): and by the passage in Plutus 1123 where Hermes, now out of work, says vuvi 8 wevav dvaBadnv évaravopat. The repeated junction with the idea of sitting really admits of no other interpretation, and the dvaBddyv of Hermes emphasises the notion of inactivity. Cf. Demetr. de Interpr. 289 éxi xpvojs KAivys KabeLouevov peréwpov. (2) On the other hand the sense upstairs is not established for any passage anywhere. It is true that a scholiast here says 76 dvaBddynv avo tods wé8as exew ert tyAod rérov kaOjpevos: and so Suidas dvaBddyv- éf’ tous dvw tots modas éxew kai Koaca. These explanations at first sight seem to jumble the two suggested meanings together, unless indeed we are to put a stop before the éxi tynAod and the ed’ dyous respectively ; but perhaps those phrases mean no more than the height, e.g. the couch, on which the feet rest. On Plut. 1123 the scholia say nothing of a tyydds rémos: they give the alternatives of legs up and feet crossed. No passage elsewhere has yet been cited in which the meaning wpstairs seems needed, and it is probable that the scholiasts had no more to go on than we have. (3) Euri- pides is evidently brought out by the eccyclema, and there is not the slightest reason for thinking that this or any- thing like it could be worked except on the ordinary level. ACHARNIANS 5 An eccyclema on the first floor is a thing unheard of. (4) The point of otx érds xwAods zoreis, no wonder you bring un lame people, is not the danger of falling down from an upper story. Why should adults be unable to take care of themselves? It is the suggestion of inability to walk conveyed by the reclining or half-reclining attitude. If the poet does not put down his own legs (xaraBaivew 408: cataBddyv 411: cf. Phaedo 61 c) and get up, no wonder his characters cannot use theirs. Line 412 presents considerable difficulty to my mind. The meaning of course must be that his own wearing of rags accounts for the raggedness of his heroes, just as his not using his legs accounts for their being lame. But then the point of this is spoiled if he is said to wear rags taken from tragedy. The rags of tragedy should come from his own, not his own from tragedy. We must not go behind his rags and start with (1) ragged heroes, whose tatters he then (2) takes and wears, and therefore (3) naturally produces new heroes in rags. The joke is lost, if the thing does not begin with him. To avoid this, I have sometimes thought of writing cis rpaywdfay for é« tpaywdias, éx and eis being liable to confusion (EK, EIC); and taking it to mean ‘to write tragedy in,’ ‘for tragedy.’ But I am not very well satisfied with that. [I think now that it may be better to suppose not that Euripides is dressed in rags as the editors usually consider and as the scholiast says, but that éyes means have about you, keep by you, as the following scene amply illustrates. €x tpaywdias is then right and éoOj7’ éX(e)ewnv may be a quotation or burlesque. But the joke still fails in the way I have indicated, as the raggedness does not begin with him.] 464 aaipyoe. we THY Tpaywdtay. Not any particular tragedy, but the tragic art in general. So in 470 gpodéd por ra Spduara. When it is said in 399 that the poet dvaBddyv woet tpaywdiav, it means that he is writing tragedy, not a tragedy. So tpvywdiay zoey in 499: Wasps 1511, 6 opixpdraros 3s THY Tpaywdiav Tore? : Plato Lach. 183 a ds av otyras tpaywdiav Kadds rote. 6 ACHARNIANS 525 veaviat KAerrovat. The compound verb used in the corresponding sentence (527 dvre€éxAepav) suggests that we should write veavios 2 f KkA€TTovct. 549 Should we write doxovs, TpoTwrhpas, Kddous avoupever 4 641 ratra rowjoas roAAGY ayabdv uittos Syiv yeyévyrat kat Tovs Onpous év tails méAcow Setgas as Sypoxpa- TovvTat. The poet sometimes and to some extent takes, or affects to take, the side of the suffering allies of Athens, oppressed, not indeed by the generous Athenian people, but by its unscrupulous and dishonest servants and supporters. Cf. such passages as Peace 759 rowtrov idiv répas (Cleon) od katédeto’, GAN’ trtp bay wodeuilwy dvretxov det Kai TOV GAXwV vyowy. 1b. 639 and 936: Wasps 669: Knights 1319 *"AOnjvats kat tals vycos éixoupe and 1b. 1408. Taking such passages into account and also the significant fact that he goes on here to describe the admiration which his conduct will excite among the allies (643-5), I should suggest that rots Syuous be read in 642 for rots Sjuous and joined in construction to dyiv. If his benefits are confined to Athens, there is much less reason for the allies’ admir- ation than if the benefits are distinctly and truly stated to extend to them too. It is not at all against this that the preceding lines have spoken of the envoys of the allies trying to cajole the city, and of the poet exposing them. 646 In the earlier part of the parabasis the poet praises himself, as he so often does, for the courage and true patriotism which he has shown. When the allies bring their money to Athens, they will all be eager to catch sight of the brave and honest poet : otrw & adrod rept ris ToApys 45n Téppw KAéos Fret, Grexai Bacrreds, Aaxcdammoviuv rhv mpecBeiav Bacavifwv, Hpornoev mpOra pev K.7.A. From this no doubt come the words of the Life § 9 ofrw de yéyover 4 yyy Tod wowtod os Kal mapa Tépoas Sujfkew cat tov BaciAéa x.7.d, , ACHARNIANS 7 If however oirw 8é is right, we should expect not ére but dare. Some one has suggested ds in the sense of doe. But, though such a s is admissible in poetry, much affected by Xenophon, and frequent in late Greek such as the Life, it is extremely rare in pure Attic and not (I think) ever found in Aristophanes, nor is there any reason apparent why it should have got changed to ére. If, on the other hand, dre is right, otrw 8é¢ refuses to harmonise with it. Recognising this and remembering Hg. 530 otrws nvOnoev éxeivos, Elmsley changed oitw 8€ to otrws, and others, ¢.g. Blaydes, have followed him. But Hg. 530 is not parallel and odrws is not suitable, for there the ofrws clause winds up the sentence, whereas here dre «.7.A. carries it on, so that otrws would have to perform the impossible task of looking both backwards and forwards at the same time. If otrws refers to the poet’s fame among the Greek allies, the Great King’s knowledge of him cannot be brought in to illustrate it. I have little doubt that Aristophanes wrote évtws 8é. Truly his fame has spread wide. In the first place otrws and évrws are frequently ex- changed for one another by copyists. Which word is the right one in this or that passage is immaterial to the argument, but it will be found that the MSS. often vary between them: e.g. Plato Huthyd. 305: Theaet. 198p: Laws 708d: Critias 106a: Ken. Hell. 7. 4. 3: [Arist.] Ehet. ad Alew. 1422 a 21. In other places the more familiar ovrws has driven out an original éy7ws altogether, and the latter has only been replaced by modern critics. In the second place the emphatic évrws is quite at home at the beginning of the sentence. Cf. such passages as Plato Soph. 236D dvrws, & paxdpte, éopev ey ravrdract xarery oxepea: Phil. 448: Laws 708 v. Finally compare the words of Xenophon Hipparch. 5. 9, which resemble our passage as much as if they had been written to illustrate it: dvrws yap ovdtv KepSadedrepov év modem ardrns, 67 OTE ye Kal of waides Grav Tailwut wocivea Svvavrat dwatay «.7.A.: unless the ézdre clause should be joined with what follows, not with what precedes. On the use of dvrws in Aristophanes see Starkie’s note to Wasps 997. 8 ACHARNIANS 655 A few lines further on he pretends that the Lace- daemonians will try to appropriate him, GAN iets tou pH wor’ apiAO’, ds kopwdjoe Ta Sikora. So the editors after Tyrwhitt, but the MSS. have adjoere, adijueb’, or ddan? (Van Leeuwen). There are reasons for suspecting both the grammar and the metre here, and the two suspicions confirm one another. As to the grammar, if roe is to be joined with a jy, payroe is the established order, not ro: 47. See, for instance, Soph. O.C. 1407 prjrot pe mpds Oeiv—py p’ dtysdonré ye, and 1439 pajrou pw? 6ddpov: Aesch. P.V. 625: Plato Rep. 438 a. Then the division of the anapaestic line should coincide with the pause of the sense at d¢70’, not come awkwardly attaching és to what precedes and cutting it off from its own clause. We can easily remedy this defect by writing os cwpwdet for és kopwdyoe, the present tense being at least as suitable (631). This leaves us a syllable short in the first half of the verse, and then we see in a moment what the poet wrote, GAN’ dpeis ov tot py wor’ adn’, as kopwdel Ta dikata, just as seven lines below he writes (from Euripides) ob py wo’ G\G. The dadyoere of the MSS. does, therefore, in a way, represent the true meaning, and may somehow have grown out of it. . In the next line (656) dyciv 8 ipas wodda Siddéew, «.7.d., Hamaker has with reason demurred to dddéew side by side with ra BeAtiota Siddoxwv in the very same sentence and suggested éri Spdcev. Before finding his suggestion in Blaydes’ note, I had thought of zoijoe. 683 lopulovres Se yypa Ta Ni ce tovBop gov 5 O€ ype Te (Oy Tpooéorapev ody Spavres oddev ef uy THS Sikys THY HAvyyV 6 b€ veavias aut orovddeas Evviyopetv els TAXOS Tale. EvvdTTWY oTpoyyVAoLs Tos PHyacty. The difficulty of 685 is well known. Elmsley and others following him think we may make veavias accusative plural or read veaviay, and translate ‘but the other (the adversary) having taken care to have young men (a young man) assisting him in the prosecution (or prosecuting for him) engages us and strikes, etc.’ But (1), if the prosecu- ACHARNIANS 9 tor could call in a young spark to help him, so could the defendant : (2) nothing can get over the awkwardness of language by which the prosecutor, after calling in an assistant, is himself said to carry on the fray, when the meaning must be that it is the veavias who wate: évvartwv etc., because the gist of the whole passage is the opposition of young and old. Merry rightly points out the difficulty of describing the prosecutor as ‘doing per se what he is really doing per alowm.’ But neither is Merry’s own explanation at all satis- factory. He explains that ‘the young barrister’ (6 6: veavias) is only anxious to ‘advance himself personally’ (éaurd Evvyyopeiv). This strains the use of the dative and does not give us a really good sense. What is there to show that this is the young man’s one desire? He is not represented as spoiling his case by display of cleverness. He does just what is effectual, that is, he bewilders and silences the old man opposed to him. What proof of self- seeking is there in this ? The only emendation that deserves consideration is one made by a modern Greek scholar and adopted by Van Leeuwen as emendatio praestantissima. Kontos proposes 6 8& veavias dv Tw orovdacas Evvyyopy. This is certainly ingenious, though the hypothetical éay is not very satisfactory to me; but I think we may perhaps do better. With the insertion of one letter I would read c x / * F >. we , n 6 de veavias éx’ ait@, orovddcas Evvyyopety, taking éx’ air@ in the sense of ‘ matched against him,’ ‘ told off to deal with him.’ So of the champions who are matched against others in the Septem we read (447) *. &. 8 a) > aA > , , 2 >” avynp eT QuTw, KEL oTopLapyos €OT ayay, aldwy réraxrat Aja, and again (620) Spus 8’ én’ aitd pita Aacbévous Binv eyOpdéevoy mudupov dvrird£oper, and Eur. Heracl. 837 dvip 8 én’ dvdpi ards exapréper dyn 10 ACHARNIANS So in Thuc. with rerdyOa: 2. 70. 1. Adyous mpordépover . +. TOs oTpaTnyois ... Tots ért odion reraypevois: 3. 13. 3 ai 8 (vies) eb’ jyiv reréxara, and in Theaet. 172% the adversary in court dvdykyy exwv épéorynxev. In these passages there is no doubt a verb added, but it does not seem necessary. We might possibly attach éx’ airg to éwvnyopeiv, if we cared to do so. For myself I would rather point it off as above with a comma. No one should take exception to aivé on the ground that it has no one to refer to, a8 zpocéorapey is plural. It is common in Greek to pass thus from plural to singular (we, they, a man, etc.) or from singular to plural ; and Aristo- phanes in particular indulges in this confusion of numbers to a surprising extent, e.g. Thesm. 789 el kaxov éopev, ti yapeil? Huds, elrep dAnOGs KaKdv eoper, kdayopevere pnt eSedOety pyr’ exxdpacay dddvat ; Let anyone study Clouds 988-9: Wasps 552-558 and 564-568: Peace 639-641: Frogs 1071-1076: Eccles. 663-672, and he will not hesitate here. For other examples of airdés referring to plurals preceding cf. Peace 745: Thuc. 3. 17. 4: Dem. 21. 76: Plato Phaedo 62D and Prot. 319p: Ar. Poet. 26. 1461 b 29, uv. Vahlen. In the lines following here (688-691) the singular is actually used. 691 tovr’ dpAwv darépxopat. As line 689 has just ended with dgAav drépyxeran, is it too bold to propose dzciAeoa here in the place of dwrépyopaty which may very well be an erroneous repetition? he dzodkgoat in another sense of 692 does not seem to me against this, at any rate in comparison. 717 kégéedavvew ypt 76 Aourdv Kdv dbyy rus Cyprody Tov ‘yépovta. TG yépovtt, Tov véoy Se TO véew. Tt seems hopeless to try to make sense of xdv diyn ris. But Blaydes’ dy (or dAG) and Van Leeuwen’s cgady are too far from ¢vyy to be at all probable. The editors do not seem to notice the scholium on this passage: xdy éfeAavvew xen Kav guy Cousodv, id yépovtos Toro rdcyewv Tov yépovra. There is no room in the verse for kav é&eAavvew— éfeAavve ACHARNIANS 11 must be right from the parallel in Antiphanes—but the scholium may point to something like «ay (or dv) guyy tes énpwot. [So, I now find, Paley.] 1088 Seervely kataxwAvets moat 7a 8’ dAXa mav7’ éotiv waperkevacpeva, kAtvat, Tpdmelat, K.T.A. ab wdépvas Tapa, K.T.A. 1093 épynorpises, Ta Pidtab’ “Appodiov, kadai. Many have been the attempts to deal with the last line, in which ra ¢. ‘A. can hardly be right and xaAai is dubious. Taking 76 ‘pidrad’ ‘“Apyodse’ as a probable correction, equivalent to scolia or songs in general (perhaps we might keep the plural ra in this sense), can we make anything of kadai, which would then certainly need altering? It occurs to me as ‘possible that it may stand:for waAa, used again as in 1088 (cf. 1114, etc.), going with éoriy tapeoxevacpéve. and still more with wdpa in 1091. 1196. Perhaps dp’ ci for the impossible av et of some MSS. (others yap «i, some «i, R av). KNIGHTS Way do English editors go on giving us characters called Demosthenes, Nicias, and Cleon? There are no such people in the play at all, and it would greatly vex Aristophanes to see his satire so spoiled. From an editor with Merry’s sense of humour the comic poet might have looked for better treatment, and now’ Hall and Geldart in their new Bibliotheca Oxoniensis text have done the same thing. The foreign editors are usually too skilful to make this mistake. Neither Kock, nor Von Velsen, nor Van Leeuwen falls into it, though Zacher does. Blaydes, too, in his monumental Aristophanes knows better. But the smaller editions which in this country will be in every one’s hands continue what I must take ‘eave to call the absurd tradition. Cleon, Nicias, Demo- sthenes do not appear in the Knights. There are three slaves, two like those in the Peace without any distinctive names at all, the third named Paphlagon, which is as mucha name as Syrus. No doubt they stand in a way for the distin- guished men in question, but the point of the whole thing is that it is a sort of humorous allegory or imaginative burles- que, and attaching the names of the real men to the three slaves dulls and blunts the humour in an unpardonable degree. Paphlagon is not Cleon: he only holds towards his master and his fellow slaves a relation burlesquing that of Cleon towards the people and the other prominent public men of Athens. We must not confound the caricature with the thing caricatured by putting the true names to the fancy picture. As well might Dryden in his great satire have written Shaftesbury and Shadwell for Achitophel and Og. KNIGHTS 13 15 (OIK. A] dA’ ceive Gappiay, elra, Kayes oot dpiow. 16 OIK.B w@s av av pou NE~etas dpe xp rE yeuv; 17 OIK. A GAN’ ovk et pou 76 Opérre. ds av ody Tore 18 etroun’ av adro Oyra KopievpiTtkas ; 19 OIK.B py pol ye, py pot, py Stacxavdsioys. This is the order of the lines in all MSS. Recent editors, seeing that they are rather incoherent as they stand, have put 16 before 15, and it is certainly the case that 17 gains by following directly on 15. But what seems to me the true order of the five lines has not been suggested. Line 16, the quotation from Euripides, is the key to the difficulty. In 18 A asks how he can express himself in subtle Euripidean fashion. He then answers his own question by this quotation (16) mis dv x.7.A., and it is this quotation which provokes from B the exclamation of disgust, py poi ye, «.7.A., with the reference to the Euripi- dean manner in dtackavdixioys. The lines will then run: (OIK. A] ddd’ etre Gappay, : clra Kayo coe dpdow. OIK. B dAX’? obk vt prot 70 Opérte. mas dv ovv Tore elon’ av aitd Sira KopwevpeTiKas ; was av ot pou A€€eras ape yxpr Néyetv; OIK. A py poi ye, wy pot, py Stackavdixions. If this transposition be adopted, some slight change must be made elsewhere, if we are not to alter the speakers through the rest of the scene. We might give 11 and 12 to B and so proceed, or Aéye ov in 13 might be given to A, od pe ow «7A. to B. For the transposition compare perhaps 258-265, where it seems best to put 264-5 after 260. 32 motov Bpéras ; éredy Hyet yap Oeovs. Several attempts have been made to supply the missing syllable. Has anyone thought of éredv yap jyet Geovs? Cf. Eur. Hee. 800 vopw yap robs Gots yovpeBa. 5 a ee ee 2 87 Te€pt TOTOU youv ert oot. That yoty is right and that the words should be taken as a statement, not as a question, appears from Plut. Mor. 14 KNIGHTS 179 ¢ mdvv yotv cou wept ris tov “ENAjvev Spovoias 6 Adyos éoriv. In 89, on the other hand, xpowvoyurpoAjpaios (1) ef should, I think, be printed as a question. 188 GAN’, dyd6’, ob8& povorxny éxicrapac TARY ypappdrev, cal Tatra pévro. Kaka KaKas. Blaydes quotes the imitation in Procopius Bell. Pers. p. 70 c: ob yap dAAo oddey és ypapparirrod dorrav guabey ore PN ypdppara, kal tadra Kaka Kak@s, ypdyat. Gibbon in a note to his fortieth chapter has cited the historian’s de- scription of John of Cappadocia, adding the comment ‘a forcible expression.’ Apparently he did not remember that the words were a quotation from Aristophanes, and his editors, as far as I know, have not pointed it out. Quintilian 1. 10. 18 must be confusing the Maricas with the Knights, when he says that in Eupolis Maricas qut est Hyperbolus nilil se ex musice scire mist litteras confitetur. In spite of Nub. 553 this is more likely than that one poet took it from the other. 339 GAN’ abrd wepi tod rpdrepos eiretv mpGra Siapayodpa. As airé is wanting in construction—would not the poet have written airod d& wepit—and as V and other MSS. add rotro to it, perhaps we may cut out wepi rod and read GAN’ airé rodro mpérepos cireiv, taking wpdrepos eirety as explanatory of rodro. 401 kal 8&8ackoiuny rpooddew Mopoipov tpaywdiav. I suggest Mopoivw as having two advantages, It fits mpoogsew better (xopoicw tradwv, Frogs 366) and it enables us to take zpaywdia in the general abstract sense, tragedy or tragic song, in which Aristophanes commonly uses it. Cf. on Ach, 464. 413 tmepBarciabal o’ ciopat rovroor, i) parnv y’ dv dropaySadias cirovpevos Towotros éxtpaelny. After much reflection I find it impossible to believe that the optative can be right here as a substitute for a past tense of the indicative, and I suggest that we should read éxrpadgels jv. KNIGHTS 15 The first thing is to show that there is no parallel forthcoming for such an optative. In Homer indeed a use more or less similar (imitated two or three times by Virgil) is not very infrequent, but no argument lies from Homer to Aristophanes. In Homer we find the subjunc- tive as a future: it does not follow that Aristophanes could use it so. Even tragedy has no such optatives, unless it be in one echo of a Homeric phrase hereafter to be mentioned ; much less comedy. An idiom with which it is natural to compare or confuse our passage is one fairly common in Herodotus, which may be illustrated from 1. 2 “EAAnjvev twas... pact... dpmdcar tov Baciréws THy Ovyarépa Hipdany’ einoav 8’ av ottot Kpijres and 1. 70 adrot d& Sdpior A€yovar ds... dwéSovTo Tov KpyTHpa év Sdpw, iuitas S& dvdpas mpiapevovs avabeival pw és Td “‘Hpatov' taxa 8 av Kal of daoddpevor A€youev aaixdpevor és Sadpryv as draipefeinoav irs Sapiwv. (For other passages see Stein on 1. 701 and Gildersleeve’s Syntax of Classical Greek § 487.) To this idiom there is an English one (found I think also in other modern languages) exactly corresponding: ‘these would probably be Cretans,’ ‘tho vendors would perhaps say,’ instead of ‘probably were Cretans’ and ‘perhaps said.’ It is distinguished by two things from the phrase before us in Aristophanes: (1) it occurs always in a simple direct sentence, not in a more or less complex one as here: (2) what is more important, the optative is not, as in Aristophanes (and in Homer), an alternative for the indicative. In hardly one of the passages of Herodotus could we use a past tense of the indicative with dv. otro. 8 av fjoav Kpares would mean something quite different. If therefore we find in Attic a passage or two parallei to those in Herodotus, they cannot fairly, even when an indicative would have been ‘possible, be pleaded in defence of the line in the Knights. Thue. 1. 9. 4 otk dv otv vijcwv ew rOv meprorxiOwv—atdrat de ok dv ToAAal elev—iretpdrys dy 1 Stein seems to distinguish 1. 2 from the other passages, not making his meaning very clear, but probably thinking with Goodwin (Moods and Tenses, 238 and 443) that it means ‘these would prove to have been Cretans,’ i.e. would now prove, if we could go into it. It seems better to take this passage like the others. 16 KNIGHTS éxpdre,, el pi te Kat vavrixdy efyev may be classed as such, though the optative can quite well bear its usual sense, ‘would not prove many, if we added them up.’ Antiphon 4.2.5 was dv émBovretvoapu aire, ef py cai éreBovdevOnv x’ abrod (which Blass alters, needlessly I think, to ras av ereBovreved. re abrG 5 tt py Kal «.7-A.), Show’ or ‘why should I plot against him, if he had not plotted against me,’ is exactly parallel to Herodotus 2. 11 xod ye 89 év 7a mpoavaicywpevy xpdvy...ovk av xwobein xddAros Kat roAkG pélwv eu rovrov; Other examples ,are Menexenus 240d év Tour dy dv tis yevopevos yvoin olor dpa éervyxavov dvres K.T-dey which clearly refers to past time; Laws 677B ot rére mepipuyovres THY POopav axeddv opel tives av elev vomits, where see Stallbaum. In the Thucydides passage and in the Laws the indicative would be as impossible as in Hero- dotus ; in Antiphon it might have been used, and with a slightly different meaning in the Menexenus. (The words of Lycurgus 138 xai rév rerpaypever of rovotror av perdo orev} are not really parallel, for peraoxovev = eBéAorey preracryetv.)* One or two Attic passages may also be quoted in which I take it that the optative is defensible on ordinary grounds as referring to future time, though the indicative might have taken its place with some change of meaning (referring to the present) and might perhaps have been more naturally used. Such are Soph. Ant. 505 rovrois rovro waow dvddvew déyour’ av, ef py) yA@ooay eyxAyor PdBos, on which Jebb has no comment, and Eur. Med. 568 oi8’ av ob dains, & oe py xvifor déxos. They mean virtually ‘Every one would say so, if the fear of Creon were removed,’ and ‘ Even you would not maintain it, if you could get rid of your jealousy.’ The optative in Ar. Hq. 1338 éue yap vouifos av Oedv is just the same. In poetry this slight irregularity or straining of expression is pardonable enough. The important point to notice is that the speaker really has or might very well have the possible future in mind, and this clearly separates such passages from the 1 Quoted as an instance by Platt in the Journal of Philology, vol. 26, p. 87. 21 have noticed a few examples in late Greek: Didymus on Demosth. 8. 1 and 13, 59: Plut. Mor. 8338 twice: Arrian Ind. 5. 13: Pausan. 1. 9. 5 and 8, 12. 7: Polyaen. 2. 5. 1. KNIGHTS 17 one before us. So in Phaedo 89 c kai éywy’ dv, ef od etyv kai pe Stadiyor & Aoyos, Evopxov av wownoaivny «.7.A.: the optative is not irregular: it means ‘if I were ever to find myself in your place, I would.’ But in Eur. Suppl. 764 gains av. ei wapyod’, dr’ hydra vexpovs (before which a line is lost, if we are not with Lobeck to delete 764 itself) I do not see how we are to justify the optative except by regarding it as an echo of the gains dv which occurs several times in Homer. We might perhaps compare it with the Homeric ¢% of8a, which Attic drama, even comedy, uses regardless of hiatus. But as far as I know, it is found nowhere else in Attic. Lastly there are a few places in which (if the MSS. are right, which is not by any means always certain) one clause is indicative and the other optative. The expla- nation seems to be that the speaker’s point of view shifts. He might have spoken in the past throughout. He might have spoken in the future. Both being possible, he mixes the two together. Thus in Lycurgus 66 ef ris éo. vopov... eEadeiperey, clr’ daoAoyoito...dp’ ovk ay dmexreivar’ abrov ; éy® piv olpar dtxaiws, eimep euéAdere al robs dAAovs oglew, and with a deliberate repetition of the confusion Lysias 10. 8 ef pév ris ce elo watpadotay 4} pntpadoiay, Hélovs av abrav dpdciy cou Sikyv’ ci O€ Tis elron GS THY TeKOdTAV q tov dicavra érumres, wou dv abrov &Ljurov Seiv elvar. The same explanation will apply, if the readings are right, to Phaedrus 2514 ads Oedv o€Berat, cai ei pr edeSiee tiv rhs oddpa pavias Sdéav, Ovo. av as aydApare: Alcib. i. 111 B, ri & ci BovdnOciper cidévar..., dp’ ixavoi dv Hiv Foav Si8doKnador of woAAoi: Pseud. Plat. wepi du. 373 DE: Diog. L. 6. 59 OavpdLovrds twos Ta év SapnoOpdky dvabypata ep, ToddAG ay ely mreiw ci kal of wy) cwOevres dveriOecav (where however ety should probably be jv). To this type of sentence I should be inclined to assign Iliad 2. 80 > t x ¥ > a » mH ei pév Tis TOV dvetpov Axalay GAXos Evierre, petdds kev haipev kal vordiloipea paddov. and Odyssey 1. 236 2 ‘ » ¥ CP 2 F émel ov Ke Oavdvte ep OO GKkayoiuny, > x e Ls , 5 , af . ee 8S x ei peta ols Erdpoior Sadun Tpwwy évi dip, aS o 2 or > ‘ ¥ A. - He hirwy ev xepatv, éret TOAELOV TOAUTEVCE. c 18 KNIGHTS In the first Nestor mixes up ‘if any one had said’ and ‘if any one were to say’; in the second Penelope says in effect ‘I should not grieve so much; I say I should not have grieved so much, if’ etc. Such passages seem to me very different from the xaf wi xev &@’ dadAouro..., ef wy dp’ 6£) ydnoe type, where a future contingency cannot possibly be in the speaker’s mind.! Coming back now to the Knights, we shall find that the words of line 413 are quite different from the usages we have been examining. The optative cannot be taken, as in the latter of them, for an ordinary optative referring with or without some inconsistency to the future ; nor has it, as in the former, any idea of what would be likely (= was likely) to occur under real past circumstances. To regard éxtpageiny as = éxreOpappevos env or pavetny (Gilder- sleeve) is to invent a licence for which no parallel is adduced and which puts an undue strain on the possibilities of colloquial Attic Greek. Rather then than acquiesce in anything so irregular, I think we ought to add to éxrpadedyy one letter and turn it into éxrpageis Hv. First let us assume that those two words go closely together and are the analytical equivalent of éferpdgyv. Such an equivalent seems quite admissible, though probably not to be paralleled from Aristophanes himself. Consider the following instances of an aorist participle with ciué or yéyvomat: 1. Epic, ete. Hymn Herm. 92 kai re dav pty wv etvat Kat kwdds dxovcas: Anthol. 12. 234. 3 (Strato) dvOos yap kal xéAXos icov xpévov éorl Aaxdvra. 1 In Eur. Or. 1132 ei uty yap els yuvatka owppoverrépay Elpos peGciuer, ducxdrchs bv Fv dédvos- why do Goodwin (1.7. 508) and Kiihner-Gerth (576 c) make pedciuev optative? It is the indicative, like dveinerv Wasps 574, dveite O.T. 1405. Xen. Cyn. 12, 22 (which Goodwin cites) is probably wrong. Latin verse, and even prose, sometimes use the present subjunc- tive instead of the imperfect, and sometimes an irregular combination of the two, as above. I do not know anything in Latin similar to the efncay & by obo: Kpiires use, nor any present tense in apodosis really like éxrpapelny. KNIGHTS 19 2. Tonic Prose. Herodotus 2. 10, 4 pya drodeédpevor peydra ciod: 3. 27. 3 ot 8& edpalov ws opt eds etn paveis (Cobet Oeds emihaveiy) : tb. 120. 1 txd Kupov karactabels fv Sapdiov vmapxos ’Opoirns: 4.127. 1 od8€ te vedrepdv eis rorjoas: 7.194. 3 euedre od 76 Sedtepov Siapvyav ~vecOar (epreccoOar Reiske, GOGos éoecat Cobet). Melissus fragm. 12. (Mullach) kas av peraxoopyfey trav edvtwy te ein ; 3. Tragedy. Soph. O.T. 90 otre yap Apacs | ovr’ obv mpodetoas eipi : 957 airdés or od onunvas yevod: 970 otrw & dv Garvey ein “E éuod: 1146 od cwryoas éoe. ; four examples in one play, to which add O.C. 816: Ant. 1067: Aj. 588: Phil. 773. There is a dubious instance in Aesch. Suppl. 460, and what seems a clear one in Eur. Suppl. 511 éfapxéoas jv Zeis 6 Tinwpovpevos. 4, Comedy. Menander JIncert, 475 M. 684 K. Grav déyys pev odd, pavOdvys Se wy, 76 cov didagas Toipov od pabwy éoet. Herodas 3. 88 Seipov & adxpis Atos Svs 7 (so Meister and Crusius for dvoy). 5. Attic Prose. Antiphon 3. 4. 4 6 waSorpiBys av dzoxretvas airov «in, and almost the same words in 2. 3. 8 (ef. O.7. 970 above, but some editors <6> 6r’, what we do admire 1s how, when etc., the verb after ds being dveBpvaéav in 602. Compare a few sentences of more or less similar form : Ach, 654 kal THS VacOU pey exeivns ob dpovtifovo’, GAN’ iva todrov Tov TrowTHv adéAwvrau. Wasps 482 dra viv ney ovdev dAyeis, GAN’ Grav x... Plato Lysis 204 D kal & pev karahoyddyy Supyetran Sava. dvta ov mavu Te ded eat, GAN’ éreday Ta. Toujpara jypav emixerpyon xatavrAe. Ar. Hth. 4. 7.1127 b 12 i 8 Gerd tevos, 6 pev ddéns } tyws ob Alav Wexrds..., 6 Se dpyupiov... dayypovéo- repos, though not elliptic in construction, is worth citing for the od Aiav. Cf. also Plato Rep. 421 a: Symp. 198 B: Herodas 3. 43. 658 Ore Sy "yvwv Tots BoAcrous ATTnpEVOS. He is not definitely defeated but only in process of being so, and he actually recovers his ground in the very next line. Read then 7rraevos. 712 aXX’, & wovnpé, cot pev oddev weiPerar. As Demos is at present quite unacquainted with the sausage-seller, there is no point in saying that he does not 22 ENIGHTS heed him. How can you heed someone you never heard of? Read the future zeimera, corresponding to the futures in the two preceding lines. The mistake is a common one, e.g. Prom. Vinct. 333. It occurs again, I think, in Clouds 1422 ovkouv avyp 6 Tov vopov Oels ToOUToY Hv TA TpGTOY, domep ov Kayo, kal A€ywv ree Tors radatous ; éreHe is certainly possible, meaning that he had habitu- ally to persuade; but grec, matching Geis, seems more likely. 783 émt rato. rérpats ob ppovrife. oxypas oe xabypevov otrws. The parallels quoted for the accusative, being practically all poetical (Ar. Rhet. 2. 9 1386b 281), are not very satisfactory. Should we read cides for otrws? On con- fusion of « and ov see Bast. Comm. p. 760. 1045 ev otk dvadiddoKe oe TOY Aoyiwy Exav. Surely dvadiddfe. éxav, if -he can help it, does not suit, a present tense, except of habitual action (e.g. Xen. Cyrop. 5. 1. 16). 1196 If rives is rightly given to Paphlagon, the lines should run without any stop at gpyovra.. éxewou yap as ey’ épyovra: TIA. tives ; AAA. mpéoBes x.t.r. 1207 rf od Siaxpivers, Afju’, odrepds éort vev avip apeivey mepi oé ; Ajpwe without 4 is rightly doubted as too unceremonious. Everywhere else—eight or ten times—it is & Ajje and even & Anpidiov. The alterations suggested are not very satisfactory, and therefore I may perhaps propose érepos, & Ap’, eri vav. 1385 dye vuv emt rovrois Tovrovi tov éxadiav \ 75° 2s gy ¥ 4 5 KL TOLO EVOPXNV OTTEP OLTEL TOVOE GOL. éomep cannot be used thus for és for doris. Read ds meptoice. So Elmsley corrected the unmetrical ofdmep KNIGHTS 23 iBpifopac in 727 to ofa mepwBpitoua, and Bothe elrep éAatvey in Vesp. 1050 to ci apeAatven (from the scholium). Thue. 4. 9. 1 the MSS. give aimep Hoav, the scholiast seems to have had ait zepijoav. Plato Prot. 3428 Heusde restored @ mepiciow as against o: mepéeiow in B and grep ciolv of T. Sometimes the error is the other way. CLOUDS 75 viv obv GAqv riv vixta hpovTiLwv 6500 pilav nipov drpardy Saipovins breppva, qv jy dvareicw tovtovt cwljocopa. A great objection to 6500 (for which Blaydes suggests and Van Leeuwen reads podXis) is that there is nothing in the lines preceding to lead up to it. They suggest no difficulty, no need of a 68¢s. The use of dpovri{w with a genitive (usually in negative clauses) seems also rather strained here, as it has to mean meditating, trying to find, a way, not caring about. It would bea very slight change and yet would make the passage materially smoother, if we read 6ddv piay nbpov, drparév Saioviws treppva. 145 ss dvyper’ dpri XaipedpOvra Swxpdrys yvAAav drdécous dAAotTo ToOds abris wddas: Saxodoa yap Tod Xaipepavros thy dppiv emi tiv Kehadyy THY Swxpatovs adyAato. All the editors seem to give dAAoro. Both R and V however, with other MSS., are said to have dAoro, and that (or perhaps dAa:ro) must be right. The question is not a general one, how far fleas jumped. 148 shows that it refers to the particular flea by which Chaerephon was bitten : Saxotca yap «.7.A. is inconsistent with anything else. The question of Socrates therefore (if it was not rather Chaerephon’s, as Piccolomini thinks, reading Xawpedav rév Swxpdryv) was, how far a flea had jumped. So apparently the scholiast took it, who wrote in his note éxécovs (é8as) andnoete. 179 As the text of Demetrius gives indriov (not Goindrov) and Epict. 4. 11. 20 is quite consistent with it, CLOUDS 25 we should surely omit the article. The passage is of course meant to be nonsense, but 7é is wholly out of place. 320 kal wept Kamvov orevodexeElv. Perhaps wept tod te xamvod orevodecyeiv. The quantity of the a in xamvod is a real difficulty. 520) ovrw vixyoatpié 7’ éyw Kal vouiLoipny codos, ds tas Wyovpevos etvat Oeatas deftovs kal TavTny copurar’ exew TOY E“aV KupMdaV mpatous nélwo’ dvayeto’ buas, ) Tapéoxe por €pyov mAciorov: lr’ dvexdpouv tm’ avdpav optikar - qrryOeis, ovk dévos av. It seems impossible to make sense of zpdrous dvayedo’ ipads. It is understood (1) of the first edition of the Clouds brought out at the Great Dionysia, as if Aris- tophanes claimed credit for having first produced it there, when he might have produced it at Piraeus or even (Teuffel) Aegina. Cf. Prologue to Henry VIII, line 23, as you are known The first and happiest hearers of the town, Be sad as we would make ye. I question whether, except a doubtful statement in Aelian and one or two other stories of very exceptional circum- stances, there is any evidence of new plays by Athenian poets being produced anywhere but in Athens. Certainly that was the ordinary practice, and the poet would never have drawn attention to the fact that he had in this case actually conformed to it. What else was he likely or even able to do? The words are also taken (Blaydes, Kock who reads zparynv) vf (2) the second edition, to which our parabasis belougs. This is however entirely inconsistent with ¢fr’ dvexwpovy x.7.A., which must refer to the same occasion and -is admitted to refer to the first performance. There is a further difficulty in the word dvayetoa. Whether it means ‘ give a taste’ or ‘give a second taste,’ the editors ignore the fact that the idea of a taste is quite out of place. A taste of a play would be given by the performance of one scene or the recital of one pios. To 26 CLOUDS exhibit a play entire is not to give a taste of it. This objection is fatal to some other theories and proposals that have been broached. When we read the passage, the sense required seems to be something like ‘I thought my cleverest play, the play which cost me so much labour, would be sure to meet with your approval: that you would be the first persons to enjoy it.’ But how are we to get this? Pending some better suggestion, I propose very doubtfully zpurovus #éiwoa yaviio’, or ‘ydvuc6', ipas : ‘I expected to delight you, or you to be delighted, first.’ The active yavotv is not found in this sense, though the passive, especially in the participle yeyavepevos, is well known. No doubt therefore yavdcar is dubious, though not out of the question. ydvucba is unobjectionable, but less near the MSS. As for Aris- tophanes’ use of the words, cf. Ach. 7 rai6’ ds éyavéOnv and Wasps 612 rovrouww éyo ydvvpa. Piato Phaedr. 234d ddxers ydvucba. ixd tod Adyov. Alexis 186 érvyavicas in its physical sense. 776 Srws drrootpépais av avriSixav Sixqy. Perhaps dvriAéywv, -8ixdv being due to Sixyv following. dvriAéyw occurs repeatedly in this part of the play, 901, 938, 1040, ete. 995 Kai py wept tobs cavtod yovéas oKxaioupyeiv aAXo Te pndev aicxpov oiv Gre THs aidods péddras Taya’ dvarrAdrrew. So R, though V and one scholium in R have évarAjoev. Among editors there is much doubt, some writing 8 mt. . pédder .. dvardjoev. dvawddrrew dyahpa is found elsewhere in the 'sense of making, moulding, an image (see the passages in Blaydes’ note). An image too cannot well be ‘infected’ (dvammAdvor does not mean stain) and in that sense the verb is always found elsewhere with a genitive. The real difficulty about réyadp’ dvarddrrew seems to me to lie in the article, which has no business there. I would propose therefore péAdes rot ayaAp’ dvarAdrrev. The youth will make himself an aiSods dyadpa. Cf. Tac. A. 16. 32. 3 habitu et ore ad exprimen- CLOUDS 27 dam wmaginem honesti exercitus, and the dydApar’ éperqs in Plato Sympos. 222.4: also Phaedr. 252, though that is less like. ; 1047 émioxes: cbfds ydp a’ exw pecov AaBov dduxrov. Nowhere else is dguxros active in meaning, and it is unlikely to be so here. Perhaps dgvuxrov stands for AaSyv duxrov (cf. Plut. Numa 15 éadwxores ioxvpav wat dduxrov dAwow), like Oepud AotoGar (1044), rodAHy yewpyeiv (Eccl. 592), rydv ryv paxpdv (Vesp. 106). dduxros is also possible. 1129 toopey rHv vixra Tacav: dor tows Bovdyjoerat kav év Aiytrrw trxelv &v paddov 7 Kpivat Kaxds. The text is right enough here, as far as I can see. But there seems to be a confusion of thought, which the editors do not clearly point out. The Clouds are threat- ening the judges with the consequences of an adverse verdict. In various ways and on various occasions they will visit them with bad weather, dare x.7.4. Now here the poet might logically say either of two things: (1) you will have such bad weather that for the future (rvyety and xpiva. must of course be future) you will wish to undergo anything and live anywhere rather than give a wrong verdict : (2) you will have such bad weather that you will wish yourselves removed from Athens, much as you love it, and transplanted to the rainless skies of Egypt. In (1) the climate of Egypt, in (2) 4 xptvas xaxaés has no proper place. What Aristophanes actually says is, if my analysis is right, a quite illogical mixture and confusion of the two things. But I shall be very glad if he can be vindicated by better interpretation. It is not irrelevant here to point out that the analogy drawn in lines 534-6 of this play between his comedy and Electra is somewhat faulty. In the first place Electra does not ‘come seeking’ for her brother, as the comedy does: it is the brother that comes to her. Secondly the comedy is to recognise its spiritual brethren, a capable and congenial audience, by something or other, as Electra recognised Orestes by the hair. The only thing on the part of the audience that could lead to such recognition under the 28 CLOUDS circumstances would be their applause and favourable reception of the comedy. But any one can see how very lamely this answers to its intended analogue, the lock of hair. What would really answer to that would be some quality or action of the audience as unconnected with the comedy as the lock of hair with the person of Electra. It is a further fault to make the audience akin (rdSeAgod tov Béorpuxov 536) to the comedy. The*audience may be akin to the poet, but not to his work. This criticism is not pressing a logical point too far. It is just in the finish of details that perfection of classical work consists, and any good judge must feel that the con- fusion here and in 1130 is a serious blemish. 1171 &s Woopal cov rpata THY xpotay iddv. viv pe y’ ideiv ef rpdrov éLapvytiKds. It is very unlikely that zpdéra and. zparov in successive lines are right. idwy and iseiy are awkward too, though in a less degree. 1366 éya yap AioyvAov vou.ifw tparov év rountais, podou wAdwv, afvararov, ordppaka, Kpnuvorosdy ; There should be no doubt that these two verses are a question, indignantly uttered by Pheidippides. In a statement made by Strepsiades the second line could hardly pass as ‘a stroke of humour, but there is no need to have recourse to transposing 1365 and 1366 with Hermann, Kock, and Thompson. To the precisely similar questions of indignant surprise quoted by Blaydes on Vesp. 1159 add the following: Vesp. 1188 éyw 8& rebedpyxa mrwzor’ ; ovdapot (for so it should be written): Soph. Aj. 1130 eyw yop dy wéayu Saidvev vopors; Plat. Rep. 344 © éyo yap olpat, py 6 Opacipayos, routi dAAws éyev: Huthyd. 297 B ddeAgds yap, Edy, eyd eiur EdGvdypov: Charm. 163 a eyo yap mov... Tod dportcynxa: Xen. Mem. 4. 2. 12 pi otv.. od- Stvapat eyo «.7.A.: Sosipater 1.38 ap. Kock 3. 315 and Strato 2b. 361: Menander Ilepx. 226 (Van Leeuwen’s text): Herodas 5. 4 éym “Apduraiyv ryv Mévevos Spy. | yuvatka: Lucian 67. 2. 2. Observe throughout the em- phatic éys. Probably the obscure passage in Xen. Symp. 4. 45 is to be explained in the same way. CLOUDS 29 1415 kAdovor watdes, marépa 8 ob KAdeuw Soxels ; Many have been the suggestions for adding on three syllables at the end to make this a tetrameter. I think the addition desirable, but I would not add anything like Cobet’s od xpivo. or Herwerden’s zpooyxew. Though some such word as xpivae seems logically necessary, the poet might venture to dispense with it under the circumstances : the parody is more ‘perfect without it. Possibly Soxé could be used for think right (Antig. 1102%) as ofua: and Hryotpae really seem sometimes to. be. Suggestions are rather idle, but I will contribute mine—icws 8€ or tows ovv, going with the next verse, dyes x.7.A. 1418 It may be that we should read rots yépovtas <évras> and simply omit rod véov. Cf. Ach, 222 pndé wep yépovras dvras expuyov “Ayapveas: Vesp. 277 1d ogupov yépovros bvros: Eur. Bacch. 189 émAcdjoped’ 75s yépovres ovres. WASPS » 65 GAN zor juiv NoyBiov yvdpyv éexov, tpav pev abrav odyi Sefuwrepov, Kopwdias St hoptixys coparepor. Perhaps Sefwwrépav and codwrépay, agreeing with yrvapnv. 318 GAAG yap obx olds 7’ cip’ dev. For ade, which has been seen to be nonsense here, read evdev. In Plut. Mor. 515 F eddew is now always read for the MS. ddev (otkor pév dda tupdAyv). So in Xen. Cyn. 6. 18 read ddioxerat for cipioxera. 564 of pév y' droxAdovrat weviay abrav Kat rporriBéaow Kax& mpos Tots otow, ews dvav av iodon toiow éuotow. So V: the other MSS. omit dav. Editors sometimes dvusv, but most of them give the passage up, or materially alter it. It presents three difficulties: (1) dvév or dvudy: 2) the position of dv, or, if we read dvodcy, its omission : {3} the strangeness of making Philocleon speak of his evils (xaxd) or miseries, when he is dwelling on the proud, almost royal (549-551), position that, as a dicast, he occupies. No real explanation of this last point is forth- coming. For the change from plural to singular see note on Ach. 683 above. After much puzzling over the passage I am inclined to make the bold suggestion, which will get rid of (1) and (3) together, that for dvuév we should read dyaois, imagining éuav to have arisen perhaps out of dv, as Van Leeuwen says, after dyafois was lost. dyafois gives excellent sense (‘makes himself out as wretched as I am enviable’) and is WASPS 31 at any rate in perfect keeping with 577 cat rdyadd po pépyno’ dxwv ddonets Tis EAAdSos dpxew and 601 oxépu & dxé tév dy a0 Gv oiwy(!) daroxAjets Kal kateptixes. For the antithesis of the sentence cf. Thuc. 6. 78. 3 rots atrod kaKkots ddopupbels tay’ av tows Kal Tols guois dyadots mote Bovdybeiy adOis POovica, and Dem. 29. 20 zpos rav évrwv ayabGy ipiy x.7.d. There remains the question of dy. Some might prefer to write éws...dvicdoy (or dvcwO7, as is needlessly suggested), but the omission of dy is very improbable, though perhaps in anapaests we ought not to call it quite impossible. In Peace 32 read éws cavrév AdBors, not AdOys. Of the two alternatives I would rather choose the other, that dv is out of its proper place. There is at least one other example of this in Aristophanes, namely Frogs 259 édcov ) ddpuyé av Hpav xovedvy 80 jpépas, not to mention the Doric ody &s médas ka ’xwvr. tat Tpijpees, Which seems a very probable restoration of Lysistr. 173. In Notes on Xenophon and Others p. 118 I have given many examples of this post- ponement of. dy with some sort of relative, but it must be admitted that the case of a conjunction like éws is not necessarily the same. Possibly éav déyafois dvicdon, trying to make them equal. 612 rovrowi eye ydvypot Kat py pe Sexoy | és oe Brepar. Editors «od py or kei pj (with dSeyoe). Perhaps xév and af with what follows. 922 For ws dvr’ ad woAv read dv7a 8) woAv. ad has no meaning here. Cf. the interchange of av and 57. 967 © Sapo’, éX€er Tovs Tadarmwpovpevors. Starkie in his admirable edition (1897) reads aiSot for the unmetrical éAée. Perhaps I may put on record that I made the same suggestion to the Oxford Philological Society at a meeting on November 2, 1894. I should prefer to regard éAda as a gloss on aidod and not as a corruption of it. With aidod the line reminds us to some extent of Nub. 1468 vai, vai, xaradéoOnre warpgov Aia, in which the last three words are of course borrowed from tragedy. There is a certain mock solemnity about both. For this 32 WASPS use of aiSeicGo: may be compared such passages as Od. 3. 96 pndé ré p’ aiddpevos perdiocoeo pyd’ édeaipwv : Eur. Med. 326: Hec. 286: J.A. 1246: Antiphon 1. 26 and 27. 1015 viv adre, Aew, mpoooxere TOV vour. adre, even in anapaests, may be doubted a little on the ground of rarity, but still more because the meaning again seems inappropriate here. Clouds 575 dedpo rév votv mpdooyxere suggests Seire, but that word too is absent from comedy and almost from tragedy. It might however find its way into anapaests, especially if there is any parody or imitation. 1022 joxjoas should probably be yvoyjoa, matching xéacOa. An infinitive is distinctly called for. 1112 8éfor re? The new point, independently presented, calls rather for 8é. 1193 wAevpay Baburdryv Kal xépas Kai Aaydva Kai Owpax’ dpiorov. Aaydve is sometimes read, but the greater difficulty, indeed impossibility, of the poetical form yépas for yxeipas in comic iambics is hardly noticed, except by Starkie. Is it not a case, such as often occurs, of words transposed ¢ Read xai Aayova (or Aaydve or Aaydvas) xelpds Te Kal. 1216 = ddwp Kara xeupds: Tas tpaméLas ciapépew. In this Barmecide’s feast (évvaviov éoruipeOa) why is the infinitive ciogépewv used? There is nothing to account for it Probably Lenting’s ciopépa is right, the subject being understood, 71s, the slave, etc. (It might be objected that the task would require more than one slave, but cf. the fragment (69. 2) of Plato’s Lacones ri ob tpéxwv od ris tpaméfas éxpépes;) For the verb so used cf. the first fragment of Xenophanes: Nv yap 8) Cdredov xafapdy Kai xelpes drdvruv kal KUAtKes* TAEKTOUS O° dpdiTibe oTeddvous, Where déyudirifa is a certain correction of dudurebeis. Still closer is Alexis 261 xara yewpds ed66n° tiv tpdmefav nw éxwv (1.2. Yxev & SodAos éxwv). The Homeric oivoxocver WASPS 33 (Od. 21. 142) and Theognidean oivoxocirw (473) are better known. 1397 Muprias | rijs ’AyxvAdwvos Ovyarépos cal Zworparys. The epic form 6vyarépos is scarcely possible, for there is nothing in the context to suggest parody or burlesque. Everything is plain and prosaic. Should we substitute yevouerys? Even that might be thought elevated in style; but Cratinus 328 has "Ioxopdéxov yeyovds in an equally prosaic fragment. Ovyarépos would then be a versified gloss. PEACE 24 For darep write dcarep, not with Blaydes drimep. In Plat. Rep. 390 B I have suggested the alteration of as to dca, omitting d. Dio Chrys. 1. 9 as is an erroneous variant for dcov. 83 pH por coBapds yapa diav. coBapds is more likely. Such words are much preferred, especially for verse, in their adjectival form. So Nwod. 406 déperar coBapds: Plut. 872 c&s coBapds... ciocAnrvbev. 139 TOUTOLGL TOIS avTOLTL TOUTOY XopTdcw. It is such a fixed thing in Greek that 6 airdés comes before, not after, otros (raita raira, not raira raird, etc.) that perhaps we ought to find some way of altering this verse. We may think of totrdév ye rots adroioe tovros xoptdcw, contrasting the beetle with Pegasus (135). Cf. however Thesm. 1060 év rGde tatTG yupiv. 180 = dBev Bpotod pe tpoceBaN’ ; Svat ‘Hpdxdes, Toutt Ti ore TO KAKOV ; Write 1éGev Bporod we tpooeBad’— ; vat x.t.d. Merry sees that the sentence is thus imperfect, not elliptical. Perhaps other editors see it too. If so, why not write it accordingly ? 208 iva. py Brérotey paxopevous buds ert. It Brérocer is right, it is a very unusual instance of the poetical sense and construction of BAérw. I doubt whether there is another in Old Comedy, except the doubtful xar’ airiy fv Bréras tiv elcodov, where 7 may be read, in PEACE 35 fragm. 388 of Aristophanes. Did he write émBdérovev or cigBAéroev here? There is no appearance of parody to justify BAérw with an accusative, but of course there may have been something of the kind. 273 } mpl ye Tov purrwrdy ypiv eyxéat. ovyxéat would seem a more suitable word and it is not uncommon for év and ovv to get confused. For 4 I would suggest as. In the argument to the Ajax should not # in }) mpos dvridiacroAny be as? 479 éoot y' abrav exovrat Tod EvAov. If this refers to the prisoners from Sphacteria, may not évéxovrat 7a ~vAw be suggested? The loss of ev after wv might lead to the corruption of 7¢ évAw. 605 mpata piv yap abtis hpSev Pedias mpdéas xaxas. Reversing the order of airs Hpéev, as the metre requires, I would then read airy for airjs and take jjipfev as used absolutely. 674 motos Tis oy elvat Soxet Ta ToAEMLKE 6 KXedvupos ; The tenses of.the context point very clearly to efvae ’Sdxet. Cf. pp. 178, 312, ete. 729 add tht xaipwv- ques 82 téws rade TA oKEdN TapaddvTes tots dxodovbors Sdpev ovlew. Certainly 8épev and rapaddvres cannot be right together. Cobet approved (N.L. 51) of Hamaker’s ride creviv émodvvres, in which an article would surely be wanted. oxevy is confirmed by 552, 886, 1318, and by ratra in 732. Blaydes reads rade ra oxevy xatabevres, though it ought rather to be xaraOéuevor (886). Since these are the first lines of a parabasis, the right word and the least possible change is aapaPdvres. So immediately below in 735 abrov éayver zpos 7d Oéarpov mapaBas év trois dvaraioros: Ach. 629 oitrw raphy pds 70 Oéarpov: Hg. 508 mpds 76 Oéatpov rapaBivar: Thesm. 785 Hpeis Toivey Huds airas eb AcEwpev tapaBaoau : Plato Comicus, fragm. 92 od av wapéByy eis A€éw roids? éxov. D2 36 PHACE 1070 e yap py Nuipar ye Oeai Bax ééardtackor, pdt Bdxts Ovytods pnd’ ad Nipdat Béxw abroy, (The sentence is apparently here interrupted.) The repetition of Nigga: Bax, ushered in with pd ai is unintelligible. It would yield a fair sense, if we read pyre. . pyr’ for unde. . wy’, taking the second line as repeating the first in a rather expanded form: ‘ that is, if neither .. nor. .’ BIEDS 63 otrw tT (or ovTw ’arl) Setvdv oddE KAANOV A€yeLv. You should not call us such dreadful names, 7.e. épviobjpa. Literally it is Sevov and not xadcv to do so. xdA\ov is the idiomatic comparative, used as in Plat. Rep. 410 p padaxdrepot yiyvovrat 7) ds KdAALov abrois: Ken. Oecon. 7. 25 od Kdkidv éote poBepay elvar tiv puxyv. odK dpevov is commoner. 79 TpoxtAros épuis otToat. I commented above on the mistake of calling Paphlagon Cleon. In its way it is equally a mistake and equally the spoiling of a joke, that this servant-bird (dpys dodA0s 70) should appear as a dramatis persona bearing the name of Tpoxidos, just as his master does that of “Evo. If he really were a trochilus, cannot the editors see that there would be no joke in saying of him here, rpoyiAos dpus otrooi? In the lines which precede he has been describing how he runs (zpéxe.) for this or that thing, and so Euelpides says, ‘why, he is a trochilus, this fellow.’ The one thing then certain about him is that he was not recognised by the spectators as belonging to that kind. He may have been any bird in the world, real or imaginary, except a trochilus. 480 oix dwoddce «7.4. should perhaps be made a question. 525 kay Tots iepois mas Tis eh’ ipiv épvidevris iotyot Bpoxous. iepdv is too often distinctly a building to mean here simply 38 BIRDS holy ground. év roiow dypois is often read. Rather nearer would be é rois épyos. For épya fields cf. Xen. Cyneg. 4.9: 5. 15, and elsewhere. But it is not a common Attic word. It has already been conjectured in 579, where a long a in éypéy is doubtful. 538 kdmetra katerkéoacay Oeppov toiro Kal’ tpav abrav Gomep KeveBpeiwv. ; For aérév, which means nothing, should we read wdvyrwv ? Cf. p. 42. Index, s.v. atroi. 604 Fw 3 apdrrwo’, ody tyteia peydAy toiT’ ori; cad’ tobe, K.T.A. In spite of 731 rAovOvycedav (sic), | edSarpoviay, Biov, cipyryy, where eiSatpoviav may well be a gloss on wAovOvyieiay, it is exceedingly doubtful whether the last syllable of tyea can be long. Innumerable cases of substantives in -ea from adjectives in -ys are against it. A simple remedy, if the words are wrong, would be to suppose this another instance of words transposed, a thing that often occurs in verse, as is proved by the metre, and to write odx byie’, Hv eb mpdrrwow, peydAy K.7.A. The words in 731-2 might also be rearranged, fiov, >. Ft > a , eipnvyy, eddapoviay, trovOvyieay. In 1536 (see below) the difficulty is the other way. 753 ei per’ dpvibuv ts dpav, & Ocatai, Bovrcrat StaAdkew Lav Adews 76 AocTov, ds TGs rw. No one that I know of has demurred to this, but the sense is surely most unsatisfactory. We want ‘If any one wishes for a pleasant life, let him join us birds,’ not ‘if any one wishes to lead a pleasant life with the birds, let him come to us.’ In the first place, there are not various communities of birds, between which the Athenian yearning for a pleasant time might make his choice. The chorus are the birds, all the birds that are, and there is no other set of birds elsewhere for any one to join. The chorus are not comparing themselves advantageously with other feathered peoples. The things they say of them- selves prove this, if proof were needed. Secondly, ‘if any BIRDS 39 one wishes to join the birds’ is not the thing to say. It is Aristophanes’ own fancy that there is a bird-community and that men might join it with advantage. He must not put the case of people having formed the wish to join it independently of him and before he made the pleasant, wise suggestion. He can only have said ‘if you want to enjoy yourselves, join the birds.’ This being so, I do not see how per’ épviwy can be right. It also strikes one that this is the only passage in Greek in which SiawAcxew is used without an accusative. It occurs some half-dozen or more times, but always with some word like Biov (Herod. 5. 92: Laws 8064) or jpépay (Aleman 16. 4). 76 Aowrdv here is adverbial as usual—so L. and S. and the editors take it—and not the object of the verb: I doubt whether it is ever used except adverbially. We may say therefore that, while per’ dpvidwy is not wanted, we do want an object for the verb. Then we remember the line (Wasps 1071) el tis tpdv, © Ocaral, rHv eunv idov dicw, and think that the line in the Birds may very well have run in the same way down to Oearaf; the missing words might come at the end. Finally we arrive at something like el tis tuav, & Oearai, BovrAcTas Tas Hyepas Stardéxey Lav Wd€ws 76 AouTdy, ds Huas ir. Of course no one would venture to say that is exactly what Aristophanes wrote. I only submit that it makes good sense and that the vulgate does not. [Van Leeuwen has made independently a very similar suggestion. ] 821 I think these lines may be fairly taken to mean ‘Is this Nephelococcugia of yours the place where the wealth of Theogenes and Aeschines is to be found?’ ‘The best account (xéAAicrov) is rather that this is the real Phlegraean plain, where the gods drew a longer bow than the giants.’ 1340 Zouxev ob WevdayyeAys el’ dyyeXos. So the MSS. All recent editors have adopted Bentley’s yevdayyeAjoey as well as the obvious dyyedos. No doubt 40 BIRDS Wevdayyedjs is a form contrary to rule and wrong, but @ verb in the future tense seems to me questionable, as, the news having already arrived, the future has to mean, not will bring, but will be found to have- brought, false intelligence. The quasi-parallels adduced by Van Leeuwen on Plut. 290 are far from convincing (Eur. Hipp. 1418 comes nearer). I would therefore suggest wevdayyeAqoal y’ dyyeAos. See Bast pp. 710, 727 on the confusion of v and y. 1536 That BaowAefa, who becomes the bride of Peithe- taerus, really is a personification of monarchy or sovereignty and not, as we might suspect, BaciAea the queen, appears both from the express statement of the scholiast (cwpa- Toro adits TO Tpaypa ws yuvaixa.., evo d& abryy dbava- giav xadovor) and also from a passage in Dio Chrysostom 1. (wept Bactrcias) 73 (67 R) foll., to which the editors do not seem to refer. Hermes there shows Heracles yuvaixa edad Kat peydAny and tells him airy cot paxapia Saipwv Bacwrefa, Avds Bactdéws Exyovos. As she has with her Aixn, Eivopia, Kipyvy, and Néjos, and as they subsequently find riv Tuparvida . . rpooroovpévny Kai ddouovotcay abriy TH Baowieig, we cannot doubt that the latter is really and truly the abstraction. This being so, two curious points arise: (1) how the personification of this abstract idea got so much into Greek mythology that according to Diodorus 3. 57 Basileia was sister of Rhea and daughter of Uranus: (2) how Aristophanes is able in 1536 and 1753 to give the word a short final syllable. 1710 olos ovre Tapdarys -dorip isety Aamwe xpvoavyc Sépw, «.7.d. The Greeks were much less particular than ourselves about avoiding the repetition of a word, as is indeed exemplified in this very passage by the occurrence of tndavyés in the next line close upon ypvcavye? in this. Yet it is hard to accept déue, when 1708 has just ended with 6ABios Sduors, especially as neither sense nor construction is quite satisfactory. I incline to think that it is an error due to the word ddpuos being in the mind or before the eyes of the copyist: an error, that is, of repetition, like many others. If so, some such word as ¢Aoyi was probably the original. LYSISTRATA 97 apiv déyew 8’, Suds Todt | émepjoopal re piKpdv. Toot TO puxpov would be much more usual, and the con- fusion sometimes occurs. 289 xSrws mor’ éSapmpedoopey Tour’ dvev kavOnAiov, ds enod ye TH EvAw Tov Gov eEurwxarov GAN’ duws Badioréov. There is no clear meaning for zoré here, nor for the causal as. Should we not read cal ras wore as a question? They are stating the difficulties; and this gives dA\’ duos more point. In Wasps 1222 dérws is probably a mistake for was, xaos being the answer to it, as dAnOes ; shows. 553 xdr’ évrygy réravov tepmvov tots dvdpdot kal powadt- opos. ‘évrnép B: evrééy RIC: evordéy Hirschig’ Hall and Geldart. But évrééy seems an illegitimate form, évryéy very questionable as to use, and évord& scarcely appropriate. Perhaps éuryféy) may be conjectured. Pind. Pyth. 2. 91 évéragav EAxos Gdvvapoy éa mpoaGe xapdig. and Isocr. 1. 46 ai Airat tais ASovats waparempyact are enough to vindicate the metaphorical use. 634 I shall be (says the chorus) another Aristogiton : autos ydp pou yiyverat Tis cots exOpas wardfon tiode ypads THY yvdbov. airé for adrds seems unmeaning, and rai7é is not much improvement on it. I would suggest mdyra... yiyvero in 42 LYSISTRATA the sense of ‘is everything needed,’ ‘carries the day,’ etc. mdvra and aird, wavrwy and airav, etc. are sometimes confused. See above p. 38, In line 486 of this play, xai pay adtGv rotr’ émbupod vi tov Ala rpita wbécOa, airar, which has nothing to refer to, has been very plausibly altered to zévrwv. 896 ddéyov pede oor THS KpoKYS hopovpéerys td tév dXexrpvdv ; ‘ The scholiast’s explanation of dopovpeévys by Siahopoupevys, Staozwpévys seems questionable. If gop. is right, it must mean carried off. Perhaps pOepovperys, since dépu, pOeipw (Stahepu, SiapOeipw) are easily interchanged. Thue. 7. 48. 5 POepeicOa. aitrav 7a mpdypara. But except in the impre- cation ¢Oeipov Aristophanes uses Oeipw only once, and that in lyrics (Av. 1068). 1150 ot Adkuwves... moAXovs piv avopas Oerrarav drddecay, moAAovs 8’ Eraipous ‘Inmiov cai Evppdxovs, Euvexpaxodvres TH TOO Huepa povor. fuvexpaxeiv is not found elsewhere, nor does it seem a probable form. Even if it were open to no other objection, it is not likely that Aristophanes would have used it immediately after fvyyzydxous in the preceding line. I take it that under the influence of gvppdyovus it has been erroneously written for another word, probably évvex- movoovres. This source of error has often been pointed out. Of course some other word, ¢.g. Evvexdpaudrtes, is possible. THESMOPHORIAZUSAE 103 «paid: for warpid:? apamis seems admissible in an Agathonian lyric, since we find it in Euripides. 148 eyd 88 tiv eoOAO dua yvauyn hops. Can yvopuy dispense with an article, or should we read dpa tH yvopyt THI is very like CN. 217 } py 'midi8ovae “pwavrov ddedov more. # Should, I think, be 7. Of Hecl. 145 por ay yeverdv Kpelttov jv. 679 avtav dtav AnOy tes K.7.A. Perhaps rovrwv for airév, which ought not to begin a line, especially after a marked pause. 794 Sober nipere ppotdov 76 xaxdv kal pH KarehapBdver’ évdov. évSobev and évbov together seem weak and unlikely. We might read oixo6erv. 795 «av xaradapOdpev ev ddXorpiwv wailovcas Kal Komacat. It is doubtful whether év jyerépov in Herodotus and és hpergpov in Hymn. Herm. 370 (perhaps Od. 2. 55) are enough to guarantee this in AtticGreek. Has év éAXorpiv ever been suggested? I have also sometimes thought of év dAAj wv. 802 pos exacrov. 1 xa’ exaorov, as in Birds 564. 870 py Wedcov, & Zed, rs émiovans éAridos. AS émiovons has been questioned, it may be worth while 44 THESMOPHORIAZUSAE to point out that 4 péAXovea éAwis occurs once or twice in Demosthenes (18. 89: (26. 21]) and elsewhere. 1061 Should xairy be xairG? There does not seem much point in the former. 1181 hépe Ooiudriov dvwbev, & réxvov, Todi. dvebev being unmeaning and a verb required, we natur- ally think of dvafapev, but it would have to bear some meaning that the dictionaries do not exactly specify. 1189 AaBE Goiudriov gives the reverse. FROGS 301 i fyrep epxet. Sedpo Seip’, & Séocrora. It seems to have escaped notice that i6’ rep epye: is not addressed to Dionysus (nor spoken by him to Xanthias, as Van Leeuwen gives it) but to the imaginary goblin Empusa, with whom the slave is terrifying his master. This appears clearly from the parallel passage Lysistr. 833 & worvia Kimpov cat KuOypwv xai Tadov pcdeovo’, 10’ dp0iv yvrep Epxet THY Gddv. 369 tovros atd0 KabOis dravdd Kathis Td tpirov pdr’ drravda 3e/ 7 a eLiatacba ptiaraiot xopois. rovros avd@ V and the text of Aulus Gellius: rovros dravdd R and most MSS. Two things are clear: (1) ai8d and dzavd cannot be combined in one and the same sense : (2) dravdd forbid is quite out of place. It is ludicrous to suppose that dmav86 ééiotacbor is a confused expression for dzavdd (4%) gurodar eivar. I cannot see any plausibility in rpwvdd (ter) read by Blaydes and after him by Van Leeuwen. The true reading really stares one in the face : Tovrots avd Kadfis éravdd KavOis Td Tpirov par’ éravda. éravda tell again does not happen to occur elsewhere, but that is quite immaterial. dyravdé seems to occur only in Soph. Hl. 1478: xarav8o only in Ant. 86: cwavdd only in Ag. 943 and once in Theophrastus. émi and dé are quite easily confused in compounds: thus in Thue. 5. 50. 1 drouéoo. must be corrected to éropdca (Stahl). 46 FROGS 456 Soot pepunped’ eb- oeBy re Sijyopey Tporov K.T.A. Perhaps gcot ’nepujmed’, had been initiated and then went on living (Stjyopev). 467 Ss rv kiv’ Hpav e&eAdoas Tov KépBepov daqéas &yxwv KdroSpas dxov AaBav. He did not drive Cerberus out; he dragged him out. Should we change éfeAdcas to éAxioas ? 504 H yap Oeds a” ds ervOe0" HKovt’, ebBéws émettev ApTous, K.T.A. The goddess and her servants cannot have known already, for Heracles (Dionysus) has only just arrived and the servants only just become aware of it (503 & iAraé’ axes “Hpdxdes;). Read then jéovra, when she heard you © were coming, which is in every way better. 553 — xal kpéa ye mpos Tovrourw dvdBpacr’ elkoow dvnpiwBodtaia. So apparently the best MSS. with others ; some however av’ HptwBorr.jaia, and in this form Pollux quotes it 9. 64, if Bekker’s text may be trusted. Both seem impossible. ‘At half an obol apiece’ is either qyiwBodrawta or dy’ jpiwBddiov. Van Leeuwen zav@’ for dv’. Did Aristophanes write cixoow ev jytwBortaia with a comic exactitude as to the number ? cf. 49 kal KaTedtoapev ye vais Tov Todepinv } Swdex’ } tTpeis Kat Séxa. and 1129 rovrwv éyew peyeas 1; EY. tAcly Sddexa. Cf. Merry on Od. 9. 241 otk dy rév ye Svw Kat elo dpagat... dm’ obdeos éxNooaav. We may also think of Gp’ or Hv. 647 AIA. xai 89 ’mdraéa. ATO. kata was obk éxtapov ; Perhaps od8’ érrapov. FROGS 47 689 kel ris Huapre ohadels Te Bpvixov raraiopacw, éyyevérOoan pypi xpivat Tots GAtoBotcw Tore airiay éxOeioe Moat Tas mpdTepov dpaptias. aiziay éxOeto. is unexplained. ‘Setting forth a reason why ’ gives no good sense. That would be justifying or at least explaining their offence, whereas Nica: tras mpdrepov duaprias (note zpdrepov) clearly contrasts subsequent con- duct with earlier. They are somehow to redeem, not to extenuate, their error. We want something roughly like the words of Thuc. 2. 42. 4 xal yap rots raAAa xelpoct Sixacoy Thy és Tols Todguous trip Tis watpidos avdpayabiar mporibecOar, or 1b. 3. 56. 6 Sikatov jay tis viv dpuaprias, ei dpa npdptytai rt, dvribetvar thy Tote TpoOvpiav. This being so, Herwerden’s airfay éxdto1, adopted by Van Leeuwen, is no improvement. Proceeding on the occasional confusion of airios and détos (corrections of this kind, a£ for aur or ait for af, most of them quite certain, have been made in Ach, 633: Andoc. 2. 12: Lys. 30. 35: Menand. 427: D. Hal. A.B. 19. 16: Clem. Strom. 2. 14. 435 P: Dio Chrys. 31 (14). 12, and 34. (17). 23: cf. Xen. and Others, pp. 254 and 314) I think it possible, though I would not say more, that dé’ dvrexOcio. may be what Aristophanes wrote. Cf. the dvreivas in Thuc. 3. 56. 6. dvrexOetor would mean of course, not that they had performed worthy deeds, but that they had produced or set them forth as arguments in their favour. éxri#évac in this sense is familiar in Aristotle, and we have the testimony of all the MSS. toithere. Cf. Plut. Titus 21 civ’ Adpixavod Sxymlwvos éxrBévres mpadryta Kai peyaAopvyiay and Ages. et Pomp. Comp. éxxeipévav trav Biwv. The éxbeows tpdcbev tov érw- vipov is also to some extent in point. That Aristotle happens to have the phrase airias éxriGecOa is not very important, if the phrase is shown to be unsuitable in this place. The poet is arguing that the offenders should be allowed a chance of retrieving their former fault, not of showing how they came to fall into it. But I confess we should have expected a word expressing performance rather than a somewhat scholastic term for statement. It is of course natural to suggest the simple dyriGetor, but then how would the ex get in? 48 FROGS 788 ekvore prev Aioxvdov, dre 84 KarqAGe, KaveBadre Thy Sekidv, Kaxelvos trexapynoev ait@ Tov Opdvov. It is quite impossible that éxeivos should refer, as some scholars suppose, to Sophocles. First, the pronoun must mark a change of subject: the subject of ééBade could not be continued with an emphatic éxetvos, not even to dis- tinguish him from Euripides. (That would require éxeivos pév followed by (say) Evpimidys 5é.) Secondly, ioxwpeiy expresses retirement from ground previously held, not a refusal to put forward a claim. Sophocles could not ézoywpety from what he did not own or occupy. The only possible explan- ation then seems to be that we must press the sense of id in trexdpyoe and make it mean that Aeschylus partly withdrew, made room for Sophocles beside him. So the scholiast tells us to take it, but it must be admitted that troxwpo does not seem limited elsewhere to a partial with- drawal. 814 foll. In the four stanzas here beginning, and too long to quote, I cannot doubt that Dobree was right in thinking that 822-825 should follow on 817. ¢pifas 8 airoxdpov «.t.r. is strictly continuous with éupara otpofy- cero. It is most awkward for 818-821 to come between the two passages in which Aeschylus is described, and it much obscures the subject of gpigas «7.4. But I think Dobree was wrong for once in putting 818-821 after 825. They should stand at the very end after 829. First come lines (814-817, 822-825) which depict Aeschylus bearing himself in the fray like a Titan or a mighty beast of the forest: then a stanza (826-829) setting forth the deft and subtle art of Euripides ; finally one (818-821) which gives the conflict and the contrast of the two. In a note on the song in As You Like It iv. 3 Aldis Wright says ‘in the folios the last stanza is printed as the second,’ printing himself from a MS. in the Advocates’ Library. In 826 read &Oev 3€ on the other side, not w&Oev Sy. 835 & Saipovi’ avdpav, pp peydda Alay Aye. héyew peydda should mean use boastful, arrogant language. FROGS 49 But the words to which it is applied here are nothing of the kind. Euripides has said of Aeschylus dmocvepvevetrat mparov, wep ExdaTore év tais tpaywdiaow érepatedtero, and to these words peydAa Aéyew seems quite inapposite. A few lines previously however Euripides has said some- thing which might well call forth the rebuke : 2 Ry AF , s f KPELTTWYV Yap Elvat pnp TOUTOU THV TEXVYV. Should 835 follow on this? 905 GAN’ ds tdéxoTa xpy Adyew: odTw 8’ drws epetrov > ms MN ae Sp) vA ¥ doreia kal par’ eixdvas yO’ of’ dv dAXos etrot. No edition that I have looked at explains'cixévas properly. (See however a note of Frere’s on Birds 970.) Thus Kock Der Chorfiihrer verlangt dasz ste unverbliimt die Wahrheit sagen: Blaydes imagines, similitudines. Quibus indul- gebat Aeschylus multo frequentius quam Euripides : Merry doreia implies ‘smartness’ and ‘neatness’; either of which would be lost by the use of metaphor (cixéves) or commonplace. But any one who remembers what Aristotle says with truth about metaphors (Poetics 22. 1459 a 7), that a good use of them is a mark of fine natural parts (cdpvia) and not a thing to be just taken from somebody else (wap’ dddov Aafeiv) must be surprised to find meta- phors put side by side with things oi’ ay dddos etro, Moreover the scene that follows is full of metaphorical expressions, e.g. 941 foll. The truth is cixdves here are not metaphors, but comparisons. The reference is to the rude humour of comparing a man to so and so, usually some- thing ridiculous. This seems to have been a rather favourite, if unmdnnerly, pleasantry at Athens. See for instance Wasps 1308 ely’ abrav ds 18 -qxacey Avoictparos, 2 a o ov: i (Bpvyt Kock OLKAS, @ WpeTNvTaA, veoTAOVTW TPVYL vye Koc as, & mpeoB HUEY » Tpuyt (Ppvy ) KAnrnpi tT eis Gxuppov arodebpaxdtu. 6 8 dvaxpayav dvryKac’ abrév mdpvoTt a n~ a f > J Ta Opia rod TpiBwvos dmoBeBAnkore BWevédrw re 7a oKevdpia Staxexappéever, E 50 FROGS with 1413 also: Av. 804-7: Ken. Symp. 6.8 ov pevrot Sewds af... cixdfew... GAN’? Guws... cv abrov py eixale: Plat. Meno 80a the famous comparison of Socrates to a torpedo-fish, followed by 30. yryvdoxw obmep Evexd pe qeacas. MEN. rtivos 8} ole; 30. iva ce dvrexdow. eye dé TovTo olda wept wdvrev Tay KaAdv Ste xaipovow eixaldmevor : and Plat. Symp. 215 a the still more famous comparison of him to a silenus-image: Aristotle De Gen. An. 4. 3. 769 b 18 8 woAAdKis of GKwmrovres cixdLovot TOV jun KaAdv évious Tous pev aiyt dvowvre wip, tovs 8 otf xKupirrovre: Tryphon on eixacpds in Walz’s Rhet. Graeci 8. 751: Hor. 8.1.5. 56 foll.. What is meant in the Frogs therefore is that great men like Aeschylus and Euripides are not to indulge in these easy and often vulgar jokes at one anothet’s expense. They are to contend in a way more worthy of their genius. 949 adN Breyer H yury ré por x@ SodAos ovdey Frrov xe Seordrys x} Tapbévos x7} ypais av. I can never read these lines without wondering whether xa Seordrys should not be rot dSeordrov, or possibly 7 Seordrns. ‘The master’ has no business to come in third and on a level with the rest without even an icws or an e€ tons to carry it off. Also ovdéy Frrov might stand if there were no x@ Scomdrys there, but with those words surely it is very odd. 964 -yvdoes dé rods rovrou Te KapoUs Exatépov pabyrds. Perhaps these words should be datives, rots . . . wa@yruis. Cf. Eccles, 780 yvioe § dad tév xeipav ye Tay d&yaApdrov and Prom, V. 51 éyvexa toicde. 1028 éxdpyv yotv jvix’ Arovea wept Aapeiov tefvecros, & xopos 8 ciOis rw Xelp’ wdi ovyKpotoas etrev iavot. Many considerable and improbable changes of this im- perfect line have been proposed. But two slight alterations will put it right. In the first place we must adopt wapdé (Welcker) for zepi. The mistake is easy and has occurred in several places. In the Persae nothing is heard about Darius, but the dead Darius appears to Atossa and the FROGS 51 chorus, and they do hear from him about the further mis- fortunes of the Persian forces. Certainly we find no iavot of the chorus at that point, but there is an ode of lament- ation, and a little further on, when Xerxes returns in person, there is an abundant use of éroroi, aiat, oot, oiotot (954, 966), in iy, id is, to which iavot comes sufficiently near, The MSS. of Aristophanes may have preserved the word better, or it may be the comic poet’s own semi- burlesque variation. There is no great difficulty in his attributing it to the wrong scene, even if we felt sure that the chorus did not really utter it earlier, when the ghost of Darius spoke. The scholiast, supported by the Argu- ment, assures us that the statement in 1026 as to the Persae being later than the Septem is the reverse of the truth, and that would be a worse mistake than the other. But, supposing this less serious difficulty to be got over, what are we to do with jvik’ yxovea, words which will not even scan? Surely we are to read something like jvixa y’ ykovoay, the subject of the verb being the chorus and Atossa with her attendants. (On the Greek stage great persons constantly have attendants, and even humble ones have slaves, although their presence may come out only accidentally or not at all.) What delighted Dionysus was not the news, not what Darius said, but the state of grief and alarm to which the Persians were reduced by it. qvixa 7’ Hxovaay and 6 xopds 7’ evOus would also be possible, if ye after yoty is thought objectionable. 1054 Tots wey yap Tadapioww gore SiddoKxados darts ppaler, Trois yBadow Se roenrat. The construction and meaning are usually more or less misunderstood. ‘ Boys havea teacher who tells them things, grown men the poets.’ dors and not ds is used because of the éort. ear doris, not éorww ds, is regular, and occasion- ally we find the same phrase with a substantive added, e.g. in the pseudo-Platonic aepi dperjs 378D eicty ypiv Soxipacrat ofrwes Spvres kpivovor. Cf. p. 71 below. With mowmraé understand not cict diddcxadro, but either eioi— oizives Hpdovow or dpdfovow alone. E2 52 FROGS 1180 i6 3) Ady’> od yap podariv dAN’ dxoveréa ” TOV cdv TpoAdywr THs 6pOdryTos THY ery. It had occurred to me that dAX’ dxoveréa might conceal d\A& yevoréa. I should however hardly have mentioned this (cf. Plat. Symp. 198 B rot xdAANous toy dvopdtwy Kal byudrov ...dxovwv;) if I had not subsequently noticed the mysterious scholium yvdvat mpd Tod dxotcat. yvavat points to some other reading, unless the scholiast really wrote akodoat mpd Tov yvavat. 1202 = zote’s yap otrws dor’ evappdrrew array, \ f % 4 x , kal kwodpiov Kal AyKvOtov kal OvdAdKiov, év rots iap,Betorcr. That @vAaxoy (Fritzsche) should be read for @vAd«iov to get rid of a final tribrach seems to me proved—if proof were necessary—by the frequent use of AxjKvOos side by side with Axnxd@ov in the passage following. This occurs repeatedly where the defender of @vAdkioy would have to contend that AnxvOov was admissible: 1214, 1224, 1227, 1231, 1234, not to mention’ 1216 where R has the deriva- tiveform. These cases prove that the poet was ready to use AyK«vOos rather than put AyKvb.ov at the end of a verse. I have sometimes wondered whether Ayxv@ov, on which everything afterwards turns, is rightly placed as the middle word in 1203 and whether it should not stand first or (as Aj«vOov) third. 1415 Tov erepov Aa Bay aret, Owdrepov ay kpivys, iv éAOns pay paryv. In 1163 Hirschig pointed out that éd6civ is a mistake for jxev. The point of the whole passage there is the distinction of the synonyms #xew and xariévar, and this is lost if éAGciv is stupidly substituted for jew (EAD ety pev eis yqv é@o6’ Or pery watpas). Here in 1416 is it not clear that the same error has occurred? The sense should be ‘that you may not have come here for nothing,’ and this is ty’ yxys py pdrnv. The meanings of iy @\Oys are (1) ‘that you may come,’ (2) ‘that you may go.’ The first would be unmeaning. As to the second, the sense of ‘go’ is comparatively uncommon, and FROGS 53 indeed hardly possible where any ambiguity would ensue. Also pdrnv can only apply to the coming, not to the going : ‘that you may not go away for nothing’ would be nonsense. It may be pointed out on the one hand that MSS. some- times vary between the corresponding parts of 7xw and HAGoy, e.g. Thuc. 1. 18. 2 FAGe and jee: 8. 65. 1 FAPov and . feov: Xen. Hell. 1. 3. 11 the codices #AGev, the papyrus fragment jxev : and on the other that in the scholia we find parts of 7xw glossed by the corresponding parts of 7AGov, which fact suggests how the latter may here and elsewhere have driven out the former. See for instance the scholia to the Plutus 357, 828, 841, 1179, 1190, 1201 and to the Prometheus 284, 299. They take such straightforward forms as jets : AGEs (Plut. 357). Of course in Attic the two words are in most cases not really equivalent, but this the scholiasts did not know. [I think now that the vulgate may probably be right. It closely resembles Od. 3. 315. uy row xara dvta padywou kripata Sacodpevor, ov S¢ tyvaoiny dddov €AOys, where éAOys certainly means have come. For aorists so used see 1. 1. 555 py oe wapetry and 10. 5388 py te wdfwo., and optative Od. 21. 395. Plat. Theaet. 189D ratr’ otv iva py parnv Oappions ddinus seems an Attic parallel. | ECCLESIAZUSAE In the opening speech of Praxagora ydp in line 3 is not easy to understand, and in line 6 éppa x«.7.A. occupies an impossible place in the middle of descriptive lines. The imperative cannot have been meant to stand between the statement éyes and the statement SyAoctpuer, nor dyAdcopev to introduce two lines only. 6 would come well after 1 or 2. But any change is open to the objection that it leaves three consecutive sentences (3, 4, 7) with a ydp. 44-5 jpév is very awkwardly placed. Should it change places with oivovt (Meineke 4 pyv for quav). Cf. on 286 below. 171 toils Geois ev edxouat “~ - + z tuxelv kaTropbwcaca Td. BeBovrcvpeva. This must be the beginning of Praxagora’s speech .in character as a man. The sense and the antithesis of pev... 5é show this. But then the feminine kxarop$décaca is wrong. The error is, I think, one that occurs elsewhere, a confusion of nominative singular feminine with dative plural masculine, there being only the difference of a vowel. Read xaropOdécact. (So for instance Bryan cor- rected Plut. Marcell. 7 (end) rovros pev jv cipjvy perpiov tvxotca to tvxotct.) The action is much better ascribed to the gods than to the speaker ; indeed it could hardly be claimed by the latter without some arrogance. On tae other hand rvxeiv, as applied to the gods, seems to me very questionable, and I should be glad to find a parallel or an alternative. ECCLESIAZUSAE 55 179 érérpepas étépw' mAciov’ ert Spade kad. As these words are addressed to the people (iptv 174, SeSoixare, dvTiBorct’ 181-2), we should probably read ererpewad’. 253 Praxagora will say of Cephalus first aapadpoveiv, then pedayxoAay, then 7a, TpvBALa Kak@s Kepapeve, THY 5é mod ED Kal KaAds. But this last would be complimentary and out of keeping with the other statements, so that it can hardly be right. Our doubts are strengthened by the fact that ed xai does not appear at all in the Ravenna MS, I suggest 7a rpvBAta Kadds Kepapevety, THY 5é woAW KaKOV KaKas, in which the reason for the loss of xaxév is obvious. Kaxds Kakas, kadds kaAds are familiar Aristophanic phrases. 280 Perhaps we should read ovrws or eivs for aibray, which is very obscure. 282 as elwf” éxet Tois py mapodow épOpios eis THV Tikva bramotpéexew exover pnoe watTadov. This use of eiw6e, ‘it is a usual thing,’ seems unknown. In Plut. Humen. 15 pydevds éumodav dvros, dorep elwOev dvtt- KaOnuévev toAeuioy we understand domep ciwhey eurodav elvai 7t.. Also éxet is entirely superfluous, the Pnyx being mentioned by name in the next line. Did Aristophanes write ds ciw6ds dv? Cf. Eur. Hec. 358 otk eiwfds dv. 286-7 jas is much too emphatic as it comes. Read perhaps yas det peprvnwevas... A€yew. Cf. 44 above. 469 ov b€ ye vy Ata 8pa raid’, iy dpioras Te Kal Kis dma. I do not think the context supplies any meaning for Spa rad’, and—what is more—v7 Ata cannot be used with an imperative. Spas taira may be suggested, raira being ra THs Toews, attending assemblies, etc. The point is his contingent reluctance to do something which it is at present his very object to enable himself to do. Cf. 56 ECCLESIAZUSAE perhaps for the corruption Hq. 1019 (8p for Spgs) and for the sentiment Soph. fragm. 669. 581 GAN’ od péAAew, GAN’ Ere Oar kai 89 xpijv Tais Savoiats, és 7d raxtvew xapiruv peréxet mreiorov maps. Toit Oearais. In 581 d\Ad rérecOor and rijs Stavoias have been suggested, Bentley who made the former suggestion afterwards acquiescing in the latter (Blaydes): Perhaps dX’ émféoGar may deserve consideration. Cf. such passages as Plato Gorg. 527 D ériOnoopeOa Tots roAuTiKots. 622 Fighting epi rod pi) évyxarabapOety is surely not a possible expression : the negative has no business there. Is yy a mistake for 54? 628-9 Instead of inserting some words like kai rots peydAos in 628 to make up for the necessary removal of of davAdrepo, I suggest that 628 should end with raior yuvaéiv, and that in 629 between mpiy and tots alcxpois three or four syllables have been lost. Such a halting line as ratot yuvarti, mpiv ay tots aicxpots Kai Trois puxpots xapiowvrat, though it has Elmsley’s authority, is not one we can ascribe to Aristophanes. 838 re for ye? 1070 rotro and rovrov cannot be right together. Read perhaps totr’ ad, zoAd Totro 70 Kakov éfwAéorepov, the second rovro emphasising the first. 1091 Possibly dydorépats for dudorépas, but I think it is hardly necessary, though a difficulty has been felt. PLUTUS 119 - x 8 = ida > , i eee 2 6 Zeds pev odv eidas Ta ToUTwY pap’ ep’ Ei mvOour’ av emitpivere. There is slight MS. authority for o78’ és instead of cides and for éry or éry ei instead of é’ el. The latter of these is certainly to be rejected and the former to be received. With oid’ és the sentence runs quite properly except for ra tovtwy papa, which is , Wrong in sense and doubtful in grammar. Read 7d y’ otro papa, and every- thing comes out clear and correct, 2.2. 018 ds eu’ émerpipecev av 6 Zevs, ci tay’ otTw pGpa zvOorro. The order of words is involved, but not more so than in Clouds 246, Thesm. 811, and one or two other places, 262 6 deordrys yap prow tyas idews dravras puxpod Biov kal duoKdAov Ljoew drahAayevras. XOP. éorw de 8) ri kal wé0ev 76 rpaypa Tous’ 6 dyow ; 265 KAP. éxwv ddixrat Sedpo mpecBurny tiv’, & rovypoi, pyTGvra, kuddv, OALov, puodv, paddvra, vwddv" ola Se v7) Tov otpavdv Kal ywAdv avtor elvat. XOP. & xpvodv ayyedas éxay, was his; wddw dpdcoov foot. SyAots yap abrév owpov Hee Xperia éxovra. 270 KAP. mpeo Burucay bev obv Kaxav éywy’ exovra owpdv. XOP. piv d&ots devaxicas Has draddaryivat aLqpuos KT, Two things strike us as strange here. One is that the chorus, on hearing of the broken-down old man whom Chremylus is bringing with him, should break out rapturously with & xpvcdv x.7.A. The second is that they 58 PLUTUS should then so abruptly change their tone in consequence of line 270, which adds nothing to 265-267. Are not 268-270 out of their proper place? & ypvadv «.7-X. should follow immediately on 263, in which the chorus are promised a complete change of life. (adéAw gpdrov pot refers partly to the words just spoken in 261 ovKovy wdAat dyrov Aéyw ; od 8 abrds odk doves.) Carion corrects their joyful expression with apecBurixdv «.7.r., and in answer to their request (264) for further elucidation explains apec Buricav xaxav by the lines 265-267 about the zpeoPurys. Then in indignation they address him pév détots x.1.X. 523 tis yap tAovTaV eedycer xivduvedav wept THs WuxAs THs abrod ToUTo TOA! ; The Ravenna codex has xvdvvevev, which I would accept with the change of zojoa: to zoujoas. xwduvevo with a participle is more proper than xwdvvevwy appended to a verb. (In Gorgias 521, which is a little different, I would read xwdvvetowv, but see Laws 649 E-6504.) Cf. the certain correction of Republic 333 & abeciv éuroujoa to AaGeiy . . Eumroujoas. 891 as 84... Stappayeins. as with optative of wish is doubtful for Aristophanes. cai 67 seems not unlikely, as ws and xaif are sometimes confused. Cf. Index, s.v. «ai. THE GREEK COMIC FRAGMENTS I should have fused the three sections of this paper into one, had not I. and II. been written considerably earlier than III. and on a different plan. Passages of a single author may therefore be found in two or even in all three sections, L A Fracment of Epicharmus (268 in Kaibel) is quoted by Clemens Alexandrinus Strom. 6. 2. 21 side by side with a saying which he ascribes to Chilon. Chilon said éyyva, mapa 8 dra: Epicharmus is reported to have written éyyva dras (or dra) Ovydrnp, éyyta 8 Capias. Lorenz (Hpicharmos, p. 264) says this gives good sense and need not be changed, but, as the three last words are evidently the end of a trochaic tetrameter, we shall rather agree with Ahrens (De Graec. Ling. Dial. 2. 461), who tries to throw the whole into metrical form—éyyvas dra ori Ovydrnp, eyyta Se fapias. It is hard however to see what meaning he attached to these words. The sense obviously requires an inversion of the relations thus stated: guarantorship, suretyship, giving bail for another man (éyyva) is the child of infatuation (dry) and loss is the child of suretyship, (‘he that hateth suretyship is sure’). With this much improved sense the line appears in another conjectural modern form, the exact history of which I do not know: éyyta Ovydrnp pev dras, éyyvas 8& fopia. It is plain however that the words are still not quite right, for there is no proper correspondence in the pe and dé. Any one who began with éyyva dvydrnp péev dras must have gone on with another predicate of éyyva (udrnp 8 Capias, tikxrer dé tapiav, or something of the sort). It follows that we may confidently restore . . éyyva pév dras, éyyvas 8¢ Capia as the real form of the verse: the only question is as to what 60 THE GREEK COMIC FRAGMENTS preceded éyyva. In the first foot the anapaest Guvydryp is probably inadmissible, as in Attic tetrameters; could réxvov stand there as a trochee? Not in Aristophanes ; but Epicharmus has such lengthenings: e.g. Adyvos (35. 8), paxpés (57) and ra mpd rod (253). Téxvoy may therefore be the missing word. But, of course, there are many other possibilities. For the general form of the sentence compare fragment 148 A. & pév Buatas Boiva (2) ex 88 Goivas moots éyévero. B. xapiev, ds y' enol (Boxe). A. é« 88 réctos KGpos, éx Kdpou 8 éyével” tavia, x.t.d. Teleclides enumerates in well known lines the things of which the Athenians have made Pericles master (Meineke 2. 372: Kock 1. 220), moAewv Te hopous airds Te modes Tas pev Seiv, Tas 8’ dvadver, Adivafre’xn Ta pev oikodopely, Ta SE abTa wéAw kaTaBddAEv KT. , Kock gives a long list of the changes proposed for ra Se aird, none of which is satisfactory, and himself suggests 74. Se répradw ad. Perhaps ra dé wdvra would do, zavra going, of course, closely with xaraBdAAew. For aitrd—ravra, cf. Index. The pe and 8 show that different fortifications are meant, but most of the changes ignore this. T[déAw is contrariwise. I have noticed two fragments of Plato where a slight change is needed to restore the poet’s hand. We find in trochaic tetrameters (M. 2. 620: K. 1. 605) my . e € F € a n elgaow jpiv of vopot TovTowwt Totot Aerrots dpaxvioss, Gy rotor Tolxos } pddayé tpaiver But the second syllable of dpaxvios is short, and Porson therefore wrote Hetv of voor elxact Tovrots Toiot Aemrois dpaxviors ay rotor Toixos y Pddayé idaiverat, (cf. Xen. Mem. 3.11. 6), which Kock modifies into etfacw hyty of vdpor | rovrourt Trois Aewroiow dpaxviosw & | x.7.A., While THE GREEK COMIC FRAGMENTS 61 Meineke in his larger edition and Cobet (Obs. Crit. in Plat. Com. Rel. p. 61) seem content with the original reading. Meineke in the smaller edition and other scholars have suggested dépayvidious, and in the line of Cratinus 190, épayviwy peoriy éxeis THY yaorépa, they would read dpayvdiov (M. and K. Gp’ dpaxviwy). This is ingenious and may be right, but I am inclined to think that after Xerrois the article rois was lost through similarity of letters. We thus get rovrourt Toit Aewrois | dpaxviors av rotor Tolyous H padayé ipaiver without any further change of reading or metre. Against Porson I may appeal to an emendation of Porson’s own (Adversaria p. 41) in which he put right Ign % aN , > > if idov yap airov mpéoBvv otk qverxeTO XN 2 Xx x wo 2 a m, fy ob 7d Aourdv dveow ev POirois exe by writing qvécyero pi od. The other passage is a fragment of the Hyperbolus (M. 2. 669: K. 1. 644) 2))\? & 74 s rn ¢ , ry\Z GAN’ Srrdre pev xpein ‘ duytadpynv’ A€yev, x ‘§ , dey 3 2 Om SZ édace ‘Oytwpnv, drdre 8 ciety b€ot 627. 7 OS\7 90 éAtyov,’ ‘Acov' EXeyer. The rhythm of the last three words is so halting that we may surely put in an dy to help it, ‘ ddéyov,’ ‘éAlov’ edeyer. It is very doubtful whether Aristophanes would use Aérw with an accusative of a thing seen or looked at. (Peace 208 above) is a poetical use, and if it occurs now and then in Menander we may remember that his Attic was said not to be of the very purest. In later Greek it becomes common. When therefore we find in a fragment of the Nijoou (M. 2. 1109: K. 1. 493) # ‘ i 2 ON 7 -~ A. riot A€yets; ciotly 8 Tod; B. aidl kar’ abriy 7) BAgras THY eicodov, we are warranted in conjecturing 7 BAéres or iv BAérys. A fragment of Alexis deals with the conditions of pleasure : 62 THE GREEK COMIC FRAGMENTS 2 2 ~ #: ~ ~ gy ovK tore Tals TWAcio Tao” THY TEXVOV OTL > > I o an c a ovK apxiTextwy Kiptos THs OOVTS 4 ft ? 2 x x nw t povos Kabéoryk’, GAXAG Kat TaY xpwpEevov - - Aa a nw vo oupBdAXcrai Tis, av Kad&s xpovrat, pepis (M. 3, 451: K. 2. 351). Read ody dpxiréxrwv with Bothe, but this is not all. DSupBdrdrgerat is not active in meaning, but passive with pepis for its subject. He does not mean that some of the people contribute to their own pleasure: it is some of the pleasure, which is contributed by them. But then rév xpupzevov is left without construction, until we turn xai to xéx. It would also be possible to read p¢pos, making oupPdddcrar middle. A slight and obvious error in a line of Philemon occurs in M. 4.11: K. 2, 486 2 a a} , S. Os. émovd yap éort Kuptos mev eis avyp, -. x * a ¥ >” , tovtuv 8 Kai cod pupiwy 7’ dAAwv vépos, where ydép and pév should change places. Of course éuod pe is contrasted with rovrwr dé, and pev is absurd where it stands. Cf. on Cratinus Junior, p. 66. In a verse from Philemon’s Ilaides will be found peréyew dvayxy Tév Kaxdv yap ylyverat, which shows that we need not be afraid of putting ydp late. , _ Less easy to put right is the fifth line of the following (M. 4. 22: K. 2.496), which must be quoted at some length for the point to appear : of Pirdcodot Lyrotow, as dxyKoa, mept ToUTd 7 avtots ToAts dvadotrat xpdvos, ti cot dyabdv, Kovde els epyKé TW ri éotw. dperiv kal dpdvyciv dace Kat 5 A€yover mdyra padAov i ti rayaborv. év dyp@ StarpiBuv riv te yiv oxdrrev eyo viv nipov: eipyyy ’oriv: & Zed didrare, THs emadpodirov kal diravOpaov Oeod. The philosophers were always expounding ri rdya0dv, and therefore these words (line 5) cannot be right. What the speaker means is not that they do not profess to tell you THE GREEK COMIC FRAGMENTS 63 ti rayabdv, but that they have not really got hold of it. Their dyadov is not the true dyafév. In other words A€yovor wavra pGdXdov 7 adtd Tayadov. Ci. Ar. Hecles. 643 for the scansion. In line 3 we should probably read ri rayaév éort; with Brunck, or (more rhythmically) réyafov 6 re éoré. Why has no one corrected the same poet’s teOvykev vids ) pyrnp Twi AON no» a , , 7 vy A? Grov ray dvaykaiwv ye Tis (M. 4, 34: K. 2. 505)? “AAAwy ray dvayxaiwy tis is not grammar: read dAAos. Turning to Menander, it is not hard to see that in the verse (M. 4. 141: K. 3.71) ok eotw ovder, rdtep, év avOpdrov dice peiLov Aoyirp0d 7G SiaPcbat mpdypara we should read Aoyiopa rod SiabdcOai, 1.€. rod Aoyrua diabécOar. For the order of words, cf. Nicolaus 1. 42 (M. 4. 580; K. 3. 384) dcamep éxew taddOrpta. Tov Seurvotvra Sci, t.¢. rov téAXOrpia Seurvodvtra. The speaker goes on to show that you can do anything and become anything 7o AoyicacGa. What possible construction is there for To diabér ba? The beginning of another fragment (M. 4. 231: K. 3. 158), clr’ ov dtxaiws mpoowerattahevpevov ypdbover Tov UpopnOéa mpés tats rérpats, our ears tell us must be altered to mpos Tais wérpats ypapovor Tov IpounGéa. And in a third (M. 4, 234: K. 3. 161) 6 ravta Bovdnbels dv avOpwros Toveiv aavr’ dy yévouro: tAovaLos Tpdmov TWh, | ¢ f , , if madw dirdcodds Tit pabjae. xpwevos* 76 ody’ iyaiver Teva Siattav tTpocpépuv, the general connection shows that we must read ty:avel. The future is constantly parallel to the optative with dv. 64 THE GREEK COMIC FRAGMENTS In Diphilus a cook is addressing an auxiliary (M. 4. 394: K. 2. 553). He begins oo ea mapaddBu a obdapoi, Apdxwy, eyo én’ épyov ob a diaredcis THY Huepay tpareLorroay év dyabois woAAois xvonv. Tlapadd Bo ay be right, but I rather suspect it should be wapaBddw, ‘I will never risk you.’ Cf. a fragment of Posidippus (M. 4. 521: K. 3. 343) line 14 dv & adn Owov + , \ A * cavtéy tapaBdAys, Kal mpooekdapeis dare, which is also the counsel of a cook. TapadaBwv éavréy in Euseb. P.H#. 15. 2.1 is corrected by mapaBadoy éavréy in Athen. 354 B (the same sentence). II. In a fragment of Pherecrates (Meineke 2. 294: Kock 1. 171), in which éxxapvBdioat or éxxapuPdioa is apparently used of the act of swallowing down, it is strange that no editor has restored éyxapuPdioa, as éyxéw has in several places been restored for éxxéw. Plato draws a simile (M. 2. 664: K. 1. 640) from a boys’ game: cis 8’ dudorépwy dotpaxov airoicw dvinow «is pécov éatds, | Kav pev wirtpot Ta Nevk’ erdvw, pevdyey Taxd Tors érépous Set, | tos dé Sidxeuv. Blaydes in his Adversaria in Com. Graec. Fragmenta, 1890, has anticipated me in suggesting that Plato wrote dotpaxoy avrois dvw ino’. But Cobet also condemns airryot, and the epic form cannot fairly be defended in such a context, though Meineke and Kock tolerate it. Here Blaydes’ xiv wiv wiry td ye Neva’ erdvw or Ky pev ye 7a. é 7. igs not convincing, nor is the ré\Aevx’ proposed by W. H. Thompson in his‘ note on Phaedrus, 241 3, é\Aevxos being a quite uncertified word. I think it probable Plato wrote xiv pev aimrovrs ta deve’ érdvu, understanding 7, as in Plat. Rep. 370 & dv dy aibrois xpeta. This is really a more accurate way of putting it than the other, though our idiom admits the other way freely. The THE GREEK COMIC FRAGMEN‘'S 65 white side or mark should not be said to fall: it is the shell which falls, and as it falls (wérrovr.) the white is uppermost. Some verses of Plato preserved by Plutarch refer to the tomb or monument of Themistocles at Piraeus : 6 ads d& Top Bos ev KAAG Kexwopévos Tols éurdpos Tpdapycts éoraL TavTaxod, Tous éxmhéovrds 7 elowAdovrds 7° Swerat, xadrérav ddd 7 tov vedv Oedoerat. Kock (1. 651: cf. M. 2. 679) condemns zpécpyors because it means not quod quis alloquitur sed actio alloquendi, but though unusual it need not be wrong ; cf. the concrete use of such words as guyy, réxvy, Evupayia, dxor, etc. in Greek and similar uses in Latin. We may remember particularly Peace 148 yt. . Hipurisy | déyov wapdoyys xat tpaywdia yevn with Juvenal’s ut pueris placeas et declamatio fias. But is mavraxod right? A monument cannot stand everywhere, nor can it be hailed everywhere, but only where it stands. It may however be hailed from all sides by men coming and going many ways, and therefore I would conjecture mpdapyots Exrar ravraxdbev Tols éumdpors. We may suppose that wavraydev was corrupted to raytaxod and that then the order was changed. For the general meaning cf, Eur. Hel. 1165-68, but that does not help us to the actual words. Kock’s mepicyyos is too unlike mpéapyots and leaves untouched the difficulty of zavrayod. In Aristomenes (M. 2. 731: K. 1. 690) éred) rods mputdves mpooydGouev we should perhaps write éred} eis Tovs mpurdvers, for even in poetry proper (as distinguished from verse) tpocépxopnar seems not to be constructed with the accusative of a person. Or can oi rpurdvets be taken as equivalent to a place? Cf. yéw 8& Bakyxas, tbimus Afros, etc. When some one in Antiphanes (M. 3, 105: K. 2. 90) is contrasting the hard lot of the comic poet with the easier conditions of tragedy, he cannot really imagine himself a tragic poet and say Oi8trouv yap dy povov | ba, rdAXa wavT’ icaow: 6 warip Adios, | wytyp “loxdory «.7.d., because he is F 66 THE GREEK COMIC FRAGMENTS pointedly contrasting ‘we’ comic writers (line 17 jpiv 8 tatr’ otk érrw) with the others. It would seem therefore that $ should be 7, %.¢. 6 ours of the line before. So immediately afterwards av zdéAw | ely tis ’AAKpéwva. The nominatives 6 waryjp etc. are idiomatic. Cf. Meister- hans § 82. The hard thing for comedy is that it has to invent everything for itself—zdvra Sei | cipetv, évéuara awd, ra Suwoxnpeva. | wpdrepov, ra viv wapdvta, Tiy Karactpopyy, | THv eicBodjv. But no one can justify 7d Supkyéva, and texovo- pypéva (Kock), 7a Sueynuéva (Herwerden), 7a yeyevnuéva (Blaydes) are all most unlikely. I do not know whether ra Siaxeiueva will be thought any better. I should under- stand it to mean ‘the previous conditions,’ or the way things had been disposed and arranged (SiariOévat, diaxet- oa.) some time before the action of the play began, ra ew tis (kwp@dias) as Aristotle might call it, e.g. the exposure or loss of a child. Cf. Plat. Crat. 440 D otrws oleoOar cat ta mpdypara Siaxcicbar: Hdt. 9. 26 povvopayjoa emi Scaxetpévorre= emt pyrois: Hesiod Scut. 20 as ydp of dcéxetro, Geoi 8 éxyudprupot Foray: the so-called Xpvoé "Evy 1 ébavdrous piv mpGra Geods vopw as Sidxewrat | tina. Atoxnuéva cannot mean this, but would have to mean ‘things administered’ or ‘managed,’ which is not the appropriate idea. For confusion of Sidxerpas and Sioxotuar cf. Isocr. 8. 74, Plut. Arat. 45, Athen. 46 5. This passage of Antiphanes, in which the Chremes and Pheidon of comedy are contrasted with the Peleus and Teucer of tragedy, was perhaps in Horace’s mind when (A.P. 94-96) he contrasted Chremes with Telephus and Peleus, Add Ar. Ran. 863-64 and Hor. Epod. 1. 33. Cratinus Junior (M. 3. 378: K. 2. 292): A. dvOpwros ef Sydoveri kal poy exes. B. xara pév [Adrwv’ odk oida, broved 8 exe. Hermann’s irovoé 8 éyew has been accepted too readily. Read xara rév [Adrwv’ odk ofda pév, irovod 8 ey exe. Aleais (M. 3. 493: K. 2. 385), describes Sleep —- THE GREEK COMIC FRAGMENTS 67 ov Ovytds 008 aBdvatos, GAN’ exwy Twa ovyKpacw, wore pyr’ ev dvOpdrov pépet pyr ev Oeod Lav, GANA PieoGai 7 dei kawvas pOivev te THY Tapovotay mddiv. The difficulty is in the fourth line. In the first place read xawds, which is much more picturesque than the adverb, But what are we to make of ryv wapovoiay? The structure of the sentence almost requires p@ivew to have the same subject as dvecOa1, and ryv wapovsiav péiveay is a very improbable expression. Kock conjectures riv rapotcay ad wdAdw, 1.€. yuépav. But, when we fall asleep, our sleep usually passes away (fOiva) not that same day but next morning. The poet’s point should be that sleep is always being born one day and dying the next. We can -get this meaning by the slightly greater change of ri ériotcav ad méhw, 1.6. yucpav. We find 7 émotca hyépa, ‘next day,’ in Thue. 7. 74. 1, Plat. Crito 44 a, Ar. Keel. 105 and other places, and the omission of juépa may be paralleled from Eur. Alc. 784 tiv atpiov péedAdovoay et Bidoern. Later Greek has 4 émotca (without jyuépa) frequently, ¢.g. Pausanias 2. 19. 3: 3.14.10. About the accusative case there is, I think, no difficulty, for jépav is very often used without any notion of duration, e.g. Ar. Pax 1313 ds odxt wacay jpepav | tAaxodow zorw évrvxetv TAavapevors épyors. The same poet has a very similar description of love (M. 3. 495: K. 2. 387): éorw yap ovre Ondvus odt’ dppyy, wadw ovte Beds ovr’ &VOpwrros, ovr’ &BeATEpos ott’ adbis eudpuv. Does not ovr’ atéis show that in the first line the comma should be put after mdéAwv, not before it? Cf. the Tvapae povoortyot 283 Kaddy 76 -yypav Kal 7d pH yypav wddy. After describing Love’s curious compound nature, he goes on: kal tadr’ eyo pa THY ’AOnvav Kal Gods odk 010 6 Tt eativ, GAN Gpws exer ye Te Tovovrov, éyyvs T elpl Tovvduaros. As he has just enumerated no less than seven elements in Love’s character, éxe: yé te rovodrov seems unsuitable. I F2 68 THE GREEK COMIC FRAGMENTS conjecture either éye: wé Te rowodroy or more probably (by. a reverse change to that suggested above in a line of Antiphanes’ Ilodyors) éyw yé 7. So in Phaedrus 239 c we find avyp zywv épwra and in Republic 575 a & épus . . tov xovra airdy. . déet éxit macay roApav. Perhaps roivéuaros should be rod vopatos. There is another well-known fragment relating to the drama in which a small correction needs to be made. Tim- ocles (M. 3. 592: K. 2, 453) speaks of the relief and distraction afforded by it to an overburdened mind: 6 yap vods Trav idiwv AROnv AaBov apos ddXoTpiv Te Yoxaywynbeis ribet pel? Hdovis dr7AOe tadevbels dpa. But yyaywycioba! ru, not zpds tw, is the proper construction, as we find for instance in the parallel passage on the same subject in Marcus Aurelius xi. 6 ols éni rijs oKyvis Wuxaywyeiobe, tovTos py axOecHe ext tis peiLovos oxnvis. Indeed zpds is quite impossible in such a phrase. What Timocles wrote was no doubt apés dAAotpiy rt’
YoyaywynOeis wdGer, according to the common idiom of ecivat (ytyveoOat) zpos Tw, to be engaged in something, to give one- self up to something: e.g. in Diph. encert. 13 wpds 7G AaBetv yap dy 6 vots TaAN otx pa. There is no difficulty in the com- bination of present and aorist participles. Cf., for instance, Andoc. 1. 106 vixjoayres paydpevor Tods Tupdvvous ‘having conquered in battle’: Lys. 3. 6 éAOay emi ri olkiay rhv éunv vixrwop peOvov ‘having come drunk,’ or ‘when drunk’: ib. 13. 67 wapadpuxtrwpevdpevos Trois ToAculous AndOeis ‘caught signalling.’ Ar. Pol. 8. 3. 12 émiBovdevovres pupabévres : Aesch, Ag. 363 rév rade mpagavr’ én’ "AdeEdv8pw teivovra maAat tofov: Il. 8.514. So here it is ‘diverted by attending.’ Theophilus (M. 3. 626: K. 2. 473) Tov Tpopea, Tov awrypa, bu dv elSov vopous “BdAnvas, €uabov ypdppar’, éuvyOyv Oeois. Kock makes no remark on ¢éov. It is curious that Meineke, who suspected it and suggested the by no means suitable e«fxov in its place, did not recognise the confusion of cidov with dev, or 7dy. Cf. Hur. Med. 537 THE GREEK COMIC FRAGMENTS 69 ‘EAAas’ dvi BapBdpov yOoves yatav Katotxeis Kat Sikyy éricracar vomos Te xp7po Oa : Theognis 54 ovre Sikas ySecav ore vopous: Anthol. 10. 62. 1 od Adyor, od vépov ofS TYxn. Callimachus (18. 2) appears to say of a shipwrecked man vaiy dua Kal poy cider droAAvpévyy. But it is so awkward for a man to be said to witness his own death that perhaps the poet really used Hoet, not «dev. A fragment of Philemon (M. 4. 26: K. 2. 500) may be improved by an insignificant change. A justly proud cook soliloquises, parodying Medea’s nurse : ec g 2 a a 3 ~ ds (wepos pwoimnAGe yp Te Kovpave Ae€~au podovTs TobWov ds éoKevaca. ) i. > a. AO , yy > * aA vy tiv “AOnvav, 40 y' err’ edypepety 2 a >. \ e * * S , év dracw: ixObs draXbs olos yéyove pot, 5 olov wapardetr’, od mepappaxevp.evov a ca ” 2 aft tupotow ovd avobev eEnvOicpéevov, ce} ~ GAX’ olos iv Lav, Kdwros dv rovodros Fv. We sympathise with his achievement, but we cannot allow him to express it in bad Greek, like the ofov of line 5. Bentley thought totev would do, but rotos is not used in prose or comedy, rowtros as in 7 being the Attic word. Kock would read érrév from line 7. I would change only one letter of ofov and read éyov as in 2. The cook speaks of the conversion of the fish into a dish. There is no need to suspect the nominative ix@ts drados, though Philemon might also have used the accusative. The cook concludes 2 , y 3 \, y¥ \ aOavaciav nipyka’ tovs 45n vexpovs og 2 a a ~ 9 Grav éogppavOacr, Toes Civ mad. To make the last line scan Casaubon wrote éray pévoy, Porson éravrep. Perhaps we should read vexpois | dvras, Gtav édodpavOdor, 7.£.7. So Herod. 1. 113, rov 8% éwurod (aida) éovra vexpdv AaBov eOyKe K.7.X. Cobet did not quite finish the emendation of © KXéwv, taicat pivapav dv dxv7is To pavOdvew, dvertxovpytov geavTod Tov Biov Ayre ody, 70 EPICHARMUS (M. 4. 30: K. 2. 533) when he wrote Ajoes for Ayre. We must also write ceavr@ for ceavrod. If Philemon had used the genitive, its proper place was between rov and Piov. The exceptions to this rule are few in the extreme. A thoroughly sound suggestion of Meineke’s is ignored by Kock, when he writes (2. 504: M. 4. 32) eon a¢ vos , ee 5 qpov 8 doa xai ta odpar’ éori Tov apiOyov Kal’ évds, Toaovrous ear Kai Tpdrous ideiv. Kad’ évds is nonsense and Meineke suggested ‘nonne xa6’ éva scribendum?’ no doubt meaning it to go with what follows. The point is that, whereas all the animals of a species are alike, in the human race there are as many characters as individuals. Quot homines, tot ingenia. You find, taking men one by one (xa6’ éva) that each is different. Cf. the xa@’ &a in Philemon’s remark on the nuisance of friendly questions as to how you are (M. 4. 15: K. 2. 490): S Agt 3 a nd, 4 ToAU petfov EOTL TOV KAKWS EXELV KAKOV ‘ poe a a“ 2 t 70 Ka’ eva rao Tots émtoKoToUpEvols Selv Tov Kaxds exovra was exe A€ye. IIl. The references in the following are for Epicharmus to Kaibel, for the rest usually to Kock. EPICHARMUS. 21 aparov pév air’ éoOovr’ ors viv, droPavors, with variant i8ys and dmofdvys. Probably drofavei, though Sos may be right. The optative dwoddvors could hardly dispense with xa. 35. 1 ovvderrvéwy TO AGvTt, Kadéoat Set povov. The nom. pendens is improbable. Read ovvdervéw. ib. 8 Avxvov 8 obx 6 rats por cuppeper, éprw 8 GdicOpdlwv re Kal Kara oKOTOS Epnpos. EPICHARMUS 71 If épypos, he is alone and probably does not even keep a slave. For ody 6 mais then read ody! mais. Cf. Juv. 3. 286. 171. 6 (8oxet) 76 ya | dyabdv a (MBS. 76 88) rpayy’ elev Kal’ QuTo. Sense seems to require t: mpayua. ‘The good’ is a thing, just as the flute-player is a man. The extract actually begins dp’ éorw atAnois 7 mpaypa. ; 254, 5 8verddatoros dy tovs dAAovs eirahaictous drodavel. Kaibel alters dv rovs to airés with Heimsoeth to get rid - of the spondee. Surely dvomdAacros dy tis is much more probable. 257 éorw dvOpdrw Aoyiopds, eort kal Oetos Adyos avOpwrw wépuxe tept Biov katactpodas. 6 0€ ye Tas TexVas aract cuvererds Heios Adyos, €xdiOdoKwv aids abrovs dre oteiv Set cvuudépov: ov yap avOpwros téxvay tw’ yupev, 6 S€ Oeds Tavrav pepe. In the second of these lines—which are very unlikely to be genuine Epicharmus—the missing syllables are, I think, dors. Cf. on Frogs 1054 (p. 51). In 3 ras téxvas has no construction, We might read either ras téyvas (ris téxvys) Or Tats réxvais. In 5 Kaibel would read pdvos for ravrav hépe, but perhaps the omission of ravray would be better. 284 4 dé pedéra vows dyabas trove Swpeirar Pidors. Wilamowitz ive. Something like xadd seems probable. In Stobaeus Flor. 90. 8 three verses are ascribed to Epicharmus, though they have much more the air of the Middle or New Comedy and Kaibel does not even print them. They run as follows. o ; oe > t Doe x AD mviyop’ érav evyéveay ovdey dv Kaas Aeyy tts, abrds Svoyevyjs Ov TG TpdTw- tis yap karémtpw kal TUPAG Kowvwvia ; 72 CRATINUS For oidtv dv, which is evidently wrong, Ahrens conjectured oddevar, t.¢. Lovdevav, contemnens. Apart from any other objection, this seems to give a quite wrong meaning. What we want is that a man shall speak in praise of good birth, though far from well-born himself; that would be like a blind man who insisted on praising, perhaps on having, a looking glass. For oid dv I would therefore suggest eidoyav. Nothing is commoner than confusion of ¢« with o, and A with 8; y and v are also frequently exchanged (Bast. Comm. p. 727). Cf. Eur. fragm. 52 edbyéveav ei Bpdretov evrAoyjooper. CRATINUS. 138 tives ad wévrov Karéxovo’ atpat; védos ovpariov 70d’ Sp@pau. as ay padAov rots wySaALols H vads Hudv weOapy7- Crates 17 068’ é& dyopas ob8 Tdxwvas qomnodped’ odd’ aAXGvras. Are not the editors rather too easy-going, when they admit on the authority of Hephaestion these spondaic endings? In the second passage nothing is easier than to invert the order of dAAavras and rdxwvas, reading oir’ dddGvtas Tomodped’ obre takwvas, for ore... odre seem also required. Cratinus is not quite so easily corrected, but one may perhaps suppose that he really wrote something like 4 vais quay ds reOapyy paAdAov rots mydadioc. For final ws cf. Cratinus 298, and it is helped by the poetical character of line 1 (xévrov, adpat, 6pHua), though obviously the first and second lines have no immediate connection. The spondaic ending is made even less probable than it would otherwise be by the fact of the fifth and sixth feet also being spondees. Athenaeus 3. 868 gives us another apparently spondaic line (authorship uncertain), rovAuzé- cov, onmiddptov, KdpaBov, dotaxdy, dotpetov, Which Meineke proposed to treat much as I would treat the line of Crates, putting zovAvrddeov at the end of the line instead of the beginning. Whether Doric marching anapaests were some- times, as Hephaestion says, spondaic, is another matter. 240 Read érucréryv or érixrerov for rixrerov, CRATES—EUPOLIS 73 CRATES. 15 GAN’ avribes Tou éyw yap K.T.A. The crasis of ro: and éy# in spite of the pause between them seems unlikely, nor is ror much in place with the imperative. Perhaps we should read 7é8’. PHERECRATES. 10 elra apds Tovrowow HAovY dpOpiae 7a. ceria dore thy kopyv banxeiv Oryyavovcay ras pwras. Ovyydévew does not take an accusative, and Kock—whose conjectures are often extravagant—would read zepiayouvodv. By the slight change to ris caus (or possibly ris wvAns) we can get all that we want. Fragments 67 and 68 appear continuous and should be put together. 132 radpbarpw rovry wepyndrropey TH Tov matdiwv. The metre being Eupolidean, as in the parabasis of the Clouds, the line is still unmetrical, till we read wepiparrdépeda, though in the Clouds I do not see an actual example of the tribrach. Evpo.is, 90 od yap pa THY Mapadaue tHy ewny paxny K.T.r. The rather curious phrase tiv éuyv pdayny has been doubted and vixyv riv éeuyv suggested. But Plut. Mor. 496 F is closely parallel and confirms the text: odk éreidse tiv Sadapiva NeoxAjs ryv OeutcroxA€ovs ovde tov Eipupddovra MuAriadys tov Kivwvos. The story in Plut. Cimon 8 per- haps indicates that Miltiades was thought to claim too much of the credit. 357.8 um pOovel6’ drav tis Hudv povotky xaipy vewy. For véwy, which is certainly wrong, Kock suggests évvav. I would propose zovdv or roav, possibly but not necessarily reading also povoujv. Cf. Phaedo 614 povotkiy orev twice over. v and z get confused. On 76 o@ Ady in line 6 see Xenophon and Others, p. 61. 74 ARISTOPHANES, PLATO ARISTOPHANES. 294. 3 Ths Tacw avOpdroicw Hipyvys pidys mist Tpopds, Tapia, cvvepyds, éritporos, Ovydryp, ddeAgy, Tavta tTadr’ éxpHTs pot. éxpfjro is certainly wrong, but there is little use in proposing edxovro (rather yiyovro) as Kock does, as though the two words were likely to be confused. Moreover, qiéxovro (i.e. ebydpevoi pou wavta Tard we éXeyov Kock) would not even give good sense. The meaning must be that they called her by these names, and this strongly suggests eipyro, though it does not seem certain. 488 ev 8€ cot dpdcw, y y , s 2 mM oe . OOW TA. KATO KpELTTH OTL wv Oo Zevs EXEL. The speaker goes on to show by one instance how true this is. Surely the poet wrote évi, not & In Hryzxias 393 a éote wA€ov wdvtwv movypdtaros 7} dow TAOvodTaTOs is it not clear that we should read zAéovi? 678 oda ToravTi Kal TolauTi Kai Sedpo cynpaticavtas. Herwerden (Collectanea 1903), citing Blaydes’ violent correction 7. 7. oxypata Sevpi Kdxeioe oxypaticavtas, pro- poses himself to read xd@’ (xat 20’) érepa. I should rather think that the line is quite right and that it was followed by something which gave the antithesis to detpo, e.g. kaxeioe pad’ at. Cf. passages cited on Antiphanes 202 below, p. 78. Prato. 91 évq for é seems to me likely. 187 kairo. jTéempaye TOV TpdTuV pev aka, abrov d¢ Kal Tav oriypdtwv avdéia ov yap Totovtwv eiver’ dotpay’ nipéby. If rpdrwv is right, we must explain the obscure antithesis of the lines by saying that Hyperbolus’ character deserved it, but that at the same time he was too insignificant and mean a person to be so dealt with. ‘Deserving it’ however implies that ostracism was or might be regarded as a punish- ment. On this point Plutarch contradicts himself, saying PLATO—THEOPOMPUS 75 in Aristed.7 poxOyptas oik Hv KoAacts 6 eEoorpaxiopds, yet in Nic. 11 and Alco. 13 calling it a xédacis or KoAacpds. But, though we have no reason to think it was usually considered to be a punishment, Thucydides does expressly say 8. 2 that Hyperbolus was ostracised 6:6 rovypiay Kai aioyvvyy modews. I am therefore on the whole inclined to think that tporwv may stand, though Kock’s zporépwv has plausibility. Cf. Isaeus 5. 14 wadov déta tév éavTod Tpdérwv. AMEIPSIAS, 9 Yaxpares, dvopov Bedrrior’ ddAcywr, wodAdGv 88 paradral’, nKels K.T.A. Anything about Socrates is so interesting that perhaps I may be allowed to urge the adoption of Dobree’s éAéyw and woAAG, which to my mind are clearly right. If the speaker says that Socrates is ‘ best of (a) few men,’ he must mean either that 8. is the best of his own small following, the Socratic school, or that he is ‘best of few’ absolutely, 4.e. no better than most men, just as obdevos duetvwy for instance means ‘as bad as anybody.’ The latter sense is most unlikely, the former somewhat pointless. Dobree’s parallel from Xen. Mem. 1. 6. 11 & Sdxpares, eyo roe dikadv ce vopilw, copov dé odd’ Grworiody is very close, and the datives make excellent sense : ‘ you may be morally a trifle better than other people, but you are certainly much more foolish and useless.’ CaLLias. 21 There is no need to assume with Kock and Herwerden that « in érvos is lengthened contrary to comic practice. There must in any case have been something preceding it, and this may quite well have ended with a short syllable. THEOPOMPUS. 59 GAN’ 4 pev HideiOua ovyyvepny exer td Tay yovatkOy otoa KatarAyé THY TEXVYY. Kock with his usual boldness would read 6. 7. y. réra karawAné yevouevy. I agree that quid stt xarawAné tH 76 POLYZELUS—ANTIPHANES réxymv nemo facile dizerit, but that does not justify so extravagant a change of text. Here again, as in Phere- crates 10 above, the slightest of changes will give us quite as fair a sense as we can demand in a mere fragment. Read 19 réxvy, governed by ovyyvouyy éxe, which may either have its common meaning or signify has a fellow- feeling with. POLYZELUS. 3 For rair’ Zor. rpia Onpapévovs (Kock with Meineke tavti 7a tpia) ratr’ qv 7a Tpia, may be suggested. DEMETRIUS. 1.4 Perhaps péyas xal | Aapapés. ANTIPHANES. 20 kpéas 6€ Tivos HotoT’ av éoOious; B. rivos; eis ebTéAciav. Tay mpoBarwy per K.T.A. Since we should expect xpéa rather than xpéas and eis ciréAecay is strange as an answer to tivos, I suggest xpeéa 8 noir’ dv éOios tivos eis edréAciav ; B. rév wpoBdrwv pev K.T.A. 40. 4 Insert perhaps something like xai rpuda. 44, Something may be missing between lines 4 and 5, and ddpet altered from dopety. 52. 15 rodpradw need not be called in question. Just as in 12 and 13 the two opposites of wine and water were mentioned, so here sweet-smelling things and the reverse. 161.9 = xopioas 8 e&€Barev olxade ? 190.4 Poevixidys d¢ Tavpéas 0 6 pidraros, dvdpes méAat dpopdyot Tovodroi Ties, olor xataBpoxOilew ev dyopa Ta Temaxn, épavres €£€OvynoKov ext To Tpdypart. The editors and others have made attempts on the second line, so as to reduce it to something like decent ANTIPHANES 77 versification. But their labour is thrown away. It never was verse at all. It is nothing but a commentator’s prose note, first telling us who and what Phoenicides and Taureas were (of which he knew probably no more than Antiphanes indicates), and then writing rowtrof tiwes to explain ofo. Kock may read odd révrwy épopayictatol tives, but wdAaz tells its own tale. Two or three similar adscripts that have found their way into the text may be noticed. Nicostratus 10. 4 was pointed out by Dindorf and is bracketed by Kock. Cratinus Junior 1. 5 was recognised as an adscript by Dobree, and Eubulus 107.19 is plainly Athenaeus, as Kock says. We read in Eubulus 119. 5 foll. ov daci wore KAynOévr’ eri Seirvoy mpds pirov twvés, eirovTos aiT@ Tod Pirov, Sryvik’ av elxoot Today pretpodvTt TO TToLXEloV 7}, Kew, K.7.d. Scholars, including Porson, have gravely tried to bring this into order, but pds... rod fiAov is a mere insertion, the removal of which sets everything right at once. xe depends a little loosely on «AyOévra, and that is the reason why cimdvtos . . . didov was put in, just as pds Pidov twos is another explanatory gloss on xAnOévra. Perhaps éayvix’ dv should be qvix’ dv. But apds betrays its origin, for no comic poet of this date would have thought of using zpos for izo, a construction confined to serious poetry. In the lines ascribed to Philemon (109) Grav EevicOys ev wdc TpOs Tov idoy K.7.A. it is likely enough that 70d giAov or something similar should be read, but this is only a further proof that they are not really his. The misuse of zpés belongs to later Greek. The form of the same story in Menander 364 y v Os KAnOeis more eis Eoriacw Swdexdzrodos pOptos mpos THY wEAHVHV ETpEXE K.T-A. tends to support the omission of apés... rod piAov here. KAnbeis Swdexdzrodos is exactly parallel to KAyOevra . . . dr7- vin’ dy K.7.A, 78 ANTIPHANES Probably the same account may be given of the first line in Philemon 102 (whether the lines are his or not). , SoA es 2 s 1 TpOTEaTL be THM TEVYTL OTLTTLA 4 tiv yap wevytuv wiotw ovk exer Adyos. As though any one would have written line 4 to come close upon line 1! 1 is a prose version of 4, and we need not trouble about the elision of in the dative or anything else. There must however have been a line or lines in which the zéys was mentioned, since otherwise imdpyy in 2 would have no subject. But possibly 102 is a combination of two fragments, and then the case of zpdccors dioria is different. 196. 13 mos yap yevour’ av, & mdrep, .. « pytup dduwvos, hv py GAG tpls Tapavopwyr ; Assuming with Kock that there is a hiatus of two feet in the first line, I would suggest mas yap yevour’ av, & warep, pyTwp adwvros, jv 4XG ye py tpis Tapavdpuv ; or qv pe) Tpis ye Tapavopuv da, and in the hiatus of the next two lines — kal pny axpiBds wdpyy éyvwxévat 76 pnb .. . GANG 8H Aeye we may very plausibly insert ofdy éorw. 202. 14 draipovoat 5€ puxpdv, od ToAv, Tod pyr’ exeioe pyre Seipo wavTedas, ovtws dveppimov adore k.T.A. If there were an infinitive with rod, it would no doubt be right enough so to express the mind of the doves not to go decidedly one way or the other. But can an infinitive be thus left to the reader or hearer to supply, and the article rod remain alone? I hardly think so, and therefore venture to suggest ré. Cf. (L. and 8. s.v. Seipo) Ar. Av. 425 wdvra kal 76 THdE Kal 7d Ketoe Kal Td Scdpo: Hur. Phoen. 266 Oppo ravraxy Sioveréov Kdkeioe Kal To Sedpo: 1b, 315 éxetoe Kal 76 Setpo Teptxopevouca. ANAXANDRIDES 79 264 py Oupacw 76 cdma Napmpivev Oére, épyous 8& kabapois Kai rpdrats THV Kapdiav. (A very doubtful fragment.) ypdpacw, pippacw, Bap- pacw have been suggested for dupacw. I may add vippaow. N would fall out easily after H. vizmara occurs in another comic fragment (Dromon 2). 278 avdpds 8 éveyxetv drvxiav 6p60d tpézov. 6p06 tpé7w Grotius. The accusative would be a slighter change and, I think, equally good Greek, though less common. ANAXANDRIDES. 1. 5 Ido not profess to understand lines 5 and 6. It is however clear that one déyew is due to the other, and the first seems most likely to be wrong. On that assumption, which is the usual one, I would suggest ééAes to take its place. 17 & 76 okOALOY ebpuw éxetvos, Satis Hy, TO pev tyiaive mpOTov ads Apictov ov dvopacev 6pOds: Sedtepov 8 civar Kaddv, tpitov S€ wAovuTEty, TOUO’, dpdis, Euaivero. dpas (or Spas ;) should perhaps be dépicas, and eépwy in the first line cirdv. We find 6 mpéros cipwy x.7.A. two or three times in these fragments (Anaxandrides 30: Alexis 148 and 185: Eubulus 72: Menander 14), but with ro oxdduov it is less appropriate than «irwv, for which cf. Alexis 27 6 mp&ros cimov «7... and Kock’s *Adéorora 115 (3. 429). Cf. pp. 186 and 224 below. 33. 5 émt tiva § &S ddAn réxvyv, & xpyote ov, Ta oTdpara TOV vewrépwy KaTaKder’ 7} bio pos éore SaxtvAwy TotovTod: ; It has been remarked that émi riva réxvqv is not well constructed with xaraxderat, but we may notice further that xaraxderat itself is a surprising word in this context. It means burnt down and does not appear ever to be used in the sense here needed. We should get both sense and grammar, if we might exchange the two prepositions and 80 EUBULUS read xara rive... Téxyvqv... 7d oTdpota... émixderar; ‘what other art sets their mouths so aflame?’ daxrvdwv, for which yaorpiSwv, Sairadéwv, and Satrupdvev have been suggested, seems perfectly apposite and good, but possibly ro:ovroci ought to be roaovroci. 34. 4 dy may be the missing word. 41. 21 Perhaps @yBas rats értamvdots. 52 The dAdq in line 9 seems to indicate something lost before it, for it has no sort of point. It may however be due to the add’ dAaBev in 11 and have taken the place of a word quite different, e.g. «i 3’. In 7 read at6ts, as has been suggested, or possibly av rts for ards. Evsvutuvs. 9 Erepot Se Pcotor ovupremAcypevor peta. Kapa Bou ovvect, and fare badly, since Carabus eats everything before they get achance. ‘rots Geoias Casaub. OeouréyPpoicr vel Oypiorce Herwerd.’ Kock. Perhaps we may conjecture something like érépots 82 rots, or érépoior by, Geoior, taking érépois in the sense of adverse, as in Pindar P. 3. 34 daiuwv 8 érepos és kaxov Tpéfais éSapdooard vv, just as it means evil in the Platonic and common téov Odrepov woeiv. Cf. on Timocles 31 below. 25 A line is lost between 1 and 2, containing some words with a dé. 90 iocropetre is not a question, but like the other futures semi-imperative. 115 Kat yap wéow KdAALov, tered, Tpepew GvOpurdv éot’ avOpwrov, av éxy Biov 2 Xiva . . orpovbev x.7.X. Kock rightly points out that 7éom cannot stand as an exclamation in place of daw. It would be distinctly inter- rogative, not exclamatory. Is od yap cogs too bold a guess? That of course would be a question, and a question (not an exclamation) seems needed by ixeredw, as for instance in Frogs 745. NICOSTRATUS—EPHIPPUS 81 119. 11) erera pdvat. Probably ddoxew rather than érerra 8 ddvar or ddvar 8 éretra. NICOSTRATUS. 5 dv ovk av Katadpdyouney ucpav tpidv 704 KarecOiovres ob SddeKa. ye. The last line has somehow gone extravagantly wrong and any restoration must be guess-work. Cobet would read tpiav KateoOlovres dvres Oddexa, Porson having previously proposed 7. x. od? 8. We might think of something like katecGiovres obde Ska ye Kal TpLOV, comparing for ot8? déca ye Eur. LT. 570 088 of codoi ye daiuoves «.7.4, and for thirteen as a humorous number Ar. Frogs 50 xai xaredicapév ye vais tov rodeuiov } Sddex’ 4 tpeis xai Béxa (kai. . ye and odd€.. ye are parallel to one another) ; or, to bring in 7#8y, xareabiovres ode Sék’ Hd Kat TpLov. yon p AMPHIS. 17 6 warip ye rod Liv ear avOpdrous dypds. Probably 6 Sorjp. 33. 4h ev yap él rod cuvrerdxOar did TéXOUS Ppdvyors ovaa. K.T.X. For éwi write did, parallel to 7 88 da 76 yn . . dtadedAoyé- oOain line 7. émé could only mean something like applied to, as in 3. 3 émi rod pabnpatos yap éoryKws 5 vods. EPHIPPUS. 5. 15 Levyn tpoodyew | unvdv éxtd cuvexads Exardv. Neither Meineke nor Kock notices the use of the genitive for duration of time. The grammars ignore it, and I know no other instance before Lucian and Pausanias, for such uses as Ar. Lys. 280 e& érév dAovros are wholly different. But the present passage can hardly be wrong, though it G 82 ANAXILAS, ARISTOPHON would of course be easy to write éxra phvas. After meptayyéAAew re in line 18 something is probably omitted. 15 GAN’ aydpacov eitehas: dav yap txavov éo7t. B. ppate 84 wore. The last words have been a good deal corrected. Per- haps dpage 89 76 ri. ANAXILAS. 35 Pollux speaks of a lady’ s dress (ra Tepivnoa) bordered by purple vycou oxnpa qowvoa ™m Teptppon Tov xpdparos. kat tadr’ tows vaoov “Avagidas daexdAe (doxadet!), ei py érepov Tt Upacpa. SyAot, Stay elry kal was yu”. dotep Oddattav vacov dudtevvuras ; in which Meineke reads @dAarra. But dyduevyvras would cer- tainly ill express the relation borne to an island by the sea which contains and surrounds it. One naturally thinks of domep Oddarrav vioos, but that does not very well fit the words of Pollux. It is possible however that he misunder- stood, and after all he only says tows. I am not sure that kal ws should not be xai tws without a question. ARISTOPHON. 4 dy Tis éorid, wdpee mpdtos, Gor’ non waAar Cwpos KaAodpa. For the few syllables missing before fwyés Grotius supplied apa véwv. He was right, I think, as to the meaning, but for the form I should prefer Tols véos, comparing for the dative Epicrates 5 ‘wat, mai’ kadetofat wapa métov, Kal air’ dyevetw pepaxvAdAtm tui. Besides Plautus Menaechma 77 we have Antiphanes 195. 10 Kadrotod p’ of vedrepor dia. Tatra wovTo oxymrdy, and Alexis 178 xadoto. 8 airdévy wavres of vewrepot Tapdoitov troKépic pa. 10. 7 A man describes himself as éAaiw pajre ypjobat pO Spav Konoprés, .e. he is like dust, equal ‘to dust, at not doing so and so. pi xpjoOo. édaiw is simple, but what is EPICRATES, ALEXIS 83 #4) pv? It seems usually regarded as wrong, and there are various conjectures, of which the last is Herwerden’s pay’ rt xpyoOat pnd’ dvap (Collectanea, p. 117). But it admits of defence. I take it to mean not pi épav ZAaov, but pi dpav éXaiw xpwuévous. Probably the only ordinary opportunity for seeing this was in the gymnasium or some such place. The whole expression would therefore mean one who neither took such exercise himself nor went where he would see others taking it. The athlete’s oil was as alien from him as from dust: there was no more mixing and commingling in the one case than in the other. EPICRATES. 3. 21 ideiy pev avTyy pad eote Kat wricat, eSépyerat te mavtaxoo’ Hoy TLopevy. 4 wtvoo. (Herwerden) is no doubt right, but what are we to do with wiopévyt Herwerden suggests zeropévy, and there are other inferior suggestions. mouévy occurs to me as possible, and it is nearer to the MS. than any. It would mean at call, when asked or summoned. Lais is willing. In line 14 read y’ év for yodv. 11. 6 Tdde pou TivuTOS, el Te KaTELOWS Hes, AéLov mpos yas. Porson Adfov dépe zpos yas, Cobet Adfov mpds ’AOnvas, Erfurdt Acfov 8) wpds tas yas. Against Porson’s insertion is the order of the words, for ¢épe should precede A€é£or. Perhaps we might read Acfov, dire, mpos yas. In 32 certainly read roraiode with the MS., raicde being impossible, as it does not refer to the place of the speakers. totaicde rovety Totatr’ ampemés, Towatocd’ amperes tT. 7., and one or two other variations are equally possible. I do not like Dindorf's division of rotai-ra between two lines. ALEXIS. 22. 1 & tiv ebruxy vaiwy Idpov, dABte tpeo Pv. etrexy seems unsuitable, edavyq (Kock) unlikely. May not Alexis have written cirvxég with a lengthened before a 2 84 ALEXIS the liquid y, as in I], 15. 463 edorpepéa vevpyv: Od. 6. 105 dpa vipat, etc.? In line 3 we have the epic form paxd- pecor, and edruyys is quite suitable to the context, the praises of Paros. 62. 3 drroBeBappevas els ovxt tabrév pa Ala, tiv aitiy pipov. dio 8 éxdornv. Scribendum utique iSiov (Kock). I would rather read iS{ws (Plato Laws 807 B) or idia. 107. 1 Insert perhaps éyyis. 116 bv’ éori, Navoivixe, rapacirwv yey, év pev 76 Kowvov Kal Kexwpwdnpevor, of peAaves qyeis' Odrepov Cyto yévos, cenvotapdcitoy x pécov Kadovpevor, voeis ov Td yevos Kal Td Tpaypa ; Meineke did not apparently doubt Zy76, but the speaker describes the class in question and asks Nausinicus if he understands (or notices) it, so that {y7 is not at all suitable. Blaydes reads {mAG, Kock fyAwréov, etc., but there is no indication that the man is not content to remain among oi péAaves. Herwerden @drepoy 8 A€yw. ek pecov is also unintelligible, being usually applied to removing a thing éx pécov. év eow (Kock) is not very much better, without some new reading for fyrw, and Herwerden’s éupeérpws unlikely. It is reasonable to think that the two cor- ruptions may go together. Much the easiest correction of éx pécov is eis pecov, for éx and eis are very easily confused (K and 1c), Does gyré suggest anything that would naturally go with «is uévov? Perhaps we may read @drepov & Axe yevos . . . eis peor, 1.2. ‘has become well known’ orin the commoner sense of «is pécov eAfeiv ‘entered into the competition.’ 8 je is not however very near to Lyra, and I do not suggest this very confidently. [Headlam’s dmpogirovs for wapacérovs in line 4 and rots pév... peya- Aovs mpoceverpe, Tots 8 €Adrrovas in 11, 12 should almost certainly be accepted.] The meaning of 11, 12, though ALEXIS 85 obscurely expressed, is that, as fortune gives us greater or smaller lives, so she gives us greater or smaller men to deal with. 124. 18 oddA@ y’ ducivwy, ds eorxas, 76’ dpa Aoyoypddos 7 pdyepos. TA. 6 déyets ob Nevers, réexvyv 8 dvedifes. Cobet explains the odd é... A€yers as quod vis decere non dicis, 1.e. quod dicts non significat id quod tw putas: ‘you don’t succeed in saying what you are trying to say.’ It would much more naturally signify ‘you don’t mean what you say.’ Of. Ar. Rhet. 3. 11. 1412 a 33 7a 8 rapa ypdppo. Nipeeeny motel odx 3 Neyer Aéyew and a few lines before & 22) ra doretd éoriy éx Tot py O pyot Aeyev. 125. 7 There is no need to alter rijs délas. dzrodidpmevot is understood with it from darodar’ éXdrrovos Fs ele Tunis. 162. 5 Read dareviow and atrots (Kock airois and diarrewidev) for the MS. demvodpev airois... The three are the younger people, the two are the old woman who speaks and her old husband. The old fare a little better than the young. I do not see how diarewapey and ovyxowwvodpev can stand together, as Kock makes them. 163 tpayypar aicbdvoua. yap ort vouileras Tots vupiow petiodor THY vindyV A€yes mapéxe, duntas Kal Nayda Kat KixAas. Tovroict xaipw, Tois be KexapuKeupevots y Q a“ y a * # Gpowrr kat Cwpotow Wdup’, & Geol ; For Aé¢yess Herwerden proposes dei, as Meineke did long ago. Aéyew would give good enough sense. JI should take trois v. as its subject, going with vouifera:: they bid (their slaves) give them these things; but the order might be vopifera, A€ye mapéxew Tois vupdiowr, ‘it is usual to bid (the slaves) provide bridegrooms with’ etc. In the last line there are the fairly plausible emendations oiddp’ (Jacobs) and payddp’ (Kock, 1.¢. pySapca yafporue). dyiap? seems not impossible. For & 6eoi cf. Nicostratus 5. 1, ete. 86 ALEXIS 212. 4 obr’ dy dSixotr’ obdéey obdels ob6 EBpiLorr’ av Exdv. This, says the speaker, would be the case, if man did not need food. For the last words, which make neither sense nor metre, Ahrens conjectured and Kock reads ov6’ wBpifo rav éxav, Ahrens also proposing déiKot y’ for adixotr’. But éxdv would be entirely superfluous, for ufprs must in all ordinary cases be intentional and the exceptions need not be carefully provided for. This suggestion therefore will not really do. I should look to éxav itself as the seat of the corruption, for it can hardly make sense in such a line, the passage having nothing in common with Aristotle Hthics 5. 9. 1-7 on 76 éxévra adixeicba. av, though quite unobjectionable, is unnecessary, and so we have the three syllables dy éxav on our hands. We might perhaps turn them into dxaxos dv, just as reversely in Dio Chrys. 38. 16 od xaxdv dvra stands for od« av évra. No innocent, inoffensive person would be wronged, insulted, outraged. dxaxos occurs in Anaxilas 33. 240.6 det od xalpes, & yivat, p’ aiviypaow. There are several suggestions for emending this, none, I think, satisfactory. I would add the bold conjecture that for det od xaipas we should read dzoXeis ob xaipove” : as in 173, 15, Antiphanes 222. 8 dod p? obroai: Menander 533. 1 dwode pe 7d yévos: Strato 1. 30 Savoe? p’ drodAtvat ; Theognetus 1. 1 dyOpwz’, drodeis pe: Phere- crates 108. 20 ofp’ ds dzodes p’ évraifa StarpiBovo’ eri, where the expression is exactly parallel, and so Agathon 13 drodeis pw’ épwrdv, Eur. Cycl, 558: Aristoph. often, e.g. Ach, 470, Wasps 849, 1202. Cf. also Antiphanes 52. 5 2 i ? > * Grroxrevets apa p’ ei py K.T.d. 245.13 There is nothing wrong with 4 ¢udormia 8 Safuovos as an attribute of love, who has been called rod daiuovos already in line 5. He has various attributes such as belong to various people and things, and this one is such as belongs to his own semi-divine nature. He demands ty, like Aphrodite in the Hippolytus (8, 1402, etc.), Dionysus in the Bacchae (208, 321, etc.). The ry of *Epus is épws, as indeed it is also of Aphrodite. CLEARCHUS—DIONYSIUS 87 Nor do I see anything amiss in 254, though it has been a good deal called in question. 262. 2 Read diaxpagdpevos ydioy Biov, ‘when he might secure’ (or possibly accomplish) a pleasanter life.’ 267. 4 Heaven itself does not rank before a mother : d0ev & mpATos ovk dmraidetTws Exo iSpvcad’ tepdy pytpds, ov delEas cadiis metas. For 6 zpéros I would venture to suggest apo rov tis. apo tov is usually said of what used to be or not to be, but it can also refer to a single event, e.g. Ar. Poet. 15. 1454 b 3. CLEARCHUS. 5 olvos Ado Buos, bv abrés éxoinoev 6 Mapwv pot doxd. Meineke read éuo/, leaving a most unrhythmical line, or conjectured éferotyoev. But why should Maron, who lived in Thrace, be the ideal producer of Lesbian wine? Lesbian wine should rather be compared for its excellence to the famous vintages of Maron : that is, we want ofov for év, and should then put airds after éroincev. Kock quotes parallels for this use of pot Sox (instead of Soxet), but does it seem probable ? AXIONICUS. 6. 12 movypos dv Te xpnoTos elvat pyoi Tus. Read 7: for re, as re is now read for 7 in Timocles 4. 5. The point is not that the man is of a bad sort, but that he is bad at some specific thing, e.g. a game. It is an uncommon but not an unparalleled use of xpyords, facilitated by zovnpes. Dionysius. 2. 15 ovrws ef’ Hptv K.7.X. ‘So with us’ or ‘in our case.’ But this needs 颒 jpar. ef’ jpiv is in our power. 88 HENIOCHUS, TIMOCLES ab. 21 I suggest eB ofS: raird to fill the gap. 4 Kydicddwpov dacw émixadovpevov fs + 3 ? i? I, ‘A ‘ mAdvov tw’ év ’APyjvats yevéerOat, THY TXOARV eis TODTO TO pepos Tod Biov KaTaxpupevov' TovrTov évruxdvTa pds TO aYLoV dvaTpexetv q ovyxabeivar ry xt 7H Baxtypia. Kock makes the excellent restoration TOUTOV TAX OVTA... jovyxa, 88 xariévar ri 7H Baxrnpia, most of which I would gladly accept. But I do not like taking taxty évra separately and still less disregarding the first ry altogether, and I would therefore read HOvxa. karvévas THY ext TH Baxrypia. dvarpéxev depends on raxvv. ryv is tiv éxi ty Baxrypia (686v), the way of walking with the help of a stick, 68és not unfrequently having the sense not of a road but of a journey, a going, etc., and being indeed the substantive corresponding to iévat. HEn10cuvs. 5.17 Snpoxpatia Oarépg dvow? éoti, TE 8 dpioroKparia Oarépa. It is very unlikely that Heniochus would write the article with 6arépa, impossible that he should separate them by dpirroxpatia. Blaydes and Herwerden are there- fore right in trying to correct the verse, but their sugges- tions are not at all convincing. I would rather read dpicro- kparia totvoua. If we emphasise dpicroxparia, the repetition of évoua is not disagreeable. The copyist has but repeated the wrong word. TIMOCLEs. 1.1 wOs ay pey obv cacetev TBs } Kvwv ; Neither pév nor oty appears apposite. mis dv pe viv Meineke. és dy tw’ otv Kock. Perhaps ras 84 rw’ av, 30 Perhaps a question, co. going with gatveras XENARCHUS, THEOPHILUS 89 31 Tols mev TeOvewow EAeos emretkys Oeds, tots Coot 8 Erepov dvocidtatos bOdvos. For érepov Meineke suggested éedpos, and Herwerden would write épywv. Haupt’s érepos adverse (see on Eubulus 9 above) seems to me better than either, but it does not account for the termination in -ov and dvocustaros is awk- ward with it. I would therefore read the partitive geni- tive érépwv of adverse powers. Kock has proposed éy6pds and Headlam éx6pév, but the word is not nearly so suitable. We want something to carry on the half literal, half metaphorical idea of ésvetxys Geds. If compassion is a divinity, envy should be so too, and no doubt Timocles has in his mind also the @Odvos of the gods, rd Getov wav éatt dOovepdr, etc. XENARCHUS. 4. 18 as 8 or’ ely eor ol’ Spav7’ ely cadias, del 8 rerpepaivovra Kai poBovpevoy, Sedidra, ev TH xerpt THY Yuynv exovTa. Read oir’ idciy gor’ ovr’ i86v0’ Spav cadds. There are many cases of two adjoining words getting thus confused. In the third line I would suggest Sedidta 7’... exe. dedidra is not a mere pleonasm along with doBovpevor, because that goes with épav and S8edid7a with éyew. Whether yuxyv should be altered is another matter. For the change of éxovra to éyew cf. Alexis 116. 6, where it is clear that éyovra should be éxov. THEOPHILUS. 6 od oupdéepor véa ort mperBiry yuvy. Gomep yap dkatos ovde pixpoy weiPerat évi mpdadiv, To weiop’ aroppygaca Se ék vuxros erepov Ayev’ exove’ ebpioxerat. The meaning is apparently not that of the parallel passage in Theognis 457 foll., which must have suggested these lines. There the wife does not obey the paddle, as the dxaros does, nor remain at anchor, dmoppyéaca de «7A. Here the dxaros itself, to which she is compared, is supposed not to obey éi wySaAt~. Presumably the dxatos 90 PHILEMON is towed by a bigger vessel and has no paddle of its own: this is not enough, and it breaks loose. So the wife will not follow in her husband’s wake, but she breaks loose too. As far as I can see, the critics do not quite make this out. 12.8 A lover says of his mistress :-— qv ear’ idely pdtoy 7) 7d Oewpixdv éxovow july diaveweyw éexdorore. ipiv Stamovety Herwerden, tyiv Sdiadadeiv Kock (iptv Grotius), understanding tiv to be the spectators and the verb to apply to the actor acting a part before them. Both verbs are unnatural, and the whole idea of the actor thus dropping his character in the play and addressing the spectators is inconceivable. Even in Aristophanes, where it does occur, it is introduced in a very different way. The passage is an obscure one and presents great difficulty. With considerable doubt I would suggest the very slight change to éxovoar (cf. on Ar. Eccles. 171 above). Her personal attrac- tions make her pleasanter to see than if she had (or than a woman who had) theoric money to distribute. It is very far-fetched, but the sense must apparently be that with any reading or interpretation. PHILEMON. 4 av. ’-els dravras yupes avOpwrous, SdAwv: ot yap déyovot rodr’ iSely tpOTov vopoy, Syporikev, & Zed, mpaypa Kal owrypioy, (kai pou Néyew Todr’ éoriv dppoordy, SdAwv) peotiy 6pavra riv wédw vewrépwr, tovrous 7 éxovras THY dvayKaiay picw dpaprdvovtds 7’ eis py MpoajKov Hy, OTHOaL TpldpLevov TOL yuvatkas Kara TOTOUS Kowds aract Kal KaTecKevacpevas. A good many changes are needed here. In 1. 2 Kock would make 3éAwv and véuov change places. This may be dispensed with, if we write ot ydp...mp&rov as a paren- thesis. Then what is the point of 4, this is suitable for me to say? Should it not be coi, anything Syporixdy being characteristic of Solon? The line need not, perhaps should not, be made parenthetic. In 8 ro. is unmeaning, and it is PHILEMON 91 hard to see what it can stand for: Meineke’s tore would be very feeble. Read perhaps orjjoai oe mpidpevov and omit roe as a stop-gap inserted when ce was lost. oe seems required, for it is difficult to supply. In 9 I suggest xowovs and xare- oxevacpevous for Kowds and xareckevacpévas. Kata TOTOUS can hardly stand alone. With oir’ idety cf. Dem. 21. 60 ovdels TuTore TOT’ cidey TO wACOveKTypa: 25. 41 obtos mpayp’ édpaxev puapov kat dvaidés: 57. 49 obdev édpa rovrwv. It seems to mean thought of. 22 kav Soddos 7 Tes, ovdev Frrov, Séo7rora, avOpwros ovtds éoriv, dv dvOpwros 7- I think there should be no doubt that C. F. Hermann’s ay dvOpwros 7s is right, if you have proper human feeling. Cf. Menand. 761 és yxapiev éor’ dvOpwros, drav avOpwros 7 : Ter. Ad. 107 et tu illum tuum, s¢ esses homo, sineres nunc facere: Cic. Phil. 2. 16. 39 homines, quamvis in turbidis rebus sint, tamen, st modo homines sunt, interdum animis relaxantur. Herodas 5.15 éydus, Tdorpwv, 4 oe Ocioa ev évOpwrots illustrates rather dvOpwaos otros éoti (Petron. 57 homo inter homines sum and 74). 28 See p. 115. 73.7 4 Ady 8 exe Gorep Ta Sév8pa TodTo Kapmov To Saxpvov. rovro seems unlikely. Perhaps Philemon wrote dozep 7a Sédpa Kaprov, ottw 76 Sdxpvov (or ra Sdxpva, which Plutarch’s text gives as against Stobaeus). In 31. 7 Dobree corrected rovrois with great probability to otrws. There is no objection to the dactyl in the fifth foot; cf. e.g. 76. 4 ei Adyou Tis SuxeAucd. 93.5 qv 8 dy cicevéyxnta piow éxacroy, ed0bs Kal vopov TavTyy éxel. elcevéyxntar contributes does not seem right, nor even brings to bear, if we translate the word so. Did not Philemon write éfevéyxnta, produces, develops ? 95.3 I think do should be id. The éyevnOy twice following seems to show that the meaning of 3 is not ‘from the time of,’ but éyev76y br6 ‘was created by,’ and in this sense dao would be wrong. 92 DIPHILUS 96.5 vovi dé dtadépovra trois Kad’ jyepay Saravjpacw 84 rov Biov 6p pdvov. ‘vy tov Ada Sip Bentl. vi rév A’ cicopd Heim. 897’ airov eloopd K.’? But the words only want rearranging : vuvi 82 tots Kad’ Hugpav dpa udvov Saravypactv 84 Siadepovta Tov Biov, or Saravijpacw épa povov dy. Cf. on Menander 537. 101. 1 éya | Alov pev evar rHv N.? 183 Read ody dre or od tePavpox’ ci. There is no point in Kock’s ovxér, for ére is quite inappropriate, as far as I can see. His No6w for vouw does not commend itself to me. Is the simple dative in referring to a play ever found in Diodorus? It is common later. DIPHILUs. 17.5 For adrfs read ary y’. 32. 6 av droAave totrov 7 Tov Biov. Read roi Biov. 60. 3 radautwpdrepov ovdey ear. yap a , : 7 a > * THs yaotpos, cis qv mparov éuBdAois, GAN’ ody Erepov ayyetov. Meineke’s Bpwrov éuBadeis adray will not do, because the first thing mentioned is soup, which is not Bpwrdv an eatable ; nor Kock’s dvr’ av éuBddous dua, because there is no question of putting in everything, or even two things, together. aavra however may well be right, and I should suggest cis qv mpata pv mévr’ éuBadgcis. The pev has no regular dé following, but cot rpooribyus réAda. takes its place. There is nothing wrong in dAd’..dyyciov without a preposition. Cf. for instance Ar. Plut. 399 3 * A , 9 a, , . XPE ovK Cort Tw Ta mpdypar’ év TOUTY. BAE. ti dys ; ov T@ peradodvat ; Theoer. 1, 117 odxer’ ava Spupds, odk dAcea MENANDER 93 62 dyvoeis év tails dpais Gre éeoriv, ci Tis wy ppdcet’ spbds bSov i) wip évavce’ 7 SiapGetper’ BSup. I doubt whether 4 zip évatcee can stand without a i. In any case it is singularly awkward that this clause should be negative in meaning and the next positive. Should we read px wip évatce’, } 8. 0.2 MENANDER. 6.3 Read 8€ for ydép. mepirxehécrepov certainly means ‘more firmly.’ (23 iiov émBvpdv pdvos pot Odvaros oSros patverat > ed W x f a a, edPdvaros, éxovra. woAdas XoAAdSas Keio bar mayvY, K.T.r. Meineke takes émOupdv as a nom. pendens and supplies Oévarov with idtov. Kock conjectures vy A’ évOvpoupéve pot. But is it not pretty clear that iSsov éiOup&: pdvos «.7.X. is what the poet wrote? ‘I have a private desire of my own.’ Cf. Meno 77 0 ri érOupetv Adyets; ‘what sort of desire do you mean ?’ etc. 70 mav7’ éotl TG KaAG Adyw iepdv: 6 vots ydp éorw 6 Aadyjowr Oeds. wavrayy and révry ave conjectured, but what is 7@ card Adyw? and what is the general meaning? I take it to be that a man’s (everyman’s?) soul is a iepov trav Kxadav Adywy, in which reason gives the divine response: yvyi oixyrypiov dainovos, sayS Democritus. tév KaAdv Adywv is obvious ; the other words must be conjectural. yoy dé mavrds éort is open to the objection that probably an article would be needed with yuyy. . In the Classical Review 12. 433 I commented on the considerable fragment of the Tewpyds edited after Nicole by Grenfell and Hunt (see now also a later text in Kretsch- mer De Menandri Reliquiis Nuper Repertis). I briefly record now with a few expansions and additions most of the suggestions I then put forward. 15 ri pirrdryy | ds odk dv adixjoaiy’ dv: od yap eioeBés 94 MENANDER 24 év roicd’ ey viv elu. 29 A stop after dv. Then perhaps yape?.. xdpyy ; 31 Kara Tuxyv. 34 xaddv y’ dv ein vy Ada is ironical and a continuation of Philinna’s words. 40 dca dépoper 5% taira rdvra, omitting the rdv6’ before dca. das at end of 39 is very doubtful. 42 od oe kabewpwv. 44 The papyrus gives paAAov Se mpdgewv eoopévwy, ay of Geol «7.4. Omit oi. 55 foll. read by G. and H. as follows évratda xpelas yevopevys avT@ Tivos kndepdvos of prev oikérat kat BapBapor élno’ éxeivos ory oipaley paxpay, éAcyov azavres: 6 5€ ods vids, otor[et vonicas éavtod matép’, aropbacals maw Hrepev e&érpiBev arevilev hayeiv mpooépepe rapenv0eil’ 6 wavy pavrus éxet - Galvr’ dvéoryo’ airov émysehovpevos. In the third line éyoe (¢ doubtful in the papyrus) is supposed to mean viait, he is no more. But there is no authority for any such use, nor any probability about it. The translation of éorw oipdlew paxpay by ‘we can do nothing but raise a long lament’ is very unlikely in view of the ordinary idiomatic sense of oipwfew paxpa (so it should be, not paxpdv) go and be hanged. I conjecture therefore first that e€yo" éxeivos should be éfjs éxetvw.. drayres, the phrase é&js wdvres or daavres being regular, and next that what all the slaves said to their apparently dying master, when he asked for anything, was cotariv (7.¢. coi éorw) olpa- Lew paxpd. Then I would read olor dv (ei being only conjec- tural, as is the ending of the next line) vopicas: dvop$acat, or perhaps dvopOdceu, 1.€. vopicas dvophacat dv otov dv éavtod watépa. Finally, keeping the zapeyv6etro of the papyrus, but again substituting ¢ for ¢ in the last imperfect line, I would read and supplement thus—zpooéepe, wapeuvOeiro, mdvu pavrus éxew Odbavr’ dvéoryo’ adrév émtpedovpevos. MENANDER 95 65 cxodhy rpiBuv. 69 depxo]zevov suggested here (not by G. and H.) is not good Attic for Menander’s time, the participle épydpevos not being in use. Perhaps diadeyopeévov. ; 71 wdOy)w’ érabév 7 kowdy is questionable, the usual phrase being wdcyew re without rdOnya, e.g. Dem. 18. 291: 19. 195: Aristot. fragm. 15 (45): Ken. Symp. 1. 9. kowdv however makes a difference. G. and H.’s text goes on kat xdpw thls emtedeias ds er’ ex wavrtds Adyou déolv aibrév dodoivat, pdvos 7’ dv Kai yépwv vow eoxe THY yop Tald’ tréryyTar yapety. For és the papyrus has w and what G. and H. give as a doubtful y. I should guess it was an: and read xdpw.. wer’ éx mavrés Adyou Selvy abrév dwoSotvar. vodv éoye is ‘had a sensible idea.’ (Kock) Fragment 128. 7. Probably atry for airy. 203 édeciv 8 exetvos euaber cirvxdv pdvos. drvyév Herwerden, who also reads éAewovs for éxetvos. The text as it stands makes sense, but not very satisfactory sense (vereor ne versus corruptus sit, Meineke), and I would suggest guabe SucruxGv pdvov. ci- and dvc- are sometimes confused. Cf. Index. ibid. yl’ éorlv dperiy tov drorov devyew Get. Kock would read réSiuKxov, which is not near enough. Perhaps rérovev. In Dio Chrys. 6. 16 dzovérarov has been ~ yead for dérordérarov since Casaubon ; in Ar. Met. 1. 2. 982 b 13 MSS. vary between dzépwy and drdruv. 285 irdpovoor ety’ abrov wavy, > , yoo , ’ a oP dxovapar’ eis Tpupyy te tadeverO det. raxovopara seems probable and should perhaps be joined to the preceding words. * 96 MENANDER 291 tavTopardy eat ds Zouxé wou Oeds , x na 3 Z f, owle: re TOMAG TOY dopdtwv mpaypdruv. adparwv Usener, which few will approve. Why not dopictwy in the sense of things that are or seem uncertain ? 301 If in line 5 is right, the lines preceding should be a statement, not a question. If they are made a question, 6’ should perhaps be y’. 302 olor Aadovpev dvres of rpradbAro dravres o§f pucdyres ep Eavrois péya. In the absence of any satisfactory explanation or emen- dation I may suggest that a stop should be put after dvres (cf. on 23 above), ‘What shadows we are,’ as Burke says, ‘we men who are talking here.’ Then oi should be & (cf. 404. 1) or possibly 7. 325. 7 yuvy toduteAys éor’ dxAnpov ovd’ ea Gi tov AaBovd’ ws Bovdrer’s GAN Eveori te ayabdv aa’ atris, waides’ éXOdvr’ cis vécov rov éxovra Tavtyy eOepdrrevorey émipedds, K.T.d. It does not seem possible to concur in the view that matdes are the good that comes of a wife. In that case the speaker states this in one word and after implying that there is only this one good immediately goes on to dwell in three lines upon a second, the attentions namely that in various ways the woman pays to her husband ; and this is not even introduced with an and or a besides. Evidently then this is the dya@ov 71, and zaises would seem to be a vocative, although it is true that we have several second persons singular further on. Cf. the aides who are addressed in Nicolaus 1. 41, and the plural in Menander 482. 1 wavoacGe voiv éyovres. Eveors may be right, & meaning in the matter, but it seems odd and we may think of d\N & éori 7. In line 3 should Avmjcavra be Aumycovra 360 xpyotods vouiLopevous epodiov dodadés els Tdyto. Katpov Kal TUYNS TaTaY porHy. Although Grotius suggested ypyorois vopfopévors long ago, Bentley’s objection seems to have prevailed with the MENANDER 97 editors, that the participle would be wrong and the infinitive required (xpyorovs vouilerOat), because the épddiov would otherwise have to be something distinct. He would hardly have argued thus, if he had called to mind such constructions as mpére, ounpéper, etc. with a participial dative, e.g. Dem. 22. 37 dow oupdéper paddov Hiv Karayvod- ow (1.€. karayvavat), Xen. Oec. 4. 1 uot mpéror ay padiora eriysehopevy (4.€. ériyseheiobau), and Hell, 4.7. 2 «i dciws av éxou airG py Sexoueve (1.e. py déxerGar) ras orovdds. He thought a verb governing the accusative had been lost. Kock writes ypyorss vépipds re voids. But no doubt Grotius was right. The text of Stobaeus has rixys porny aca, and poryy macay tuxns would be at least as good as rdyys Tacay pony. 473 ov Tévy TL yypdoKovow al Téxvat Karas, dv py AdBwor tporraryny hirdpyupov. ‘yap mparrovo. Herwerden, yap 6d¢AXovow Blaydes. The sense required seems to me, not that the arts do not flourish, unless their representative (for that is what mpoordrys means) is gain-loving, but that, unless this is so, they do not afford him much of a livelihood. I would therefore read, what is also nearer to the text, od wdvu rt yap Bockovow ai réxvat kadds. They do not maintain him well, unless he practises them in a mercenary manner. 481. 8 wavyyupw vopiody tw’ elvar Tov xpdvov = - Gv dype rovtov 7} “midynpiay, ev @ bxAos, dyopd, KNéwrat, kuBetat, SiarprBai. av mpa&tos amins katahicets, BeAtiova 2 AQ oY 3. ‘ > iz edd’ exwv dHdAOes, €xOpds oidevi: € f X , 3 a 6 mpoodiatpiBwv S¢ cxomidoas arddece kak@s Te yypOv évdeys Tov yiyverat. Three or four things here are or seem wrong. First é& o in line 9 should be & 9: the relative clearly refers to maviyupts and ériSypia (dzrodyuia in the parallel passage of Alexis (219. 10), but émidpuia is better as indicating the place), not to the time. év gy. may or may not be right. T incline to think it is. av wparos dains (or mparos aréAGys) katadvoes means nothing, Meineke av par’ admins és (sic) xatadvces, but there is no point in the mention of a H 98 MENANDER karddvots or im, and mpér’ is even more wrong than mparos. mpéos (Preller) may seem tempting, till we remem- ber that it means early in the day or year, not before other people. The comparatives following point to a comparative here, and we should probably write wpérepos, which often gets confused with ap&ros. The best I can make of katadices is to suggest xarav’oas in the sense either of hurrying (évicas) or of having despatched your visit, but there seems no exact authority for either use. Porson was probably right in changing oxomidoas (or xomidoas) drdAcoe to éxorfacev (the a being long) dodéoas, and we must make dtrodéoas refer to épddia. 482 atcacbe vodv déyovres: obdty yap wAéov évOpdtwos vows éorw, dAN’ 6 THs TUXNS, ely’ éori rotro mvedpa Oetov etre vous, tobr’ éort 76 KuBepvav amavra Kal oTpEehov. & ris réyns, 1. vots cannot very well be right. The expression is almost or quite absurd and not to be defended by line 3, for it is one thing to say that chance really is vovs and quite another to talk of the voids ris r¥xns. More- over, etre... etre is inconsistent with assuming the vois. Should we not then read ddd 76 THs TUxns? Cf. Dem. 4. 45 76 ris réxyys cuvayovifera: and 7b. 12: Eur. L.A. 1404 7d Tis TUxns...vord: Alcib. 2.1474. whéov in 1 is doubted, but I am not sure of its being wrong. 531.13 (Of man in comparison with other da.) dobevérratov yap bv poe peylcros oikovopetrat mpdypacw* Gray wéoy O¢, wActoTa ovTpiBer Kadd. For oixovoyetrat, which makes no sense, we have such unsatisfactory conjectures as d:amovetra: and oixodopetrat. Considering the general sense, may we not feel sure that oixo- represents dyxo-$ éyxos is just the word for the superiority of man. Cf. such passages as Eur. Androm. 320 pupiowe 89 Bpordy ovdty yeySou Biorov dyxwoas péeyav: fragm. 81 tarred... héyew, és dyxoy 6’ ovx d&vw Brerew Teyys and 825; Soph. fragm. 856: Herod. 6. 126 dco odio re abroice Hoav kal watpy ééwyxwpéevo.: and very many others. For the correction cf. Reiske’s éyxov in 4 wéds eis éXdrrova oikoy MENANDER 99 owyxra (D. Hal. Ant. #. 3. 11. 7) and Eur. Ion 15, where yaortpos oixoy has long been corrected to éyxov. As to the latter part of the compound we cannot be so sure. déyxo7o.® occurs in the sense required (éyxomoueiy Kat ératpew rov Adyov) in a late writer and dyxomoeirar may be the word here. 532. 6 roy 8 tpdrov adits Tis yamoupevys pel? Hs Bidoeror pyr’ eerdoat pyr’ idetv. ceiciSctv Meineke. I should prefer ciSévat or Sudeiv. A few lines below (10) I suggest ray . . . kaGedoupévyy for the genitive. 536. 6 dyer should, I think, be éyes, matching émAdBoro (Porson for émAdBor) in line 9. 537 6 pev "Exixapmos tots Oeovs clvar A€yer dveuous, Dowp, yijv, K.7.d. Dobree was certainly right in demurring to rots Geovs. Epicharmus did not say that the gods were winds, but that the winds were gods. Dobree left out rods and put in a ydp after wey. I think @eovs and dvéuovs should exchange places. 538 Read dy for érav in 1 and érav d8ouropys for ds Gdouropeis in 2. ds when can only be used with a past tense or a historic present. In line 7 men were proud of their birth or wealth or fame, kar’ ovdey adtois Tavs’ éripxecev xpdvos. Possibly _xedvw or év xpévy finally. F. W. Schmidt reads ypdve (Kock) but joins it with the next line. jépov is much too poetical. «dr is Kock’s correction of the MS. kai: perhaps it should be xaérow. 541. 8 Wyttenbach’s cis 3 Sei, quoted by Headlam, makes excellent sense, if we may understand it of time at which, but not otherwise. 542, 3 Join icws with cvvyywvitero. 544 mapdderypa Tovs vpovs AaPeE. ¢ , 2 A ap oa , Grav ddywo’ ixOdv éxeivor, Sud twa aitav dxpagiay Tobs médas Kal yaorépa > a ” s Ce 3 XN eQr oisotew, éXaBov caxiov, elr’ eis Ty G8dv éxdGicay abrovs. ’ 100 MENANDER Kock is right in reading oiSdo.w, but surely not in ékei xai for éxetvor. I would rather suggest dia re rv for did twa (rv and rwé getting confused sometimes) or insert 7’ before éxpaciav. Grotius cfra caxiov édaBov cis 6 Sov exdbicav atrovs. 549. 5 mapyyope 8¢ 7d Kaxd. 8: Erépwv Kaxav. There is not much sense in this admonition. Read kadév and it becomes very different. ra xadd (552. 2) and xadd without an article often stand as a substantive. For the correction cf. on ’Adéo7ora 412 below. 557. 4 7a 8’ tia mpooriéact Trois dAXOTpiots. The second syllable of é\Aorpéous cannot be long. Hither rots d. is an adscript, corresponding to 74 Tv wéAas in 2, or it belongs to the next line and something, e.g. droBeBdy- xéres, has been lost before it. 573 dpyh mapadoytopds ror’ ovdels Pera. One would think the poet must have written Spyj Aoyopds ovderore wapapverat. Cf. 407. 3 adN eyyis dyafot mapamépuxe kai Kaxdv. mapa has strayed from its place. 580 Perhaps ei yap 6 deaxdrns darohdXex’ (or dzrAeo’) abrds mavta. 666 7 woAAa havepds & Set per’ dveidous AaBeiv. Grotius ddd for a dei, and see Kock for other suggestions. Surely @ per’ dvetSous Sei. 714 bray yépwv yepovti ? yodpnv 586, Onoaupds éwi Onoavpov ewropilerat. Meineke éxropiZerat, but Kock rightly challenges the construction with éri @yoavpdv. émi(or éu)poprigerat suggests itself. 734 Perhaps éorovdale . 1092 odr’ éx xepds meBévra Kaprepdv ALGov paov katacxely obr’ dad yAdoons Adyov. MENANDER 101 The comparative jgov is unmeaning and must be wrong, but it is harder to correct than the jdov in Monost. 470 (see below). Perhaps we should write fddiov for xaprepdy, since the hardness of the stone has little to do with it, and then égorw for faov. Notice the parallel from Plutarch given by Kock. 1097 ay déyvodv m1 rapa twos Oédys uabeiy, 76 TOD pabyrod mpHrov, exe Thy oryyy. Possibly 76 tot padyrod. rpdagopoy, cvynv éxe. A genitive is found with 76 zpéodopov in one or two places. 1098 6 iy yeAwros aétos dv 7 yéAws, K.7.A. Simpler than any change proposed would be the change of order, av 7 yeAws 6 py yéAwros a£tos, (cf. on 666) and this is also better than one line ending with yéAws and the next with xardyehus. Tue New (1907) Menanprr Fragments, The following suggestions on the fragments of the four plays found and published by Lefebvre are reprinted with very slight change from the Classical Review xxii. 48. "Hows. 26 xeppdriov ef ovvyypevov | jv da révev cot, Toor’ enol Sodvas téws | owe trép cov, my ceavtov mpdypara | cis TyAikatr aye. 33 py Katap@ pou. 65 apagew. *Exirpérovres. 1 Invert the two speakers in this line. 66 Kav | jv apodosis. aBev dv. 85 rovdi Adyov . daattet’t 92 dr. 98 Fv Sep rovTov A€ywy jew 66 122 tv Biov 8 Cf. 126-7. 157 airés dca dAaBdy eyes. 174 dwdSaov? Cf. Thuc. 1. 129. 183 mpoomaifas 2 belongs to S. 260 ais wiv yap &padAov kdpais abrdbe xOdpav (or Avpav) ératfov. 262 Kat pad’ od. 263 aida dy tis Fv. 304 off ciOvs. 340 roractixoy 76 yivatoy, womwep éprerd (or -dv), @.e. creeping things that go round an obstacle in their path, Plut. Lycurg. 9 éwei yaderas édpa zpodexo- pévous Thy avrixpus adaiperww, tepiyArAOev 68@ Kal Karemo- Airevoaro «.7.A. Cf. in a way Goldsmith’s saying that 102 MENANDER Burke wound into a subject like a serpent. 355 py’ dpa 6p with infin. as in Thuc. 8.60. 386 If rovovrow can stand for r. 7, then read ri ydp ris av vopiccey dAdo 76 yeyovds. 403 ratty’. 404 AoWopetr’. 420 atriv 8 Setas or Sei€es? 424 Guod y elev. 476 wis; A€yas .. . Biov all spoken by O. 478 fpdv of Geot. 487 cuvrpiBou oe ras Tappyoias, part of S.’s words. Cf. Soph. 0.7. 1478 xai oe ricde ris 6500 Saduwv dpewov } "ue dpovpyoas rvxor: Eur. Or. 1407 éppots ras dovxov wpovoias: Ach. Tat. 4. 15. 6 cot wey dyada yévorro rijs Siaxovias. 490 Geoi! belonging to O.? 515 éza- pwrrépus ouvqxe?t 516 odderi yap ciriyyua .. . eyever’. Tlepixeipopevn. 5 xal rod woAcuov. 28 pepynvur’. 31 ind tovrov. 38 7& Aowra 8 adrds iwordBo mis dv. 63 Leow tpémov tw’... abra. They do have a life of tt. 81 ris éo6’ 6 8 atrios; 82 rd’ dv iows or taxa Téws, 83 odv od. 103 perpiws with Aéyew. Cf. Ex. 19. 107 dros? 108 apis bedv. 110 evdupa6’ of edaived’ fvixa TAvKépa AdBor, or daivel’ vir’ dv 4 xépy AdBy. 111 TIAT. od yap éwpdxewv wore Zywye and jv for éor’. 147 zpaxOyjoerat rodro 7d yéAowov. 149 otyava’> éyd dé 0” dyap’. Sapia. 15 I have thought (very doubtfully) of «ai orepavovpevos Oe. 33 SovdAn Kal rode (ér’ ota TOnvicw 1). 66 wreparéov | abrév rapayayev éor. Tod padeiy & det? that I may learn from him what I want. 84 76 y e&vOadi. 93 mpds pe, vy Ala. 105 ueyé dye? 109 4dy ye, 2.€. orifs oe. Hoy so used in 155. 115 jpebiopevos. Cf. Timocles 30. 117 fv av ext ris airs Siavoias viv er. (zpiv is not used in common Attic, except as a conjunction.) 118 éuol 7 éwider’ ok dopevos. 124 xarédAaBev. 125 eEepydferat. 127 wore or ww (for ror) and perhaps rév pnd’. 136 xat rariyynpa pév. Cf. "Eq. 419, 423, 426, all referring to an irregular birth. 146 ris 6 yépwv; 159 Perhaps «dv 7 xai (as in the MS.) with aposiopesis or interruption, even though, 1.e. in spite of what passes or of your remembering what has passed between us. Cf. Ajax 520-1: Eur. Hec. 828-9. 178 Possibly érepa:, the con- fusion being common. 196 éuBpovrycias. Ar. Ach. 833 probably zoAvtpaypoovvys. 209 dyaiv euapyoev dreddv (1), tov tidoiv 8 drrdpevov dov dpricev. 235 devd y. 244 ovk axyxoas Aeysvrwv; 260 dia Kevys vor and 308 otrus épuyes. 261 wraidas. 263 088 adv ei odarro. tis adrov. MENANDER 103 264 dOvpia, 265 cixds Fv. 270 cupmecetv. 274 ewovs yiyvopat. 276 odddpa. 289 weprowréov. 294 pour mdvv. 306 76 wav. 322 miBavds etvar pavOdvev? 327 diaxive, if the anapaest is right. 328 dyovo. ydép. 329 ra Ovpaé’. 336 dvrusiows. 337 omredoov. 339 dmévat. Tvdpat Movéorixor. Along with the fragments of Menander is usually printed a large collection of Tvéuo. povdcrixor, or apophthegms contained usually in a single line, traditionally derived from his plays. Some of them come from him, some from other comic poets, many from tragedy, others from other sources, and many no doubt are late. The MS. evidence for them is at present so uncertain (see Kock’s Preface to his third volume and also his article in the Rheinisches Museum, Vol. 41) that it might be the part of wisdom to follow Kock’s example and not deal with them at all, the more so since their absence from his edition makes it difficult to ascertain what scattered conjectures have been proposed for the improvement of the text. I will however set down what has occurred to me in the course of several readings. I give the text and numbering of Meineke’s larger edition of the Comic Fragments (4, p. 340). 6 Gmav 70 Képdos adtxov bv héper BAGByv. dav 76 xépdos would naturally mean not all gain but the whole gain. Read dravr. 30 avip dpioros ovk av ein Suoryevys. A superlative is quite out of place. Perhaps dvyp 6 xpnorés. Cf. Meineke’s probable conjecture of ypyord for dpiora in Aesch. Sept. 183 9 ratr’ dpiora cal rode cwrnpia «.7.A. and Orelli’s dpiorov for dypyorov in Aen. Tact. 2. 1. 53 apxnv vopule tov Gedv poetic Gar. Perhaps ¢ofeicba: is a gloss on 76 Sediévan. dpyyv seems imperfect and a genitive may have been lost before it. Or should we read dperqv! 104 MENANDER 60 Bpadis pds pyiy éyxparis pépetv yevod. éyxpatas pépwy | 65, 66 Biot pay oddels dv rpooupetrar Biov. Bios xékAyrat 8 bs Bia wopieras. These two lines should be joined together and or read for és. It is not clear from Meineke’s note whether there is any MS. authority for dru. 73 BonOes tot Tots Kadds cipyaopevors. Probably o° cipyaopévos. In 601 fjret rorety ed Tots KaAds eipyacpevots read rods Kadds o” elpyaopévous. 114 Sixatos elvas waAAov 7 xpynords Oérc. ‘Corruptum xpyords. Sententia postulat felix aut dives’ says Herwerden. May not the sense be or resemble that of the English proverb, ‘be just before you are generous’? xpyores might express either generosity or a weak kindli- ness not always admirable. 115 Set rods Prroitvras riot, od Adyous Exe. It is hard to make any sense of od Adyovs. Should we read eddAdyws # 119 Sixatos tot iva Sixaiwv Tvxys. Non est antiqui poetae says Meineke, but Sikacos tafe tv dixatwy iva tuyys might be. 158 éavrov obdels buodoyel Kaxoipyos dy. The meaning should be ‘ no one admits to himself’ (apds airév? atrg ydp!): a statement not more untrue than Juvenal’s se tudice nemo nocens absolvitur. 167 eixatadpdvyrds éore ovynpds tpdzos (avyndds Blaydes). It will be another example of a well-attested error, if this stands for dxaragppdvytos (Hirschig ob xaradpévyros). The point may resemble that of Eur, Med. 319-20, or that of Ar. Ran, 916, MENANDER 105 178 cipeiv 76 Sixaiov mavtaxGs ov padiov. This being rather pointless, perhaps we ought to write elrew for eSpeiy, understanding cireiy to mean speak out, dare to utter. Cf. p. 79. 182 zore Kav Kakolow NOovis Te weTpov. Perhaps pérpov rt 8’ gore (or, if the line is tragic, éorw Tu pérpov) Kav Kaxotow Hdovis. 185 edvodyos GAAo Onpiov tev év Biv. This may be the remains of something like kdéxiov oddév éory 7} | edbvodxos dAdo K.T.A. 191 £90t rpovexdvrws bs paxpay eyyds Brerwv. és paxpdy? or doa paxpdy (te. err)? If apoo- exévrws might have a dative, rois pwaxpév would give fair sense. 194 Giv Bovddpevos py mparre Oavarov déta. Oavdrov y déia Meineke. Oavdrov ’mdéiat 6. ra&ta? ra 6.4.4 203, 204 70 rovnpa thy plow Siactpédet. HOos Tovypoy evye Kal Képdos Kaxov. "H6n and 400s should be 6 and éos. 206, 207 9 ydtordv eore TOv brapxdvTwr Kparely. yoiorov éorw edtvxovvTa vody ExeLv. ‘Pleasantest’ is a strange word to use; yet cf. a fragm. ascribed to Democritus (Mullach 248, Natorp 5) 780 pydev érrodéxeoOan jv wy Evepépy. ‘ Best’ (dpicrov) or ‘ difficult’ or ‘uncommon’ would make sense, but I do not see clearly what the Greek word is likely to have been. Was it éiordy or eOioréov ? 209 Heer TO yhpas aca airiav pépov. mwacav aixiav? I should take it in the sense of 552 & yijpas éxOpov cupdrov dvOpurivey, dravra ovAdy Ta Kaha Tis eipopdias, x.7.A. and 648, an imperfect sentence (v. Meineke on 32), dxavr’ davite yipas, ioxiv odparos, t.e. in the sense of ‘sans teeth, sans eyes, sans taste, sans everything.’ Democritus calls old age éAdKAypos mHpwcts, 106 MENANDER 221 you cvomav 4 Aadely & py mpéres. Here again 480 must be wrong for two reasons, and this time it is easy to see what the original word probably was. Read é\od (EAOY for HAY), for éAod 7 comparing Dem. Ol. 2. 22 éXoiunv 7 and other passages. 229 Oediv céBov Kai rdvra pages evOéws. For évéws we should read civ Ocois or ov Obed ‘with heaven’s help.’ ’Ev and ovv are often confused. 235 Oyoavpds éort Tov Biov Ta mpadypara. I conjecture ypdypara for mpdypara, as the words are sometimes confused. See Schifer’s Greg. Cor. 1. 185 note and elsewhere. For the sense compare such other yvapar as 312 Api wépuxe rao waidela Bporois, 403 6 ypdppar’ cidws Kai mepirrov vodv exe, 438 6 ypappdrwv arreipos ob Brera BA€twv, 652 Baxrypia yap éort wa.deia Biov, 657 dimAotv Spdow of pabdvres ypdupara, with the saying which Diogenes (5. 1. 21) attributes to Aristotle, xédAducrov épddiov 76 yypa tiv matdetay ercye, and Diodorus 12. 13. 2 (end). This seems much more likely than Kock’s attempt (Rhein. Mus. 41. 90) to connect the line in meaning with Aesop Fab. 22 (98) 6 pidos SyAot bre 6 Kdyatos Onravpds éott Tots avOpurrots. 246 buoia peyiory TE Ged 76 edoreBetv. 76 y Meineke. Possibly 76 uy doeBety or wy SvcceBetv. 249 Ovyrés weduxas Toiricw repd BArev. rovmiow is of course the future, and why should a man be bidden to try to foresee the future? O@vyris weduxws points plainly to the exactly opposite precept and we must conclude that in the previous line (or perhaps in the same, being a tetrameter) wy was found ; cf. 362. In like manner 327 A€ovre ovliv 7 yvvaixi cupBiodvy presupposes something missing, ¢.g. édod, and 613 py pdvos étictac’ dAAG Kal Sodvat MENANDER 107 pido is seemingly imperfect ; perhaps also 680 xadov 75 vyew 7 TO TONG Kpatmadav. 257 loos tof waor Kav brepBadAys Biv. Read Bia. Cf. Soph. Aj. 130. 262 ixavov TO vikav éml tev édevbépwv. €or. TH y' édXevOepy? Or we might try something like kadov 76 viKav emi ye Tov éAevOepwr, as in 299 Kaddv TO viKay, «7.4. But the sense is uncertain. 270 ixavas Buooras ynpoBookav Tous yoveis. Here, if not in 262, the confusion of ixavéds and kadds seems plain. Read xadws. Pisces may be right. 299 Kadov 76 vixdy, drepvixay 8& opadepov. trepéxew 8 odk dodadés? wav seems accidentally repeated in tmepyixay. 318 Asi woiov pév, ddumia 8 dppos Biov. Read Auynv tAotov mev dpeos, Tod Biov 8 dAvmia. Cf. Meineke’s note. 328 Addet wérpia Kal uy Adda. & py oe Set. Perhaps pérpia Aadyjoes py AadOv a py oe Sel, or p. AaXeis, ef wy AaAEls A py oe Sei. 331 AaBE mpdvoray rod mpoayjKovros Biov. Read Aafetv rather than with Meineke dood. 337 pubs diddoxer ypdppar’, od diddcoKados. MicOés seems:unmeaning. Perhaps pdxfos, as in the next line: poybely dvdyxy Tovs Oddovras edTuyety. 355 paprore AdBys yuvaixa cis cvpBovdLav. Meineke yvvaixas, but the plural is less likely. Perhaps yrvaixa py AdBys wor’, or eis may be an error for zpos. 108 MENANDER 362 py ydper yuvaixa Kod dvoi~es Tabov. I would either read pi | ydwer yuvatka Koi droits radov, or still inserting oo: regard the line as part of a trochaic tetrameter. There are many such commingled with the iambics in the Sententiae of Publilius Syrus, and some of these yvGua: may very well have had that form originally. So 318 above. 364 put) mpbs TO Kepdos <76 ody> det reps Brerew! 367 peta Oixacov det SiarprBas voter. Rather pera rav dixaiwy StarpyBas det rovod. 374 véperw puddrrov pydty trepdpovav ? 376 vixg yap det diaBoAy Ta Kpetrrova. Should we alter vx@ to some verb meaning ‘aims at’ or ‘assails’? Possibly Sdxveu. 385 vv pev dvaraver, qyepa 8 epyov zrotel. Read dvaravow. Is roe right? Perhaps dépe. 424 oddels 0 voeis pev older, 3 52 motets BACret. This could only mean ‘no one knows what you think and sees what you do.’ Read 6 voeis ev ovdels otdev, 8 58 motets BAgre, 1.€. ‘what you think, no one knows, but (he) sees what you do.’ The change from odde/s and nemo to a positive idea is common enough: e.g. Plato Symp. 1925 oud dy els éfapyyfein... ddAG... oor’ av «7.A.: Hor. S. 1. 1. 1-3. For the meaning cf. Publilius Syrus Sent. Append. 11 (Ribbeck Com. RB. Fragm.) Cernuntur facta, nemini animus cernitur. 428 ols pev SiSwou, ols 8 dpaipetrar rdxy. ots 8’ (perhaps merely a misprint in Meineke) should certainly be ois 8’. 430 6 pdtv cidas obdey eFapapraver. There might be more truth in this, if we wrote eirdy for eldds, MENANDER 109 439 oddels rip eis Xpypata. didovs éravoaro. Meineke says in wip cis latet genitivus substantive dolorem vel malum indicantis, and rvpdés or muperod has been con- jectured. Rather, I should say, the words conceal an aorist passive participle in -es, and the well-known 6 py Sapeis dvOpwros ov taideverat (422) suggests that dapeis is the word. This very day (Nov. 6, 1896) it is reported in the Times that Mr. Justice Wills thought ‘a punishment so érrevoc- able as flogging should not be added to the existing punishment.’ 44] Spxov O& Hedye Kat Sixaiws Kddixws. If this means ‘in every honest or dishonest way avoid taking an oath,’ it seems little better than nonsense. Why should one avoid it so anxiously? Why dishonestly? We might read kai dicatwy xddikwv, understanding that an oath adds nothing to the word either of an honest or of a dishonest man, since ovdk dvdpés épxot tiatis GAN’ dpxwv drip : or detye Kat Sikatos Kdduxos. 446 ToAAovs 6 Katpos ovK dvras Tovel Pidovs. Perhaps zoddods 6 xaipds, obx 6 voids, moved pidovs, circum- stances (or interest), not real feeling. Transposition of the words is easy enough, but emendation gives a much more pointed sense. 447 TOAAOL pev evTVXOUCW, oO Ppovodor dé. Either, as sometimes happens, pév is much out of its place, or the writer must have gone on with something like ToAAol 8 Ppdvipor Suervxodcr Tov Biov. 448 MpaTTEe TA TAVTOD, py TA TOY GAAwWY qpdvet. Probably zpdrrwy: perhaps zpdrrew and ¢poveiv. 452 matnp ody 6 yevyyoas GAN’ 6 Opépas ce. Herwerden introduces the very poetical word 7éde (rarip o & Opépas, odx 6 yervyoas wéAa). This may be avoided by some such turn as éoriv o° (or a2 8 jv) 6 Opepas, ovx 6 yervjcas, Tarp. 110 MENANDER 456 movypes cot. Tas axdpirros avOpwros. Many turns are possible: was tis dxdpiotds y' dvjp, was dxdpiaros dv | dvOpwros, etc. 465 porn ’orw judy 6 Bios, dorep 6 Lvyds. It seems strange to say either of human life or a balance that it actually 7s a pory. Perhaps we should read éy for}. Cf. on 249 and 362. 470 pdov dépew Set ras tapectaoas Tuxas. faov should perhaps be apdws. Cf. 739 dpovodvrds éore fnpiav mpdus pépev. 478 and 629 curnpias oynpetov auepos tpdmos. Though this is found twice, there seems little meaning in cwrypias. Some word signifying ‘ prosperity ’ seems wanted. It might be edypepias, which with jyepos would give us a play upon words. «vernpias would be doubtful prosody, and edmpayias is hardly near enough. 523 tavos Savoy dvOpurois KaKov. *Yzévowa has been suggested, but I fancy the original line was trvos mepitros Sedov avOpwrois xaxdv. Cf. 88 yéAws deat pos év Bporois devov xaxov. Cf. Plato Laws 808 B and Diog. L. 3. 39 against too much sleep. (I find from Kock in B.M. that Meineke in his editio minor proposed tavos 8 dxaipos.) So Webster in the Duchess of Malfi 1. 1. If too emmoderate sleep be truly said To be an inward rust unto the soul. 536 dpdvnpa Aurapéy obdapds [dv]adioxerau. Meineke Xaprpdv. Perhaps idapév with 8’ or y’ before it. ‘A merry heart goes all the day.’ In Isocr. Hp. 4. 4 (p. 228 below) I suggest idapdéraroy for the unmeaning \ryupdirarov. 538 xpvods 8 dvotye. wdvta Kaidov wvAas. Meineke is inclined to accept xiv AfSov méAas. As far as the sense goes, one would rather expect adj AiSou amvAas. . MENANDER 111 559 Os TdvTa TYnhs ore TAHVY TpdToU KAKO. Is idows’ for rizys too bold a conjecture? For the short ef. 326, 577, 622. 590 iros dirov Seduevos odk ext Pidros. Unmeaning. The author wished perhaps for a candid friend and wrote didos pidov peddpevos odk Eorw idos. Cf. Syrus Sent. 10, amici vitia si feras, facias tua. 606 loyve copia kdpery, xpdve 8 uy. For xpévw we might read Adyw (xpdvos et Adyos ob similitudinem compendiorum quibus scribuntur, . (fre- quentior est nota inversa °) et . passim confunduntur Bast in Schifer’s Greg. Cor. 1. 33, note), meaning ‘mere power of words.’ Cf. Index. xpvog would also give a good sense. 612 pndérore wAovTEL, Odvaroy evAaBovpevos. In wishing for something like t8piZov in place of rAovre, Herwerden curiously misses the point. It is that a man is wise in not accumulating wealth which he may never enjoy. 636 Wouxis dpxaias ovdey ore yAuKvrepov. *Axepaias Meineke: ikapas Schmidt. Perhaps dicaias. The three letters a p x are known to get confused respect- ively with 6.x. But are just people always the pleasant- est? 644 dvaravois gore Tov Kaxav ampagia.. This might be taken to mean that a man can escape troubles by doing nothing at all, and some one may compare Syrus Sent. 377, nil agere semper infelicist optumum. That however means rather that a man had better not try his luck, if it is habitually bad: and here avdravois must mean cessation, not avoidance. When in trouble, a man cannot get out of it by mere dmpagia. Was the real word drapagia, meaning that your troubles cease to exist, if you cease to concern yourself with them Not to feel them is not to have them. 112 APOLLODORUS, PHILIPPIDES 647 aird oe Siddoxe: tod Biov 7a wpdypara. Perhaps &8défe. I have also thought of rév Biov, but the genitive may very well stand. 653 BpaBeiov dperis éori edrradevoia. Excellence does not however bestow a good education. It is good education that bestows excellence. Perhaps therefore we should read something like Bpafetov dpery éorw (or dpery Bpafetdy gor) edradevoias. 682 Aemriis yé ror Liv Kpcicoov 7 AapTpas KaKds. For errs yé ro. Blaydes writes Autés xadGs, Schmidt AerrGs yap ed. Aapmpas xaxds also seems to me doubtful, and I suggest Aurobs yap ed Lav kpeirrov 7 Aaympods KaKds, the adjectives going closely with the verb. 711 feyxet tapovons Tis TUXNS TA mpdypara. Is péyxet anything but a blunder for tpéye. ‘things go on swimmingly’? Cf Pind. Pyth. 8. 32, 76 8 év woot poe tpdxov irw, where however tpdxov is not always taken with iro. APOLLODORUS. 3 rots yap pepywvdcly te Kal Avroupévors daca vue eouxe faiver Gon paxpé.. Tt does not seem to have struck anyone that éowxe, which with ¢afvecOa: is very weak, if not unmeaning, should be ” eiwée. 13. 7 Should dzorvyeiy be émitvyeiv’? airdv in that case could not be right and might represent rdvrwv. Notice the repetition of was (4 times) in the lines immediately following. PHILIPPIDES. 26 ovk cor ‘eucbvcbnv, wdtep’ Néyovta , ‘qpaprov’, dare mpds éue ovyyvepys TvyElv. If this were the merely superfluous dove (Soph. Phil. 656: Plat. Phaedo, 103 8, etc.), it would almost certainly come before the participial clause and follow close upon SOSIPATER—ATHENION 113 gor. Read ds ye mpos gue. So for instance in Plat. Prot. 348 c the MSS. vary between dore wor (unmeaning) and ds y’ epoi. SosrPatEr. 1. 56 I suppose something to be lost after ris 3 7. EUPHRON. 11.15 ovdey 6 pdyetpos Tod ToinTod Siadépe’ 6 vots ydp éorw éxatépw TovTwy Téxvn. I cannot see any meaning in the last line, though it does not seem to have been questioned. What sense is there in saying that their reason is art? Is 6 vots a mutilation of épdvous, like-minded? In D. Hal. Lystas 15 (486) of Adyot is plausibly restored for époAoyei. PosIpIPPus. 19 ovrw moAvmouv éoriv 4 Avay Kakov. Various suggestions have been made for roAvrow. I have sometimes thought we might strike off ro and read ovrw Tt AvToby éotw y Avmy Kaxdv. It would be an identical proposition of the type of ‘much human nature in man,’ etc. Menander says (667) otk gore Avans... dAynya peilov tav év avOparou dice. 26.6 oxeddv "r ? ATHENION. 1.21 ds 8 jpeo adrois vorepoy, kal Tovs das mpoodyovow non TOV lepay yeypappevuy, 7a Tatpia Swatnpodvres. Surely zezpaypévev, ‘but not till the sacrifices are over.’ In Lys. 14. 2 rpoyeypaypévov is a mistake for werpaypévwv. Cf. p. 106. I agree with Meineke that something has been lost between 24 and 26, unless 25, 26 themselves belong elsewhere. Possibly they might follow 43. 26 is too like 16 to come anywhere near it. 114 NICOMACHUS—’ASéo7rora Nicomacuus. > a f 3 2 ft 4 2 epyadés cor é Bio BeBuoxdra : : Tovs Tov POovotytuv rdavras 6pOarpovs Aabeiv. There would have been less desire to alter éy Biv, if scholars had remembered, what most of them at any rate must have known quite well, that in latish Greek {ios sometimes means the world, society, etc. ev Biw BeBoxas is therefore a man who has lived in the world, as we say, and not in retirement or seclusion. Po.iocHus. 1 \6y should probably be é6éAy. *Adéarora. 108. 3 Read airéy for airds. 115. 6 aévrwv éori! 120 jrrwv éavrod topyidiw tpicabdLo éavrév otrw mapadéduxe(v) ? qrrwv éavtod requires an dy or yevdpevos, and the latter word just finishes off the line. 123 Perhaps év péow yh xetpévy. 212 The verse which Tiberius quoted to Agrippina (Suet. Ted. 53: Tac. Ann. 4. 52. 6) should be, I think, el wh Tupavveis, Téxvov, dducctobat Soxets ; I do not know why Kock gives it to comedy. 404 ioxupdv éort wovctos dyvoovpevov éxwv kAnpovdpov. Read kAypovopov exwv. 412 7 8 dpery povy kal 61 kadod rod odatos Katadaiverat. Is it not obvious that xadod should be xaxod? C£. on Menander 549. 5 above, p. 100. *Adéamrota 115 463 od dia tpamedlns 008’ ep’ (ad?) Eorias pidovs. 1206. 4 7G rovoupévy ? 1224 et yap Has 7G Ged Over Grav yen Katopirrytat Tadw, odx GTav yapy. As the last three words are fairly certified by the metre, we should probably read éray | yuvatka xaropitry Tis, odx étav yay. ADDENDUM. Philemon 28. There are storms not only for those who sail the sea, but also for those who stay at home. * A tS + ee / , xoi mev wAdortes eviol’ tucpay piav H vixra xetpacbevtes ei’ ex Tod KaKod A a cutypias éréruxov' 7) TO rvetpa yap abtovs Th o@Lov Hey } avy ywjv- > ‘ ~ a? > mw > ? € , éuol 88 radr’ obk got: odk els Hucpav , s , 3 Syn a ¢ xetpdlopar piav ydp, eis 76 Liv 8 ddov, a %, Q a“ XN a Z del 76 Aveta Oar SE peiLov yiyverar. So the fourth line stands in Meineke and Kock corrected from 76 o@lov pevy ebavy Aysnv. The article with mvedya and oéfov does not seem right, nor the imperfect 7jxev side by side with gnomic aorists. Perhaps the poet wrote something like 7 76 mvedua yap airois tows dvijxev, ‘either the wind drops or a haven comes in sight.’ Cf. Alexis 15. 17 dowep wvpetos avijxey and Herod 2. 113 od yap dvie 7a mvevpara, with 4, 152, Lines 5-6 give the antithesis to dvijKev. THE DICTION OF ARISTOPHANES. Tuosr who have attended to the language in which even the iambic parts of Attic tragedy are written know how remote in many ways it is from the speech that in the fifth and fourth centuries was actually used by the people of Athens. The poets differ of course among themselves. Aeschylus keeps furthest away of all from the usual language of living men, Sophocles and Euripides draw nearer to it. It is even said sometimes that Euripides in particular takes it as the staple of his own vocabu- lary, as though he wrote in’ the Attic of his time and only threw in an old-fashioned poetical word now and then to grace his verse. This however is far from being the truth. There is one standard poetical language which all the tragic poets known to us from plays or fragments habitually used, and Euripides like the rest. Once care- fully noted, it is quite unmistakable and quite unlike the language of the purest Attic prose. Many of the words belonging to it are so common in poetry that we are almost as familiar with them as with the usual words of prose and sometimes perhaps do not even notice that prose abstains from using them. Another thing that makes unobservant readers think the language of‘ some tragedies not very different from that of prose is that, side by side with this regular poetical diction, the diction of prose is also largely admitted into tragedy, and no doubt a good deal more by Euripides than by Aeschylus. Many prose words are quite unfit for tragic use, but that ‘ selection of _ language really used by men’ which Wordsworth declared to be the proper language of poetry, and which is certainly very often its most effective instrument, found frequent THE DICTION OF ARISTOPHANES 117. place in tragedy along with the rarer poetical words which no one in actual life ever thought of using. Thus many simple familiar words were as much at home in Greek tragedy as in ordinary life, but most of them had poetical synonyms, which the tragic poet used as well and often in preference. There was no reason why he should not make use in his iambics of mais and yj and oikéa, éAOeiv, and deyew, dxove and dépav, and he did actually use them freely enough. But with equal, perhaps greater, freedom he used réxvoy and xOdv: ddéuor, ddpara and oréyy: orefxev, éprev and podciv: avday and guveiv: KxAvew, BAérev and Aedooew. These and fifty words like them are the words that mixed with more every day Attic expressions form the real staple of tragic diction. In lyrics the proportion of ordinary words becomes much smaller, and the use of distinctly poetical, often quite rare, words much more marked ; and the effect of this is of course heightened by other things.1 With many and varying differences on which we need not now dwell the language of tragedy presents great resemblances to the kinds of language used in other forms of serious Greek poetry. The differences would bear more careful investigation than they have yet received. Many words for instance are common to epic and tragic poets, and yet there are other epic words metrically suited to tragedy that seldom or never appear there. So there are words belonging to lyrical poetry, actually used in the choral lyrics or monodies of tragedy, but seldom or never admitted into iambics. Putting however these fine distinctions aside, we may say that tragedy and other forms of serious verse have to a large extent in common a diction notably different from what for brevity’s sake may be called that of prose. There cannot, I think, be any other European language in which the difference is so great. In Latin it was quite trifling in comparison. The actual 1 Aristotle Poet. 22. 1458 b 18 foll. seems to bring all such words generically (e.g. 6Alyos=uixpés) under the head of yAa@rra, though we cannot suppose that they were all to be found in the everyday language of this or that Greek community. His account of them is obviously imperfect, not even recognising archaism. Cf. Rhet. 3.7. 1408 b 11 foll., where obpavoyhens, meAdpios, TARvar are given ag examples, the last from a fervent phrase of Isocrates, 118 THE DICTION OF ARISTOPHANES vocabulary of Virgil and Horace, Ovid and Propertius is infinitely nearer to that of Cicero than the language of the three Greek tragic poets, of Pindar and Bacchylides, is to that of Isocrates and Demosthenes, or even to that of Plato. Passages of some length occur in which there is scarcely a poetical word to be found. Latin is not the least like Greek in this respect. Roughly speaking, it has no special poetical language. So with modern languages as far as my knowledge goes. We have in English many words seldom or never used, at any rate now, in prose, but English poetry does not really present a diction so largely, regularly, consistently peculiar as Greek. The same may be said, I think, of other modern languages. Certainly it holds good of French, which in this matter much resembles Latin. The foreign reader may come across a fair number of words that are strange to him, but he will often find that they are not specially poetical. Their occurrence is often due only to poetry being descriptive. In descriptive prose they may be found too. We have been speaking so far of serious Greek poetry and of tragedy in particular. Is this special vocabulary found in humorous verse also? and if so, to what extent ? In other words how far and under what circumstances, if at all, do Aristophanes and the other ‘comic poets’ appear to have made use of it? Do they admit it freely and without distinctions? Do they on the contrary adhere strictly to the living speech of their own times and give us like the orators, or even more faithfully than the somewhat formal and elaborated language of most orators, a faithful image of Attic as it was really spoken? Or again do they to sume extent compromise, holding usually to the language of life and prose, but departing from it either capriciously or in accordance with certain general prin- ciples? With explanations both the second and the third alternatives may be said to be true. The first is certainly to be rejected. Let us confine ourselves for the present to the only writer from whose hand we have complete comedies, remembering at the same time that his comedies or most of them are of the old school. What is true of him is not necessarily true of all his successors down to Menander and THE DICTION OF ARISTOPHANES 119 Philemon. About Aristophanes the all-important fact to bear in mind is that there are usually four or five different levels upon which various parts of a comedy are composed, four or five different strata below or above one another. Of these strata the common iambic trimeters are the lowest. Probably iambics were not the earliest comic metre, but in our comedies they may be called the standard or normal metre more than any other. The players speak iambic trimeters when there is no particular reason for their speaking something else. In this standard or normal metre the diction is in an ordinary way almost entirely that of actual life and the poetical vocabulary above described has no natural place. At a level distinctly higher come trochaic tetrameters, a metre of more move- ment and excitement, as we know from tragedy also; and into them, sparingly but unmistakably, are admitted words a little above common speech, not used in the ordinary course of things and seldom or never—except for special reasons presently to be stated—allowed to appear in iambic senarii. There are two or three other forms of verse which go more or less with trochaic tetrameters, such as iambic tetrameters, and dimeters both iambic and trochaic. It is very hard to draw a line between these four species as regards the vocabulary allowed. They are all freer than iambic senarii, and perhaps the dimeters are the freest, for they have a touch of the lyrical about them. A very well- marked interval separates from any and all of these the anapaestic measures, mainly tetrameter, but sometimes dimeter, of which Aristophanes is so fond that the commonest and finest of them was called after him. ‘The rise and long roll’ of the ordinary anapaestic verse raises it a long way above the rhythm and general effect of prose, and its poetical character in this respect is matched by a corresponding liberty in the matter of diction, restricted however within limits by no means wide. We find very many words, and a few forms of words, in anapaests that would be out of place in any inferior metre. Finally, ‘as our highest stratum of verse, we have the lyrics, varying infinitely among themselves in subject and language, character and elevation, some still humorous and common, some purely poetical. We may say of them as aclass that 120 THE DICTION OF ARISTOPHANES they admit the diction of poetry in considerable though varying degrees, and of a few, especially those religious or semi-religious in character, that they admit it almost as much as the lyrics of tragedy. The best examples of pure lyrical poetry written, we may be sure, with real enjoyment on the poet’s own part are the stanzas in which the Clouds make their coming known (N. 275 £., 298 £.) and again the charming lines in which the many tribes of birds are called together (Av. 209 f., 227f.). We may regret that chance or the poet has not given us more examples of his powers in this way. In them no resource of poetical diction is neglected. Other lyrical passages are less beautiful and some are a little conventional and perfunctory in their devotional character : others express various emotions in appropriate but not striking style. Now and then a touch of burlesque is added, for the reader or hearer never knows quite where to have the poet and may find him slily laughing when for a minute or two he has seemed serious enough. But in all these the doors of diction are thrown open and many of the words most familiar in Attic tragedy come crowding in. It must be noted however that in some plays and parts of plays, the first and last for instance, Acharnians and Plutus, the lyrical passages are both few and brief. The ordinary comic iambics are at the opposite end of the scale to lyrics, and it is from them that the gradual rise must be measured. Careful study of Aristophanic lan- guage has therefore to start with these. There are two objects which such study proposes to itself. One concerns Aristophanes and Old Comedy in general, of which he is no doubt a fair representative. Weseek to know just how far at this lowest level the comedy of Aristophanes keeps to the real and regular speech of men, and when, if ever, and how far it may rise above it. Our other aim is not study of the language of comedy in the light of what we happen to know about the Attic in general use, but better ascertainment of that common use by observation of the comic poet. In other words we seek first to know how far Aristophanes confines himself to ordinary words, and secondly to learn more securely from him what words were ordinary, and the latter study helps us in the THE DICTION OF ARISTOPHANES 121 former. At first sight this may seem a method that can only proceed by circular reasoning. In using such and such a word he is following ordinary speech. How do we know that it belongs to ordinary speech? We know that from his using it. If his plays and the comic fragments were indeed all that we had to go by, this would be a fair and a fatal criticism. But it is not the case. We have all the extant writers of Attic prose to help us and a few minor aids into the bargain. It is by a very careful combination of the evidence afforded by the prose writers, who vary considerably among themselves, with that furnished by the comedies and comic fragments that we can to a large extent study and settle the two things simultaneously without begging any questions or arguing confusedly, even if not illogically, from one to the other. From observation of the prose writers, especially the orators, we may be disposed to think that a certain word was not in ordinary Attic use. If we find on search that it does not occur in Aristophanes’ iambics or only occurs in one of two or three peculiar ways, our confidence is distinctly strengthened. In like manner an impression derived first from comedy may be confirmed by reference to prose writing. The two support, supplement, or check one another. The delicacy and difficulty of the task consist largely in this, that for certain reasons the occurrence of a word in prose writers or in comie iambics (not to say trochaics, etc.) does not by any means prove it to have been in common Attic use; and again that the absence of a word from both comedy and prose cannot always safely be taken as proof that it was not in such use, since we have to allow for the imperfection of the record, that is, for the relatively small quantity we have of Attic prose and comedy belonging to the best and most truly classical times. It would take us too far away from our immediate subject to enquire minutely how it is that words may occur or even be common in certain Attic writers without being entitled to be regarded as good regular Attic words. It is perhaps the less necessary to do so, because in another work I have briefly dealt with this subject in general, and have also in particular shown very minutely how curious and in the strict sense how unattic is the vocabulary of the Attic 122 THE DICTION OF ARISTOPHANES writer Xenophon. It is enough to repeat here that the use of a word by Xenophon or Plato, by Thucydides, and among orators by Antiphon and even Andocides must not be taken to guarantee it as belonging to ordinary Attic speech. I am going to Jay before the reader a mass of minute observations about the use of words, for it is only in these small, dry, and unattractive details that the matter can be studied. It will appear from them, I think, that in comic iambics, so far as our knowledge enables us to judge, Aristophanes seldom under ordinary circumstances makes use of words raised in any degree above the language of common life. His characters talk as the Athenian talked. It would even be putting their style too high, if we said without qualification that it was the language of the Attic orators. They are talking, not making speeches. They have the ease and freedom and naturalness, the colloquial- ism, coarseness, vulgarism, slang of Athenian streets and houses. The language of oratory need not be elevated, but it is almost always serious and careful, usually in some degree formal, with periodic sentences and well-chosen words; and in many cases, without at all passing the proper bounds of prose expression, it is dignified and impressive. Even with such artists as Lysias and Hyper- bolus the simple and apparently impromptu everyday style which they put into the mouths of their clients is not the talk of Dicaeopolis or Strepsiades, Xanthias and Dionysus. Under ordinary circumstances then Aristophanes makes no use in iambics of fine words. But there are circumstances much commoner with him than with most modern literature of a comic kind, in which this restriction is thrown off and many expressions used which are more or less elevated, sometimes highly poetical. This occurs. when he is parodying or burlesquing serious poetry. It is only careful study of the minutie of language that shows how frequent with him this is. There are of course plenty of passages where it is on the surface for almost any reader to see, and the Greek scholia not seldom draw our attention to the fact and indicate the particular play or poem that he has in mind. But, when we read the comedies with the trained perception that THE DICTION OF ARISTOPHANES 123 these linguistic observations give us, we find far more occasions in which from the language it becomes clear that he is burlesquing, not making characters speak in their own everyday manner. We are not to suppose that he is always thinking of some specific passage in a specific play ; he is often only using tragic or dithyrambic language at large to produce a ludicrous effect, just as Ancient Pistol will now draw on actual expressions of Marlowe or others, now break into tragic fustian of his own. With burlesque of one kind or another the comedies teem, and, unless we have an eye for this, we are constantly in danger of thinking that he is using his own language, when he is as a matter of fact heightening the fun by the use of a style proper to serious verse and tragic situations. Often enough he emphasises it and gives us a further means of seeing what he is about by the rhythm of his lines, which will then not have the free, rough, irregular feet of comedy, but be carefully constructed with the smooth, solemn, measured iambi and spondees of tragedy ; or he indulges in a touch of tragic prosody, the lengthening of a short vowel before mute and liquid: or with poetical syntax he omits the article that common speech requires. Besides the many passages where some burlesque effect is intended there are just a few even in iambics where the poet becomes serious for a moment and has something to say which he really means, cS , > ‘ , LKO ts Kat . To yap é LOV Oo be tpvy@oia In such a case the language may rise a little too, as sometimes in life when men feel strongly, and may aim not only at 7d Sicaov but at 76 ceuvdy also. In iambics however this is quite rare. There are occasional places again where religious terms and expressions have to be used. These are naturally of a serious and often elevated tone, standing apart from the context, which may be mean or jocose. We must also make exception now and then of iambic lines combined and mixed with others of a lyrical or semi-lyrical sort. The character of such a passage is heightened throughout, and the iambics themselves may sometimes admit an 124 THE DICTION OF ARISTOPHANES expression which would not otherwise be allowed. Finally we have to recognise that in all languages comic verse indulges now and then in some license of words or form that would hardly pass in prose. It is excused by the necessities of metre and in modern languages of rhyme, and the very associations of verse help to suggest it. In our own language we shall find Hood and Barham, Calverley and Gilbert in purely comic verse occasionally allowing themselves such a license, and a small percentage of such uses may be conceded to Aristophanes. Just as comic iambics employ quite freely the longer forms in -ow1, -aice of the dative plural, unknown to Attic prose except for a few places in Plato, and sometimes, though I think much more sparingly, the -yeofa form instead of -yefa in verbs (e.g. érpuxopecba Ach. 68), or again the optative -ofaro for -owro (e.g. Hg. 662, N. 1199), or even a tmesis in compound verbs (e.g. N. 792 dad yap ddotyat: Pl. 65 dxd o° dha: Av. 1456 xar’ ad rérwpat); so too no doubt in the matter of vocabulary the author must be allowed some slight and rare indulgence. In the other metres the possibilities of language enlarge in the sort of ascending scale which I have sketched above and shall endeavour to show in much detail below. We do not pass at one bound from the prosaic level of iambics to the poetic diction of song. We have the gradations between, mainly trochaics and anapaests. The lyrical language is so different and so free that for our purpose it may, I think, be set aside. A careful study of it in comparison with the language of the tragic and lyric poets might possibly yield interesting results, but I have not here attempted anything of the kind, though a list will be made out presently of some of the decidedly poetical words which occur in Aris- tophanes’ lyrics and not in the other metres. This will bring out the contrast and illustrate the variety of vocabularies which one play may contain. The license in trochaics and the considerably greater license in anapaests are matters mainly of detail. But it should be said in general terms and should be thoroughly understood that even anapaests, except for a very few lines (perhaps only Av. 209-222 and 685-689), fall much more below the average level of tragic iambics than they THE DICTION OF ARISTOPHANES 125 rise above the average iambics of comedy. This is true even if we take account in them only of the more elevated lines or passages, burlesque of course excepted. But, as these are often intermixed with lower and comic elements, the total effect is not that given only by the bits most serious and heightened in style. When the poet has for a line or two been in earnest, there is at once a change from grave to gay. Putting the lyrics out of sight—and, though many of them are full of poetical words, this is not the case with all— we may say that the highest tone attained by Aristophanes, and that only for a minute at a time, is like the tone to which Dryden and Pope sometimes rise, the tone, that is, of elevated verse, not the tone which belongs to genuine poetry. He constantly falls below them, because they are almost always serious and, even if simple, dignified. But, if his terminus a quo is lower, his terminus ad quem or highest level (lyrics excepted) is about the same. The higher levels of Moliére and La Fontaine might also be compared with his, though perhaps he sometimes rises a little above them. The contrast between his meanest iambics and most dignified anapaests is almost like that felt in some plays of Shakespeare, say Henry IV or A Midsummer Night’s Dream, when we pass from very colloquial prose to the passages of less poetical verse. The more poetical, such as those plays also contain, soars of course high above him. Pope’s own expression, prose swelled to verse, verse loitering into prose, describes him fairly, though not in the way Pope meant it as a description of bad prose and bad verse. In the following lists I have as a rule, though not always, ignored all words used in an obviously burlesque way which no reader can well miss. They throw no light on the question how far his own language, as distinct from burlesque and parody, admits words above the level of the orators and real life. But there are cases constantly occurring where many readers may not catch the burlesque tone, and some where even instructed ones may not detect it; such as these I have tried always to point out. Paratragic is a word in use and convenient to cover two or three different cases: the case of direct quotation, the case of partial 126 THE DICTION OF ARISTOPHANES quotation or imitation of specific passages, the case of language meant to sound tragic or dithyrambic without any particular passages being aimed at. I have used this word rather than burlesque as a brief characterisation of many uses, implying by it that Aristophanes is not writing in any style of his own and that such words must not, as far as he is concerned, be included in the vocabulary of comedy. They belong to that of tragedy, epic, dithyrambic, and are used of malice prepense to produce a ludicrous effect, which the Athenians would feel instinctively and we by care may come to perceive. In the estimation of his own style they may be set aside, except that they indicate how prone to burlesque and paratragedy in general he was. To repeat the former illustration, they are no more part of his own language than the bombast of Pistol is part of the language of Shakespeare. There are also whole passages of some length, partly or entirely burlesque in their character, which I have practi- cally or absolutely ignored for this reason. Such are Ach. 418-489 ; Hq. 1015f. (hexameter oracles) and 1232-1252 : Pax 82-172 anapaestic and iambic: Av. 1238f., 1706f.: Thesm. 776-784 anapaests, 855 f., 1055f.: Ran. 470-478, 814-829 hexameter: Eccles. 1-15. I have thought it in most cases unnecessary to call attention to such compound words, whether substantives, adjectives, verbs, or otherwise, as are pretty clearly the poet’s own coining or at any rate not in ordinary use. It is true that they may and do form a feature of style, and they are very numerous. Sometimes they betoken a sort of excitement that raises the language a little, while at other times they are merely comic. But usually they tell their own tale and need no pointing out. Adjectives for instance made up with the privative 4 and used perhaps once only are not worth giving, when we are not trying to frame an exhaustive statement. My lists are fairly full, but even in more important words, I fear, not com- plete, and these particular classes they practically pass over, though for special reasons such a word has sometimes received attention. Metaphors are a marked feature of elevated and imagin- ative style, but they do not come within my present scope, THE DICTION OF ARISTOPHANES = 127 unless themselves expressed in elevated terms. Dialectical words and forms are of course excluded. When a more or less poetical word occurs for whatever reason in non-lyrical metres, I have, while indicating its general character, not thought it worth while in most cases to examine how far it appears in Aristophanic lyrics. Silence on this point is not to be taken as indicating that it does not so appear. As stated above, I am not studying Aristophanes’ lyrical language, and I have therefore done all that seems necessary if I have brought out the fact that the word is poetical. No one can be better aware than I am—for these minute researches show it at every turn—that it is often difficult or impossible to decide whether a word ought or ought not in strictness to be called poetical. Plenty of words certainly ought, and plenty are unmistakably words proper only for prose ; but there are also many ambiguous in character, at least on our information, which we cannot put satisfactorily into either class. In the lists poetical will not be used quite consistently throughout. It will mean sometimes that the word does actually occur in poetry, sometimes that its character is in some undefined degree poetical, that is, that it never or seldom or only under certain conditions appears in prose or ordinary comedy. Poetical must not have a too precise meaning attached to it. It should also be understood that in the lists of words prose sometimes stands for good Attic prose. If a word is said for shortness’ sake not to occur in prose, this will not necessarily mean that it may not be found in Ionic prose or in the Attic or Kowy of a later date. It is a little difficult sometimes to distinguish vocabulary from syntax, and a few things which I have mentioned, e.g. uses of pi}, re, ds, may be thought to be more properly syntactical. Though I have not had occasion to quote him fre- quently and though his contributions on words mentioned are not very numerous, I owe a good deal to the late Dr. Rutherford, especially of course to the New Phrynichus. I also owe something to Hope’s Language of Parody (Baltimore, 1906), though it did not come into my hands until almost all my material had been got together. Toa 128 THE DICTION OF ARISTOPHANES certain extent we have gone over the same ground, though our real subjects and aims are different. He studies Aristo- phanes’ use of parody and paratragedy ; I have been studying Aristophanes’ own style or rather styles of writing and have held that the paratragic expressions do not really enter into my subject, except so far as I have had often to point them out by way of putting them aside. The more burlesque and paratragic a passage, the less I have to say to it and the more Mr. Hope. But I have taken from his very useful dissertation a few words or points that had escaped me. I have got some help too from the notes of Van Leeuwen to all the plays and from passages in Neil’s notes to the Knights, in which he pays particular atten- tion to the diction. Dunbar’s Concordance has been constantly in my hands. It is almost incredible that only twenty-five years ago and by the Clarendon Press there could be published a concordance which took virtually no account at all (herein like Bétant’s valuable lexicon to Thucydides) of small words, such as particles, prepositions, conjunctions, pronouns, and which arranged words rigidly under their first letter, so that, ¢.g. ropiZeww comes in one part of the book, érépice in another, while cat érdpie, if it occurred and had been written xdmdpice by Dindorf, would have been placed in a third. Butin spite of these and other exasperating things it cannot be dispensed with, until a thorough lexicon-concordance shall appear, a great desider- atum towards which comparatively little has yet been done. Owing to the omissions of Dunbar, the old index of Cara- vella is still of some use. It shows for instance the facts, very small but curious, about diy and ry (see below), which it is impossible to ascertain from the so-called Concordance. The lexicon of Liddell and Scott often fails us in our need as to these niceties of language and to the ordinary reader is often extremely misleading. It constantly fails to distinguish between prose and poetry, a vital point, and again (as does Veitch) between serious poetry and comic, and it sets down a word equally as ‘ Attic * whether it appears in Lysias or Xenophon, Demosthenes or Sophocles, Aristo- phanes’ iambics, or an Aeschylean lyric. Its collection of references also is often sadly inadequate. Fortunately we are much better provided now than twenty-five or even THE DICTION OF ARISTOPHANES 129 fifteen years ago with special indexes to various authors. All the orators except Isaeus are so equipped. It is only such works as these that render studies like the present in some degree secure and not too laborious for the patience even of a scholar.. If ‘index learning turns no student pale,’ it enables a good deal to be accomplished that could not be done at all, or at any rate well, without it. I have added to the examination of Aristophanes a similar examination of the language of the comic frag- ments, divided in the ordinary way into Old, Middle, and New Comedy. Though I have spent due pains on this, it is less elaborated than the part referring to Aristophanes, and I have arranged it differently. I know well that as regards all the lists, Aristophanes included, there must be many sins of omission and commission, a good deal of error and oversight, not to say ignorance. Those who have conducted any such enquiries themselves will be readiest to make allowance for others. ABBREVIATIONS. Besides some abbreviations that explain themselves the following are used :— A =Acharnians i =iambic E =Equites id =iambic dimeter N =Nubes itr =iambic tetrameter V =Vespae tr =trochaic P =Pax d =dimeter Av= Aves an =anapaest L =Lysistrata dact = dactylic T =Thesmophoriazusae hex =hexameter R =Ranae 1 =lyric Ec =Ecclesiazusae p =poetical, poetry Pl =Plutus par =paratragic fr =fragment conj =conjectural gen, etc=genitive, etc met =metaphor A BC and Dare the four lists, divided according to metres. Cases of very obvious burlesque and parody are, as a rule, not given here. If a word is used in iambics as well as in more poetical metres, an account is almost always given of it in A, although its uses in the other metres may be more numerous and noticeable. So again, K 130 THE DICTION OF ARISTOPHANES if it occurs first in trochaics, I have usually dealt with it in B, although its anapaestic use may be more frequent. It follows that B, C, or D alone will not give an adequate idea-of the peculiarities of diction belonging to this or that metre: they all require to be largely supplemented from A. . % Lyric uses of words in Aristophanes, though often given with a view to bringing out the poetical character of a word, are also often ignored. No inference as to non-occurrence must be drawn from absence of mention. The same holds good of hexameter uses, which are usually burlesque. Along with noticeable words will be found a few noticeable forms of words and also a few noticeable meanings. The Comic Fragments are dealt with subsequently in separate lists. For that reason parallels from them are not cited here in the same way as from tragedy or prose. The index will assist comparison of them with Aristophanes. For the plays the references are made according to the text of Bergk, for the Aristophanic fragments according to that of Hall and Geldart. A,—IampBic TRIMETERS. &Bvocos noticeable, if not in Alduvnv..dBvecov R138, at any rate in dpy’piov ra4Bvccov L174 dyavés (not a prose word) L 886 an L1109 V 1467 dyavédpwv an Av 1321 dyopevw, rare in Attic except in old formulas tis dyopevew BovAcrar; A 45, Spav cal dyopevew prose T 306, and the old phrase o’x jydpevov; A41, therefore unusual in N 1456 ci Sra radr’ ov pow rdr’ Hyopevere; R 628 ayopedw tui with infin another old formula? an T 786 dyrayopevw an R1072. The aorist form #yépevca P 107 (xatayopevon) unusual (cxdp) deivov R146 an inten- tional contrast, like A119. defvws or déévaos very rare in prose. dact N275 1 R1309 par ’AGnvaia for *AOnva. unusual (except in inscriptions) ’A. TloAids Av 828 old-fashioned, and so probably & worma Séomow’ ’A. P 271; in E763 along with peddovoa in an; but in Av 1653 éni- kAypov elvar THv ’A. Soxeis; no special reason is apparent. Cf. under Aogias 4p only once in orators (Isocr 5. 43) and Thuc (5. 26. 2), Xen twice’, Plato often, Aristotle a few times, not Herod, p N 731 V 140 P538 Av 1196 T607 trT530 itr E436 (five of these are the imper dpe as usually in Plato, Eur several times d@pycov) : not in the comic fragments? ai rdkav P1706 aiat only in burlesque A1083 1084 T885 1128 an L961 IAMBIC TRIMETERS 131 1 T1042 aiytadds V110 has perhaps burlesque effect aidyp, aifépios rare in prose, partly no doubt because there is little occasion for them. aifyjp is par (quotation or otherwise) Av 1183 T 272 1099 R100 311 892 an N 265 T 43511068 1N 285570 R1352, Even T 14 is somewhat elevated in tone, as other words and the tragic rhythm show. aifépios Av 1277 is certainly exalted. an Av 689 1 Av 349776. To Av it is specially suitable aifés p in T 246 has probably a paratragic effect like crip detvwv (R146) above aixdAAw E48 rév Seomdryv yKaAX’ éGdzreve has no such appearance of burlesque as 211 and T 869-870, but we hardly find the word out of poetry aivat® P1020 R476 both par aivvypos for alvyywa R61 is par. Once or twice in trag and Plato drys like edxAejs and xAéos quite uncommon in prose. L853 od yap dxXeés Ttotvoua To ody x.7.A. is elevated in tone: notice rhythm and phrase 8a ordya. 1 Av 944, Cf. edxAea in D édts R1364 1 fr. 506 hardly ever in prose, Isocr once, Plato once? Xen two or three times ddurpia A 907 (used by Boeotian). Once in Soph, and dArrpds is only found in poetry. Cf. duaprwrla P415 Eupolis 199 do (technically thresh Xen and Plato (?) Theages) seems otherwise out of the way. R 149 4 pytép’ AAdnoev prob old phrase: cf. rarp- and pyrpadoias. In T 2 the use not quite clear: fr 661 ddoay xpi) Tas yvdbous wants context dpa as preposition (not in the most scrupulous Attic) T 148 par R 512: cf. dua in D épaddivw P 380 par, like next line dpivo help fairly common in Thue and Plato, Xen twice, Dem once from a or7An, Isocr once. duivare, -dfere aS a Cry N 1823 V 197 probold. The form jyivafov elsewhere only in p. Other parts of the verb an E790 V 383 tr E 577 itr Pl 256 1 Ec 573 ava (of place) not used in pure Attic prose, though Xen affects it. A796 and L 1002 are Doric. The phrase dvd xivxAov, which is only partly local, Av 1379 and half 1R 441. So dui appears only in lyrics. Cf. on dua and on ovy Gvadvopar TISe R 1460 etrep dvadice: rdAw seems p ava, dvak, évaoca common only in addressing deities or heroes. L 706 dvacoa apayous rodée is par (notice mpayos, duo1, etc.) and so an P 90 @ déomor’ ava, ws mwapamaies (cf. N 264 K.2 132 THE DICTION OF ARISTOPHANES V 875) dvaraidedw E 1099 par dvaTtepa met Av 1439 1449: once or twice in Herod Xen Plato, Menand Epitrep 461 dvarvvOdvopat P 693 dact Av 403, once or twice Herod Xen Plato (€AaciBpovr’ ) dvappyyvus éry E 626 all three words par: cf. on pyyvums in D dvacxeros not cited from any prose author but Herod: A 618 Pl 419 898 are expressions of excitement and half burlesque, tr E 1305 P1179 1 Ec 941 dvexds=dvw V 18 fr 188 and in other comic fragments. Pind O. 2. 24. Cf. dvéxafev Aesch and often Herod av@’ dv p phrase Ec 17 par tr A 292 évtipipos T 17 par dvriryswpoduar P 134 par and tr 1b 626. Eur Thuc Xen once apiece drraidAy, daidAnpa p N 729 1150 burlesque effect drrovd4 p (for drayopedw) E 1072 of the god, poetically called Loxias: an R 369 prob éravdé dreprore p A 374 droneppypiltow V 5: p. epic ao- tporos Ec 792 religious adxrw0ev Av 1184 Pl 674 used very little in prose. adrurépo N 771 tr Av 1475. Though drwrépw -rétw occur a few times in prose (Plato, Dem 4 times, 3 with yevet), Toppwtépw -TdéTw are far commoner dpa, not interrogative Av 1688 p dpyadéos V 1279 L 764 Pl 1 an N 450 of a person T 788 hex R 1532 1 E 978 of persons L 324: Hom but not trag, oratory once (Aeschines 1. 61 of a person d. tiv dy), Xen once (é. mp&ypa Hz 6. 4), not Plato apyyo L 459 (2) T 696 Pl 476 in the perhaps set phrase oi« dpygere; par itr L 303 Xen and p dppyxtws L 182 emphatic dpxéAas E164 p word and form, with orixes Aadv in 163 certainly par: cf. 159, 176 dorevaxri Ee 464 in very low company: cf. on deivwv dotpamrnpopo P 722 from Eur dravpwros L 217 prob religious and ritual dre (Thue Xen Plato, but in oratory only Dem (?) 42. 24) L. 418 T 456 par? R 671 He 37 257 tr P 623 634 Av 285 R 546 drepduav A 181 and 1V 730 (not in p (1), but Laws once or twice) drnpds V 1299 drynporarov xaxev par avpa rare in prose R 314 half par, 1438 and Av 1717 completely so, an Av726 1P945 éi7G L717 par adaros L198 1880 1148 all Doric, 1 Av 428: 3 of the 4 exx. are ddaros IAMBIC TRIMETERS 133 (or darov) ds. Notin good Attic prose ? axnvia fr 20 looks par, Aesch, éx4v Theocr Baiv’ dy Oékas A 198 prob a phrase, for Baivw is even less used than xwpé in common speech Bods A 2 an N 1013: onlyp Pipa stride E 77 BAdBos for BAdBy R 1151: rare in prose (Antiphon Xen Plato Aristotle). Cf. Bréros N 1176 Bracrévw L 406 par: (cf. Aesch Sept 594) tr N 1124 Av1479 dyBdAacravw id L 384. Uncommon in prose (Xen Plato Aristotle) BAérw see an object is quite p at this date and there- fore surprising in P 208 iva pa Brérovev paxonevous tuas ere (par 1). Fr 388 aidt Kar’ eri qv Bréras rHv eioodov might well be 4 BAdmes or fv BAérys Bov6urd Pl 819 religious : Bov@vros Av 1232 also par Bovdn plan, scheme E 3 is not atall usual in prose, where BovAcupa is the word (a dozen times in Aristoph), Andoc 1. 61 probably the only instance in an orator Bovdurés Av 1500 par 1 Bpadive Ec 1140 itr V 2380 Ec 493 500 tr T 661: p and a few times in Plato Bpéras E 31 religious Bpords almost unknown to prose (Republic 566 p, twice in Aristotle t): used by Aristophanes only for some particular effect. See Hope Bpixw Av 26 tr L301 itr L 367 P1315. Hippocrates: not Attic prose Bptw N45 1R329p Buorpd A 959 tr P 1146 Av 274 L685: Homer Theocr Opp (Sta)yaAnvifo E 646: yadnvn, yoAnves mainly but not exclusively P yavotpar AT (éyavdOnv) rare: yeyavw- pévos Plato: érvyavdoas Alexis 186.10. The first lines of A have several unusual words: cf. on Bares yatpos R 282 from Eur yepairepos -ratos E 1004 N 1395 L 1126 Ee 473 tr A 286 E 1301. Neither ynpacds (Antiphon twice) nor yepaiés is usual in Prose, which prefers apea Burns -Tepos-taros. Thuc has yepauds once, Plato seldom, Xen much oftener yepovraywyS E 1099 Soph par? yovat Ec 3 par yins Av 996 perhaps technical : 1 230 SapSdzrw (epic) R 66 burlesque: so with the epic puoriAdo in fr 409, and so no doubt an N 711 8610 Sd. E 230 tr V 373 quite rare dénua A 1059 dar. dey.? Se(Aauos several times in phrase oipo. SeéAaos (always at end of line), four times otherwise N 12 od divapya 184 THE DICTION OF ARISTOPHANES SetAatos eddev: 709 1504 P1850. The prose exx seem to be Lys 24. 23 (conj and doubtful): Aeschines 1. 172: 3. 77, both 6 8eéAatos in close connexion with a verb (e.g. éxxorets 5 SeiAatos tots SPOadpovs) as in our four passages. Its use in Aristoph therefore, as in prose, is very restricted. N 552 SetAatov xoderpdo’ (Eupolidean metre) again with a verb deypaTo R144 p and once Herod Se€todac Pl 753 Xen once or twice dikn (1) justice quite uncommon in prose, perhaps A 362 riv diknv Swpiow, but more probably there the case for trial, the hearing: an N 904 és d9ra Sikys obons 6 Leds x.7.A. ; The phrase é& &xy (occasional in Plato, not in Xen, and seldom, if at all, in orators) N 1332 1333, and half a dozen times in tr. (2) déeny 8& xodvys T 18 is only conj but may be right. 1516 are p in style and Plato has déknv with gen a few times Siolyoua Ec 393 T 609 par (& Zed) Sidmra cat xarérra A 435 par dixa with gen fr 473 par (Sioxw) forms of the aorist euoxafov N 1482 V 1203, a few times in Plato. Cf. under dpivwy above Soca think (often Plato and Xen, Thuc sometimes, very rare indeed in orators). Besides was or wécov Soxeis ; the use occurs a few times (e.g. Av. 1653 R 188), one Ionic P 47, one Megarian A 775, two par N 1415 T 194 deAos P1 1158 used because of ‘Eppys Sédtos: R 1143 quotations: once in tr L 226. Otherwise it and kindred words occur only in 1 and hex. Scarcely known in prose. ddoAov eipdvay L 169 almost technical : cf. Av 633 Thuc 5. 18 Sova Av 1183 par 1 Ec 954 dvoxdbapros P 1250 of a deity duc- pevys V 1160 par T 757: rare in prose, as is ebuevyjs Svopopos Av 7 par? dvatnvos only once in good prose (Dem 19. 255). In comedy 8vcryve and & dvoryve regular and no doubt colloquial: T 878 par fr 461 burlesque: twice in tr Av 354 L 652: burlesque 1 BR 1332 dvodop T 73 itr R922: Ken once or twice, Herod (once), p dvoxeiuepos R125 dapara fr 268 certainly par Sdpov gift (common prose word Swped) P 424 1206 T345 P1849 tr Av 1104 an fr 81 1 Av 937 eyxipynpt, see Kipvnpt éyxov® A 1088: id and itr 4 or 5 times: not in prose éyxepd (Plato Xen Dem IAMBIC TRIMETERS 135 fairly often: Isocr twice, Lys never, though émixeipd is common in them) A 365 half par P1717 an N 476 T 777 par 807: éyyepyrys an Av 257 e6dAw, Oé\w wish (common in Herod; in Attic prose always noticeable except in set phrases like 6 é6éAwyv, ds av eOédy, doov Odes) apparently in fr 100 éédw yewpyetv with the rejoinder ¢ira ris ce xwdver; A 426 T 908 R 1468 it is par id Av 407 kadeis 8% rod cdvew Oédruyr ; (notice cAvev) an P1613 Oddo wAourav ciwxetofar 1 V 5361 Av 929. See also 6éAw éxatt P 699 xépdous exate kav ért pimds wAéo. must be old phrase, quotation, or adaptation. L 306 itr Oeév éxar: religious and probably old phrase éxdiddoxw, like many other éx- compounds, mainly p R 64 dp’ éxdiddoxnw 7d cadis } 'répa ppdow; we know the rest to be a quotation : an 1019 1026 éxvoutos P1981 992 rare p exrepouat H 1234 par L 1113 eAaciBpovros E626 par éduvtw T 598 a quite p word: 579-602 has marked par rhythm, though not language éuraitw ‘975 (semi-lyrical) Herod and p éuropevouat travel A754 (Megarian) Soph, éuropevréa A 480 par éudepys N 502 fr 68 and tr V 1103: not in ordinary prose (Xen once, Aristotle a few times) rpoceudepys fr 460 évaomidotpar A 368 daw. d. par évdcxonar receive E 632 T1129 Thue Plato, not orators ééaxovw rare T 293 Av 1198 par eEavopotpo. E1241 par éEavotyo A 391 par eazrivys P1336 339 815 the older form in the latest play, éfaidvys 1b 353 and often elsewhere éénmeporevw L 840, cuvyreporedw 843: 7. is epic éfdAAvpu seems regular in certain phrases and uses, ¢@.g. maledictions é£dorro etc. and threats éfoAG oe. Constantly implies divine agency. Cf. é&sAys. In these recognised ways common in Aristoph, hardly otherwise é€oupard P1635 from Soph e&driobe A 260 868 etc. It must be accidental that the word is not commoner in prose, where Plato almost alone makes use of it. Xen once ératw N650 tr V516 Plato often, Aristotle, Herod, poets, not orators nor even Xen éravw Av. 1126 hex L773. Notin orators, except [Dem] 60. 7, not Thuc, who once has érdvwev: Xen once only, Plato a few times; not in trag, though Eur Alc 463 1 has érdvwGe 136 THE DICTION OF ARISTOPHANES érei in temporal sense after that, when little, if at all, in orators, often in tragic narrative: Pl 660 par fr 403 (émei, since, because very common and placed at least a dozen times so as to begin a sentence at the end of a line, e.g. N 688 781 1470) érmv L1175 Av 1355 (parody of law) epic Eur H.F. 1364 prob not in prose émyAuwtroya: L37 Aesch éridydos E 38 T575 P1368 an T799 Ec 661 Herod Xen once each? Aristotle, not orators emlety tiv xoAnv T 468 Soph Eur érivoa E90 T 766 Pl45 an E 539 V 346 1050 Ec589 hexR1530 1Av405 Ec574. Not in any orator: Xen Plato once each, Thuc four times. T 766 may well be par and in the other two iambic passages a rather elevated word is suitable érowa: a few times in i, mainly imperative mood, and a few in other metres: unusual in orators éros in sense of either word or saying is not usual in pure Attic, and there is perhaps no example in Aristoph (the phrase éros apés éros N 1375 apart?) where the language is not meant to be a little elevated, metre often heightening the effect. In such cases as Av 174 & cxatérarov cipnkas eros, L467 & rdAd’ dvaddoas éxy, wpdBovdre rhode Tis ys, the intention is clear, fr 323 lacks context. In the sense of verse, line it is not noticeable except as being sometimes rather technical (E 39 508 N 541) érovpitw T1226 p épdw V 1431 proverb épeidw (some uses) E627 itr N 1375 R914 épts, épi{w uncommon in prose A 1114 R866 par? trR1105 and hex 877 an T 788 gp7w in general sense of go, advance, etc. is p, often in tragedy. The phrase 6 médenos Eprérw comes three times E673 L129 130: fr 210 6 8 HAacris cipre mpds THY KryxAida has no context: itr V 272 an 552: all three probably refer to slow movement of old men. zpocéprw V 1509 1531 probably in special sense too. xaféprw R 485 mapéprw Ke 398 511 édéprw Pl. 675: the notion of creep impossible in the second of these passages and may be absent from all. Cf. Rutherford New Phryn. 50 edo (kafevdo much commoner in prose) N12 Av 8215 an Pl 541 (1 V 318 for dSav?) evAoy® rare in prose A372 Ece454 tr H565 596 ciAoyia an P 738 eipevys L 204 Pl 636 both par religious : cf, dvopevys above edrrepos N 800 p IAMBIC TRIMETERS 137 evpvOuos P1759 par 1T 121 985 edoTOLO N 833 Aesch Soph eioupare N799 Eur eidpaivw (not a common prose word except in Xen) A 5 (but see under Bods) L165 Ec1123 tr N561 an P 764 L 591, several times in later comic fragments édnpépios an Av 687 p édrepos N 223 par éduddAdAw N1299 V 1348 P 432 fr 552 seems used as an ordinary word : otherwise p éyGaipw BR 1425 par quotation 4Ods Ec 151 par? tr Av 271 an Ec 584 Soph Eur Hippocr HArtdlopac V 772 for the sake of a pun: itr L380 an E798 qmos P 934 an V 879: seldom in prose, but once or more in Thue and Dem darvw (sic) E 1023 refers to some oracle or other verse: epic and lyric, i Rhesus 176, where oddly the spelling with a also occurs (as in Anthol. 9.99 3) with C.L.G.S. 1818 (Neil) OdArw E210 probably quoted (ai xe wy OarAPO7 Adyos) an Av 1092 in physical sense dopa Pl 1166 fr 149. 4 unusual in prose: an T952 1EH990 Av 234 Oeivw not in prose (?), and used by Homer Pind Aesch etc ; yet Aristoph has it ini only and seemingly as a common word, always in aorist, E640 V 1384 R 855 etc Gékw, the rarer and more p form for é@édw, (1) in the old religious jv Peds OéAy etc.: (2) semilyrical Av 407 xarcis 32 Tod KAvew Oedwy; (notice xAvev) cf. 1929; (8) in certain relatival phrases, ds dv OéAy, etc, and in 6 OdAwy (cf. under é6édw) it was not uncommon in prose and occurs perhaps thus A198 E713: (4) after a long vowel or diphthong in which ¢ is merged, so that perhaps we should write ’dédAw ; of this there are many exx. ¢.g. jv dt wy Gedy N 801 V 493 Ec 1017. 1216 read qapaywpely od Oédras; for od m. 6.; deomwdo Pl 9 par OnrAvdpwv Ec 110 par Ovyarépos V 1397, if right, an unexplained poetical form bundy Av 1520 technical and religious Oupaivw N 1478 prayer, tr 610 Ovpzos not really common in prose, especially in sense of anger: oftener as spirtt. A 353 may be either. 2b 450 480 483 E1194 it is par as mind in general, fr 398 as anger. Several times in tr itr and an, eg. V 383 567 649 1082. In another sense P 559 domdcacbar Oupos quiv eore K.T.A. Oupodpa: R 584 T 518 perh. par an R 1006 Herod, in Attic prose very seldom outside Plato and Xen 138 THE DICTION OF ARISTOPHANES tenor middle hurry etc seldom in any prose but Xen, who rather affects it: Plato once or twice: E625 Ee 346 tr V 423 111259 i{w very rare in prose. Xen once, Plato once or twice: R199 1H 403 Av 742 ipmepos R59 par an L552. ipeipw an N 435. The verb not in prose, the noun very rare ivddAXopat (Homeric) V 188 par? Plato once or twice immnddv p P 81 io often, but always par: see list in Hope, who points out that it comes most often in 1 xafoo.® Pl 661 par religious xaxoopos P38 132 p burlesque ? xadvrropat N 740 scarcely used in prose: Xen twice? xa\vrrov may however be a stock expression like our old ‘be covered.’ Of. ‘InaéAuros xaAvrrdpevos kapadox® E 663 par? Xen once kaptepos seldom in prose, except of strong places. T639 seems ordinary : elsewhere it has an air of par A 393 dpa ‘oriv dpa por x. poxiv AaBetv 622 xara 76 xaprepdv? T31 R464 1398 xatadovpat N 1468 par as aarpgov shows: quotation ? xatat6ahk@ Av 1261 par xataidw T 730 p caractéAAw dress 'T 256 rare kareiBopat L127 +i Sdxpvov xare(Berar; par karopvupe not usual in prose, perh more solemn Av 444 R 305 306 karomw EK 625 Av 1150 1497 Pl 13.757 1094 1209 fr 493 not common in prose, but hardly, if at all, in serious poetry: xarémic6e, not in Aristoph, uncommon in prose, but found in Hom Eur xéop A 5 par, as in Eupolis 90 xwvpouat E11 par (éy)xipvnt Ec 841 in spirited description, xpydvres an fr 683, x. simple and com- pounded decidedly p: Herod émx. once Kado P 249 par cdevos Av 810 1272 1277 Pl 772 par in various degrees, A 1184 burlesque, an P 737 E1328 1N 1024. Not in orators, Thuc, Xen: Plato twice kAylw Av 921 par (article omitted) kdvw Av 1390 ov 8 KAvwy eioe. téxa perh quotation id 407 416 433 xvepatos V 124 and 11327 R1350. xvédas (occasional in Xen) Ee 290 1 xvica P 1050 Av 193 1517 id A 1045 with Adoxw. kvnodv dyuds old ritual phrase Av 1233 an E1320 xvedarov V4 11476, both abusive as Aesch Hum 644 p, not in good prose xopapat L758 Ee 723 drow. V 213 xarax. an T46. Not in orators Thuc Herod: Xen freely, Plato sometimes, IAMBIC TRIMETERS 139 Perhaps military. éyxoduar is technical com fr 318.8 and fr 602 an Av 734 T 795, not in p nor Attic prose xopn eye V 7 par? notice dropeppnpioat in 5, T 902 par Pl 635 par koupidtoy A€xos P 844 old phrase kpadaivw A 965 par ®, Kpavaa médis A 75, erera peilw: rov Kpavadv Lyrels mwoAw; Av 123, both humorous 1 L 481 humorous too xrumos Av 1156 a little exalted, xruvrotca Ec 545 ordinary ? but éxrureiro Pl 758 par even in construction: elsewhere only tr etc: so émuxtura Ec 483. Prose two or three times xvdoy.ds A 573 par AdlveGe wacat THs KUAtKos L 209 must be par as a solemn summons (cf. 203-4): only p Adurw Xen Plato once or twice. Hence V 62 ci Kiéwy y’ eAapife is notice- able: Av 1710 Ec 13 are par: an E550 N 395 1160. Adpurropat R, 293 of something supernatural. ékAduaw (com- moner in prose) L 387 Av 1712 par tr P304. doAdurw Av 1009 Adoxw R97 doris Appa yevvatov AdxKoe must be quotation or imitation: A 410 P 381-4 Pl 39 all par Aadpvypds N 52. The verb and its deriva- tives are p ede~a, Aé~ov, AexOeis, NeAeypevos (for eizov, pybeis, etc), though common in some prose, are almost unknown to orators. Dem and Lys 74 Acxdévra only, Andocides é\efay twice, etc. Aristoph has Adéoy in iambics several times, also A¢fw subj and Adfat, seemingly in an ordinary way, eg. V 15 drap ot A€fov mpdrepos, 28 od 7d cov ad défov, Av 1668, A 1057 Setrar.. re cot Ae~at, though at other times these and kindred forms may well have a particular effect of varying force, e.g. R 1020 Aicydrc, A€fov pnd? aiOddws ceuvuvdpevos yaderauve. Aorist and perfect passive not found AevkoTAnOys Ee 387 (kovpistov) A€xos P 844 see xoupidiov. Menander 535. 8 yaprjAtov A. Aeds not in prose, except Plato twice? and Herod: regular in tragedy, where however Aads also occurs. In Aristoph there are marked traces of old phraseology: proclamations dxovere Aew (cf. V 1015 viv aire (notice) Aep apdcxere Tov vodv) and of wévres Ned, with which goes Av 1276 oi Aed after of 7. d. in 1275, Phrases like A162 6 Opavirns ». BE 224 6 révys d. P 632 921 are all of one type, T 39 1 was Accs. Aads only. in par or | and plural E163 R 219 676 Aijpa. (only 140 THE DICTION OF ARISTOPHANES Herod in prose) R 463 500: for tr see B Aieocopat P 382 prayer Adywov E122 etc” ete prob old- fashioned. Herod Thuc Eur Aristotle (but see Neil on E120) Aogias E1047 1072 appropriate to oracles : so P18 (perh quotation) with par following. Such divine names sometimes have burlesque effect. @oiBos E 1024 V1 39 81% 213 all oracular R 754 par E1240 par, with which cf. Soph O.T. 738. Clearly too & Soff’ “AoAAov is a regular exclamation. Similar is the burlesque use of TloAAds E1172 N1265 Ec 476 and of Kpavad (above). TladAds in 1 H 581 T1136 and SoiPos e.g. 1 N 595 T 112 are quite different and as serious as anything in Aristoph can well be. Other poetical names used in a burlesque way are Tpiroyevys E1189 (cf. the designations in 1172- 1177 1178), Tpiroyévea an N 989-1 L 347 Kpovééys an V 652, Zdvan Av570 L717 +i Ziv’ direts; par, And an Pl 515. The longer form ’A@nvaia (cf. above), rare in prose, may have a touch of humour in Av 1653 érikAnpov eivae tHv ’AOnvatay Soxets; E763 an P271 Av 828 it is obviously appropriate as solemn or technical. Kdmpis appears only in 1 AwByrys R 93 p Agoros Av 823 prob wrong: if right, par, for Aeros is not ordinarily used except in & Adore (Phaedo 116 D jw dv8pav Agoros) paddtrw V 973 met Soph Eur padrBaxds A 70 1200 Av122 tr R539595 anV714 1 Av 233 Plato and p, padaxds being the usual prose form peyddus E151 172 1162 (in all these with ciSamoveiv) an E 782 1N 600: 3 or 4 times apiece in Xen Plato, p pedéwy, pedéovoa L 834 in prayer an E 763 (cf. Plut Them 10) 1 £560 585 pedid T 513, once or twice in Plato, not in Xen or other prose péoatos V 1502, Menander 267 perdpowos Av 1383: the context suits an ‘elevated’ expression: Herod Hippocr zeSdpouos 4b 1197 is par py lest (= iva pH, pure final conjunction, used by Plato and Xen, but very seldom indeed by orators) V 162 P1622 itr V 247 (1) Ee 489 (2) 495 an N 267 hex E1081 141 294 + pjKos height A 909 Av 1130 pytpdbev A 478 par pnxavy means very little used in purest prose, chiefly in doy p., oddeuia p., etc.: often in Plato, not IAMBIC TRIMETERS 141 much in Xen, Herod often. pyydvnpa prob no orator but Antiphon. yyaveuor commoner than pyxavy, but a little above common diction. pmnyxavy in i N 479 V 149 L111 T 87 765 927 1132, sometimes half met or touch of par such as is well marked in A 391: a few times in other metres. pnyavnua Ec 872 itr E850 901. The verb only T 736 (observe pnxavy 4 times in T) and A 445 par tr V1106 1 V 870, dcapnyavapo id E917: cf. duyyavosin C. puvdpopat Ke 880 pujnpov N 484 an 414 P761 rare in prose except Plato porcty L 743 prayer (cf. pedéwy above, Aiccoua, etc) 984 Spartan, E 21 for the joke: otherwise 1 only poudy P 664 slightly par povos L 1127 elevated in tone: notice rhythm 1123-7 p000s = Adyos sermo, speech, words, etc., not story, fable, myth: constant in trag, not used in ordinary Aristoph i (for P1177 has the other meaning, asthe scholiast tells us, and so clearly V 1179: cf. an V 566 and 1 L781 806, while P 131 is par): even an V 725 mpiv av dudoiv podov dxovoys is a quotation prdile (L 94 posse Spartan) seems not known in Attic, found in Theocr popapo (Av 171 vy tov Atévecoy ed ye pope tavrayi) is p and this use is odd vavoTroAG Av 1229 par like T 1101 vadpapKros A95 par tr E567 p veadys fr 361 apparently not p, but rare veodaia fr'67 par 4 vepéern végos. L and 8S say ‘the more common form is védos especially in prose,’ but it seems really that végos is the less commonplace of the two, Neither occurs much in our prose, because not wanted: vepéAn Xen An. 1. 8.8: védos Plato Tim 49c [Hpin 987 4]. Aristotle has véos a good many times, vedédy only a few, but he is not a model of pure Attic. Dem 18, 188 tov .. xivduvov rapedOeiy éroinrey domep vépos shows which word is the more dignified. Both are used metaphorically but védos most, and this points the same way. vedéAy frequently in Aristoph where no special effect can be sought : not only in N, which we ought perhaps to leave out of account as the title might be chosen for effect, and so NedeAoxoxkvyia, but Av 178 194 1384-6 1502 1608. vedeAy is also metrically convenient for an and hex. véos not iniatall, tr Av295 met an Av 578 met dact N 287 P1090 1Av349 vamos foolish 142 THE DICTION OF ARISTOPHANES N 105 vytitios N 868 Homer vixoBovdos E 615, prob NexoBovAos vonpa (ouly Plato in prose) E 1203 par? N 229 743 Av195 an V 1055 1N 705: specially suitable to N voora Av 1270 perh quotation. A 29 a sense intermediate between return and the sense of mepwoord (P1121 an P 762 T 796 Pl 494) vuxtepeiotos T 204 prob used with a purpose vuotaxtys vmvos V 12 par virrw (epic) P1784 an N 321 édor7Aava A 69 affected ddvvy -npds -Ouat hardly used in orators, not in Thuc nor in Herod. édvvy twice in Xen, fairly often in Plato; the verb rare in Plato and not found in Xen. dédvvy A526 T 484 P11131: dduvypds an Pl 526 (cf. Dem prooem 15 Plato Gorg 525c) 1 A 231: ddvvdna. L164 Ee 928 Pl 722 1V 283. Some- times pain, sometimes grief, trouble olknois 'T 272 par an Ec 674, not common in prose oixtipw V 975 pathetic an 328 556 L961. Prose usually éked éis for rpéBarov (very unusual except in Xen) P 929 930 1018, but with reference to sacrifice hex 1076 an Av 566 sacrifice oiws V 1363 (the Aldine for MS. ois) prob not right 6ABos GABtos extremly rare in prose, Herod excepted. The substantive in Aristoph once Av 421, and that in id, the adjective three times in 1, twice in i, Av 1708 clearly par Ec 1131 with rpicdAPwos in 1129, rapturous in expression buixAn N 814 an 330 E 803 p Xen Plato Aristotle dppa, though very frequent in Plato and Aristotle and fairly so in Xen (not found in Herod) is only occasional elsewhere in prose and in orators hardly occurs at all (Aeschines 3 times, Hope) A 1184 Ec 1 are par: the other exx. are in other metres: see below épuorrepos Av 229 p and Plato évivnus except in Plato and Xen a rare prose equivalent for dpedd. Antiphon and Dem (28. 20 obrws dvaicbe rovruv) have it once, Isocr twice (2. 8 and 15. 264). N 1237 Pl 1062 and in Dem.’s phrase, evidently old, T 469 tr L 1033. The word is not Herodotean. xarévato 1 Ec 917 6vOunos E 706, but T 466 is humorously solemn tr V 406 455 1105, dévévpodua tr ib 501 6 (very little used in orators, though often in Plato Tp Thy TY and Xen) A 198 Baiv’ dry Gédas prob old phrase (notice Baivw) IAMBIC TRIMETERS 143 N 1345 a passage of mixed iand1, 1R 1257. ap appar- ently only in an V 396. TI have found no clear instance in Com Fragm dzws when N 60 seems unique for comedy and prose, yet the context is quite commonplace (epic trag Herod) épyy E41 an V 1030=P 752 seems to be not anger, but temper, mood, a rare sense épyta = pvorypta Or TeAeraé uncommon L 832, also an T1151 RB 356 and 1, dvopyiacros par L 898 épxwpote fr 96 metre unknown p épovw fr 523 metre unknown p éppwdG not often in prose (yauora Aypooberixdy Says Didymus on Dem col 11. 15) though even orators do use it : E 126 Ee 994 P1122 tr R1112 an E541: not common in verse, in epic and | not found? moddpn -Gno. quite p N 176 is par, as an A 659 (adaptation) and P 94 (burlesque) show: so 1 V 645 Xen once TlaAAds, see under Aoégias adaAAw A 965 par wapBacirevo. N 1150 par prayer, an 357 the same mapayoia Ee 868 par? Tave)hebpos Av 1239 par tr L 1039 Herod and p marat A 1214 L 215 trag Plato Laws 704 B wapos E 1337 perh quotation 1V 1536 not in Attic prose, Herod once matayos A 539 an N 382 P 155 L 329, marayd an N 378 384 389 fr 116 marpo. for warpis never in prose: A147 Bonbeiv rH wdtpa is comically’ solemn, T 136 quotation, R 1163 and 1427 trag adzrw N 1330 an 912, raoréa hex P1074. émum. and xataz. perh commoner (4 or 5 times in Aristoph) meddpotos, See perdpatos médov only in par T 856 Pl 772 and 1 twice mevixpos Pl 976 Plato once, otherwise p mepixadAys T 282 prayer wépvnut E 176 reason not apparent: not in prose (Eur Cycl 271) wOev for the usual zeiécas R 1168 par (cf. context) P1949 prob from old phraseology wAGtis (only in aan A 132 old? mAéws A 907 E 630 P 703 T 734 itr N 1367 tr E 281 1£E 305 R 1372: very rare in prose (ueords), Herod Xen Plato once or twice, and Plato likes dvdaews mvoy p N 161 par? Av 1397 takes up 1396, R 154 313 (addy avo7ys) an T 43 par mobewds par in A 886, R 84 not markedly so, but pathetic tr P 556 an Av 696: hardly ever in prose, Lys? 2. 73 Thue Plato and a few 144. THE DICTION OF ARISTOPHANES times Xen moAguioTyptos seemingly par A 572 1132 P 235, technical N 281 mwodopa Av 1565 an 553, rare in prose Thuc Xen moAotpat Av 181 etymological mwokuripytos extremely rare in prose (Plato Huthyd. 396 p), common enough in the plays in exclamations om. Oeoi, @ 7. “Hpdkdets, ete Cf. A 759 (otros) woAurivaros arep Tol Geol. R 851, & wodvriuyr’ AioyvaAc is half burlesque modvtAjpwv P 236 par with Bporoé: see tAjpov mopOd A 164 7a oxdpoda opbovpevos unusual construction and sense mépysos R 1429 par though not quotation anT777 1P1030 Aesch Plato Thuc wépos means seldom in ordinary prose in this general sense, Xen and Plato a few times, Andoc once, Dem only in financial sense: P 124 par and so T 769 itr E 759 an Ec 653 1V 308 with pun mérvia (re- ligious, not a prose word) often, always vocative except 1L 1286 motopat. P 830 par Av 1445 wemoriobar Tas dpevas (notice dpévas) an N 319, not prose mpayos L706 par: so prob Av 112: not prose (notice article omitted in L 706, as in tr E 568) mpadvas R 856 unique ? apevpevys fr 21 par? not prose mpoBypa Pl 759 par mpobeAvpvos P 1210 part of a par phrase an E 528 mpodrcirw T 927 par tr Av 1558 1 twice. Dem 57. 70 in very emotional ending, Thuc and Plato once or twice, not Xen mpomoXos Pl 670 more or less technical. Herod, hardly in other prose 1 R 1333 mpoppifos R, 587, like apoféAvpvos mpocenpepys fr 460 mpdoGe decidedly less common in Attic than dumpoobe, even in Plato, who uses both often. In Aristoph gumpoo6e only once, mpdobe 11 times, 7 in i besides A 43 wdpur’ eis 16 xpdobev mpocAapBdvopa P 9 mpoodpotos T 516 an V 356 Av 685 Plato once or twice mpoodepys Ec 67 Herod Plato apoadopos V 809 Av 124 an P1025 Herod Thuc Plato [Dem 59. 56] mpoowbey for réppwhev fr 676a mpéra, adverb very common in Aristoph, 30 times or more in all metres, and mpdricta 10 times: Thuc Xen Plato not common. Dem ra apéra 27. 19 and 2. 8 are not clear cases: not Lys or Isocr (or Aristotle‘). Tragedy often mputiatos -ra, Ae ch once, Soph 3 times, Eur 2 or 3, Homer etc Not Herod Xen Plato IAMBIC TRIMETERS 145 Aristotle Isocr Lys: the two Dem exx. are in doubtful speeches, 43. 75 and 48. 33. Ini A 28 1002 L 1169 BR 519 1121 Ec 749 1059 (none apparently par) and 6 or 8 A in other metres mrepoddvytos Av 1390 1402 mrepwrdos (POdyyos) Av.1198 aupTovAa ¥ 1497 Av 1580 tr V 1079 T 727 karaTupTon® T 243. Hardly ever in prose: Herod once, Xen once in other sense pimtafo Li 27 poiknwa Av 1182 heightened phrase: cf. aifyp doveirar and piyy pupy ibid Ee 4 par an N 407 P 86, a few times in prose cehaya A 925-6 of ships set on fire: usually p and so dact N 285 1604 cbévw P1912 xa’ dcov av cbévw prob more or less of a phrase. o@évw not used in good prose, even in literal sense: nor is ofévos (not at all in Aristoph) except in phrase mavri ce (cf. ‘might and main’) onavitw N 1285 V 252 ; not often in prose, nor is ordvis omddyxva of feelings R 844 par itr N 1036 an R1006 oréeupa Pl 39 par arevaxo A548 p and met otixes (Aadv) E163 par : Aadv p in form and number: cf. under dews orody (little used in prose, esp in general sense of dress, though Xen likes it and Plato has it sometimes. Lycurg 86 seems the only place in orators, for in Lys 6, 51 itis a special robe) T 136 quotation? 92 and 851 par? Ec846 trd R591 oroAos L.93 (Doric) Av46 both partly par: seldom in best prose. Dem [pit 3 times] 6. 36 9\Oe atoAw which from Thue seems a set phrase, Xen Plato a few times, not Isocrates avyyvworéds T 418 usually p ovykexpayat met Pl 853 par ovivyos Pl 945 p cupmapactarns Pl 326 Soph Ph 675 cuppéepw agree with E1233 L166 unusual. Another sense result (Ee 475 quotation of Adyos yepat- tépwv?) is also not a prose use. The middle has these meanings in prose sometimes (result, turn owt tr N 594) civ Mommsen shows that in Aristoph it is found (1) in quotation: (2) 1, once semi-l R 444 (quite poetical), and once earnest iambies L 1143 oty brAiratoe : (3) ordinary . Attic uses, ovv inclusive and civ Geis, civ 6rAots, pysevi Evy vg tr N 580 ovvav7® A 1187 par Pl 41 44 perh oracular Av 137 ordinary? (Rutherford Babrius p. 47), Xen twice cwoiroAts A 163 quotation ? L 146 THE DICTION OF ARISTOPHANES tayés E159 par TdAas. olpot TdAas occurs some 25 times, aiBoi r. a few; also & rdéAas and oftener & rddav or rdédav, raddvrare, and raddvrar’ dvdpov. Apart from these forms rdAas occurs only P 1225 ci. . rade. . Kira .. xpyoopa: tédas; and 3 or 4 times]. Xen Cyr 4. 6. 5 xdyo pev 6 rddas «.7-A. is said to be the only prose instance of it taAavpwos A 964 par P 241 par @s (dcov, 671) taxos not in common prose: Plato dr: 7. once? In Dem even dia tdxovs is found in disputed speeches only. ds r. L1187 écov +. tr T 727 re standing singly and attaching a word, clause, or whole sentence to what pre- cedes is p and also found pretty freely in Herod Thuc Xen, in orators uncommon. Aristoph fairly often, but with varying frequency, ¢.g. E twice as often as V. In Hi about 15 times (5 in the speech 624-682, 3 in 732-735), itr about 9 (3 in 457-459), id 4, an6, 17. The double Te.. ve, uncommon in strictest prose, E i134] 11125- 1127: not in V? I have not gone through other plays réxvov (1) 1 N 1165 1170 V 1518 R 211, 1322 1356: (2) par A891 T 469 half prayer 698 754-5 fr 461 585 (rexvov): (3) vocative réxvoyv and @ réxvov L 7 T 754 perh par 1062 1181 1198 fr 125 three times. This voca- tive is common in Menander ete and prob traditional. texviovov Li 889 ouvrexva T 15 piddrexvos T 752. The word is therefore one way or another curiously frequent in T. It is well known that réxvoy is not a word of ordinary Attic prose, as judged from the orators, who always use mats rexraivouo. lit L674 P1163, met E462 an A 660 quotation Tépw etc rare in prose, except Plato Xen Thuc: réproua: itr Pl 288 1 T 992 repzvds Ec 889: also A 881 par and an L 553: répyus 1 R 675 téxvacpa T 198 par (notice rhythm) rexvefw T94 itr R957 1A385 texvanar V 176, once or twice in Thuc Xen ctnod N 138 par quotation typodpat (middle voice) V 1386 and tr 372 thjwov P723 R85 PLI77 (6 cAjpov) an T 1072 par Pl 603 four times in Xen (twice voca- tive) but not elsewhere in prose. oAvrAjpwov i P 236 tAjvat still more p, once in Isocr 4. 96 (noticed by Aristotle Rhet. 3.7.11) and once Xen Cyr. 3.1.3: N 119 V 1159 fr 149. 2 id N 1387, rAyrés L 529, A SH IAMBIC TRIMETERS 147 avathds 1 P 1035 ToApnpa Ecl06 Pl 419 itr Ec 288 L284 an P94, rare in prose (Trop) teropyow P 381 par tpn passing tome Av 156 Plato once Aesch Soph, delay A 385 tpiBwv adj. N 869 V 1429 Herod Eur tpvywdds fr 149.9 an V650 1V 1537 tpvywdia A498 500 tpvywdicds 886 rpvywSoroiopovetxy fr 333.1. The word hardly occurs elsewhere, except etymol Athen 408 dq’ of 8) Kat ctpvywdia 1d mpdrov éexdA7On 4 Kwopwdia: (xopol tpvyxol A 628, zpvyodaiuwy an N 296) tpvxw A 68 an P989 rare in prose. Cf. xatatptxw epic tupBos of a man L372 Eur twice: in lit sense V 1370 p tuxnpas A 250 T 305 prayer in prose proclamation trai p A 970 if right, prob quotation or imitation, Av 1426 prob iad birepnvopewy P 53 par (epic) tréppeyas E158 par tréptaros P 52 par tricxouar fr 615 (of tpayixot Kat “Apirtroddvys Photius and Suidas) Hom Herod imva L148 Doric (Herod trvwpévos and xatutvwpévos) troto- moduoat T496 itr R958 Herod Lys 9. 4, but irororedw is the usual form géyyos fr 188 par, see D $beypa P 235 par T 267 fr 606 an N 319 364 1 Av 683 Plato once or twice / grcE Ec 6 P1661 clearly par N 1494 T 242 Thuc Plato Xen ®oiBos, see under Aogias goivios T 694 dpyv, dpeves (1) 1 and hex often e.g. E 1052 N 475 705 P 1030 1068 1099 ete (2) an R1040 P1581 (3) clear par A 445 E1237 Av 1238 L 708 T 275 R 101-2 886 (4) vois xal dpeves L 432 T 291 tr R534 (5) N153 Av 1445 may be taken as more or less par. Hardly ever in prose dpirrw with ace N 1133 p dpovricua N 155 with special ref to povTicrypiov gpodveos very rare in prose (Antiphon 5. 29) but Aristoph uses it sometimes without special point P197 T691 R94 3051 Ec 311 341 950_.L106 par and possibly A470 an N 718-9 722 T 794 fr 379 1 A 208 R 1343 gutrevw met of men and things V 1133 1V 1276 gwvd A 777 (Megar.) an T 51. Not in prose for speak, except Xen Symp 3. 13 xaipyiév A 4: cf. under Batds xépw with gen though frequent in Plato is rare in other prose. Thuc L2 148 THE DICTION OF ARISTOPHANES and Xen perh once apiece. In Aristoph often and with no special point, eg. A915 1051 E268 R 1418. Notice in id A 1232 ov xdpw xpeov E 138 oracular, fr 347 ordinary idN 1447 anP765 1 P1029. Not in most prose : Herod Plato esp in Laws and other late work Xen twice? xpyco A734 (Megar.) T 751 an N 359 891. The Megarian apart, the uses are all relative phrases (6 re xpyjlers etc) and prob idiomatic (cf. ds dy OéAy etc), for xpyfw is not in ordinary use xejua thing not usual in Attic, which prefers rpéyya, if any word at all is wanted. Plato sometimes, Xen once? Andocides once, and so on. In Aristoph +i xpjjya why or what often: also xpijua, dcov rd xpyua ete, with or without gen. 716 xpiypya the thing, the business etc N 325 V 799 834 P 38 etc. xpyya a thing L677 fr 333. xpyjyara things does not occur in him xpds L127 par xwpo go, though very frequent in Herod Thue and fairly so in Xen Plato, occurs twice only in Dem (2. 28: 19. 163) and not at all in any other orator, so that it occupies a peculiar position. Aristoph is very fond of it in trochees etc, when referring to movement of chorus and actors. But it also occurs in iambics freely and without any special force, e.g. N 509 Av 1186 L 66 738 Ec 851 : dpuds A272 P1963 fr 235 ds and as dv in order that. ds dv, hardly used in prose except by Xen who much prefers simple és (Thuc 6. 91%), occurs in Aristoph 15 times (Weber, who does not give references or distin- guish metres). I have noticed it 9 or 10 times in i (A 44 V 113 Av 1454 1509 Pl 112 etc.) and 3 or 4 in tr. @s alone is found only itr Ec 286 (altered by Dobree) 1 Doric L. 1265 1305, for in i L 183 the verb is probably future indicative as when, after that, though frequent in Dem, is almost completely absent from other orators (Lycurgus 62). It occurs a few times i V 1476 T 507 R 504 and noticeably three times Pl 653 668 688 in a narrative which may be slightly suggestive of a tragic pjows, tr N 551 P 612 632: E62 and perhaps an V 673 are not temporal as in a wish Pl 891 (? xat 8y for és 8%) usually p: cf. Menand Epitr 207. TROCHAIC TETRAMETERS AND DIMETERS 149 B.—TrocHaic TETRAMETERS AND DIMETERS. The dimeters. are marked d dyxédn mainly p, R704 met: cf. Nausicrates dyhads L 640 epic and lyric more than iambic. Cf. érayAatfw an fr 682 aiddos d R 248 éAKywos V 1060 11061 iPl. 1002 1075 are quotations of the old wadau mor qoav dAkiwo. Minow Xen often, once in Plato Rep 6148 for the pun on a name, not elsewhere in prose before Aristotle ? ' dpive, see A dyw verb V 369, Herod and p dmactia N 621 religious dropdpyvuynt A706 an V 560 1A 695 epic: eEoudpyvume itr A843 Eur Plato daromtuaTos E 1285 p, as is the verb in met sense érooraTé Av 311, Xen Plato 3 or 4 times, Dem once dpopa P 1158 aéooov E 1306 Herod not Attic (nor is dyxt) aifadia for aifddera T 704 airoidé P 607 L 687 abtompenvos R 902 met *"Axapynidys A 322 burlesque epic form like T1nAniadys Baows T 968 Plato BiBpdoxnw V 462 Xen once (Bpapo. several times) Plato xara. once, Herod and Hippocr: Attic always in perfect tense? BvOds E 607 609 d R 247 Xen once Bwodrw P 1155 form and sense (call for or to) noticeable yapnAtos, See A€xos (1 T 1034) yepaipw T 961 Xen Plato once or twice ynpoBooxa A678 Eur yovos V 1116% prob an error 1T 117 defua R 688 quite rare in prose (Thuc Plato) as are depaive and depard dyAady V 442 Ec 1157 Soph Eur Plato SiwctiABo P 567 d fr8: oriABw an Av 697 dixy see A (80xe) the longer forms Soxjow edéuyoa N 562 R737 id 1485, Herod and p ddAos see A dope (p form) V 1081 P 357 in probably old phrase civ ddper otv doridi (notice civ) dvoBovAla N 587 p (dvoBovdos not known 1) éyyerys V 1076 Herod not Attic éxromos d Av 1474 Plato once éxpréyw P 608: see dréyw in D and dé in A éurvpetw d P1137 itr L 372 very rare évduxos P 630 Plato once or twice p 150 THE DICTION OF ARISTOPHANES &vOa where Av 1485 1556 (d aoe 1T 1046 par: Xen, not Plato or orators. &ev d Av 748 L 625 (cretic line among tr) dact R 826 ééeraipw L 622 éfiro L291: ixd id E 924 éEopnd T 659 1 Av 1326 Thuc Xen Plato éropBpia N 1120 Aesch Hippocr edjyAtos d R242 evkv«dXos T 968 with xopeia and Baors épeéw V 338 rod 8 eheéw .. Taira dpav oe BovAcrat; not known elsewhere Cwrvpe met L 682, so éxg. P 310 p nrvyn A 684 Ged (not in prose, except for distinctness in phrases like Beot cai Oeat) N 1121: see D Onpopoar FE 607 Onparys an N 358 (6ypeiw usual prose) Ovyoxw for drofvyckw never found in Aristophanes except in perfect réOvyxa (which is regular in prose) and special future form refvyfm tr A325 itr N 1436 an V 654 and i A 590 (Lamachus) oip’ as reOvyées, probably par: Aesch Ag 1279 shows that we must not regard it as a colloquial form. Plato Gorg 469pD seems the only other Attic instance ixvodwoe L 1037 ddA py dpac’ ikowbe old phrase (ddux- votpat prose) xdépta Av 342 and in quotation or imitation i A 544. Not in Attic prose, though common in Ionic. Perhaps some unknown reason for Av 342 (1 Ameipsias 22) xataidw T 727 repeated in i730: ai@w and xartaiéw not prose. Cf. aids T 246 xatofus V 471 Aayds d Av 1560 Hippocr p | AciBopat E 327 Aefavov V 1066 Plato once or twice p Aéxos yauydcov d Av1758 i T 1122 par Anyo P 332 an 1328 hex 1076, mainly p though often in Xen Plato Aristotle, Thuc twice (5. 81 of time yepmdvos drfyovros) Dem 24. 98 mepi Ajyovra tov évavrov, Isocr Dem 14 Lys (2) 2. 74 Ajjpa N 457 T459 R603 899 (all d) itr E757. CfA Aurapys L 672 Plato a few times, Soph. The verb Xen Dem and often Herod Ady spear, not spear-head V1119 itr A1226 par an T826 R1016 ifr 404 prob par (Aoyxodédpos i P 1294) parevw T663 rare word in still rarer sense (search a place). Not in prose. Xen has pacrevw pedro T961 id 970 974 989 pevowd V 1080 earnest lines TROCHAIC TETRAMETERS AND DIMETERS 151 p pévos V 424 14665 Plato once, Xen two or three times pucdmodts d V 411 por) see D. Not in prose povapxyia V 474: see pdvapyos in D povodpartis Av 276 vetxos R1099 an V 867 T788 hex R 818 Herod often, Xen and Plato once, Dem? 25. 52 in a very elaborate passage. (Isocr 15, 268 it comes from Empedocles) veoypos R 1372 p. C£. T 701 éyx@ R703 an V 1024 Xen once. éegoyxa Herod Hippocr ddourdpos A 205 Herod, who with Xen has also Gdouropia ola how E269: not in prose except a few places where it is rather accusative with verb than adverb proper, e.g. Xen Cyr 5. 2.18 éoxwrrov ofa oxw- POivat Wotov 7 py éuBpos AR246 an N 338 377 Xen twice Plato once Isocr Busty once Aristotle Herod duBpws i Av 1593 hex N 287 6200 near E245 id P513fr542: Xen two or three times, Plato once, Dem of numbers about 6éukdpdios V- 430 Aesch éprBarns Av 276 mahatds = yepwv A 220 676 (V 442 prob not in this sense) Pp madiyxoros d P 390 p Herod once, Antiphon once meAdptos Av 321 par epic qémetpos d Ec 896 Soph rare in prose mep A 222 used (ynd€ wep yépovras dvras) as in Aesch Cho 504 Suppl 399 Eur Ph 1624: and ei py repi N 1183 tr L629 Xen Oec 1.14 and 7.17 Plato Ale 1. 124 B Herod 6. 57. 4 moia for aoa E606 (Epicrates an 11. 26, Eur i Cycl 333) ° mwoAvmAoxos d T 435 apéuvov met Av 321 api adverb d V 1063 seldom in prose, where zpdrepov and other words or expressions are usual: Xen Plato once or twice, Thuc a dozen times, Dem 1. 11 [7. 5] mpovcedotpev R 730: the participle Aesch P.V. 438: nowhere else ocBifo T674 1106 prob not in prose ceAnvain N614 Ionic form Emped Eur Ph 176. Cf. Plato Crat 409 B etymol orapatrw A 688 P641 i R424 par otidos P 564 itr E 852 Thuc once Xen a few times couvinps hear P 603° cvvvod d R 598 Plato often Aristotle Soph Eur , tapaéixdpdios A 315 da. X. eis taxos A686 Xen twice: cf. A 152 THE DICTION OF ARISTOPHANES SrepxodS L 693 not elsewhere in Attic. yoAd a few times in later comedy iméwrepos Av 797 p Herod Aristotle Srovpy® P430 Thuc Xen three or four times, Antiphon Plato once, Isocr Demon 31: trovpynpa. Andoc once, ézovpyfa Lys once © pirdpedos P 308 xapate A341 V1013 itr L358 1 A344 Eur once, Homer xavddvw d R 260 Homer Hippoer not trag xGés for éx6és R726 itr V 242 an N 353 fr 53. Elsewhere there is a long vowel or a diphthong preceding which would absorb the «. MSS. sometimes give x6és in prose, e.g. Plato, esp in old phrases with apany x\.86 L640 not in prose xopeia T 968 Rd 247 396 és=dore with infin R1110 uncommon in prose, except Xen apeAyoysos Av 316 Soph C.—Iampic TETRAMETERS AND DIMETERS. édyyos for dyyciov A 936. Cf. fr 234 511 deXarros L 256, not in Attic prose? adeEixaxos N 1372 (dAcéixaxe) an V 1043 épnyavos E759 i R 1429 par, -ia an Av 475; two or three times in Xen Plato [Dem 60. 12], dunyavia Andoc, éunxavé Thuc 7, 48 Xen twice. Even of things the word is uncommon in prose except Plato: see pynxavy in A dvOeuadns R 449 én. E 349 an P 764 L593 Pl 538, Andoc once, Xen Plato freely : the active at any rate is rare: the noun dvia seems confined in prose to Plato aves L 385 an E 534, Plato and Herod once Bopd E 416 1T 1033 Plato once Borov N 1427 Plato once Bpadive see A éyxova, see A éxar. see A guBa Ec 478 évavra E342 p évdvouar enter on Ke 288 with the p word réApnpa é~apmpedw L 289, dp- apevw Alexandrian p é€aveyopat N 1373 p eEedpnua E 886 1 Ee 577 Aesch Herod éravaBod Pl 292 am. r.2 érwpeA@ N 1442 Xen and Plato once p gpapot N 1076 an R1022 1 N 1304 doubtful (all these 7pdo6nv, used sometimes even in prose) : present tense an V 751 par with fr 51 hex P1098 IAMBIC TETRAMETERS AND DIMETERS 153 ebyhorria, E 837 Eur: edyAwrros an N 445 p edpnxavos E759 Plato édicnar bid V 242, Thuc 4. 1087 p éforxds V 268, Thuc 4. 108 in another sense qué N 1487 V 245 T1030 1A 336, unusual in prose for 7AuKudrns, Antiphon once Xen and Plato a few times Oapuvd Pl 292 with p réxea, Xen three or four times Oovpios E757 p Opéupa L369 1 Ec 973, Plato often Xen twice Isocr 2. 45 prob an error Owpds fr 217 (cf. 94) 11 973 p and Theophrastus im@, see é&iro in B katacroow T 560 Aesch: kdove E 361, cAdvos an N 387, notin prose koptracpa R 940 Aesch kéros P1321 1 L542 two or three times in Xen Plato who have also dxoros: R.1268-80 it refers to xéaos in the quotation. See com in A koppos L 255 Herod Aexrés Av 422, not in prose ? peOnxo E937 Eur . pépysva N 1404 an 420 1 950 (uepysvodpovriorys i 101): not in good prose (Hrastae 134.8). jeptxvd Dem once Xen several times pveia E876: a few times in Plato and orators, esp Isocrates, chiefly in phrase pveiav rovetoOo, seldom memory as here punpdovvov V 537 an 559 Herod Thuc pox06 Pl 282 an 518 525 556: pdyOos only in an T 780 par: Xen has both a few times, Plato neither, Thuc the verb once. Not in orators poxAreutns N 1397 (cf. Porson on Med 1314) 1 567, not elsewhere otxos house (t.e. building) for ordinary oikiéa fr 73 rov otkov ypepev had prob some reason, 115 obscure and doubt- ful (airjs suggests oikias): tr Av 380 prob estate, 1N 600: xar’ ofkov iA 1174 par id L 261 an P 88 was to some extent a phrase: it occurs two or three times in Herod Thuc. Cf. éx’ ofkov. Plato has oixos Phaedr 247 a (dignified) ev Ocsv oixw, Xen a few times : p very often 6pex06 N 1368, rare, not tragic ovpuos E433 an L 550 1A 669: used in prose a few times of the wind itself mayxpnotos A 936 Xen once mavypepos R 387 Aesch Soph Herod mapacrat® T 370 and ovpr. R 385. Both i Ec 9 and 15 par p mapevoadevo Pl 291 anpaivw A 842 Plato p moAda often E 411 Pl 253. Other passages like A 694 984 V 678 685 154 THE DICTION OF ARISTOPHANES R 697 seem rather acc with the verb. Xen now and then e.g. Cyrop 1. 5. 14 woAAd por cuvdvres, and Plato e.g. Phaedo 61 © ToAAG 48 evretixyKa TO dvdpi modvppodos R 448 apecBus old man A 1228 i T 146 par an Av 255 1V 1451 o¢Bw fr 569 13, 14: see D ovvedds Av 635 Herod, Plato once half jestingly Phaedo 92 c Aristotle p zéxea, P1292 p (notice form) opedjpa E 836 Xen p D,.—ANAPAESTIC TETRAMETERS AND DIMETERS. dyen Av 591, fairly frequent in Plato, once or twice in Xen. Its not being commoner may be mere accident dyjpws Av 689 Thue, Plato, p aXéxrwp V 1490. aryooe: Ppinxos is said to be proverbial, but not the other half of the line, ds 71s dAékrwp. Cf. N 666 851. Prose dAexrpudy ddoxyos R 1050 1 L 1286. Ar Pol 1. 3. 1253 b 7 wéots cai dAoxos a8 an old phrase, both being quite poetical in their ordinary use (Theaet 149 B is etymol) dpa as prep E 519 520 V 712 1A 346: see A dpeérpytos N 264 p avaxeioOat éxi (tur) Av 638, with dat Antipho 5. 6 and Eur, with és Thue once, Herod twice dvaddys V 1045 Hippocr dvdperros N 984 Eur [Dem. 25. 32] rare, évapeord R 1084 az. 2.1? évapiOpyros V 1011 pupiddes dva- piOunro. a phrase which occurs three or four times in Plato and others and is perhaps a poetical quotation, as the length of the . suggests, if not the tribrach dvapoBS V 670 aa. X17 évepwoxyns Av 697 Eur évrioy@ N 321 V 546 hex R 878 Soph avripepiftw E 813 818 epic érapivey V 597 Plato once, Hom dreipov Av 694 i? fr 250 Hom dayvys N 974 Plato twice, Hom but not trag dzroxAdw V 564 Plato once p drrodéyws V 580 L 576, not p, rare in prose dary Av 687 Plato once Hom dpityAwros E 1329 ada. d.? apifnros epic dporos R 1034 Plato Crat 406 B etymol p aiAy P 161 par atre V 1015 ANAPAESTIC TETRAMETERS, ETC. 155 hex P1270 1 N 595 L 1296, decidely p &pOtros Av 689 with pnddpevos and 702 p Baxyeia R357 Béoxw T 783 par. SsiaBdoxw an Av 486. Rareeveninp — Bioros Av 718 Ec 594 669 Pl 526, not in prose, though fioredw is found there now and then: Pury Xen Cyr 7. 2. 27 BopBopddupos P 753 da. 2X. Bpovrycixépavyos N 265 ydvupor V 612, Plato Phaedr 234 Dp. (yava) éyaviOnv i A 5 (yeyavwpévos Plato Rep 411 a) yéynba E 1317 Plato two or three times, Dem 18. 291 323 ‘yeved. race etc Av 685, a few times in Xen Plato. Prose uses it for a generation yevvacompery~s P 988 da. X. yAvkvOupos L551 1N 705 Sdios N 335 perh quotation R1022 1 897 p SdredSov P1515 1 Av 1265 R 351 fr 110, Xen three or four times, Herod dedérropat L 564 Plato and even Dem, though mainly epic (80x) Sedoxnpar V 726 ap form: cf. B Sepaivw V 1042, Plato once or twice SiaxpGpat use Ee 609, Xen once Herod often Spopatos P 160 p, Xen Cyn as a technical term dvoxoAdKapatos N 971 az. dA. 8vaKoAd- xotros N 420 da. A. éyxucdodpat V 395 699 rare éyxeipa, see A eixoBoAG fr 689 Eur éxduddoxw R 1019 1026, see A eheaipw E 793 epic Acrilo Av 213 fr 500 par epic éédxeois R 1033 dir. 2. éxaydai~w fr 682 1 Ee 575, Cf. dyAads in B érefyw intrans T7831 P943 p érixoupo aid V 1018, occasional even in orators: érixovpos E 1319 éixepa V 581 very rare in prose éxraBdeos R 1017 épddy E511 i V 1148 for a pun, Ap Rhod ciavOys N 1002 R 373 p Plato edxrea N 997. ci(Suo)xAeys and cd(Sve)kAen occur in Xen sometimes, two or three times in Plato, in orators very rare, see A and xAdos below eivy V 552 Twice in i, but T 1122 is par and Av 1286 assimilates men to birds: tr E 605 refers to horses: 1 Ec 959. wodAd. 76 would easily be lost before zo, just as before rovs (rd robs Kapmovs) in 60. 5. 20 det 82 xpypdruv. It is well known that 8é and ydp get interchanged. Logic here seems to require ydp. 26 ris airév kwrvoe Seipo Badifovra; OnBator; py dlav > > a > 7 a c #. > * nw + mucpov eirety 7, Kal cvvercBadodow Erotuus. GAA Buxeis ; ‘T wish it were not (I fear it may be) too harsh to say’ (what is only too true) is Sandys’ rendering. (1) This use of pis however not found in Demosthenes or (I think) any other orator, being practically confined to Plato, who in Gorg. 462 for instance has the apparently similar pi dypouxdrepov 7} TO dAnOes cireiv. (2) The sense is unsatis- factory. The editors seem to have a notion that the words are more or less equivalent to those of Phil. 3. 1 8éoxa wy Brdogynpov piv cimeiv, GAybes 8 7, and mean something like ‘harsh though it be to say’; but they cannot be twisted into that. 174 THE ATTIC ORATORS Demosthenes is quite fond of putting a rhetorical question to his adversary, or his audience, and suggesting an answer which he immediately proceeds to demolish. Thus we read in Mezd. 41 zoia yap mpddacts, tis avOpwrivn Kal perpia oxpiis pavetrar tov Texpaypevuv ait@; dpyy vi) Afa: kai yap todro tuxév Ade. GANA K.7.A. Chers. 17 x moujooper, av ext Xepordvycov iy ; kpwotpev AtoreiOn vi Mia. kal ti Ta mpdypar eotar Bedtiw; De Cor. 101 ti gueddov Kerevoew % TL cupBovretoev airy qoreiv; pynotkakciy vy Ala... kai ris obk Gv dréxrewey pe Sixaiws, ef «7A. For these and many other passages see Rehdantz’ Index p. 35. Anyone who has read thus far will perhaps have seen what I wish to suggest, namely that yy Aiav is a corruption of the vy Aia (MHAIA, NHAIA) which occurs so constantly in parallel passages. So Herwerden (and I myself not knowing that he had anticipated me) suggested some time back that in Plat. Rep. 6078 the absurd rév Afa cov should be rav Alay codiv. If we adopt vy Aa here, a slight further change will be necessary. I should suggest ris airév xwdvoe. Setpo BadiLovra ; @nBaior viy Alas Kav mixpov ciety 9, Kal cvvercBadovow Eroipus. Ol. 2. 2 Set roivuv, & adv8pes "AOnvator, rodr’ dn oKoreiv abtovs, drws wy xelpous rept Huds abrods elvar dogwpev Tov trapxdvTwv. 4b, 22 &edévrwv & mpoojKer Torey Sudv adTyv Kal xara pxpov. 1b. 25 pedddvrov ait ay, érépous twas edmildvrwv mpasey, > tf 3 gi aitiwpévoy GAARAOvS, K.T.A. In 22 airév is omitted by Dionysius and Gregory of Corinth: in 25 S alone has airdy, other MSS. tyudv. In all three places some modern critics have wished to omit abrovs or airay, e.g. Cobet in 25. In 2 the atrovs seems very pointless, and with the jas atrovs immediately following distinctly objectionable: in 22 I can hardly think airdy right, all things considered. Perhaps for aitovs and airav we might think of substituting rdvras and révrov, which will be found more suitable. It is curious that in all three places the word preceding ends witha y. v and a DEMOSTHENES 175 are liable to confusion (Bast Comm. Pal. p. 747), and the v might therefore absorb a z. In Plutarch’s Demetrius 30 rod xabeornxdtos é&éorn 8 épyyv a’rot should we not read zavrds for airod, he lost all composure? The final y recurs there. dmavras for adrovs is adopted by Blass after Hertlein in Isocr. 5. 120, and in (Ar.)’A6. TIoA. 39. 1 adroxpdropas dardytuv is a very plausible conjecture for airoxpatopas éavrGv. See Index s.v. airds. 3 iptv 8 obxt Kadds wempaxOau. Do not tds preceding and jyeis following point to tyiv? 14 iawfpéé ro Syitv ét Tyobdov mpds ’OdrOiovs (7 Maxedovexy Svvapus): wédw av mpds Toreidaray ’OAvvOiors epdvn te TOTO cuvayporepov. So the words stand in most editions, though Madvig’s stop after ’OAvwvOios is sometimes adopted now. I have no doubt that it is right, but it does not give us quite all we want. The scholiast’s explanation of cuvayddrepov: Rep. 400 c as oty dddw Twi yevouevov ought not to have been accepted by modern scholars in view of the use and the plain meaning of the word. After the last letters of rotro a 76 has been lost. Demosthenes wrote ébdvy ti rotro cuvapddrepov, this combination. Cf. [Plat.] Hp. 8. 353 B elre.. elte.. ele kal TO CWvapporepoy aitiav EvpPnvar yevowevyy. 17 of 88 8% wept abrév dvres Eevow kal meLéraipor ddgay piv dxoue’ ws eiot Oavpacrol Kal cvyKexpoTnevot TA TOD TOAEuOV. Even if we render davypacrot by admirable and not by wonderful, which is rather its meaning, do we not feel that a bare ovyxexpornuevor is not enough to keep it company ? That troops are o. is not enough to make them deserve the epithet @avpacroi, and we should at least expect to find with o. some heightening adverb like ¢d or xdAMora. I suggest however, not that any adverb or adverbial expression has actually been omitted, but that Demosthenes wrote Oavpactas (or Gavpaciws) as ovykexpoTnevot, as in 29. Lhe writes davpaciws av as yidaBovpyy, in 35. 16 Adyous Bavpaciws ds wiPavors, and in 37.10 davpacras ds éAvyOnv. cai and ds are very liable to interchange (Bast u. s. p. 24 176 THE ATTIC ORATORS and elsewhere), and, «af once written here, the adverb would easily become an adjective. OL. 3. 1 ode ody dAXOo por Soxodow of 7a Towatra héyovres i viv idbcow repli Fs BovdcterOe ody) riv obcay wapirrdvres iptv duapravey. As it stands, this is a very clumsy expression. Has no one ever suggested the obvious and idiomatic future BovAev- ceaGe, ‘bringing before you for consideration’? Unless I am much mistaken, a similar correction is to be made twice at least in the First Philippic. In $46 érav yop Hyfra wey 6 otparyyss GOAiwv dropicbuv Eévwv, of 8’ irép” dy ay éxcivos mpdéy mpos twas Wevddpuevor fadius évOds’ dow, idiom in the same way requires or at least very strongly suggests Wevoduevo. Cf. the end of speech 51, where Tos... wevdouévovs has all the MS. authority, but the future is recognised as necessary. In 43 efra rodr’ dva- Pevotpev ; kal Tpinpets Kevas Kai Tas Tapa TOD Setvos éAmiSas ay drooteiAnre dvr exew oleoGe xahds; odk euByodpcd’ ; odk ééyuev adrol «.7.A. the future tenses before and after, the really future meaning of av drooreiAnre, and the manifest sense point in the strongest way to fev, unless indeed we should read oijcec6e, which is less likely. If these corrections are allowed to be probable, I should be inclined to go on and say that in 32 tmdpye 8’ imiv yetpadio piv xpjoOn TH Suvdper Ajnuvy kal Odow Kal Bidby . . ev als cal 1 Is not the puzzling expression in Soph. Phil. 300 pep’, & réxvov, viv nal 7d rijs vhoov wdbns. Tatty meAd Cet vavBdrns obdels Exdy to be corrected in the same way by substituting és for nal? I can- not indeed produce any exact parallel to pépe, &s wdOns; but, if Plato could use the unusual phrase pépe, édv (Crat. 4304 gépe 5h, édv wy BiadAaxOGpuev, & KparvrAe Gp’ ob «.7.A. Rep. 453E pépe 54, fv 8 yd, édy wy etpwuer Thy todov: duodroyoduev yap 5h) x.7.A. Of. Dio Chrys. 13. 29 pépe, dv... Siadéywuat. Sannyrio the comic poet has ¢ép’ «i (fr. 8). Late Greek has also occasionally épe eimety, say, let ua say), there would seem to be nothing against the possibility of pépe, ds in a poet. The two words need not however go together, nor need there be a full stop at pdéys. Cf. the elliptical ds ay oxomol yuy Fre Antig. 215 and ws obv pnd’ ya rabrdy wd0w O.T'. 325, DEMOSTHENES 177 Aypéves Kal otros Kal & xe? otparedpar. wav0’ ordpxet, which is matched with a future in the 8€ clause (Jadiws éorar), one or other i dardpxev should be irép£e, both for symmetry and to avoid imdpxet awkwardly occurring twice. Cf. Dobree’s cor- rection of é&w to éyw in Frogs 1230, where ee immediately follows. For choice I would write iwdpgea 8 iptv, though strict symmetry would perhaps rather favour mavO imdpeer. In Cherson. 38 eiot Toivey twés, ot Tor ebehéyxew TOV TapiovT’ otovrat éreday épwrycwot ‘ri otv xpy movetyv;’ I have very little doubt that we should read é&eAéyéew. In De Cor. 99 thy épyny eis TAAN’ eere would tally better than éyere with the corresponding otre pyyoixaxyoer’ ob6’ brodoyteiabe. This corruption of future to present is one of the most inveterate tendencies of MSS. 28 éxOpov 8 ed’ yas adrods ryAtKotTov HoKjKapev. hoxjxayev is not a suitable word at all. Will anyone really defend it against what I should like to read in its place, yigyxapev? Ol. 1. 9 ydéjoapev ... Didurrov Hyets: 5 péyas nbénby and 1b. 6 yiéyudvov, T ni&yOn: Phil. 1. 104 em IbeNT aL Pil. 3. 52 eis & dyav’ dwewov jpav éxeivos oxyrat is no parallel, yArkotroy is predicative like péyas in péyas nb&70n and other words (dmeipov abfjoee Plat. Rep. 591). For the correction cf. on de F. L. 339 and Isoer. 8. 13 below. 31 dyamavres doy peradiddor... i}... meppwow. The difference of present and aorist here is much more than that of a ‘continuous series’ and a single incident’ (Sandys). édy with aorist means ‘if they have done so and so,’ while éév with present means ‘if they do’ or ‘are doing it.’ It follows, I think, that we should either read peradaor, for which there is authority, or change réuyoow to a present. The mixture of times, though not impossible, would be awkward. The parallel passage 13. 31 favours the present tense, for all MSS. seems to have peradidacr there. 34. odkoty od pia Bopopay deyets ; pire. Tust Kal mapaxphyd Ye Thy abtHy ovvracw é drdvrwv ..., va tav Kowa éxacTos TO pépos AapBadvwv drov Soil” 4 woAus Tovd’ iadpxot (or rovTo N 178 THE ATTIC ORATORS mapéxor). eeorw dyew yovyiay; olko. peévew Bedriwv.. .° ovpBaiver te Tototrov olov Kal Ta viv; oTpaTLOTHS avTos brdpxov... els rééw yayov tiv Tow... obK ~otw omov pasty ey rovotaw ra. Tov movotvTwr elroy ds det véwev. rodro mapéxot, which the text of D. Hal. 609 gives us, seems a more usual and likely expression. But, if we adopt it, [ should like to bring the later irdpxw into harmony with it by reading orpatuiryy airov mapéxwv. Though in each place the use of trdpyw as distinct from elva: can probably be justified, it needs justification and is perhaps a little suspicious. Cf. rapacyelv éxacrov atrév x.t.A. in the similar context of 13. 3-4. Great difficulty attaches to the use of the optatives after iva and no even plausible account of them has been given. Observing the aorist tenses (jyayov, «irov) which follow a little later, I should suggest that we read édeyes for Aéyes and thus remove all difficulty as far as the optatives go, just as Madvig’s xaficracay for xaficracw in Plat. Rep. 410c makes the optative possible there. Hirschig must, I think, be right in reading éAcyes for Adyes in Protag. 3508: otherwise what is the sense of kal viv ye! But then what is the meaning here of the past tense éAcyes! I do not know ; but neither does any one know what is the meaning of jyayov and etrov. The obscurity is not increased in any way by éAeyes, while it accounts for the optatives and renders tapaypyua to my mind more natural. On the other hand 1. 19 ri ofv; dv ris eizor, od ypddes rair’ elvat orpariwrixd ; may be quoted in favour of Aé€yes, which in itself is no doubt easier. 36 py wapaxwpeiv tis tdégews Hv tiv ol mpdyovor Tis dperis .. . ktnodpevot xarédumov. Entirely disbelieving that ris dperjs can depend on 7, I suggest either either ris dperjs or TH dpery. Phil. 1. 22 woGev 8% rovras 4 tpopy yerpoerar; eyo xal rotro dpdow Kai Selgw, éredav Sidte TyAtKavrny aroxpiv olpat tiv Siva Kal mwodiras Tovs orpatevomévous civar Kederw &basw. In this passage rots orparevouevous can hardly be right, because only a portion of the troops were to be citizens. DEMOSTHENES 179 Dobree thought of rois cverparevopévovs, Spengel of roAiras Tos otparevouevors mapetvat. Probably we should read Todéras Tovs oTpatevomevous <éviovs> ecivat, where the loss of éviovs will be due to the -evovs preceding. “Evior (ciotv of) is sometimes put thus in apposition instead of taking a Benitive. Thus Ol. 3. 11 rots wept trav OTparevop.evov évious: Aphob. A. 23 d0° &na papbe caradepOiivat mayrdmacw qupear Byrne: Thuc. 1. 6. 6 év rots BapBdpous & éorw ots: Ar. Hth, 9. 1. 1164 a 27, év rots Towovrots 8 éviow: Poet. 9. 1451 b 19 & rats tpaywdiats éviats. We find such words as moAAol, of roAXoi, Exacros, even oddeis added in the same way. 44 otk eipev-adrol pepe yé tit otpatiwrav oikelwy viv ; Should we write oixefw? A possible cause of error is obvious in the v of the viv, but really wy and » are almost freely interchangeable. De Pace 11 oisé Tpoomoujropat ov obdev ado yeyvaoe kal mpoataOdver Oat wAnv 80 dy byiv efrw dvor ev pev K.7.X. The editors seem hardly to realise the difficulty of 8/ av buiv etrw dvo, ‘whatever two things I shall have said.’ It is surely nonsense to say ‘my intelligence and foresight are due to whatever two reasons I shall be pleased to assign.’ The words cannot possibly mean ‘ the two things I am about to state,’ which is the sense required : dy or & dy must be indefinite. It seems likely that av iyi ctw has strayed by accident from the line before, and that we should read dv obdév dAdo yryvedcnew dy ipiv ctw Kat mpoacOdverbar wiv 81a. So" ev pep x.7.A.) In a somewhat similar way kai dao. ris ay efroe in Ol. 1. 13 has evidently grown out of dy ris efrou in the next line. 20 od8€ ye tv adrots werovnpevwy toratov éAOdvta Ti ddfav exeuv. It is not easy to understand torarov. Read torepov. 1 See an interesting article by A. C. Pearson (Class. Rev. 17. 249). N 2 180 THE ATTIC ORATORS 24 ddd’ ds ovre mpdfopev oddev dvdéiov judy abtav ott’ ora. wéAcuos... TodT’ ofuar Sety woreiy. (‘detv woely 8: defEew or Seifar vulgo: Seiv évvoety Liebhold’ says Sandys.) I doubt whether és with a future, even when resumed in a rovro, can be made to depend on zovelv. As ds comes at the beginning, the verb on which it was to depend must have been foreseen, and, whatever a writer like Xenophon might do, it is very questionable if Demosthenes would have written oveiy és with a future. (Cf. however the con- struction of drws in § 13.) Observe further the clumsiness of bringing the two verbs zouely and zpdrrew thus together, mouiy &s mpdfouev ovdev, For oueiv we may perhaps read oxorety, which would be the natural word for Demosthenes to use. Aclear case of the same corruption is Plutarch De Exilio 606 c rod 8 Ovcapévov kai ra omddyxva KwAvelv gpdoxovtos, ereriuyoey eirov ‘at Ti Kkpéas A€yet Troveis, ov Ti vowv éxwy dvOpwros,’ where Madvig restored oxomeis for motels. oxdme. (or voe) should probably be read, as has been suggested, for the unsuitable vote. in Plato Rep. 609c. In Isocr. 7. 64 vociv is a variant for zovetv. Phil. 2. 3 ota wovet 8’ as Seva kat roatra diekepydueba. I should prefer dca to ofa. The mistake is fairly common. 11 pera tadra rpdgavres tai6’ & waves del ydixovrat déyelv, déins 8 oddeis cirety Seddvyra. Here on the other hand I should like to write rot’ for 7236’. Demosthenes’ ear would probably have shrunk from 7ad6 just after ratra (as from imdpye. twice over in Phil. 1. 32 above), and ro.aira is at least as proper. 25 ri Lyreir’ ; ebyv: edevdepiar. Should we not point édevdepiay as a question ? 31 ri 84 tadra viv Aéyw;... ody Wy’ cis AoWopiay eumerdv euauTe pev é& icov Adyov wap’ tiv Toujow, Tois S€ K.7.d. For éuavrg Adyov rowjow Sandys offers the two alterna- tives that the words=iva Adyov rixw, and that they =ia Aéyov oujowpat. Besides obvious objections to both alternatives, the context indicates, I think, that iva should DEMOSTHENES 181 introduce something in itself plainly undesirable. Not merely cis Nowopiay éurecdv, but the main predicate should be of this character, as in the parallel clauses. Such a sense and one otherwise unexceptionable, as far as I see, might be obtained for these notoriously difficult words by the slight and familar correction of Adyoy to Wdyov. wdyov €uuur@ rover, bring blame on myself, would seeem to be as good Greek as oAeuov woeiv bring about a war, which occurs repeatedly, for instance in the speech about the Chersonese (§$§ 7, 8, etc.). Add such phrases as 1b. 52 hovxlay movotow éxeivw mpdtrew 6 te Bovderar: Phil. 4.°7 9 Kad’ ypéepay pactivyn... odk ef’ Exdotov Tov duehoupéevuv moet THY aicOnow edOews : Timocr. 210 xara tis moAcws Sdgay od xpyoTyv moujoere: Plat. Apol. 20D ri wor éorw Totro 3 éuol weroinney 76 Te bvopa Kat THY SiaBodryv; Laws 937 c TH vikny tO édOvre TeTojxevar. Alcid. Soph. 18 has wéyov mapacKkevalety. But gxAov wap’ iuiv wowjcw would be equally admissible and is perhaps even more likely. In Thuc. 7. 56. 4 éxAov ‘is now often read by conjecture for Adyov. Cherson. 5..To the proposals for supplementing éors dé I would add the suggestion gor, 8 . Plato makes free use of éorw otrws, and o8%, a favourite expression with Demosthenes, would seem equally possible, though I do not know any precise parallels but I/. 18. 266 de yap éorar : Soph. £1. 573 &8 jv 1a xetvns Ovpara. 64 Perhaps ypypard rw’ for xpypar’ airdv. Phil. 3. 29 dowep repiodos 7} kataBoAy wuper 4B. Should we omit # xaraBodj as a gloss on epiodos, since they mean just the same ? 48 otrw 8 dpyxaius efxov... dor’ od8& xpyudruv dvetcOar map’ ovdevds ovdev. The difficulty of dar’ od, not adore pj, here is well known. Has it ever been suggested that we could remove it by writing éyew for (or , the usual tense with ds when, after) dravra, rér’ qAmilev Ta Tpdypar’ dvaipjoerOar, kara Suépevorat. Ol. 2. 2 &s tors tv aicypav, paAdOV 8 Thy aicxicTov, pH} pévov médcwv kal térav dv Fpév wore Kipior paiverOar (or ) mpoiemévous GAAG Kat rdv ind ths TUXNS TapacKevacHevTwy cuppdxoV Kal KaLpdv. ib. 28 évratOa pév gore rdON srép dv éorw 6 wdAEMos Spérepa, "Apirodis: Kav Andy, tapa- xen’ duets Kopretore of S& xivdvvor trav epeornkdrwv id.o1, 184 THE ATTIC ORATORS pucbds 8 obk éotiv® éxet 5é xivdvvor pev eAdrrovs, Ta SF Ajjpparo, tov épecrykérov Kal tov otpatwwrav, Adpipaxos, Séyeoy, ra mot & cvrddow. Ol. 3.15 xai yop ciety ra Sdovra wap’ ipiv cioi , & dv8pes ’APnvaior, Svvdpevor, kai yvavar tavrwv bets d€vrarot 7a. pnOevra [ciof wanting in §] or perhaps
    duvdpevor. . - > i” » tb. 26 ob yap cis wepiovoiavy éexpdrrer’ aibrois Ta a y o rhs moXews, GANA 7d Kowdv avéew exaotos wero betv. Cf. Meidias 159. Phil. 1. 48 of pév wepudvres pera Aaxedapoviwy gaol Pidurrov rpdtrey tHv OnBaiwv xarddvow ..., of 8’ as tpér Bets wéroppev ws Baciréa, oi 5 ev “IAAvpiots wéAes TerxiLetv, of S& Adyous wAdrrovres <érépous> Exaros mrepiepyducHa. Phil. 2. 22 6AAG pv yéyovey radra Kal waow éorw eidévat. Chers. 32 dv pav oby tov airiov elry tis, dv io’ ére AnperOe wap’ tuly airois, gare kat Bovdrgcobe dy de Towovroy A€yy Tis, dv K.T.A. OdK exer’ Oluat Ti Toijoere. Or write tw’ for tov. Ol. 1. 21 The meaning required seems to be: he never expected to be compelled actually to carry on a war: he thought the mere rumour of his coming would bear down opposition and give everything into his hands. But this can hardly be got out of ds émv, ‘he thought to carry everything before him, as being about-to attack.’ Cf. Ol. 3. 4 darnyyéOn . .. wodtopxay : Thuc. 3. 16. 2 #yyéAdNovro al vijes Thy Tweproixida TopOodca. 2. 2 It is certain the genitives cannot be governed by mpotepevous. If the passage is right, they must be attracted into the case of the relative dv. I have noted only two places in Demosthenes (though there may be others) where anything like this construction occurs: 18. 16 érépw 8 érw caxdv Te Saoonev Cyreiv, and 49. 3 gero... od pdvov r& éavrod Koptetr Oar GAAG Kat dAdov ei Tov Seorro mpds TiydOeov wpasat trdpfewv airg. These are obviously very different from our passage, even if they can be called examples of the DEMOSTHENES 185 construction at all. dAAov and probably érépw go so closely with the other pronoun that they need hardly be brought into any relation with the infinitive. In prose moreover and even in poetry it is almost always nominative or accusative, not genitive or dative, that stands in this sort of attraction. In our passage the difficulty is greatly enhanced by the genitive continuing in réy tré x7. I conclude therefore that some such expression as deAclv, dduywpely, pndev dpovriey has been lost. 2. 28 “Audirodis wav Andfy, as it is usually written (Madvig and Blass in some of his texts omit “Apdimrodts altogether), .throws an unreasonable emphasis on the name, which requires none at all. The parallel sentence with Adpyaxos, Séyeov, ra mwAoia strongly suggests, I think, that one or two names have dropped out. Cf. de Cor. 69 non yap o épwrd avr’ adeis, “Apdiroduw, Tvdvav, LoreiSatav, ‘Adévynoov. 3. 26 By itself cis wepiovciay is in no way. contrasted with 7d xowdv atgay. There may be a surplus in the treasury as well as in private hands. Chers. 32 Is the meaning sufficiently expressed without such a towodrov as appears in the parallel clause? The resemblance in letters to tov airiov would facilitate its omission. : There are one or two other places where the insertion of a word or two might make things easier, e.g. Ol. 1. 2 an infinitive like qvAdrrecOar before py mdfyre: 2. 29 a participle with os rovrous (but cf. Thuc 8. 67. 3 éAéoOae... mpos éavrdv) : 3. 7 something after jes ; but I do not wish to propose it. De Halonneso. 13 7a cipBodra ratra yiyverar cis trodoyyy Tod pyd dpdioBytica: edAdyws buds ér. Toredaias, ‘as an admission that you cannot lay reasonable claim any further to Potidaea.’ The sense requires pyd’ dudioByrioa. dv dropped out before az. The future duds Bytyocew would be less suitable. So in 16. 5 it would seem necessary to insert dv before dyrimdAous. 186 THE ATTIC ORATORS Epistula Philippe. 5 dor éywy’ dopa ti mor éorat Kawwdrepoy éay 6poroynonré jot mohepeiy: Kal yap Ore pavepiis Scepepdpeba, Anoras eLemgumete Kal Tobs mA€ovTas WS ipas érwAX€ire, Tots évavriow éBoybeire, tiv ydpav pov KaKas érrovetre. Read dr’ davepis Siehepdpefa. The time referred to is from 346 to 340, when Philip and Athens in spite of acts of hostility on both sides were not at open variance and war. Cf. 18. 43, 44. OT and OY are much alike. 14 rpoikarovpny Kpibivat rept rovrwy mpos tas. The meaning is clearly the same as that expressed in 16 by mpoxadovpevos tyas eis xpicw édAGetv, and it follows that mpos tas should be zpos jas. This would be evident even without the parallel passage. In the Letter jpeis is constantly used of Philip. De Contributione. 19 rehecOyvar is perhaps justified as against Cobet’s doubts by Herod 3. 86 émiyevopeva 8% tatra TG Aapelw érehéwo pty. “ De Symmoris. 14 Adyov pev obSev’ eBouddunv Opaciv od8 zxovra pdrasov pjKkos ebpety. a t 36 60° 009’ dv Katapwpevos nopé tis abrois, TooadTa memdr- Bact Kaxd. The verb cipicxw seems a strange one to use in either of these places. Is it not pretty plain that we should read elvey and etre? In Cherson. 54 and Phil. 4. 56, twin passages, the former has ciety, the latter ecipeiv. Isocr. 2. 41 MSS. vary between the two. etpor has been restored very plausibly for eto. in Dio Chrys. 32. 91 dyapryjpara eiro. Tis dv mavtaxod. Cf. Index. DEMOSTHENES 187 De Rhod. Libertate. 15 éwet ‘Podios ye. . . cvyxaipw r&v yeyevnpevor. The construction of the genitive ought not to be pro- nounced impossible in prose (cf. Isocr. 18. 51: Andoc. 1. 67, etc.), but 7d yeyeynudvy would be a very simple change. Of. such sentences as Phil. 1, 42 rots yeyvomeévois... aicxuvduevos: De Cor. 291 rots yeyevnpevors dviapots obdév “6moius Exxe Tois adAas: Thuc. 3. 98. 5 trois mempaypévors poBovpevos tots “AOyvaiovs. We have aapoéw0atuv ra yeyernpevy in Dem. 39. 3. 27 Kv xat ‘Péddov kat dddas érépas wéAets “EAAyvidas. For dAAas read roAAds. Cobet Misc. Crit. p. 75 would apparently just omit érépas, but this is unsatisfactory. Goodhart’s suggestion of dAAovs for zoAXovs in Thue. 8. 66. 5 is very convincing. Isocr. 9. 5 the MSS. vary between the two words: so too Dio Chrys. 29. 4. Pro Megalopolitanis. 19 GdAG pay a ye hacw wempaxOar . . brevayria pods Hpas aromov viv pev ev Katyyopias pépet trovcicOar, BovAopevwr dé yevécOa. ditwy airav, iva tovvaytiov et Toldow jpas, Backaivew x.7.X. SS There is certainly no meaning in viv pév, for there is no contrast of times. Weil would omit »é. How would it be to insert ratra before it (viv pév)? De Falsa Legatione. 10 mpéoBes Treupat trois cvvdgovras Sedpo rovs BovAevao- pévous Tept ToD . . woAguov. The article being unusual with a future participle after mwéumev and an accusative, and the two expressions, rovs ovvégovtas and rovs BovAevoopevous, being somewhat awk ward together, I think the first tovs may well be an acci- dental insertion due to the second. 188 THE ATTIC ORATORS 12 yiyverat trav rpéoBewv rovtwv eis kal otros, ovxy ds Tov drrodwcopeveov Ta terEepa K.T.A. After droSucouéver insert dy. The cause of its loss is obvious. Cf. on Metdias 220. 16 vépov re Onoew Kal ypdipew. Oyo kal ypdiper? 99 8 a 8é + <—- © a oe >on © ~~ > , xX et O€.. . exely’ Spay dri, dvtw’ av tpeis eis tadtyv TH * ¥, a nN rdéw KateoTyoaTe ... ovTos ... Tav lowy aitios HY av KaKor. Editors seem to acquiesce without demur in the dv following dévrwa. No parallel however is cited for it (Goodwin M. T. § 557), and surely its proper meaning would be ‘anyone whom you would have appointed.’ This could hardly mean anything in its context except ‘any other man whom you would have appointed, if you had not appointed Aeschines,’ and it seems impossible that the condition should not have been expressed. Goodwin renders it ‘might have appointed,’ but that is not the meaning of dy with a past tense of the indicative. Our idiom allows us to translate 7AGev av tows ‘he might perhaps have come,’ but it does not follow that 7A\Oev dv can mean ‘he might have come.’ No doubt dvrwa kateoTyoate, Which would give the meaning here required, ‘anyone whom you had appointed,’ could also mean ‘the man that you did appoint as a matter of fact, whoever he was, but this possible ambiguity will not justify the strange use of dv. dvrwa 84 (64 for dv) is rendered some- what unlikely by the hiatus with tyeis that would arise. It is noteworthy that dr «i kat évrwoty seems to be the reading of the less good MSS. 53 as ta wap’ tudv ervOovr’ ex tis exxAyoias. Tax THs éxkAnotas t 61 dépe 84 poe Kat Thy cvppaxiay tiv Tov Buxéwv. 178 obkoty ratO’ irecxdunv ev dpyy, Tair’ éwederéa. Cobet would read ar@’ for the first radra, but [Plat.] Ep. 7. 347 radr’ éppyOy, raira ~vvwporoyyoapey and this passage together seem enough to establish the idiom. DEMOSTHENES 189 193 pero . . ri 8} pdvos oddty érayyéAXerat (asks no favour of him) 7} tiv’ év atré puxpoyryiav 7 apds abrov dndiav éveopakas. The first 7 or makes no proper sense. # and xai get confused, and the latter seems much more likely here. Cf. on Androt. 28 below. 224 éud the reading of S seems to me much better than pe, which the editors usually adopt. There is a certain emphasis on the pronoun, which is heightened by the tiv... adixodvra. In the same way I would write cé, not o«, in 244. 272 ddvruv tév “EAAjvev ra xpypata Tatra, the money for this purpose. Weil would rather like to get rid of raira. The use is much better known in Latin, e.g. Cic. Mur. 68 st t1bt istam (1.e, istius ret) rationem non possim reddere: de Am. 3 cum im eam ipsam mentionem incidisset: Virg. Aen. 2. 171 ea signa: 7. 595 has poenas: 8. 426 his manibus. Cf. however Plat. Rep. 371 E tiv tyuny tavrnv the price of this : Herod. 4. 35 ryv oroddv ravryv: and see Stein’s note on Herod. 2. 39. 7. 295 of vopilovres abrovs agidxpews etvar Tod Bidiwmou Févor kal pido tporayopever Oar. Possibly airovs should be omitted, but cf. note on Phil. 4.49. Itis hardly likely that éévo and pido: are wrong. 297 parreiay dvayvooopar.... A€ye Tas wavTeias. Read pavreéas in both places. 303 GAAG pay Ste Tadr’ ottws exet, adrds ody olds 7” dvreurety Era. Greek idiom surely needs the negative. dvrewmety 8r would not be parallel here to duduoByretv dr, with which it has been compared. 310 ‘kat rots éxydvors’ mpooypawpayres TH eipyvnv having added to the peace the words x.r.X. 190 THE ATTIC ORATORS Most edd. follow Dindorf in inserting cis after the -es of the participle. I would suggest 74 ecipyyy, nothing being commoner than this confusion of terminations. So in 87 aityy might very well be airy. Cf. 56 rporypdiat mpos tiv eipyvyy TO kat Tots exyovots, for mpds is not eis. 336 jay Ady’ ds Kaddv cipyvyn pnd ds oupdéepov: oddels yap airtatal ge Tov moijocacbat ryv ToAw eipyvyv (Blass rot mojoat, omitting 7. 7. ei): dAA ws otk aloypa Kdsoveidioros & ¥ >” 3 ¥ kat TAN’ Vorepov eEnrarypcha x.7.A. Perhaps ds obk aicypav kamoveiSiotov, 1.€. éroujoaro. «A 339 drav pev idnte Sewdry7’ 7 cdpwviav 4 TL TaV ToLOvTeVY ayaav ..., cvyxaipew Kat cuvackeiy méyras Sei. For ovvackeiv read cvvavéev, comparing the note on Ol. 3. 28. Leptines. 15 riyzay seems to me one of the mistakes due to a word that is coming immediately afterwards, in this case rij. The coming word is already in the writer’s mind. There is therefore no need for the displaced word to have resembled it closely. 20 cxepdpeba Sy Ti TovTo TH moAG, eav GmavTes ovTOL Aytovpyaou pavycera, yap oS roAAod Sel ris yeryooperys d£tov aicyvvys. The phrase oi8é woAdod Set is always elsewhere preceded by another negative, which it emphasises: otk ... ov8é modAod Sei. Is it not therefore probable that we should read avycerat yap ovdé woAdod Sei? Cf. Dobree’s very probable dv otdtv éxeivors tpoojxev in 104. 24 as dpa Sewov ef év xowd pev pnd’ Srioiv badpyxe TH TOAE, idta S€ twes wAOUTHTOVC’ areAclas ereAnppéevol. tmdpye. may very well be right ; but the corruption of the future is so common that imdpée as parallel to mAourycovc. is worth suggesting. 93 ouvieh’ dv tpdmov, & avdpes ’AOyvaioy, 6 SdrAwv rods vopous ws KaAGs KeAevet TIErat. DEMOSTHENES 19] Cobet brackets és xadds: I would rather suggest that «ai has been lost before ds, the symbols for these two words being often confused together in MSS. Wolf com- pares Aesch. 1. 20 i cid96’ ofwy vépwv ipiy Kepevor, as kahdv cal cwdpdvwv, «7.4. That passage is however very different in that (1) the sense is complete before we come to the as clause; (2) even the ws caddy could hardly stand without the addition of kai cwdpdvur. 123 wy rote dia pev Tod Tavde KaTyyopeiy ds havo 2 < - a ra r] - Z 2 z. , a ra éxeivous darpod, Ov a 8 ad karadeirew éxeivors Pyoets Tovcd 5 povov AaBdvres Exover TodrT’ ahedp. Sense and symmetry alike call for da ev 7d, not rod. 157 aicypds, & dvdpes "AOnvator, kal kaxds exwv & vopos kal dpoos POovw Twi. aicypas t Merdias. 52 (oracle) orehavagopeiv xarta marpia . . . Bias Seas Kal dpiotepas dvicxovtas, Is it possible that the unexplained idias (iGeias, dcias, dirds have been suggested) stands for Aaids (IAIAS for AAIA3), Aas being a variant, and probably the original, of dpiotepds’? It is of some significance that in Macart. 66 - the same oracle is quoted and idéas does not appear. 75 dxovw yap aitov éywye pid pdvov drA@vat Wydy, Kal TadT’ ovre kAatoavr’ ovre Senbavra trav SikacTaéy ovdevds ovTEe PiAdy- Opwrov ore puxpov ote péya ovd’ driody mpds Tots StkacTas ToinoavTa. It would be idiomatic (cf. eg. 129) and much more elegant to substitute odd¢ for the otre before iAdvOpwmov. ovre five times over, with the fourth and fifth subordinate to the third, is far from good. 78 74.8’ épya dvr’ Hv Kal Ta mpartopeva bd ToUTOU. Read certainly dé for .izé, though I do not find it recorded as a variant or a conjecture. v7d is surely impossible with 7, on which the preposition must depend. 192 THE ATTIC ORATORS 97 Kai pO éopris pire tepav pajre vopov par’ &ddov pydevds mpdvounv ToLovpevov ob KaTandretobe ; ob mapdderypwa ToLCere ; It is justly objected that xarayydiciofe cannot take an accusative of the person. oovyevov must not be thought of, because with od a hiatus would result ; and the remedy is still to seek. It is perhaps not far off. Transpose ob karaynpueiobe and od rapdderyya moijoere, 80 that the accusative will follow the latter verb. The similarity of the two clauses accounts for the error. 125 Insert something like éreftéva: after dpa tis. 153 ei pev éoriv ... 7d Antoupyeiv Toto, 76 Adye ... “Hpeis of AyTroupyodvres, Fuels of mpoeadepovres, Hels of wAoVOLOL éopev”> ef TO 7a ToOLadTa A€yet K.T.A. Weil sees that there is something wrong here, but his remedies are not, I think, right, when he proposes to omit éopev or to write ei pev ody for éopev ei. The real remedy, no actual change of letters, is to write jyels of rAovorod éopev. 157 qycpov cuppopias eyevduyv .., toov Poppiwn xai Avorbeiby «7d, elopépwv od dd vrapyotons oiaias... GXAG K.7.A. Remove the comma before ciogépwy and the real con- struction of icov becomes manifest. 174 For ras rouras jyetro, which Cobet rightly I think condemns, vais zopumrais would be less change than his éxouciro. The dative is often used, ¢.g. in the Hipparchicus. 204 od é6éAes should perhaps be ovd’ é0éA ets. 209 odk dv ciOdws elroey ‘tov S¢ BdoKavov, tov 8 dAcOpov, tovrov 8’ éBpilew, dvamvelty d5é; dv ef tis && Liv, dyamay Set’ (or éez). It is somewhat surprising that scholars have acquiesced in this use of dvarveiv. Reiske indeed sought some other explanation and Wolf conjectured péya aveiv, but most scholars seem content to believe that, though everywhere else it means simply to breathe, or recover breath (cf. dvamvon), here it has the sense of péya dpoveiy and can be DEMOSTHENES 193 used even as a climax after tBpifew.! This is to me incredible and I venture to make the suggestion that follows, though not quite satisfied with it. dvamvety must somehow go with the {jv following, for the two words mean really the same thing. Perhaps therefore Demosthenes wrote something like rotdrov 8’ tBpilew, dvamvelv ov ef tis éa chy, ayarav @&e. The place of dvamrveiy will be due to rhetorical emphasis. But I would not urge this particular conjecture ; I would only maintain that the words can hardly be right as they stand. Demosthenes may have originally written something like dvamvety ég kai Sv, and dvarveivy have got displaced. It does not seem clear that we are to take the accusa- tives rov d¢ BdoKavov, tov 8 dAcOpov, by themselves as exclamatory. Resumed in todrov, they may be the subject of iBpigew. For the dé cf. Aristog. A. 91 éredav obv tis... eiry ‘totrov O& taira moety’: Ar. Hg. 269: Ken. An. 3. 1. 15%? Libanius 14. 20 imitates Demosthenes. 220 cvyxwpycair’ ay totrov ... Kipiov yevérOar Tod Tad’ Garep otTos ee tuGv Exacrov Toijoat ; There is no reason to doubt that the correlative to da7ep is sometimes otros, sometimes 6 aitds. When the actual word is ratra, itis really iudzcit nostrd how it should be written. We have as much right to judge as the man who first accented it, a remark which applies also to the oi or of in 153. Here I think raira would be rather more natural : possibly in 193 too. 4b. dray ets 6 wabay py AGBy Sikyy. Probably . Of. on FL. 12 above. Androtion. 35 «i 88 rd Totro woijoat Trelovs pupiovs To’s ddXovs moAtras Bedriovs elvat mporpepere. rovs ddXovus woAiras has all the appearance of an adscript added to explain zAcéovus 7} puptous. 1 Goodwin, Moods and Tenses, § 787, translates ‘should be thus insulting, and should draw his breath,’ meaning (I suppose) ‘should be allowed to live.’ But with dy ef tis é@ (fv, &yaray %8e: this is clumsy in the extreme. oO 194 THE ATTIC ORATORS nw ~ a - 38 tows dvaByoerat Kal cuvepel TH BovdAy Pidimmos . . . Kai “~ x x tues dAXot, otrep éxet Si’ éavrav elyov pera TodTov TO Bovdev- THpPLoV Kal TOUTWY TOV KaKOV oay aLTLoL. ‘It is agreed’ Wayte remarks ‘that éxei here=rdre, a sense of which I cannot find another undoubted example.’ Rather than accept anything so unlikely I suggest that we read ézei for éxet. otrep... Hoav airio. is the main predi- cation and xai means also. Aristocrates. 26 rhv Boudry dixdlew eypaye kal ody arep, dv GAG, Tabeiv > €LTrEV. S itself has a\@ civar: ad@ wabely cirev is a marginal correction often adopted by the editors. They illustrate the construction by § 53 rot vouov A€yovros ed’ ols efeivat xretvat, etc. Waiving the question of the infinitive, no one seems to have noticed that drep is wrongly used. és can stand in an indirect question, as in 30 & yp racyew elpynxey but domep is (I think) always relative, never in good Attic interrogative. (In late Greek do7rep is thus misused, eg. Arrian Anab. 1. 17. 6 oxorowr7 ... drep émirndedrarov xwpiov.) Here, if right, it would have to be the indirect interrogative: a relative sense, the very things which, would be quite out of place. I would suggest that the wep be transferred to the dv which follows, so as to give avmep or édvrep. 33 AvpaiverOar 8, oat, 1%, we amowav Taira. e éori Tes 76 pey 37 a Avpaives Gar yrdpysov of6’ dre mao py pactry dv, wn Setv, uy TA TOLadTa Tovely Evel. So this passage is usually printed, but the construction of yvwpiov off drt maou is then awkward: was ohne Z Zwerfel allen verstandlich ist says Westermann, comparing . 143 rotro roivw én’ éxeivou pey eb mowoiv ob avveBy hevaxicbeiow tpiv aicyivav dpdAciv, ert rovrov St édv euot melOnabe purdgecde. tovro is variously taken as nominative, explained by aicxivyy opdciy, and as accusative after it on the analogy of rutro Hdouat, etc. In ed qovody the editors seem to find no difficulty, comparing such passages as Ol. 1. 28 imép tov Tohdav Gv eb rowitvres exovor: Lept. 110 tpets xadas motodvres ... Guewvov éxeivov mparrere: Ar. Plut. 863 xadds wolav adréAAvtat. But what can be said of persons cannot therefore always be said of things, and anyone who considers it carefully must see that, whatever the con- struction of rodro, it is very extraordinary for ed zovody to be combined with it. I cannot myself doubt that it is a mistake for eb rowtow, just as in Plato Rep. 434 p it seems to me probable that isv is for iotow. There is no reason why o® zowtow should not precede and be separated from the word (jpiv) it agrees with, as in 163 tov pev yap Kéruy eb moidv dvra y’ éxOpov tpiy Kai movnpov aroxtivvvaw 6 Wvdwv. I take rotro to be governed by cd Tovovcew. 145 woddd yeyvicKovres SpOGs tpets od 81a TéAovs adtois xpHyobe. oiov ti Aéyw; et Tes buds Epouro x.7.A. No precise parallel to ofov ri A€yw ; is quoted, nor are the editors agreed how it should be explained. Perhaps Demosthenes wrote zoidv tt A¢dyw; This use of zoios is very common in Plato, when a man is asked to explain himself (zotdv tu héyets ; etc.). Timocrates. 41 dXos per ydp eotw dracw évavrios Tots ovat (vomols 6 vopos), padurra 8 TovTw. 02 196 THE ATTIC ORATORS Should not 8dos be éAws? There is little or no point in édos, for the idiom mentioned in Rehdantz Index s.v. seems not to apply ; whereas dAws will go with dracw or contrast with padcora. 53 et ris Epo’ Spas wordpous padAov dy eixdrws rorjoatd’ driowv, tots Seopévors roils émurdrrovow, o18 dre Pyoatr’ av tots Seopevois: TO pey yap xpyorav, 76 8 dvdvopwy dvOpdrwv épyov éativ. It is surprising that the editors should adhere to avdvipwy, which is utterly inappropriate, and should not even mention Cobet’s convincing conjecture dvadév, a word that meets us again and again in Demosthenes’ denunciations. 61 ei repi mdvrwy épd, eEwoOyrouos epi ToD yd exuTpSerov dAws byiv efvar Tov vopov elzetv. Surely the second zrep/ is an accidental repetition of the first. é£wOciobai rwvos is Greek; eEwhciofar mepi twos is not. 141 Fyovpevos GBiwrov atta civar tov Biov rodro Twaddvrt. Insert dy before airé or civat, The meaning is that, if he lost his one eye, life would not be worth living. 171 The 460s of the city is not rots pév rodAots apas peraxeipiler Oar, Kodaxeve 8¢ tov def te dvvacGas Soxodvra. 8 av roves, & Tiysdxpares: 8: & woAAG paAdAov ay cixdtws wy Medyjoavres dxotoat cod Odvarov xatayydioawi otro 7 de ’"Avopotiwy’ ddeinoay. As the sentence stands, py with é6edjoavres seems impossible. It could only be conditional, and a conditional sense is inappropriate. In ‘they would condemn you unheard,’ there is no place for a condition nor for It would seem therefore that uy is probably due to the relative & being generic such things as would make them, in which case wy would be right. But we cannot give & this sense without some change or addition, which it is impossible to specify with any certainty, ¢.g. <émirydevwv> or 8 ad. A very slight change, which may I think be right, would be to read od ot rovels... be & «7A. a8 a question, DEMOSTHENES 197 Aristogeiton A. a > ~ 13 od yap qyvdow ot 6 Toijoas TL TOLOdTOY Tap’ tpiv Kat mabiv dmépxerat. Reiske remarks on rafdév subaudi a communt m1, and so Weil. But in this case should we not have to understand to.otrov too? 71 is so like 7 that it sometimes disappears before it, and it may have done so here (1: wawv). Cf 95 By wepipeivavtds Te wabciy. — 16 of 88 vopot 76 Sikatov K.7.X. Blass certainly seems right in demurring to véuo. I would only notice (1) that it is due to vduos coming both before and afterwards and already occupying the writer’s mind; (2) that therefore the original word need not resemble véyo., though very probably it was disyllabic (cf. on Lept.15 above). Blass suggests dourof or dAAou, but moAAot, xpyoroi, and many other words are about equally likely. 25 Saroytodpevos TadO’ arep odros, bru cere K.T.A. io p > Perhaps rav6’ would be better, though not necessary. Cf. on Mezd. 220. 31 pa ydp, & Zed wat Geol, rooavry ordvis avdpdv yevorro TH move ote Tap’ “Aptotoyeirovos TOV Kadav TL roca Oan. momoacGo. is supposed to mean procure or get done, but mouioOa: mapd. twos cannot be Greek. Cobet conjectures mopicacOat, Naber airjcacba. The latter seems to me the better in sense, but airjoacba: could hardly have got so changed. Perhaps for zoujoacOa, i.e. aojoachat, we should read Sejocacfac which comes very near. Cf. the necessary correction of wydicpa ... peravonoa in Aristog. B. 17 to peramroujoa: or perawojoa. (De Cor. 121 vopovus peramrouar). ‘ a 3 , a 57 TYHV TOLAVT EVEPYETYCOATAY AUTOV. rocatr ? The mistake is not uncommon. 198 THE ATTIC ORATORS 66 ris éorw 6 rHv mpos Tovs yoveas etvoray Spay mpodeduxdra TodTov iv mpos Tov Syuov viv éxew briryvetrat misTEvw. There is no occasion for any large change like Cobet’s hv... broxpiverat dAnOA elvar wistedwy, nor does Weil’s tmoyvetrar éxyew for egyew tmoyvetrar seem as likely as Reiske’s simple and easy 7 for jv. But the two present: participles épév and micrevwv are awkward. Keeping the first, as it is, subordinate to the second, I would act on the very common confusion of futures with presents and read morevcwv. Had Demosthenes written épév and morevwy, would he not have used péy and 8¢? 100 ri gdyoere; dpéoxew rotrov ipiv; kat ris & tovro ToApyowy eimeiv; tis 5 THs TodTov wovnpias per’ apas Kai xakys Sdéys kAypovopetv BovAynodpevos ; GAN’ od« abros Exacros éroyndicacbat ; odxody katapdcecbe Tois dreyndirpévors TodTO 76 TirTOv ExacTos Sidovs ds OdK abtés obrds eortu. It is difficult to make sense of dAN’ odk . . . dropndloacOat, which Weil after Reiske prints as a statement (oi« = 0d gyoere you will deny) but prefers to omit, a somewhat desperate remedy when the words are not of an explan- atory nature. There is slight authority for xarayydicacGa, which perhaps may be regarded as a correction. I would suggest the loss of an o¥ (no infrequent event and here the easier because there is already one od in the clause), reading GAN’ odk abrés Exaoros droyndicacba ; will not each of you say that he did not acquit ? or the simple change dAX’ airds Exacros otk érowndicacbat. The otros at the end may seem strange and possibly we should read rodros. But I think it is right. In any case I take the word as referring not to rots drepypiopévors (Reiske with Weil’s concurrence makes it = 6 drepygicpévos, as though only one man had done the acquitting) but to 6 THs ToUTOV Tovypias . . . KAnpovopetv BovAnaopevos above. Aristogeiton B. 4 86 Kat ras tyswpias 6 Bdrwv roils pev ididrais Bpadeias éroinge, tats 8 dpyais kal rois Sypaywyois Taxeias, brodap- Bdvov rots pev evdexerOar cal mapa tov xpdvoy To Sixasov AaBetv, Tots 8 ovk evetvar Teptpetvas. DEMOSTHENES 199 rots wev and rots d€ are certainly difficult. Reiske read rots 8€¢ without touching the equal difficulty of rois wév. Schafer calls them dativt incommodt. Weil’s note leaves it obscure how he really explains them. He says that the phrase is equivalent to zap’ éxeivwv pev... AaBeiv evdéxeoOu and ‘le démonstratif est constrwit avec évdéxerOat, et plus vas avec evetvar, au lieu d’étre construit avec aBeiv et repipetvat, adding zrapa rév xpdvor est singuliérement dit, (Some at least of the earlier editors seem to think zapd means, or should be changed to, perd.) I think it is impossible that the words can mean any- thing but ‘the one set of people may even be helped to justice by lapse of time, while the other cannot afford to wait,’ and I should explain rots pév and rois 8é as referring by a certain brevity and confusion of expression, not to the idaro. and dpxai themselves, but to the persons who have had to do with them respectively and have been injured by them. ois wév = people in the one case, rots 5é people in the other. Such a reference is irregular but not, I think, impossible, though we should look for it usually in a more compressed style, such as that of Aristotle. Anal. Post, i. 24. 86a 12 for instance illustrates it: 6 8 ri kadArov (drddesw) exwv olde Kat 76 Kata pépos’ ovTos & 7d xaOdXov ovk oldev, Where otros means the man who has to do with (in this case the man who knows) 16 xara pépos, and does not refer to 76 xara pépos itself. The meaning of zapa tov xpévoy is that time may bring facts to light, calm angry feelings, and so on. 16 Should év be added to help the construction of mwaporréov? _ I incline to airév . 23 8are ought perhaps to be Sdcere (after rapadvAdéere dus). In Aen. Tact. 16 tva décovor is a blunder for iva 8aou. The Private Speeches. 31. 14 dv pdvoy avOparuv odd rhs érwBedias déiov Fv xevOuvevetv. As a genitive seems not to be found elsewhere with kivéuvedw, the conjecture may be hazarded that a substantive 200 THE ATTIC ORATORS on which it depended has been lost. We find elsewhere xivuvoy xuvSuvevew and «iwdvvevna xwdvvevev, and such a word would easily drop out near the verb, ¢.g. immediately after it. Or epi may be missing. The genitive with gevyev, etc. is not parallel, because there was of course no éruBedias Sixy. In 41. 16 Blass writes xwédvveves tov Wevdouaprupiav for x. tiv Wevdopaprupiav, but there a Sinn would come in. It seems however just possible that the accusative should be read in both places, pevdopaprupiay and érwBcdiav, though the two are not exactly parallel. 34 arg. (ad finem) exe pev <> Exarépov SiacrodAy avepa. The similarity of v and » (N and H) often leads to error. 37. 4 éy toils épyos év Mapwveia? F and T. 53 twds . . . of TO mpayya Téexvyv Terompevor pyre ovyyvepns pyr’ GdAov twés cio GAA’ 7 TOD wHelovos. Here again the genitives seem unaccountable, and some- thing may be missing, ¢.g. pyr’ dAAov twos <#rrovs> ciciv, if cvyyvopys ATTwv could stand. 41. 11 qudAny pev yap AaBovres ... Kat Oevres evexupa. pera xXpvoiwy ov dvevyvdxacr Kexopiopevot tairyy . . ., oKynvyyv © qv éxovaw, ovde yap ravrnv AaBdvres dvadepovow. It would be hard to interpret oxyvyv here; but, when we come to 27 zapa Tod Acwxpdrous exoveay 7a xpuaia Kat Ta imaria thy yuvaix’ ~kaBey and compare 59. 35 dca fv aiTy.. wept To copa iwaria cal ypuaia, we see it to be unnecessary, a8 oxyvyy is an easy error for oxeuyy, to which iudrea directly points. 44. 17 oxapacbe @s moAAoorés eis riv tod 'Apytddov ovyyeveav mporynkuy. The adjective moAAoords seems questionable in con- struction. Could one say zparos or éyy’raros mpoorjxwv 4 Perhaps zo\Aooras, as in Ar. Hth. 10. 5. 1176 a 29 devrépws kat 7oAAoo Tas. DEMOSTHENES 201 45. 59 xafrot doris, & dv8pes "AOnvaior, xaxGv adXdorpiov khémrys trepewev dvopacOqvat, ri ay iyetabe womjoo. TovTov trép avrod ; xaxav é\AoTpiwy kAémrys is unmeaning. Sandys proposed xai tov for xaxGv, and I would modify this to (é£e)xa Trav, which gives the proper antithesis to iaép atrod. In 42 Reiske’s picOwow for picbwors and in 53 Cobet’s ra THs dUoews Sikowa (for oixeta) should surely be adopted. With the latter cf. Plato Gorgias 484 a e&édapwe 76 THs dicews Sixatov, and Hp. 7. 326 a ra qoAurixa Sixasa. 68 éxvijoeév tus dv mpooedbeiv mprov. Should not zparov be zpérepos ? 47. 4 dvaykd£e for dvayxdfor? 48. 7 wept dv obros Aéiov éavT@ efvat. e a €QAUTOU q 53. 1 088 at otrws daopos fv ovd’ dditos dor’ odk av eevpely Tov droypdiyovTa. This is well known as one of two passages in Demo- sthenes, where od is joined with an infinitive after dore not in oratio obliqua. The other passage I have dealt with before (see p. 181). Here I should suggest otk av égytpov or otk ay égeupeiv: 54, 20 ivdaddol rwés éopev Hueis cvverdeypévor, Kai épdvres ods dy qyiv SdEn Taiomev Kal dyyxopev. The connection of épdvres with the verbs seems grotesque. or épavres would be rather poetical. Should it be wepuovres? 56. 10 wvOduevos tas Tyas tras evOdde tod cirov Kabe- oTyKvias. The words 70d ofrov seem impossible where they stand. Move them a little or regard them as an adscript from § 9 or read 76 city. 16 ratra 8 judy AcydvTwy . . . Kat dévovvtww A. rovtovi THy pev ovyypadiy yn kwely..., tov 8& xpnudtwv doa piv adros 202 THE ATTIC ORATORS Gpodoyes dmodotva: yuiv, mept S& Trav dyTieyopévuv as €roipwy ovrwv xpiOivar . . .. otk ey mpowéyew A. TovTwy 2 a ovdevi. Kennedy follows Schafer in taking éro/uwy as neuter and translating it certain, as against Reiske who says it is positum in bivio and = d£ovvrwy jay A. kpiOjvat (td est éav xpiOfvat), &s éroiuwy yuav dvtwy kpiOjva. No doubt Reiske is right in making the word masculine. Omit os as having arisen from the wy preceding, and all difficulty disappears, é. dvTwv being masculine and parallel to déiowvrwv. mpoo éxewv should be zpoceéav. [In Blass’ text xdy just below is, I suppose, a misprint for xdv. | 57. 7 10 yap eis avTd To Tpaypa Tayta déyew Totr’ eywy’ € / a ft ? 397 2 a trodapBdva, doa tis... wérovd’ ddixws edetfaut. I do not see how these two things can be identical. Read rod ydp. 44 Blass is certainly wrong in adopting pydicaibe from Lambinus without any MS. authority and leaving the nominative pydeis «7.4. without construction. What Schifer says is perfectly true, I think, that padeis really stands for tis. tis has become a negative under the influence of the pydevé following it and of the tendency to double and emphasise a negation. e ris pydevi tovTwy épdicBytav .. . Yydicairo is what the orator had in his mind. 59. 105 ereita, TOUS Soxpacbévras: dvaypapyva. év ornAy AGivy kal orjoat ev dxpordde taps TH Ged. orivas! 61. 43 Kairor twés 43y Kal dv cdrvxiav Tpaypdtwv yupva- obevres EBavpdcbyoar. As in 31. 14 and 37. 53 above, so here the genitive seems to have lost the word which governs it, though here possibly the error may be different. A few lines before we have riv pey é« tov mpdgewy eureipiay yeryvonevyv, which ply TPaywaTwv <éuretpia> yuuvacbévres or something similar DEMOSTHENES 203 54 80 @ dei ce Trav éraivun Gétov civar Sdgavta Képe TIS OFS girias dveririunrov motelv. Again a dubious genitive. Any real parallel can only be found in poetry, eg. dOixros pyntijpos. <évexa> tis offs girias averttipyrtoy 4 The Prooemia. 2. 3 76 dé pnd? Sriody peradapBavew tov Shuov GAA Tods avrimpatrovras Tepteivat K.T-d. The drift of the whole passage seems to require some such word as dei or rdvrws with zepietvat. 26. 3 dOgovs tods kwduvous Toijoovew adrois. Such a use of d@gos is unparalleled. Should we read GOwous Tod Kwduvov mojcovaw avrovs ? 29. 3 rotro 5y, Totro. Reiske was practically right in rodro 8) raird, but the regular order is rairé 87 Totro. 32. 2 ob8' émi 7G Tos exOpods py SuvnceorOar Oappeiv arr’ ext TO Kav SUvwvTat kpatyoev, and 3 éyw 8 ovx amorpepouat Néyewy & Soxet pot, Kairep 6pGv Hypévous buas. The absolute use of dvvacOa: and still more that of jypevos are strange. I conjectnre something like <¢avAws> aymevous (as in Or. 18. 15 dray tyets, & dvdpes ’APnvaior, datrdws ypévor «.7.d.) OF , and rots exOpois <émiévar> py Suv~oec Oa or more simply pydev Suvycec Oat, in which dev would easily slip out before duv. 33. 2 cat yap as Sixadtaro. trav “EAAyvev éore wdAN’ eimeciv kat édpwy kal 6pG, kal ds dpictwv mpoydvwy, Kat 7oAAG To.avTa. Something like 7éAN’ ciety, or TOAN <évovT'> eireiy would seem more likely. 34. 1 wddw tadr’ cis tHv érépav exxAnoiay otro. AaPBdvTes TOUTWY KATHYOpyToveL. haBédvres, which Kennedy translates (with tatra) take the same course, can hardly be right. I would suggest 204 THE ATTIC ORATORS avaBaddvres having deferred. To avoid hiatus this should be put before ofra., and then we see that the ava may have been lost after the av of éxxAnoiav. 39. 3 BovdryOervrov ipav kal rapohwbevtav Td -yeyernpevo. There is nothing in the context that can be supplied with BovAyféevrwv. Has not an infinitive been lost? 53. 4 kal yeAdoa.. . perddwxav ipivt 55. 1 év ovdéror’ edruyjoat totro voile. Should we not write oiSerirore? In the best prose ovderore is usually, if not always, future or present. Comedy is freer. Letters. 1. 3 dorw pev obv épyov é& émorodys eupetvar oupPovry Toddois yap eiwOar’ dravrav dpeis mpd TOU weptretvar poabetv. éupeivat cvpBovdy can hardly be right, the sense needed here being only giving advice. It has arisen, I think, from the wepizetvor following in the next sentence, which would have prevented the writer from using éupeivar here, even if it were suitable. The true word need not have resembled éupetvar, and some other case of ovpPovdy may have followed. 2. 7 80 Gpidrdas weioa wpocéyew aire tov vodv ds BovAotto. mpocéxew has no distinct subject, and on the other hand és Bovdorro is otiose and weak. Read therefore ods for ais. ANTIPHON. The question of the authenticity of the tetralogies that bear Antiphon’s name is not an easy one. It should be distinguished from the question of their date; for, though if late in origin they cannot be his, they may be of his date and yet not of his writing. We are not bound, if we think them early, therefore to pronounce them authentic, nor, if we think them unauthentic, therefore necessarily to consider them late. Herwerden has impugned authorship and date at once (Mnemos. N.S. 9). Dittenberger (Hermes 17) equally disputes Antiphon’s authorship, but he holds ANTIPHON 205 them to have been written just after or during the Pelopon- nesian War by an Ionian who atticised without complete command of Attic idiom. Blass (Att. Bereds.2 Vols. 1 and 3) still inclines to think them authentic, as does Cucuel in his Hssat on Antiphon. The offences against pure Atticism adduced by Herwerden = pre prcee. (2. 1. 2), xatadoxcioOar (2. 2. 2 and 3, and 3. 7), dvaycyveoxw in the sense of persuade (2. 2. 7), ee es (2. 3. 1), eixdrepoy (2. 2. 3 and 8: ab. 3.5 and 4.4: 4. 3. 2): daedoyyOnv 2. 3. 1: “o 4 3; 3. 3.2: 4.3.1): ‘paradopBdva Jind guilty, convict (often in all three tetralogies, e.g. 2.4.4 and 11), It will be noticed that the first four of these expressions occur in T.11 only, the fifth four times in the first and once in the third, the other two in all three, and the last of them repeatedly. ’Avaytyvioxw persuade, xatradox® (not the passive), and metpaa with participle are familiar in Herodotus. The last, though not the former two, does just occur elsewhere in Attic (Plato Theaet. 190 E: cf. dpyopas with part. 1b. 1874 and a few other places of Plato and Xenophon, and vépuce dwoxre(vwy Xen, An. 6. 6. 24: also did@ and pod in Aristophanes Vesp. 1535: Plut. 645: Eccl. 502: dpvodpo: and doBodua occasionally ; probably a few other parallels). xatadapBavw find guilty is not cited from any other author. The aorist form dweAoyyOyv occurs in a passage of Xen. (Hell. 1. 4. 13) which is indeed doubtful, but Xen. is just the Attic author who, if anyone, might use it. It occurs again in a Middle Comedy fragment and, though rare itself, has of course many Attic parallels in dievoyOnv, éropevOnv, eSvv7iOnv, etc. Herodotus has other such forms not found in Attic. In the undisputed Antiphon speeches however the ordinary aorist deAoyyoduyy occurs five times. ére€epyduevor is almost unparalleled, but Xen. An. 2. 4, 25 ventures upon zapepxopevous (cf. Rutherford, New Phrynichus, p. 109: in Lys. fragm. 47 (65) épyopevwv must be a misquotation of the scholiast’s own). The conjecture ot émefepyducba is by no means convincing in 1 In the usual order of Antiphon’s speeches the T.s (tetralogies) are 2, 8,4, Thus 2. 2.2 means the second section of the second speech of the first T. 5. 2 is second section of speech 5, which is not a tetralogy. 206 THE ATTIC ORATORS view of the other irregularities, especially frequent in T. 1. Finally cixérepoy is a surprising form, since Greek, unlike Latin, hardly ever compares participles and éppwpévos has perhaps no companion in this respect. Van Cleef’s Indea Antiphonteus (Cornell, 1895) makes it much easier to study the language of all the speeches ascribed to Antiphon. A few notes may be contributed here in answer to three questions that suggest themselves. A. In the vocabulary of the T.s as compared with that of 1, 5, and 6 are there any other noticeable things ? The following words may be remarked as occurring in the T.’s and not in 1,5, or 6. Among them are a few pointed out by Dittenberger, Cucuel, or Both (de Antiphontis et Thucy- didis genere dicendt) that I had not myself noticed. It will be understood that in this and the next list I have not put down a word only because it happens to be rare. Rare words are not necessarily noticeable. dyxiora 2. 1. 6. Stayvepor 3. 3. 3. adoXos 3. 3. 4. dpa often, esp. 3 and 4. dAtyptos (five times in 4). dvopevea 4. 1, 3. avayvos 2. 1. 10. Svcpopos 3. 2. 11. avaxAatw 2. 4. 1. éyyiora 2. 1. 6. dvaxpivw question 2. 1. 9: é6édw, GéeAw= BovAopar 3. 4. 2. 3. 2. 3 and 5: 4. 2.1 and 7: dvarroxpitus 3. 3. 2. ab. 3. 1. dvaxwpel (cis Huas y mown, Tro ex =twd 2. 4.1: cf. 3. 3. 10 piacpa, 2. 1. 3). (e dpoir). dmekws 2. 2. 5. éXaccdvus 4. 4. 6. drrodvaipos 4. 4. 9. €dadpos 3. 3. 12. dpxeiy 2. 2. 2: 0b. 3.3: 4b. eXeyeryp or -rijs 2. 4. 3. 4.10: 4. 3. 6. éumvous 2.1.9: 2b, 3. 2. donpos? 2. 4. 8. éudpuy 2. 3. 2. domaipw 2. 4. 5. evOvpios 2. 3. 10: 3. 1. 2: drpepitw 2. 4. 9; 3. 2. 5: 1b. 4, 9 twice. ab. 4. 4 and 5. ebreras 3. 4. 7. dwpi 2. 1, 4: 4b. 2. 5. and = edoeBodpar pass. 3. 3. 11. 4. 5. Hrug 3. 2. 3. dwpos 3. 1. 2: 4b. 2. 12. Oavacipws 4. 3. 4. yepatds, yypaids 3. 2. 11: Oepyds of a person 2. 4. 5: 1. 2 and 6. ef. 1b. 1. 7. ANTIPHON 207 Ovpovpevov, 7d 2. 3. 3. mpaxtwp doer 3. 2. 6. knAds 3. 3. 8 and 11. mpooBory 3. 3. 8. kpecodvus 4. 4. 6 and 7. mpootpéraios 2. 3. 10: 3. 4. péroxos 3. 3. 11: 4b. 4. 6. 9:4. 1. 4: 7b. 2.8 and pyvipa 4.2.8: 2b. 3.7 and 4. 10. 4.10% arporepov 7 With subj. 2. 1. 2. puatvw 2.1.3 and 10: 76.2. céBopar 2. 4. 12. Ll 3. 1. 2. ovdAAjmTwp 3. 3. 10. papia 2. 3.9 and 11: 3. 3. cupmpdetup 3. 4. 6. 12. ow 2.1.4: 3. 4. 6. vymuos 3. 2. 11. ve*+ re Specially common. oldapev 2. 1. 3. tpguw 2. 3.8: 16. 4. 9. évivnpe act. 2. 2. 1. trairuos 2. 2. 6: 3. 3. 11: é« maAatod 2. 1. 5. 4,1. 4. Tmapadéepw vopov 3. 4. 8. yiahiid ale 4, 3. 4, mevOos 4. 4. 1. ppovris 2, 2. 2. mepidvw 2. 2. 5. piror (v. 1. yévos) 4. 1. 2. mown 2.1.3: 2b. 4. 11. Yatvw 3. 3. 5. Many of these words hardly call for special comment. They belong to the large class which we find seldom or never in ordinary orators, but sometimes or even frequently in Thucydides and Xenophon, Plato and Aristotle, and often in the Ionic of Herodotus. In Antiphon (see below) we should quite expect to find some such : the only question is as to their number. Neither dyyiora nor any word connected with dyy belongs to standard or even to unusual Attic prose, and in the same way éyyiora for éyyvrara is only cited from a document in the De Corona. dpxeiv is less used in Attic than égapxeiy: here we observe that it occurs (four times) in T.s only, while ééapxeiy occurs (twice) in the other speeches. We note the religious or semi-religious terms dAtryptos, mpoo- Tpdmrauos, KnAis, puaive, mpoo Body, to which may be added mown and évOvpios. yypaios tehevTn death in old age is a decidedly poetical phrase, nor is ynpatds itself, though Xenophontean, a word of ordinary Attic. It is very noticeable that Spav, which is not really a common Attic word, abounds in these speeches, though it is completely absent from the others. é$édw is used=BovAopar (of a decided wish) several times, but in 1, 5, and 6 only in its 208 THE ATTIC ORATORS common sense of am willing, for in 1. 8-12 this is all that it appears to mean. In 3. 4. 3 and 5; 4. 3.1 notice the form @elw. av éé euod mpaxdévrwv seems the only instance of this use of éx, which is however Thucydidean, Xenophontean, and Platonic as well as Herodotean. €lacodvws and xpacadyws have been noticed as quite unusual adverbial forms: Thucydides however has jpeLéves twice, Xenophon once, Isocrates three times : Plato pesLovus often, caAdidvws and BeArivws. Not édradpds but Kxoddos, and not 9A but FAucudrys, are the usual prose words. vymios iS rare in any prose. oidapev for iopey is the reading of the four best MSS. in Xen. An. 2. 4. 6 and occurs also in Alcid. ii. 141%. Herodotus has it a few times. zpdrepov 7 with subjunctive is found Thue. 7. 63. 1 and in Herodotus. ovv used in the poetical and Xenophontean way is quite noticeable. ddov (on which yévos is no doubt a gloss) is rare in prose, and yadtw extremely so. It will be observed that T. 1 has a decidedly longer and more remarkable list of peculiarities than either of the others. B. What noticeable things, if any, occur in the other speeches only # &Bovdos 1. 23. eipétus 5. 761 airiacts 5. 25 and 89=6.6. HKovtTa, Ta eis TOUS KuWddvOUS dxeords 5. 91. 5. 81. dxdeds 1. 21. xreivw 5. 11: 6. 4. déat (aor. inf. dyw) 5. 464 xuBepva metaph. 1. 13. drapvos 1. 9 and 10. AvoiteA® 5. 10. aredeyxw 5.19. 21, and 36. pyxavnuo 5. 16 and 22. éroduxdlo 6. 47. pnxavapa 1. 3 and 28: 5. dropolo 5. 41. 25 and 55: 6. 9, 36, and drrogevéis 5. 66. 38, arédeotos 1. 22. véuw 5. 10 and 66. BAdBos 5. 91. vikOpat with gen. 5. 87. yvopiorrys 5. 94. 6de 6 dvyp = eya 6. 9. Sogaorys 5. 94. ot (dative) 1.16: 5. 93, and eipappery 1. 21. odgiot 1.13: 6. 35. éxixeipa 1. 20. ér7np 5. 27. épevva 5. 29. 6pOovpevos 5. 7. ANTIPHON 209 mavoupy@ 5. 65 twice. TotTo pey..., TOTO O€.. 4, mapa THs Bacdvov eidévar 1. 6. several times in | and 5. éx mpoBovdjjs 1. 3 and 5. trobjny 1. 17. mpotvonodpny 5. 43. ppoddos 5. 29. dmé mpdrys 5. 56. xetpovpy 1, 20. cagpyve 1. 13. xopopird 5. 78. ovppépoucn agree 5. 42 and = as=dore 1. 28: 5. 63. 26. Some of these, like some in the former list, call for no further remark. dzodicdlw, daredéyyw, and especially érapvos (Herodotus) are unusual words. aa for dyayety is strange, and the parallel of apoojgav (Thuc. 2. 97. 3) uncertain, partly because the aorist seems not to be the tense there wanted : here d&a: and zapévrt hardly harmonise, and déa: may need emendation (édéyfar, ééerdoa, etc. are suggested). It reminds us however of daedoy7Onv etc. above. poivoyodyny for rpovvoyOyv is unusual, but occurs in Ar. Hq. 421 and in Euripides. 7a eis rovs xvdvvous HKOVTA = Ta Tept Tovs kivduvous is quite an unusual phrase, and 6de 6 dvyp=eéyd belongs rather to tragic Greek, as does also dpotdos. mapd with genitive of a thing or action is against ordinary Attic usage ; cf. Plato Rep. 461 © BeBaw- cacao. rapa Tod Adyov. as=dore is Xenophontean and poetical ; little known otherwise, except in Herodotus, who has it often. In the papyrus fragment published by Nicole and attributed by him on other grounds to Antiphon two ex- pressions at least. may be noticed which quite agree with Antiphon’s characteristics: (1) dékyv émippéroveay, punish- ment coming on me, érippérw being a poetical word, and (2) ides tod A€yew, a highly remarkable construction for prose and not cited except from Homer, where it is common. auvéypadov Sixas wrote forensic speeches is a noticeable phrase. C. What noticeable things, if any, are common to the T.s with the other speeches ? I have observed but few. It is perhaps notas well known as it should be that Attic orators are chary of using such parts of Adyw as éAefa, AexOels, AcxOjoera, etc., much preferring clzov, pyOeis, pyOyoerat. Thus Demosthenes uses P 210 THE ATTIC ORATORS 7a dexévra twice at most and no part of eAega at all, Lysias Aé~ayse and A6cxbévra perhaps ‘once apiece. © Antiphon’s speeches 1 and 6 have édeéa or @dcée three times and NexOérra once: T. 2 has édctay, Adéas, and AexPjoerat. Sox think occurs both 5. 65 and 2. 4.5: 3.2.3: this again is a use almost unknown in orators. The curiously frequent use of xadfornus with a predicative word, make, render so and so, with a similar use of xa@éoryxa, xatéoryy, is found in all the speeches except 1. {yr with an infinitive is remark- ably frequent and found everywhere but in 3. Neither of these uses is noticeable in itself, but their frequency is. éxovoiws and dxovoiws, which are by no means common, because as a rule éxwv and dxwy take their place, occur in the T.s several times and in 1. dOéui(c)ros, aibévrys, éyvetw, tywpo (active) and tiuwpds perhaps deserve mention. Final @s occurs in 6. 15 and probably ‘n 3. 3.6. In all the speeches, T.s and others alike, there is a marked tendency to use new or at least) unusual compounds with privative d, ¢.g. dvayvos, doAos, dveAenpdvus Kat dvoixtiotus. Speeches 1 and 5 as well as the T.s use oo, not tr; 6 appears to have rr. But all six have ov, not éw, in compounds, and eis, not és; in these two points differing from Thucydides. Do these lists—which no doubt other scholars might extend—help us at all? We cannot, I think, attach very much importance to the fact that both sets of speeches have in them some unusual things, nor even to some of these being identical or very similar in the two cases. Assuming that the two sets were composed roughly about the same time, it would be only natural that they should present similar features. It can hardly be said that the identity or similarity is very decidedly marked, and no doubt the greater accumulation of abnormal things in the T.s is a difficulty. It is however only in 2 that they occur very markedly. We might perhaps say that in this respect 2 stands to 3 and 4 as 2, 3, 4 together stand to 1, 5, 6. Even in 1 only érefepyduevor and eixdrepov need cause us any great searchings of heart. It may indeed be asked why we should boggle very much over émefepxdpevor itself, | when we find zapepxépevos in Xenophon, and generally why ANTIPHON 211 we should not accept some irregularities from a writer older than Thucydides, who seems undoubtedly to have influenced Thucydides’ own style, when in Thucydides himself and still more in Xenophon we find a good many things that Lysias, Isocrates, and Demosthenes never allow themselves. To this it must be answered that besides any influence or tendency that might affect Antiphon as it affected Thucydides and Xenophon there was a reason for peculiarity of diction in the two latter that did not, as far as we know, apply to the former. They spent much of their time, Xenophon in fact most of his life, abroad, and their speech might naturally be coloured by this. We have no reason to think that Antiphon ever lived away from Athens, though there is probably nothing to prove that he did not. Anyhow three things may fairly be said. (1) The recognised speeches do display some inclination to the sort of vocabulary which the T.s employ in a more marked manner. (2) The general style as distinguished from the precise vocabulary, a style akin to that of Thucydides, is certainly similar in the two cases and again more marked in the T.s. (3) The T.s are rhe- torical exercises, not genuine speeches like the others, and this may very well account for some of the difference. In works composed to order for actual delivery in ordinary Athenian courts, and presumably adapted in some degree to the persons who were to deliver them, peculiarities might naturally be fewer than in specimens of oratory written as models or suggestions with a sort of professional purpose and used educationally we do not know quite how. We cannot call them exactly epideictic, but their very brevity and compression show that they are not altogether matter of fact and business-like. Remembering then that these speeches are probably as early in date as anything we have in Attic prose except the old Respublica Athenienstum, or even earlier — for there is really no reason to say with Herwerden that they belong to late times: apart from a few things the Greek is both excellent and unmistakably of an early stamp—and remembering how little we know either of the conditions under which at that date they might be written or of the individual tastes and circumstances of Antiphon, we ought P 2 212 THE ATTIC ORATORS perhaps to regard them with some doubt, but by no means at present to condemn them. 1. 1 dvayxaiws eye. éfe. would agree better with érégetus preceding, and the mistake is very common. 3 Séouot 8 tudv ..”. gol pove drodcerpérva BonOjoas. iets ydp pou dvayxator. ods yap éxpiy x.7.A. The imaginative logic requires, I think, dvayxaio. or évayxaio. It requires, that is, that he represent them not only as close friends, but as the only close friends he has. So in 5. 32 Blass corrects aéroé to airot. Cf. on 11 below. 6 év ols pev yap aird e€ovoia Fv cadas cidevat, Tapa THs f, > > Z 2 2 2 > 2 , aA? a § Bacdvov, otk AO2noe év ots § odk Fv wvbécbat, rodr’ airs mpovbuuyOn Kairor adrd Todro éxpiv, 6 Kal éya mpoiKadovpyy, a g 4 a ae f 2 a: mpobupyPyva., Grws 76 mpaypa 7 adnOés, eregedOeiv. The last six words are obscure and awkward. Blass mentions the conjectures drws 76 zpaypa jv Kal TO dAnOés, éreLehGetv and dus 7d mpayya Hv GAnOds, éxeedOcitv. It is strange that anyone should have come so near the truth as the latter suggestion and not have seen the truth itself, dws To mpaypa hv dAnOas éreéedAOeiv, ‘so that it might have been possible really to investigate what had taken place,’ dws being used idiomatically with the indicative (asin Ar. Peace 135 éypiv oe Tnydoou Cedfar mrepov, Grrws épatvou K.T.A. Plat. ep. 378 A. @yunv Sev... dxodvew 8? daroppyruv, dros 6rt éXaxiorots oweBy | dxotoa.: Dem. 36. 20, etc.) and qv with infinitive like oi« jv just before. a > 8 ra yap yevdpueva ev TovTw ddavicbjva anOycav. As this refers to the future, an dv must be added to the infinitive. 10 It may deserve consideration whether here and in 11 Tov govéa should not be THv povéa, asin 3 dmodeigw oe TY ToUTwY parépa govea otcav. In 20 (rév éavrav dovea pera- xerprfdpevot) 6 govers is of course the poisoned bowl, as in Ajax 815 6 adayeds is the sword, and goveds tb. 1026. ae ANTIPHON 213 11 Perhaps <6é> rovrovs airots xeAevwy to match 6 6édwv in the parallel clause. 6 would easily fall out after «. So in 3.1. 2 épot 8& obk éAdoow Tod ExdvTos dkwy THY TUUpopay Katéstnoe we ought perhaps to read <6é> dxwy or dxwy. Cf. on 3 above. 23 ay buds welby. meton? Cf. 5. 95 day pdvov... reicwow. But the present tense is not necessarily wrong. 2. 1. 1 éréca piv rav mpaypdrov bird roy émirvydvTwv 2 , > a a) XN > ie a Xx. émtBovdrcverat, ob xaAcrd eAdyxecOai got av O of ikavas pev mepukores, eurreipor Se mpaypdtwv dvres, .. . tpdoowot, xaderor kal yvwoOyvat Kai edeyxXPjvai cio. The use of rpécowo. without an object is awkward and has suggested the improbable conjecture zavoupyéou. Should we not for dy at the beginning of the clause read av or & dv? It forms at the same time a more direct antithesis to émdca «7.4. There is no difficulty about the half- understood antecedent. 2. 2. 4 obde yap mixpdrepov Tis dvayKys éotxey elvat. Dobree’s icyupdtepov is likely enough. Cf. 6. 25. Possibly however xupedrepov. The probable mistake pointed out below (p. 323) suggests mpdrepov. 2. 3. 9 Adyomer 8 tyiv <éri> ei k.7.d. 2 2. 4. 10 ea 8 rev eixdrwv rpoomotodpevol we eXéyxewv,.. . govea pé pact rod avdpds elvarr 7a Se cixdta GAAa pds emod paddXov arodddekrat dvTa. Matzner’s explanation of d\Aa is unsatisfactory, alia esse 8. alam vim habere ... atque a meis potius partibus stare evict. Schéll dzavra. Should we not read aira? The pretence (zpoomo.ovpevor) is contrasted with the real eixédta. moAAd seems also possible. 3. 2. 1 Axiora Torodros dy Kat Bovddpevos civat. Perhaps 7 for xa‘. The confusion is not uncommon. 214 THE ATTIC ORATORS 3. 2. 3 cupBéBynxe Sé ror woAd mapa yvapyv TovTwy (neuter). Cf. Plat. Rep. 4904 év pev rodro opddpa otrw rapa Sdfay tots viv Soxoupevois Tept airov. 5. 16 xaxodpyos drépuyov GAN’ od Tod povov tH Sixyy. There seems no parallel for such a phrase as xaxotpyos drépvyov, I was acquitted on the charge of xaxovpyia. Reiske suggested xaxovpyias. Perhaps a word has been lost and we should read something like xaxotpyos <évdedery- pévos>. So in 9 xaxotpyos évdederypevos hovov Oikyy pevyw. 21 Gre py rpovola padAov éyevero } TUXD. For the impossible 7 Blass adopts Jebb’s ot rq éug. I much prefer the simple o?, which Matzner suggested. But he failed to point out the probable origin of the error, namely the fact that in late Greek nothing is commoner than py following a dr in this ordinary sense. How the use, or abuse, sprang up, we hardly know, but there is no doubt of it. Here 77 éuy would rather spoil the contrast of zpdvoww and roxy. 46 6 8é pddiora évOupeiobe .. . py ovv eéeAnrat TodTo tyov pdeis. évOupeiobe is the imperative. Can it really be attached to the relative pronoun, when that relative refers to some- thing not preceding but still to come? We should certainly expect def éevOvyetrGar, which has some MS. support, but may of course very well be conjectural. Blass compares the imperative with a relative in 1. 22 6 xal zromjoare, but there the relative is retrospective, which makes all the difference. Possibly for 35 8€ we should read réée Se. There is no objection to the repetition of the syllable de. Cf. Dem. 24. 5 rovde 5é: Plat. Rep. 449 B rdde 5é: Meno 700 evOdde de: etc. Or é 8¢4 We might indeed think of pddior’ dv évOupoiobe, but that seems less probable. 48 wodAAG ay Sixatdrepov tpeis expiverbe 7H éyh viv devyw tf’ tyudv ddikuws. ddixws makes nonsense of the sentence and Cobet would simply omit it. But why should anyone have put in what ANTIPHON 215 is much worse than a mere superfluity? We might reduce the last words to such a superfluity—superfluous they would still be—by reading devywy for devyw. Perhaps it should be ey 6 viv devywr. 62 > f 8e eS. € n Ss ¢€ = + a y Lf amecrépe 5 atrov tepwv kat doiwy Kai THY GANwy azrep f péyiota Kal wept wAclorov éoriv avOpaross. iepav kat 6oiwy are of course things sacred and profane, and, since everything must be one or the other of these, what is left to be comprised in ray d\Awv ? Could you divide Englishmen into clergy, laity, and-the rest, or the human race into male, female, and the rest? It is not d\Awv, I think, that is wrong, though one might have expected somebody to suggest dadvrwy, like daayra for dAAa in 2.4.10. I would rather suggest that éo‘wv is a mistake for Ovowsv. Cf. Antiphon himself 6. 4 eipyecOar roAcws iepav dydvev O@vowv (where Odvowy ought probably to come next to tepdv) and Xen. Hell. 2. 4. 20 perecynxapev tpiv cat iepdy Tév cemvoratwv Kal Ovotdv Kal éoprdv. In 6. 4 Dobree actually wished to read éoiwv for dvoidv, but the change we need is in the present passage and the other way. The MSS. of Isocrates vary at 7. 66 between trois datos and tats Ovoias. 66 28 , Se vy 9 9 nm as EL JLT] WPOOHKEL LOL PLNOEV WOT GTOKTELVAL AUTOV. Logically it should be a past tense and probably Antiphon wrote mpoo7jKe. 6. 21 rca dre tov pev vdpov ov Sikarov od mpoxabiorain Biroxpdrys xarnyopav kal diaBddrAwv cis 76 Sixaarypiov, weAAGv- tov ececOai po aydvey... dvrep evexa Tovs Adyous TovToUS Aeyou: & pevroe aitidro Kat SiaBadrdo1, padiws eLeAeyxOjororro , pevddpevos. No one can be said to have made much of the obviously corrupt words at the beginning of this sentence. The prevailing idea has been to alter at any rate mpoxafiorain, and all sorts of words (zpocijxov qotet, tpoayopevor, tpopaci- erat) have been proposed. It isso peculiar a word that in its case corruption is most unlikely. It should be recognised as a probably technical expression and the error sought 216 THE ATTIC ORATORS elsewhere. Those critics have been on a better track who removed the od before it, an obvious course to take, but this is not enough. What is meant by zpoxafiordvat vopov ? and what law is referred to? The sense still remains very doubtful and obscure. As soon as we begin to consider the sentence and look for something in it that may easily be wrong, we remember that véuos and Adyos sometimes get interchanged in MSS. and we find that rév pév Adyov here will go far to restore an excellent sense. dv Adyov mpoxabiordvat suits admirably the remainder of the sentence, the peAdvrav goecOai por dydévev and the dvrep evexa Tovs Adyous Totvrous Aéyot, which latter is almost rév Adyov wpoxaficrdvat repeated. The Adyos in question is supposed to be advanced beforehand with a view not to the occasion on which it is put forward but to one soon coming. It paves the way for something subsequent, prejudicing the audience in advance. The pév and 6¢€ mean that such a Adyos would be unfair, even if it were true, but that as a matter of fact it was not even that. Having got thus far a priori, we are gratified to find that the very phrase we have conjectured actually occurs in the Ars Rhetorica which goes under the name of Dionysius of Halicarnassus. In 5. 2 there the writer is speaking of the art of constructing conciliatory exordia so as to win over beforehand the goodwill of your hearer, in this case an individual ; and when you have done that, he says, otrw 5? rpoxaractryodpevov Tov Adyov, éfs iréov emi 7d éyxaptov tot Baciiews, and so on. The two uses tally almost exactly. As for the double negative in od Sikatov of mpoxabiorain, it may arise from a confusion or mixture of od 8. 7. and 8. od ., either of which would be possible here. od Sixarov seems the better. 40 airés would perhaps be better than ofros. ovrws airés has already been proposed, but there is no need for both words. 41 a&...Karyyopoto..., rovro Sé (or 84).., éora TEKMTpLOV. Probably ¢ for 4. ANDOCIDES 217 ANDOCIDES. 1.17 6 8& weloas wat Sedpevos peivar tov marépa eyo fv podicra. Omit xai, Seéuevos being subordinate to weioas. So in 19 6 8& mefcas... éyd woAAG tkerevoas Kal AapBavopevos TV yovarwv. Cf. 106 vinjoavres paxdpevor tovs Tupdvvous : 4. 36 rocaurdkis dywvitduevos viknoas. meioas and Seduevos may have exchanged places, but that is less likely. Seduevos is the present tense just because it is subordinate: cf. p. 68 above. ab. 21 drov av euedAev KT. dy is usually bracketed. Perhaps 54. So perhaps in 81 ws 57, not éws ad, should be read for éws av. 4b, 119 ovre ypyjpara erepa ovr’ cituxiav dvdpds édéc Oar. <érépov> dvdpss has been conjectured. Read rather érépov for érepa. For the order of words see R. 8. Conway in Class. Rev. xiv. 358-9. 3. 4 cipyun éyévero tpds AaxeSatpovious éry mevTiKovTa. Kat evepelvapev .. . Tals omoveais ern TpioKaidena. ib. 6 obroe Huiv eipyvyv éxoinoav wpds Aaxedatpovious ery TpidKovTa. The construction of éry in these passages is remarkable. Reiske added eis and I proposed at one time to insert é7i. Cf. however the treaty in Thuc. 5.47 oovdas éroujoavro éxatév ‘AOnvatoe éry and perhaps Plat. Symp. 201 p "AOnvaio .. . Séxa ery dvaBodjy éxoince THs vooou, Editors do not call attention to the accusative as they should. 1b. 9 rds Te GAAas dzrotkias Kab? Exacrov Supyetr Oa. Kad’ éxdaorny ? 2b. 22 ’Apyetous 8¢ d&yaydvres cis THv rote Gidiay, Rather rH rére. 218 THE ATTIC ORATORS tb, 24 ri oty éorw Srddourov wept drov Sei BovdeverOar ; vai: wept dv ad Ads ’Apyeioe poxaAodvrat. vai is no proper answer to ri ov éoriw trddourov ; Read ri ovv ; dorw taddourov mept Grou K.7.A. ; 80 in 1. 86 dpa ye éorw 3 a oy ac 3 a“ cs g évravOot & tL meptedcimero; and 89 évravOot éorw 6 7 brodefrerat ; [But in Blass’ third edition of the text vai is absent altogether. ] ab. 34 Read wparre for rpdrrou. 4. 15 ci xph wpoodoxay rotrov . . . Siamparrea Oat ; Stampdfeoba, or add dy. 4b. 20 can, I fear, hardly be restored, but it is fairly clear that dé after xeXevovros should be ydp. Lystas. Unless by any happy chance new evidence should become available, it seems unlikely that any great advance can be made now in dealing with the many and great difficulties presented by the text of Lysias. They offer abundant scope for conjecture, and conjectures have been abundantly made, but these are bound often to depart so far from the MS. readings that no certainty or even great probability can attach to them. The few suggestions here set down hardly touch any of the more serious difficulties. 1. 14 éfacke tov Avyvoy drocBecOnvar...cira ex Tév yerdvay évdwacBat. Cobet after Stephanus dvayacOa. Should we not read évaicacbar? The middle is not quoted in L. and 8. or Veitch from any of the best Attic writers, but it is certainly used later and there can be no objection to it in the sense here needed of got a light. 2.2 dore kad ev woAAG Tots mpoTéepois .. . eipyoOar, TOAAG 8€ kal éxeivors mapadedrcipOat. TOAAG pev KaAd..., TOAAG dé... ? LYSIAS 219 23 The Persians made careful calculation («i pév. . . ei 82) of various risks, of pév toivuy raira Sievooivro. of 8 Huérepo. mpdyovo. ov Aoyiop@ eiddres Tois ev TE Torcuw xwduvous . .. ovk epoByOncav 76 TAOS TeV evavTinv. Thalheim gives as the MS. readings ‘ciddres V X, ddvres F’ and cites a number of emendations, none satis- factory. The MS. readings and the sense required seem to point very clearly to Aoy:opa SueAdvres Cf. Timaeus 72 a mavra, Aoyiopga SieAéoOar: Phaedrus 249B & Aocpope ovvatpovpevoy : etc. Of. also on 21. 24 below. 25 The words éoryoav pv... ywpas seem clearly to have suffered some disturbance of their internal order, but any restoration could only be doubtful. 3 s 2 , € a a 5 , aD 4 THV EVTVXLAV OMLOVOOVTGY TOLS KLVOVYOLS EKTYTAVTO. ‘dporAoyotcay Rs (7.e. Reiske), sed est: per pericula ad felicitatem et concordiam pervenerunt’ Thalheim. Surely edrvxia duovootca cannot=eitxia Kal dudvoia. It may however be a question whether épofav otoav would not be better than dpoAoyotcay. Their success did not agree with their danger, but it was as great as their danger. Cf. of Gpotot, Gpotov dvramobiddvat, etc. 52 dravricavres avrot eis Th Meyapuiy evikwy paxdpevor dmacay THY Sivapey THY éxeivov Tois 4On ameipnKdoe Kal Tots otzw dvuvapévois, Tods els tiv oderépay euBadreiv a&idoavras, eis THY GAAOTplav dravTycarTes, TpdTaLoy O€ OTHOAYTES K.T.A. The speech is-much too carefully composed, whoever the author, for it to be probable that démavrjcavres was repeated in this clumsy way. It may be noticed also that rovs... é£usoavras is wanting in construction, for it is very awk- ward to take it in apposition to ryv Svvayw tiv éxetvwr. If in apposition at all, it ought to be a genitive, close to éxe(voy and not cut off by ois . . . Suvapevors. Perhaps both these faults may be removed together by reading something like ¢@dcavres for dmavrycayres and taking the latter as an accidental repetition of a word previously used. The construction will then be p6dcavres cis riv dAotpiav rots... dfwaavras, anticipating the invader by reaching foreign soul first. 220 THE ATTIC ORATORS 60 cituyys 8 6 Tis ’Aoias Bacirets Erdpuv Fyepndvev AaBopevos. . . éruaBduevos is possible in this sense (Herod. i. 127 Wépoat wey vey mpoordrew értdaPdpevor PAcvOepodvro), but is the simple verb? In [Dem.] 25. 32 «i AdBour’ éovcias is by no means the same as «i éfovgiay AdBor, but like Plato Rep. 360d rowirys éfovcias émAaBdpevos. Cf. however zpocdapBdvec Oa. Blass’ arguments against Lysias as the author of this speech do not carry conviction. When we have made allowance for the interval that would inevitably divide the deliberately dignified style of a funeral oration from the deliberately plain, almost colloquial style which Lysias adopted in his ordinary forensic speeches, there seems to be nothing in the speech which he might not have put there. If it has all the faults of taste which Blass finds in it, I do not know why Lysias should be thought in- capable of committing them. When a plain writer ventures on an elevated style, he may well fall into such faults. Xenophon in such cases fell into worse, if all the work attributed to him is authentic. On the other hand I would not maintain with any great confidence that Lysias was really the author. I would only say that no very strong case is made out against it. I do not know whether the following words have been pointed out as rising above or departing from the usual vocabulary of the orators: 5 xAéos, and xdpw with genitive, a construction hardly used in the purest prose: 7 doeBeiobat passive: 21 oreAdew: 24 déwoiv think: 27 dwabis Kxaxav (frequent in Herodotus): 37 defiotcGar: 42 épifew: 50 yepatrepor: 60 AaBduevos (this use is however not even poetical ; see above) : 64 dScAdds as metaphorical adjective : 69 edxrdés: 74 Ajyew: 75 roxets: 77 xeipioros: 79 d&yjparos? Some of these will be found in Xenophon, Plato, or Thucy- dides, hardly in the orators. Less noticeable are wevOeiv (several times, ¢.g. 2), dyyy (3), wiadvw (7), evBpifw (9). deipynoros (20) occurs also in 6. 25 (thought not to be Lysias) and 26. 4, but is uncommon. Lysias seems not to use elsewhere 51a rayewv (26), but it occurs three times in LYSIAS 221 the speeches ascribed to Demosthenes and I dare say elsewhere in oratory. of rijs jAudéas évrds (50) is worth noticing for the sense of évrds (L. and 8.) and for its position. 4.7 viv S& dpodoyovpefa mpds waidas kat aidAntpisas xal per’ olvou EAOovres: Hote THs TAdT’ ear! rpdvota ; Something like éuoroyovpeba mpds maidas kal atAyrpidas Kal per’ olvou édOeiv 4. 11 trois 7’ dddows euhaves al rovrous rotjoat. In the context rovrors seems wrong, but it would be better to substitute odrws than Thalheim’s 8v airijs. 12. 11 Geduyv airod efddid por Sodtvar, 6 8 dyaryoev pe packer, el 7 copa cdow. Pison did not say that Lysias would be content. Lysias’ request showed that he would not. He said he ought to be content, 7.e. dyamav Sejoew we. For two words thus run into one cf. Plato Soph. 257 © EvpBeByxéva for EvpBeBynxev evar: Theaet. 185 D épyavidiov for dpyavov iSiov : Phileb. 41 iorepotpev for torepov épotuev: Thuc. 8. 67. 2 dvereiv and dvarpérev for dvari ecireily: Diod. 14. 56. 6 mpoeOuporvro for tpobvpous érowodyro: etc. Cf. on Lycurgus 131 below. 12. 21 wodAots 8 emeriuouvs dvras drivovs [rhs TdAcws] KaréoTyoay. So Thalheim, but where did ris wdédews come from? Perhaps we should read dripovs tis ToAtreias, though the word is no doubt superfluous. Cf. Dem. 15. 33 dripos rod ovpBovdrcvav: Thuc. 3. 58. 6 aripous yepav. 12. 35 pabovres driaow dre 7 Sixny Sdcovew ... 7h KT.d. There is no sense in dr 7 here. The men will find out whether they are to suffer or the reverse, not that they are to do one or the other. We must read ei, etre, or aérepov. Of. on 13. 82. 12. 60 Kai modes émdyovres Kal TeAevTOVTES Aaxedatpovious 4 222 THE ATTIC ORATORS 13. 18 ob yap Siprov éxeivor obrws dvéyrot joav Kai dpiror Gore «7A. For xai read #. So in the closely parallel passage Lycurg. 68 otrws éoriv dvéntos kal mavréracw ipiov Kato- meppovynxds Blass reads % for kai. 13. 46 tiv dxpémokw quay elxov. Half a dozen aorists preceding and following strongly suggest écyov. The mistake is quite frequent. 13. 71 For ofSe (Dobree ciS<) I would suggest 7da. A past tense seems much more suitable. ec 4 ‘ oy a = 5 c, A : 13. 82 trodapBavew xpy el "Avutos aitd éyevero airios pH} drobaveiv. Should not ei be 671? = Cf. on 12. 35. 19. 20 joav 8’ Amides rod wAOD weioat Atovictoy k.t.dr. Perhaps <ék> rot wAod. So perhaps in 2. 61 <é&> katvots ktvovvors after recobevtes. 20. 3 otros d& Tivos ay evexa dAtyapxias éreOdpyce ; érEpov os WAcKiav elxe Néywv te Starpatrec Oat Tap’ ipiv, y TH cHpare motevov, iva bBpifo. cis trav tperépwv twa; GAN’ Spare airod tiv WAtkiav, 4 Kal Tovs GAAous ixavds éorw dmotperev tour. I cannot make any sense of és #Atkiay efye. Does not the speaker say ironically as jAukiav clye? ‘Was it that he was not old enough to have the right of speaking, or that he was young and physically strong?’ the point being that he was really by no means young. 20. 5 ef ris dAéyas dp£as. éd/yas is certainly strange. Perhaps, as elsewhere, it has got exchanged for zodAds through a ‘certain similarity. 20. 12 67’ eerie rH Sypociv, od clojveyxev atta 7d dpyvptov. But why should he have contributed the whole sum ? Omit 7é (due to last letters of airg), or read tod dpyvpiov. LYSIAS 223 Just below éom in tpav éore wodirys seems a mistake for #v. How can the present tense be used of a dead man ? 20,24 dore rH Ged re Tas Sexdras efaipeOqva. mwrdov 7} tpidxovTa pvas Kal Tols oTpariwTats eis TwTypiar. After orparirats a numeral, representing a sum of money, would seem to have been lost. 21. 24 For the unsuitable #Aéjoa read probably aca (A for A). 22. 11 ofopat airots ext pev rotrov tov édyov odk eXcdoeo Oa. For the unattic édedocofae such words as rpépeobat, mopetvoecOat, katahevéecbar have been proposed. I take it to be perhaps a gloss rather than a corruption, and suggest dravrycev, as in Meidias 151 éwi ratra 8 éayvrov, os KT. In fragm. 47 (65) épxopevwy can hardly be right either. (Cf. on Antiphon p. 205 and on Isocrates Apophthegms below.) ~ 23.5 Probably padurr’ dopny eiddvar. 24. 3 ofua deiv. .. 74 Tod cHpavos dvoTuyypaTa Tois THs poxys emirpsetpacw iaobat Karas. xadois Markland, which will not quite do. Read xadois . After otc was lost, xadXots was turned into an adverb. 30. 19 dors afi mprov pev Kata Ta warpia Ovew, érera a aA £. nn # & paddov cupdéper 7H dre. For padAov, which seems without point, méAae and pddiora have been proposed. Perhaps @ perder cupdépew. 33. 4 eiddres Gre Pidovicety pev eore ed wparrdévtwr, yvavat 6¢ ra BeATicTa TéY airav. airév has been altered in a variety of ways. Such con- jectures as drvxovvrwy (Markland) and xaxés (Cobet) seem from the context to give the probable meaning, but I would suggest @avAws as coming nearer to the letters of auto. 224 THE ATTIC ORATORS Fragments. 1. 4 Write rod (for rovrov) reOveiros. 4 kat ef wey xpGro Kal exor AAKiBiddys, “A€idxov packer elvat Ovyarépa: ei 8¢ "Agioxos, AdKiBiddov. Surely épacxor, ‘they said.’ 15 7d pay epdv dm’ enod yévos dpxerat, 76 Sé coy ev coi Tavera. 5 mavoeras 4 111 ri pev Kopny prray exes, Tas 8 paoxdAas Saceias. Is xépyv a slip for xepadryv? how can long hair, or any hair, be yArAy ? IsocraTEs. 1. 5 od wapdkAnow eipdvres GANG Tapaiverw ypaiavres ped\Aopev cor ovpPovdrevev. There is no propriety in the expression rapdxAyow etpeiv, Jind an exhortation, or even devise. But cipeiy and cizeiv sometimes get confused. Seep. 186. Read therefore here eixdvres, which is much more pointed, for it carries on the reference of § 1 to Isocrates and the family of Demonicus being now separated. He must therefore not speak but write. So éraivovus cixav 12. 130, etc. 15 Fyod pddiora ceauTa mpérev Kdcpov aicxuryy, Stxato- civyy, cwppocivyv: TovTos yap amact SoKel kparetr Gan Td TaV vewrépwv 700s. Blass (Teubner 1879) bracketed xécpov, turning xpareto Bar into xocpeioba.; and this may be right. But perhaps mpérew Should be zpérovra. 25 pare pera BAGBys wepS rdv hidrwv pyr’ adretpos elvas TaV éralpwv ede. todro S& roupoes, dav pty Sedpevos [741] SeioBar mpooroly. Tept Trav pyrav as dtoppytwy dvakowov" uy TYXOV pev yap obdev BraByoe, tuxyov 8& paddov aitdv Tov Tpdrov 2 4 ErLoTHTEL. ISOCRATES 225 It is surprising that the irrelevance of wept rév pyrav as droppytwv avaxowod has not been more generally recognised. py Texov k.7.A. evidently follows zpoorog. The intervening words, which merely disturb the sequence, must be put either earlier or later; probably earlier, as they could not con- veniently come later for some time, whereas they might quite well stand before pyre pera BAaBys. 7é should pro- bably be omitted. 28 qreip@ tov wAodTov xpywaTa Kal (Blass) xryjpata Katackevalew eore 6& ypyuara pev toils dmoAavew émirrta- pevols, KTypata dé Tois KTacOaL (Blass) duvapévors. py woust no doubt be inserted, but I am not so sure about povoy, rather suspecting that here too the words have got out of their order. Did Is. write ori 8& xrjpare. pev 7. x. 8, xpnuata de 7. dé? The similarity of xrjpare and xpypara might lead to the error. 31 pydé wapd .. . wyde mapa. Perhaps pre in both places. 2. 45 cipyoopey Tovs roAAovs adrav (1.€. Tav dvOpurwr) ove tov oitiwv xalpovras Tots bytewvordrots ovTe Tay émiTydevpdTuY tois KaAAioTos ore TOV Tpaypatwy Tos BeATioTos oUTE TOY Opeupdruv rots dPeAtwwrdros, dd\AG wavrdémacw évavrias TH oupépovre ras HSovas éxovras. Opeppdrwv is the reading of the Urbinas and one other good MS. as against the palyydrwv of the rest. It has had the fortune, rare with bad readings, to be defended by Cobet (N.D. p. 154, and V.L. p. 515), who refers to the Athenian éprvyotpédo. as illustrating Isocrates’ meaning, and it is adopted by Blass. But Opéupara is rather ludicrous after émitydevuara and wpdyyara. Perhaps ano- ther word may be found, giving a better sense and also explaining better the v.l. waOnpdrwv. Isocrates must have written trav Oeapdrwv tois apedtuwraros. A word like Oéopo. harmonises much better with mp@ypa and émirydevpa, and is confirmed by Oewpotvres (48) and the ptOo. which are said to have been made not only dxovoroi, but even Gearol (49). Ch Thuc. 2. 39. 1 Oeduaros 6... av tis dav ddednOein: Polyb. 1. 1. 6 ékrabys mpés tt Tv addAwv Q 226 THE ATTIC ORATORS Ocapdrwv 7 pabyudrwv. Isocrates could also write Qewpy- parwv in the same sense, and that would account for the p in Openudrwv ; but perhaps Ocaydrwy is the more likely. The word is used in 4. 45, Oeduara mdciota Kal KéddtoTo KEKTY TOL. 5. 1 wy Oavpdons, & Bidurze, Sidre rod Adyov roujropat TH apxiv ob ToD mpdos oe PHOnTomEvoy Kal viv SetxOjocrOae féAAovTos GAAG TOU Tepi ’ApdirrdAews ypaerTos. I cannot translate these words as they stand, but con- jecture otx <éx> rod. The cause of the loss is obvious. 7. 20 Hyetobar tiv piv dkodraciay Snuoxpariavy, thy 6é mapavopiay edevfepiav, tv d& mappyociay icovopiav, tiv & efovatay Tov Taita wovely eddaipoviav. Tatra seems without meaning. sdvra, sometimes con- fused with it, is exactly what we want, révra qoveivy being a familiar combination. 54 yeydLovras ev rovovrois év ofs od BotAopar A€yew. This is certainly unlike the usual form of expression, which would be either év rovovrois ofs (or &) or éy ofs simply. Should we omit the second éy, which indeed hardly makes sense ? 8.13 drav 8 Urép ris ToAcws exxAnotdlyTe, ToIs ev TOLOUTOLS dmotetre kal pOoveire, ros 8 movypotdrous Tay él 7d Biya mapiovTwy doxeite Kal vowilere Symotikwtépous Tovs peOvovras Tav vyddvTw. In Demosthenes Ol. 3. 28 and FL. 339 I suggested changing joxyxapey and cwackeiv to niéjxapev and cwvavéew respectively. Here in the same way there seems little sense in doxeire and I suggest avéere. atéw or aiédvw is used in much the same sense 5. 120 dou 8 Idowv ddyw povov xpycdpuevos otrws abrév yvénoev, molav Twa xpy Tpoo- Soxav mepi cov youn dravras tev x.t.X.; Cf. Plato Lysis 206 A of Kadoi, éreddy tis adrots érawy Kal avéy, ppovipatos éumiurvavrat. So sometimes augeo in Latin and ratse in English (‘ while wits and Templars every sentence raise’). Or we may take avgere in its more usual sense and compare [Xen.] B.A. 1. 14 rods pev xpyotods dtyotor ... rods dé ISOCRATES 227 movnpovs avgovow. But Isocrates’ form of the word was adgdve rather than aiéw, if MSS. may be trusted. 114 jv ras adras mpdéeas ext wévrwv Tov duoluv daivwvrat yvupilovres. So the two best MSS. (Blass), but the rest appear to have éyuoiws for tay sSpoiwr. Should we explain and reconcile the discrepancy by reading rav éuolwy Solus 4 9.47 rhv wow ... obre réxvas émirrapérny obr’ gumopin Xpwpevyy obre Areva KexTNmeEryv. Rather éyopia. 54 76 pev yap hv dicet rarpis (4 quetéepa mdAs), Tov 88... vouw ToAityy érerotnvto. Surely éremoinro. 10. 8 rodpéor ypddew os err 6 rOv wrwxevdvrwv Kal dev- yovruv Bios Lydurdrepos 7 6 Ta dAduv évOpdrrov. Blass follows Benseler in omitting 6 after 4, so as to get rid of the hiatus. Kayser had omitted 4... dvOpdéruwv altogether. With 4 but without 6 the grammar of the sentence would be I think questionable. Perhaps we might regard #6 as an explanatory adscript and read by the common Greek idiom ¢yAwrérepos rév dhdwv dvOpdruv, as in Democritus (fragm. 98 Diels) évds ditty Evverod kpécowy aévvérwv wavrov: Xen. Hiero 1. 38 éfexdfovow éavtots tais tav didovvrwv troupyias: Ar. Hth. 3. 10. 10 gddpvyya paxporépay yepdvov: Isocrates himself 2. 31 13 rijs woAews GAns 700s Gmotodrar Tots dpxovow: 11. 7 wérepa rots rept AidAov Acyouevors adrov rapardéwpev; and many examples in all authors. 11. 5 drodoyjoacGa should be érodoyyoer Gan. 12. 36 dyvod should be dyvody. Otherwise there is no construction for cidas and cipyxds. a 12. 131 xarecrycavro yap Sypoxpatiay ob ryv cixy roAtrevo- pévyv Kal vouilovoay tiv pév dxoAaciay eAevOepiav civat, tiv 8 efovoiay 0 tu Bovdcrai tis moety eddammoviay, GAAQ THY Tots Tolovrols pev emityooay, dpiotoKpatig. 6& xpwperny. Q2 228 THE ATTIC ORATORS Snwoxpatiay ... dpirroxparia xpwyevny is an expression which no artist in words, such as Isocrates, would have thought of using. Read xexpapévyy, a word which Aristotle employs once or twice in the same connexion ; Pol. 6. 3. 8: 8. 8. 2 rats eb Kexpapévars oduteias. CE. § 153 of this same oration : dypoxpariay ... ryv dpioroKparia pepeypeevnv. 15. 28 dv oddey pou rAdov yeyovev. Is this Greek ? must we not read <é> dv! 16. 26 ycyvomevys, not yevouevys, seems required by the expression émdre Tdxeivwv Kparyceev, ov. . . KaTeoKaTTOV GAA . « dvdputtov. 40 apa rijs Tédews does not seem to make very good sense in the context, and I should prefer zep/. 43 indpfe. would suit both érérav Séf and the future following better than trdpye does. 17. 6 ayovpnv... cSuvevav ... yernoerGaris not, I think, possible Greek. Read xwéduvevoew (or xvduvedev av)... yevéoba. In7 again not zoey but roujcev must be read after rpocopoAoyeiv. Ep. 4. 4 ovvnpepcioo Kal cupGiadvar aévrwv ydirrov Kal Atyvpwrarov. This use of Atyupds is both unique and unintelligible. May we not suppose a corruption of tAapdérarov? Plut. Cato 25 ovvdujye tdapds: cf. Ken. Ages. 8.2 idapos .. ydioy Sinuepeve. Somewhat similar is A¢tov (Stobaeus) for iAewy Plat. Laws 792 5, and éhéoGa a variant for A€dyerPar Rep. 392 c. Ep. 9. 6 pdov éort wept rav yeyevnpevwr ebmopus éemdpapetv } wept tov pedAdvrwy vouvexovTus eimely. eirepus seems unsuitable, but neither éruexds nor dvextas (suggested by Dobree) is satisfactory. I would propose drévws. et and d- are not unfrequently confused, and mopos also gets exchanged with ovos, e.g. in Aeneas Tact. 14, 2. ISAEUS 229 In the ‘Apophthegms’ put together in Blass’ edition two or three small improvements may be made. : fpurnbets rive of iAdrova: Tov pabipwv Siadepovow etrev ‘ as of eboeBels THY doe Biv, eAmiow ayabais. Is it not plain that és should be 1 ef. Bywater (Textual Criticism of Nic. Ethics) on Eth. 1106 b 36 ds (or @) av 6 ppdvipos dpiceer. ot dvOpwrot Tore yiyvovrar Bedtiovs, Grav bed zpocépxwrrat Gpotov Se exovat Hew 70 edepyereiv Kal dAnOevewv. “Icoxpdrous. Of course Is. did not use zpocépywvra, but neither, I take it, did he say BeAriovs. BéAticro is almost required by the sense, and no confusion is more common than that of comparatives and superlatives. In zpos roy eimdvra tatépa ws ovdey GAN’ 7 avdpdmodov ovverempe TG Tatdio, ‘ toryapovv’ épy ‘dri. , dio yap av0’ évos ees dvdpdroda’ no doubt oddev should be oddev’, and perhaps ovvémeupe should be cupréupe. In the extract (6) from the Téyvy read ré (not rév) Emdpevov. Isarus, 1. 33 dare pyde Adyov trodeiipery. Read todetrew, comparing Goodwin M.T. § 591. otras éroie Bore brodcipe: would be out of the question. 4 49 a y 3 s 2 6 2 > ON a ow 64 yvevat OGOV AVALOXVLVTOTATOL av PWTWV ELOLY OUTOL. Rather dow, the regular case with comparatives and superlatives. 4.18 eixds hv ddyOets elvar ddgew Tas SiabjKas. eixds Seldom, if ever, takes a future infinitive, just as it very seldom takes dv. Read Soxety or ddgat. 5. 14 gets 8& xabyyoupevor odkeérs mpoojne K.T.A. &6 Fyovpevor (Baiter) is not likely to be right, dre being little used in common Attic prose. xai #yovpevo. may be suggested, a participle being perhaps lost before kai, e.g. kal yyovpevor, Or . The confusion of «af and xara is familiar. 230 THE ATTIC ORATORS 6, 50 iSeiv & obk éfdv airy. Read ééjv. We might think of omitting d, but ideiy needs an object. [So too Wyse.] 7 8 ¢ fos , 2A 2 / . EWS OU EUTOPYTELEV EKELVOS TAPYUPLOV. éws o8, if right, is very unusual. It occurs Herod. 2. 143, but probably not elsewhere in Attic. The occasional use of pexpr od in Thuc. and Xen. is parallel, but their Attic is not of the scrupulous kind we expect in Isaeus. of may be a dittograph of ei-. 9. 21 odSérore, referring to the past, should probably be ovderurore. Fragm. 32 ypi rods vopous pev TiHcoGar opodpas, mpaorépws 82 xoddLlew 7 ds exelvor KeAevovow. I suspect that cpodpas is due to zpaorépws and that Isaeus wrote ogodpors. Lycureus. 13 apos 3 Sef kat Spas droBdérovras pH exerpérery tos éfw Tov mpdypatos A€yovow: ovTw yap ~orat Tois Te KpLvoméEvots My aA SS. % * - , + Cet N dvev diaBodjs 6 dywv kal tots Sidkovow AKLoTa, ovKoparTelv Kal tpiv ebopxoraryy Wihov eveyxety: advvarov ydp éoru avev Tod , a ie: # t - ‘ a Adyou pH Sixaiws Sedidaypevous Sixaiav Oérbar tiv WHdov. Multi multa coniecerunt, says Blass, velut dvev rod dddyou G. Herm., dvev rod Adyov Nicolat. An dvev rod Adyw (quod facile in Adyou abibat) Bixatws Sedi8dx0ax 4? take dvev to be a careless repetition of the dvev preceding (dvev diaBodjs) and to have taken the place of ééw, ‘after listening to irrelevant and unfair speeches.’ So 2éw rod mpdyparos in the first clause. We might compare Maine’s remark about an English jury in Popular Govern- ment, ‘there is a rigid exclusion of all testimony which has a tendency to bias them unfairly.’ 7 dixacws is in antithesis to dixaiav. 28 kai tatra 8’, & dvépes, énod Oewpyoare, ds Sixaiay tiv eféracw rrovoupévov Tept TovTwr. LYCURGUS 231 kai tadra refers to what is coming. Should not zoovpévov be zo.otpart He wants them to observe how scrupulously he is proceeding. 29 Read rév trav wavra (for tov Tapa Tav) cuveddtwv édeyxov. [So too Herwerden. | 30 Keeping robs idfous xivddvovs, add perhaps tropévwv. CE. trduewey following and otdéva kivdvvov tropetvas in 78. 32 obtrwot dé diadroyilecbe wept tovTwy wap’ tpiv airois. tivas ddvvavov Av TH Seworyntt Kal Tals wapacKevats Tats Tod Adyou Tapayayely ; kara piow Toiviy Bacaviduevor racay THY GAnGeav wept wavtwy Tov dodikypdrwv Euc\dAov gpacew ot olkérat kal ai Gepdmawvat, GAA TovTovs Aewxpdrys Tapadodvat epuye, kal Taira otk adXdorpious GAN’ abtod dvras. tivas dé Suvardy elvat Soxet Tots Adyous WuxXaywynoat Kal THY bypornta avtdv Tov HOovs trois Saxpvous eis eXeov mpoayayérbat ; TOUS ducacrds. évradda Aewxparys 6 mpoddrys THs Tatpidos EAjAvOEY, obdev erepoy 7) hoBovpevos py ex THs adris oixias of 7’ éfedey- xovres TG Epyw Kal 6 eLeAeyxdmevos yévyrar. There seem to me to be at least two things wrong here. First the question rivas ddvvarov fy x.7.A. has no answer made to it, although xara vow tolvw «.7.d. is incoherent without something to precede it and although symmetry with rots d:cacras, which answers the parallel question, requires a corresponding accusative. Insert therefore as an answer after wapayayetv something like rots dovdAovs. Secondly the words ovdty érepov... yévytar appear to be entirely irrelevant where they occur. They have no sort of connexion with rhetorical artifices addressed to the court. Clearly they refer to the domestics mentioned in the first half of the passage, and their place is after airod évtas. In Blass’ text will be found one or two other similar transpositions (§§ 14 and 438). There is said (Rehdantz) to be no parallel for oidty érepoy 7 instead of the usual ofdév d\Ao 7, and this suggests the possibility that Lycurgus wrote oidév érepoy (or mpdrepov, §§ 92, 129) doBovpevos 7. But the text is probably right. 40 épav 8 fv... yuvaixas édXevOepas . . . dvatins atrav kat THs TOAEWS Spwpevas. 232 THE ATTIC ORATORS épwpévas may be defensible in itself (cf. for instance Lys. 3. 6: Eur. Androm. 878) but as governed immediately by épay it is surely intolerable, and dvaiws édpwuévas is an odd phrase. How to deal with it is another question. If it is an ordinary corruption of some more or less similar word, é3upopévas (which is suggested) or éAopupopévas is likely enough to be right. Oddly enough D. Hal. 4.2. 8. 44.2 76 py ev dvipdow aowyPeow Spacba at any rate illustrates épwpuévas, while 2b. 39. 1 karaAurotoat Tis oiKot povis 7 cbrperés cov... dAopupdpevac might be quoted in favour of the change. Anything rather than Rehdantz’ dpvopévas. If however it is due to épav preceding and to isety and épav following, all in one sentence, it may represent some quite different word no more like épupévas than in 13 above dvev is like éw; and then conjecture is almost idle. 57 ¥déus 8 av abrod mvOoipyv ri’ europiav cicaywv Xpnotpdrepos éyever’ av TH oder TOD wapacyeiv 76 copa tdéat Tols otpatynyois Kal tovs émidvras dpivacbar pel’ bpav paxopevos. Tov wapacyxeiv is so awkward a phrase, when the speaker had only to say 7 wapéywv, that I would suggest tiv’ éuropiay eiodye ypyoipwrepoy . . . TOU Tapacxeiv... Kal... dpivacbar... paxouevov. Or we might read 7 7G wapacyeiv, taking 7 as lost after e. 67 ob rotro Aoytiode, «i els ears pdvos 6 dvOpwros GAN’ eis TO Tpayyo.. Blass writes ‘dAA’ ofov 7d ap. Bk., ddd’ ei (Tayl.) 7d zp. péya (Dobr.), ddd’ cis 76 tp. = Frohb.’ It is well known that cis and ris are sometimes confused. Does not eis here stand for ri? 73 xal, 7d Kepddaov THs vikys, ob Td év SaAapive tporatov dyarjoavres Exrnoay, GAN’ dpovs tois BapBdpors wygavres . . . owOyKas érojorarto. ay. éor. is such an unnatural phrase ‘did not set up the trophy, contenting themselves with it,’ that Blass follows Meutzner in omitting gorycav altogether. Should we not LYCURGUS 233 rather exchange verb and participle, reading #ydéryoav orjoavres? Many passages have been or are to be restored in this way. See Index, Terminations. T4 rods kaxovs pucety Te kal KoAdfew, dAAws Te kal AewKparny, bs obre Cercev obre Noxon ipas. Is there any parallel for such a use of dddws re kai? It introduces adverbial expressions, especially as, especially when, especially if, etc. Can A., a simple objective accusative, be placed after it? or should we read dAdovs te kal? I confess rovs re dAAovs xai is rather what one would expect. 76 ipiv yap éorw dpxos... duvvely TH warpids Kal dpcivw mapadocev. For dzefvw read peifw. The words of the oath, as § 77 purports to give them, are tiv warpida Sé ob« éAdoow mapadwow, wAciw bé Kal dpetw, where dpetw means stronger. Cf. also 78 tin 8 dv thy watpida Tapéduxe peilw mpodocia ; Make the same correction in Plutarch An Seni etc. 786 ¥ Thy ray maAaav xdpw eyeipey Kat mouely dpeivw (read peffw) kal pdvipov. In Soph. 0.7. 772 7G yap dv kal peifov | Aeaup’ av 4 cot; I would make the reverse change, Kayteivove (p. 292 below). 93 Callistratus, who was told by the god at Delphi ér dv 26y 'AOjvale revSeras Ty vouwv, returned there and was put to death, dixaiws' 7d yap tov vouwr Tois HdiKnKdoL TYXELY tyswpia éoriv: 6 b€é ye Geds dpbGs awédwxe Tots AoiKypeEvors KoAdcat Tov alriov: Sevov yap av ely, ei TaiTa onucia Tots eioeBéor Kal Tots Kaxoripyots paivorro (patvovra. MBSS.). If Bursian and Blass are right in reading ripwpias for tyswpia, I think we must go a step further and insert another rvyeiv either before or after rots 78ixnxdou. Clearly it would very easily get lost. But the last words of the passage are the real difficulty, ei tavTa «7.4. We may probably | accept ¢aivorro or some- thing like it, but raird onpcia is very obscure. Why should not the same signs be given to pious men and evil- doers? or what is meant in ‘this context by ‘the same signs’? The real point is that the god, as was his way, used an ambiguous or equivocal expression. The speaker 234 THE ATTIC ORATORS must be defending him, not for having given ‘the same sign ’ to righteous and unrighteous, but for having given a ‘sign’ or intimation which might bear more than one meaning, the sense intended varying with the character of the man to whom it was given. Just then as it seems very possible that one rvyeiv out of two was lost above, so here I would suggest ei ratra onpeia... paivorro, t.e. gave the same signs in the same sense, did not make the meaning vary with the man. radrod, or radrod ra, onpeia, might also be suggested, but raira rairod oypeta perhaps expresses it best. 106 xarédurev yap airois éXeyeia moujoas. Are the words in their right order? oujoas ydp . . karéAumev ? or éAcyeia . . . karéAure roujoas ? 123 rods émixepyoarras tiv... cwrnpiay dmocrepety ? 131 ds trép tavrys paxovpevor 7 Koy . . . cvvaTvxodvTes. Dobree and editors cwarvyyjcovres to match payovpevor. Possibly ovvarvyeiy dgtodvres. Cf. on Lysias 12. 11. 140 Hyotpar 8 eywye oddé’ [dv] otrw peydda tiv rdw edepyeTyKévar wor’ ééaiperov déioiv AapBdvev ydpw Ty Kare Tov mpodiddvray Tinwpiay, ovd’ odTws avonTov doe K.7.X. At the beginning of 139 he has expressly said that certain persons on the ground of their own public services are trying to beg L. off. It is impossible therefore that the words before us, in which he declares that no one could do such a thing, should be right. Blass accordingly brackets aéoiv; but how did it get into the text? I think as a corruption of déiov eivar, which makes excellent sense (‘no one’s past services are such that he 7s entitled’ ete.) and has the further merit of smoothing 038’ odrws dydyrov, where an clva: is rather lacking. déwdv would make sense, if it could mean ‘that the state should think him worthy,’ but that does not seem possible. It seems to me that something has been omitted in 73 Spovs rois BapBdpos mygavres tots cis tyv éXevbepiav Tis “EdAddos. AESCHINES 235 105 dore trois dvdpetordrois Aaxedaipoviors év Tots éumpoo bev xpdvois TroAcuovot mpos:Meconvious dvetAev 6 Beds K.7.A. Surely for prose rots dvdpetordros A. is not good Greek, or is at any rate very unusual Greek. An epithet is not thus attached to a proper name, or very seldom. It seems likely that yevopevors or otot, going with dvdpeordros, has been lost. After one of the plural datives oto. would very easily fall out. [Naber has recently suggested dvdperarta, omitting Aaxedaxpoviors. ] So again in 18 dowep rq marpid: peyddas ebruylas evayyeArlopevos it is hard to believe that the dative really depends on cirvyias, pieces of good fortune for his country, when any hearer or reader would naturally attach it to eiayyeAtcopevos. Here too some participle like yevouevas or cupBdcas has perhaps been lost. Two very small points in conclusion. In 2 rdv pre mpodévra. poy’ éyxaradumdvra might be better than Dobree’s pare mpoddvra pyr’ éyxaraAudvra. Cf. the réy in 1, which is also bracketed, I think needlessly. In 96, 76 xupiov . mpocayopeverOa trav edocBdv yxGpov, is not Tov wanted ? AESCHINES. In Tomarch. 29 ry wore, dzép Hs Ta STAG pH Teo 7 Sid Sediav py Suvards ef érapdvat, pyde cvpBovdrcdew agiov. The two alternatives in the relative clause seem hardly to make sense, as 5:4 derAdav would surely apply to ra éada py tera too. Change 7% to xai, for which it does some- times get substituted and vice versa. De F. L. 177 drysou 8 ex cvxopavrias <évres>? In Ctes. 4 75... kypvypa... ‘ris dyopevev Bovrerar tov trép wevTyKovTa ery yeyovotwy kal wadw év péper TOV GAdwV ’AOnvaiwy ;’ So the editors give these words, but the quotation really ends at yeyovérwy. This is both clear in itself (for the whole as one question or proclamation would be ludicrous) and made still clearer by what Aeschines says elsewhere 236 THE ATTIC ORATORS (in Timarch. 23) pera taira émepwra 6 kijpvé ‘ris a. B. 7. bm. é. y.3' éwedav 82 obror mavres elrwot, tér’ dn Kedever A€yew tov dAdAwv “APnvaiwy tov BovAduevoy ois eeotw. In Alcid. Soph. 11 we have the abbreviated form ris dyopevew BovAcrat Tov ToALTOV ; 153 yéverOe 84 ror puxpov xpdvoy tiv Sidvorav py ev TE Sucactynpiy GAd’ év 7d Oedtpy. ‘Malim rq Siavofa, ut § 157, 186, Plato legg. 683 c Bait. sed cf. 1.§ 179 (éredav 8 dad tHs dmodoyias drooracbire Kal Tas Woxas éd' érépwv yevynobe).” Ziirich ed. I think the dative should be read here and in 1. 179. Cf. Aristotle Fragm. 58 (Rose) ei tis jas ofov eis paxdpwv vycous TH Stavoia Kopioeev: Isocr. 6. 110 ei tats dtavotats domep wapestaras iorre robs yovéas kai tovs waidas: Polyb. 8. 15. 1 adda S& rH Stavoia wept Tov Aicoov Kai tov "Axpodcooy dv: Lucian 30, 5 émi tot abrod déous viv ta Aoyiope yevowevovs and again mpods ddiyov obv TH yvopy els "Axpdyayta wap’ ue arrodnuycartes. 184 In the fourth line of verse, zpérot Svopevéwy ebpov épnyavinv, I cannot but think etpov is a mistake for the more forcible and appropriate eiAov. Cf. p. 279. Theocr. 10. 17 efpe Oeds rov aAutpov is only partially parallel. 218 dpxet ydp pot puxpa Kal peldvov aicxpds od ériBups. There is nothing aicypév in desiring more than a little. Read yA‘oypws, a much more suitable word and one which is elsewhere confused with aicypds, e.g. Plato Crito 53x where the MSS. and editors are divided between aicypas and ydoxpws émbupely Gv: Dem. 37. 38. Cf. Cobet Collect. p. 510. 231 ef tus Tay TpayiKov TowTav Tov pera TadTa érecayovTwy moioete K.T.X. The future éreoafdvrwy seems almost necessary. I have sometimes thought that in § 6 odcerat would be better than o@ferar, as expressing what the early voyobérys fore- saw; see however Veitch s.v. who says ‘we know no instance of fut. mid. odacofar being passive.’ HYPERIDES 237 235 ovbdels wdmore érefero. . . wpty av peiloy Tov Sixacrnpiov ioxtoy. Cobet apiv icytoa, but in this and similar cases a present meaning easily suggests itself, oddels twrore éméBero for instance being roughly equivalent to ovd«is ore émitiBerar. So e.g. Isaeus 4. 27 ovre drodedyuyxacw ovdapy mumore, OTot av pun) Upeis Tpooraéyre. Letters 1, 4 dvrervel. Probably dvrérve. All the tenses are past. 5. 6 kal AoWopiat als édowopovpnv ? 10. 1 kat & pev efdov airdOe ypddew eet Soxet &pOovov exe THY VAyv ciwtyow. Perhaps ypadew. HYPERIDES. In Athen. 2 ottws ds éouxev ekiotyow avOpwrov vow épws mpooAaBav ‘yuvaikds..... tay (the five dots repre- senting the number of letters apparently missing in the papyrus). Blass cites a number of conjectures, xapdiav, aipvAiav, kakiay, otktAtav, mavoupyiav, dmuortiav. Has anyone thought of zovypiay 1 18 dowep imd........ év rodootpaBy ciAnypevor. The missing eight-letter word may be Oypevrot or Onpevtav. 21 6 dorepos mpidpevos 7 6 waar KeKTNMEVOS. Should not dorepos be iarepov? 23 kal raira Kai mpds buds airov hac pédAay déyew, iva 39 Soxoin pérptos etvar. If the optative is right, it is another instance of an irregular sequence like Ar. Ran. 24: Dem. 22. 11, in which the speaker’s mind changes the time relation. ‘His idea was, when he resolved on this,’ etc. Cf. particularly Ar. Av. 1524 émorparevoew pao’... ei pip wapeser... iv’ eigdyotto. 238 THE ATTIC ORATORS The conjectures yapicOjvat, xovdicOivat, dvruradeiv, pde- Oa, cwOjvar are given by Blass, who adds on Kenyon’s authority de kit. p non vid. dubitari posse. Has ddercicGar too many letters ? Epitaph. 26 imép dv drdvrwv obrot révous rover S:1a8dxous TOLOUPEVOL K.T.A. This has the appearance of a verse quotation, movous Tovay <8} 1> Siaddyous rovovpevot. Cf. Eur. Hec. 588 Avan ... diddoxos Kaxav kaxois : Suppl. 71 dyav ... yowy ydors Siddoxos (quoted in L. and §.). Fragm. 192 éwedav émi Sterés HBdow. The aorist #Sjcwow, after they have attained manhood, is indispensable. DINARcHUSs. 1. 42 rpia rédavta AaBov peréypade kal pereckevale Tov vopov Kad? éxdaorny éxxAyoiay, kal Ta pev erdre dy ciAnpe THY Tyunv, Ta 8 daroddpevos ovk éBeBaiov. érdAec seems hardly right. If D. meant ‘he offered again for sale what he had already been paid for,’ the again and the already could hardly fail to be expressed. éréAe must, I think, stand for érofe, and otk éBeBafov (=ov« érole) decidedly supports this. ab. 72 értBrépare 8 ext riv OnBaiwy wodw. éyévero wos, éyévero peyiary. mwédts is unmeaning, but its proposed omission unsatis- factory. Where did it come from? Read rather méAat, which gives excellent sense. Cf. generally § 75. ab. 87 ry Tovrov dAyOciag or something similar ? 4b. 110 eis 70 ris méAcws oop’ droBddpaytes Kal Thy mporepoy Sdgav imdpxovoav airy. DINARCHUS 239 No parallel is cited for such a use of cio. Read cyjpa. Cf. Eur. Androm. 1 ’Acidribos yas cxjpa, @yBaia wdALs, where the schol. uses xakAdmicpa and xopmos. So mpdoyynpa in Herod. 5. 28 Midyrtos, rijs "Iwvins tpocxnua. In [Dem.] 61. 12 the MSS. vary between odparos and cyyparos: Plut. Luc. 8 cxfpva is a certain emendation for cdo. 2. 9 Gr’ cis 7d Seopwrypiov Td mpOrov arynxOy . ., Toatra TOAMYGAVTA, Tole aiTous. atrod Blass with other editors, meaning, I suppose there: but this would be a misuse of airod, which is not = simple éxel. adrovs seems to admit of defence as referring to rovs Secpras implied in Secpwrypiov. So eg. Ar. Hth. 3. 11. 1118 b 19 kadrotytar yaorp/papyot, ds rapa 7d Séov wANpodvTes airy, €.g. Thy yaorépa, and many other passages. 3. 16 Perhaps kat zpis jv av (for 437) dixaiws eCyuswpevos. RHETORICA AD ALEXANDRUM. I quote by page and line of Hammer’s Teubner text (Ehetores Graect), 1894. 10. 10 iety 8 eori mdvras rods dpiora Tov “EAAjvev moXttevopevous Adyw TpaTov 7 Tots épyous vyyvyvopevous. For apérov read mpérepov. The writer goes on to say that careful consideration before action is an dxpétoAts cwrnpias, and that (16) ravrnv émdpOyrov oinréov, od Thy éx TeV oiKodo- pnpdtov aopaly mpds owrnypiav elvat vopicréov. oinréov is senseless, for the important point is that it be impregnable, not that we who hold it should think it so. Is it not an easy mistake for romréov, made still easier by vopsoréov t 16. 14 e ydp should be ei dé The Hibeh papyrus as well as the codices has yap, but it cannot be right. 19. 17 ei otrw xadas ras éoptds ... cupBe Byxev? 21. 18 Something like def d¢ airav tiv Odow . Cf. 15.7: 16.12: 22. 22. 22. 5 rots 8% wAovrodow cis Tas Kowas AyTovpyias éxovoiay dracav diAotimiay éumroujowo. There is little force in dracav. Read dracw. Cf. p. 54. 23. 19 apds rHv Tdv woduTiKOv Kadokdyadiav. No doubt tay wodirav. [Ar.]’AQ. IloA. 8. 4 the mistake is the other way. 26. 1 dtefdvras 7a Kara tov médenov druxypara tots avOpdrois OF dtuxnpata 1? RHETORICA AD ALEXANDRUM 241 4b. 13 48 eipyvy rods pev HrTwpevors aeley (répuxe), Tods X a zt. # x an 6& vuxdvras dv ex tod woAduou éxrycavto dwoAavew <éav> or 1 38. 9 ré€yw 8’ olov véos éoriv. Insert ei after ofov. 52. 9 80 épnpiav Trav Aeydvrwv. Here and in 69. 10 and 19 épypias rv cupBovdrevdvrev the participles should probably be future according to the usual idiom. Cf. 87. 16 didwy daopiav tav d&ywvioupévev trép abrod. 59. 12 xp 8& Kat rots éydpace Tots péeoows xpyoOa, Kat pare Tots paxpordros pyte Tots BpaxutTarous pyre ToAAOIs epi ye évds, GANG perpiots. With pajre wodXoits repé ye évds some- thing corresponding like pyr’ éAlyois wept woAAGy seems wanted, as in the antithesis of paxpdrara and Bpayvrara. 70. 6 dray tis jovxiay mpos tous pydey adicotvras }) robs f + nv oe ~ 2 - a kpeitrovas cupBovrety 7 cipyvyv toretcOa aicxpdv. Sense seems to call for xpeirrovas ovpBovdrevy. od ovpBovrevw is sometimes, e.g. Herod. 7. 46. 1, like od dye 1b. 22 ragopev dé mas ; Perhaps tus. 83. 3 trepBadrAav tavtas exeivor 2 90. 26 éyere should be ééere. 93. 7 otdeis yap oldev évOparwy ci tL ToLWWtTOV aiT@ couweurécot This is the sense needed. Just after add 8¢ to drddawve. CRITICAL WRITINGS OF DIONYSIUS OF HALICARNASSUS. In the following notes, which deal mainly with the De Compositione Verborum, I have followed primarily the Teubner text of Usener and Radermacher (referred to as U. R.), of which vol. 1 appeared in 1899, vol. 2 in 1904. 1. 5 ei pédAovor wy wav . . . Adyew pnd’ cixy ovvOjoew. Considering (1) the rarity of such a mixture of tenses, (2) the frequency in MSS. of this particular error, the writing of present for future when they are very similar, we should probably read Adfev. 4b. eis 8 ToUTO Td pépos 0 Set TpGrov vows doxeio Oar. Write véovs. The dative with the passive present, though of course possible, is unlikely in Dionysius. doxa can take two accusatives, and two are therefore right with the passive after dei. ab. 6 éav 8 éyyévytai por oxoA7. Probably éav 5 yevytar. Is there any parallel for such a use of éyyiyver Oa 1 ib. 7 wotav xparioryy abrav evar re(Bopar. 21. 146 ras pevror yevixdas abtis Siadopdas tavras elvae me(Oouat povas. I suspect zeéOoua: in both these places—in spite of Oyoerat following almost immediately in the latter—should be rieuot, as in 21. 145 cidixds pev Stadopas rodAAas opddpa elvat TiHepuar. Schafer in this third passage would suostitute DIONYSIUS 243 weOopat for 7éMeuar (see his note), but it seems more natural to say ‘I make three classes,’ ‘I lay it down that there are many distinctions,’ than to say warmly ‘I am convinced’ that there are. Such uses of riféval, rifecOar are very common in Plato and Aristotle. So again in 26. 215 yédAwrtos aétov riPcuan. . 4. 29 rots pev otv dpxaious dAiyou Selv maar moddy émirndevors HV abrod.... xpdvw d vorepov wavtdracw nuedyOy kat ovdels dero Sely dvayxatov avro elva. For dcty after gero, which can hardly be right, U. R. rather strangely suggest Norov now. The error seems very unlikely. It is more probable that deity repre- sents an infinitive governed by d. civat, @.g. ovdels wero émityndevew d. abrd elvat. émirndevey suggests itself from émirndevors preceding, and the partial similarity of wero might occasion the mistake; but no doubt various verbs would do as well. 6. 39 ely ri pera rivos dpuoldpevoy mépuxe Kadyv kal ydetav AnperOat cvfvyiav. It is very improbable that anyone would put a future infinitive after répuxe. A great number of infinitives must be found after it in Greek literature: is there another case of the tense being future? Perhaps we should read Ti peta Tivos d. wépuKe kK. K. Hy. Anwerar o Two or three times over in this chapter we have both Aywera (wérepov Anwerar and kata wolas éyxAioers . . . Anerar) and wépuxe with present infinitive. 1b. 40 I think zés should be repeated before dzoxpotcat, or some equivalent in its place. With xai tpirov «i x.7.A., we cannot carry on the force of the wis above. A few lines below (41) I cannot see why U. R. should alter mas ovk dpewvov to was ed 7) duewov. ov« duewov is a familiar expression, practically = undesirable, to be avoided, better not, and gives just the sense here required in contrast with émrydews. 9. 50 redeia yap av % Adéis Hv. . . 7d 8 perpov NOikyTO Kai ovk dv doxev iy viv exee xdpur. After ov in pérpov has not ay fallen out ? R 2 244 DIONYSIUS 11. 55 ri weipay abriy mapéEopoe pdprupa iv ody olov re diaBddrew rots Kowois raDecw Sporoyousevyv. Schafer expresses SiaBdArAev by reicere, agreeing with the old translation quam arguere potest nemo. This may be right, but 7v may also be the subject and 8. mean, as it sometimes does, deceive. 13. 71 dowep yap Hdetd tis yiveror A€és, obrw yevvata Tis dpa. For dpa U. R. suggest érépa. Perhaps d\dy, as ddAd and dpa. are certainly sometimes confused. 15. 87 peiCov . . . éotar. . . kal ért Bpayeia péver. Obviously jevei. So too probably five lines above. 1b. 89 ob rHv abryy. MSS. vary between od and ore. Perhaps o@ ru. 18. 112 ra yap dvouara Keira Tols mpdypacw ds ervxer. The MSS. have ékxerat and éyxerrar. Perhaps émi- or even cvy-xeirat, as there was probably some reason for the éx and éy. 1b. 118 twép dy érépwOé por Syrotrat capécrepov. dedp7Awrar may be conjectured. ab. 126 In the quotation from Hegesias a man is stripped naked and dragged about the rough ground : wAovjevos 8é kakois wept oAAds tpaxuTyTas Expalev. miAovpevus KaKOts seems possible in itself, but strange in combination with mept ToAAXs Tpaxvtytas. Is it too bold to suggest that it stands for ciAovjevos or even EAxdpevos KaKas 4} 20. 136 dorep drav evOvpdpeba pndiv ddws Huds Tapdrrew pnd mapadumeiv. So the MS. which U. R. call F; others have raparry and wapadvry. I am inclined to suggest rapdrrov and TapaAvrouv. 22. 167 yevouevy should be ywopuévy, I think, as twice above tiv yivouevny and ai ywdpevat DIONYSIUS 245 25. 198 By another trifling change péedAAo. should be written pédXet. 4b. 199 and 203 Is é& dvaraicrwy fvOuav really right, or should it be dvaraiorixav ? ib. 204 Did D. mean to call the De Corona the finest of speeches or the finest of Demosthenes’ speeches? In the second case read dv éyo xpdticrov dmodaivopa: mdvruv doywvr. 26. 214 cis 8¢ déd Statpetv x.7.d. As there is a as just before, it would be neater to write & here. ab. 224 In the last words of the book dv seems entirely out of place, and U. R. ought not to have introduced it. It is easy to see how orovdaiay (sic) grew out of orovdaia. I add a note or two on the other writings. De Imit. 428 oixovouias <évexev> or ? Cf. 430 lines 3 and 7, Otherwise the various genitives seem without construction. Ad Pomp. 1. 750 Probably éx’ (for év) aérats going with StariOepevos. : 4b, 3. 766 ypdior we epi abray 2 ab. 3. 776 The defre. inserted by the edd. should surely be detrerat, if they mean 7s left, remains. ib. 6. 783 Slightly alter the order and read Sypaywyois Te kal oTpatnyois. 4b. 6. 785 Kai pou Soxel was . . . 6 pvOevopevos év “Aidov . eeracpos eri tv exe Sixacrav ovrws dxpiBns etvar as 6 dtd THs Ocomdprov ypadis yuyvdpevos. A pointed sentence has lost its point here by an unlucky accident which does not seem to have been detected. We have to read otras dpi is. Ars Ehet. 1. 1. 225 oikripavrwy tov Oedy 76 dvOpdreov éxirovoy yéros! Otherwise the words are more like poetry than prose. 246 DIONYSIUS 2. 1. 233 Read ye for re after yuyais. 7.6. 277 iva py Kai Yoxp cat odpart ddA TIN padAov Soxbow TH ArTy Kexpyoba. Sauppe may have been right in omitting xai before yxy. In any case should not padXov be pévovt See Demetrius 103 below. 9. 1. 322 dméyer dpOds reve 4b, 5. 331 dédyovow pev ra évavria, mpdrrovow St rd évayria. évavria so repeated is hardly Greek. The first seems an accidental anticipation of the second, representing some such word as mpooyjKovra. 1b. 8. 348 roils 5é Sypors mixpdrepov. It is hard to believe this should not be Sypérais or re dye (made plural by the influence of tots BaciAcdor). D. could scarcely follow the use of Il. 12. 213. de Or. Ant. 1. 446 I incline to think odctas should be oixias, to which dpxew is more suited. So dsocxety woAeus in the corresponding part of the next sentence. The same confusion occurs in MSS. of Lysias and Isocrates. Lysias 3. 459 wepi rav émoroXxav abrot Kat Erat plikay kal Tov GAAwv. Surely épwrixdv, as Sylburg suggested long ago. Well known from the Phaedrus. 4, 462 ds ye (ds re, Gore) ovdey rots Sia yxepds exouor Tov avdpa obre axaipoAoyias ovre dcadetas Sdgav NaPetv. U. R. conjecture and read ddfeey dv for ddfav. défav AopBdvey is a good Greek expression (e.g. Ken. Cyrop. 1. 6. 22 ef 89 weioars eraweiv oe wodXdods Sts Sdéav AdBors), but AapBavew re (undiv) dxaipoAoyias is perhaps more questionable. In any case I would suggest retaining Séfav and inserting something, e.g. ddfav dAaBeiv . Isaeus 4. 592 mrayrés pddiora, ought, one would think, to be either rdvrwv padiora or wavrés padXov. DIONYSIUS 247 Demosth. 2. 956 4 8 érépa desis Hh AuTH Kat apeATs Kal Soxotca Karackevyy te kal ioydy tiv mpos idiudryv exe Adyov Kat duoidryra moAdovs piv exxe Kal dyafous dvdpas mpooraras. As this stands, zpés is unintelligible nor can xatacxevy, icyvs, and épodrys really be coordinated. A little reflection however suggests that after Adyov another word parallel to dpodrnra, probably oikedryra, has fallen out. What D. says of this style is that its affinity and similarity to ordinary speech is its xaracxeuy and icyvs. For oixedrys mpds cf. eg. Ar. Pol. 1262 b 19 tiv oikedryta tiv mpos adAnAovs. 1b. 23. 1026 76 pev ody éxdrdyew . . . ef TL Kdxiorov eipyTat . obk oxipalov: 75 8 é& dudorépwy pdAtora eddoxipovyTwr, ravra wap’ GdAnda Geis eLerdfew Ta Kpeitrw TodTo ofa elvat Sixatov. There seem to be two mistakes here. First xd«iorrov should be xdxtov (like paAdov perhaps for padiota, above). The superlative is too strong, and the confusion of the two forms is by no means uncommon. Secondly 76 8& «.r.d. yields no proper sense or construction. U. R. cite the conjectures zap’ dudorépors and zap’ dudorépw, but these do not really mend matters. I would read 7a 8 ex rév dpdorépuv padiora ebdoxtpotvta, Tatra x.t.A. The participle was probably miswritten under the influence of the genitive preceding it. The nominative 6eés is not strictly grammatical with otro eiva: dikaov, but this irregularity occurs elsewhere. Thuc. 2. 813 xat ov8 otros Huds 6 Aoywpds elon Oev 4 tb. 9. 826 rév apd airod . . . % KaTd Témous pepilovTwv Tas dvaypapas 7} Kara xpdvous edrapaxodovbyTovs. Perhaps the adverb etwapaxoAovbyrws. So in 37. 908 Hppnverpevoy odk ebrapaxoArovbyTws. ib. 51. 940 For the ot7w and rodro of the MSS. I suggest abiré. obrw leaves rowotvres without any proper object. 4b, 52. 942 mapdéopey xa mapefduefa? Future tenses seem called for. THE TREATISE sep) tyovs 7.4 speaks of men dd Siaddpwv éxirpSevudrov Biwv Oyrov jrusdv Aéyuv. It is difficult to find any satisfactory meaning for A\éywy (Rhys Roberts’ languages is not I think possible) and, if it were right, it ought to come earlier in the enumeration with words more akin to it, not after nAucav. I conjecture xpévwv. It is known that Adyos and xedvos are apt to be confused, and belonging to different times is a very good point to make in the context. The agreement of these people is then spoken of as 4 éf dovpddvev os kpiots kal ovyxardbects. ds has been doubted and seems unmeaning. Perhaps we may read ofo-a, as ws and ovs are liable to confusion, or omit as as having grown out of ww. 9.7 qyty pev 8voSapovotcw dadkerar Ay | caxdv 6 @dvaros has all the appearance of a quotation from poetry. 1b. 9 thy rod Ociov Sivapww Kata thy délav éxdpyoe Kasédyve. Toup éyvepwe for éxdpyoe. Perhaps dveyvdpice, av being lost in défar. ab. 10 & ere rod rourod Kai tov évOpwrivuy mapa- Oépevos 4 mép. would be easily lost before rapa, but aepit rod x.7.X. is more natural. ab. 11 “Opunpos pév évOdde odptos cuveumvel ... Setxvvor §? Spws da THs ‘Odvaceias K.7.A. #év seems very much out of its place. Write éOdde piv “Opnpos. Tlept “Tryous 249 4b. 13 od yap ért tots “Ikuaxots exelvors rompacw itcov évrad0a owler tov rovov (“Opnpos). moujpacw is clearly wrong, but rafjyacw (Wilamowitz) is not very probable. I think zvevpaow may be suggested. We have just before rijs pev “Ikiddos ypadopevyns év dxuy mvevpatos: cf. the use of mvedua in 8.4 and 33.5, and cuveutrvet above quoted in § 11 of this chapter. We might also think of vonpacw, comparing voyoets in 8. 1 and 15. 12, évvota in 9. 2 and 3, véyua in 12. 1. 4b. olov broxwpotvros eis éavrév "Oxeavod Kat rept ra Mia perpa. Hpepovpéevov. For pérpo, besides the passage of Aelian quoted by Weiske, cf. 14. 1 ras puxas dvoice: tus pds Ta dvedwro- rotovpeva pérpa and Philostratus Vit. Apollon. 189 & 8 wept thy cehyvyv act daiverbat ... tadra (radra ?) wept tov ‘Oxeavoy ol8a: 74. yap éxeivys dvicot pérpa cuppiviOwv ary Kal cuprdnpovpevos. Aucpovjevov is certainly right for MS. épnuovpevov. Cf, Liban. 1. 32 juepotyruv tiv OdAarrav : Pseudo-Callisth. 1. 1 yijs pérpa xatadaPdpevor, Oardoons kipara kabypepwodmevor: Pausan. 10. 11. 4 ei dééeral ve Arius 76 Boop: A. Gell. 2. 21.2 now fuit et clemens mare. ab. 14 rapeEeBnv & cis 1006’, ds epyy, iva Seta ds eis Aijpov éviore faorov Kata tiv draxpwiy ra peyadopyy mapa- Tpemerat. The writer is arguing (11) that peydAys piccws trodepo- pervs 798y iSiov eoriv év yypa Td PidduvOov and applies this to the Odyssey, where there is an év rots pudddect xal driorois mAdvos (13) and where 7d pudixdv gets the better of 7d mpaxtixdv. He illustrates the eis Ajpov taparpémerar by ofa Ta wept Tov dokov Kal Tovs éx Kipxns cvopopBovpevovs (or év Kéipxns ovopoppovpevovs) «.7.A. No one seems to have noticed the word facrov, and one might almost think that the editors fancied it was facra and went with the verb. Very easy nonsense does not seem to give any proper meaning. It is however not difficult to put right. Ob- serving the context as given in the above quotations and observing also the last letters of the word before facrov, we see in a moment that repdcriov was what the author wrote. He means that the Odyssey is full of répara, and 250 Tlept “Trpous that these mythical wonders are the form that in the decline of genius érixés djjpos (Frogs 1005 rpaysxéds d¥jpos) naturally takes. Strabo 1. 2 repeatedly uses reparoAoyia, etc. of Homer and the epic. For the loss of one re after another cf. 16. 2, where the MS. quotes Demosthenes in the form ody juapte, & dvdpes “AOnvaior, 1.2. ody Hudprere : Babrius 9. 4 érépifev for érepérelev. : 10. 3 riv poxiv, 75 cbpa, Tas dxods, ryv yAGooay, Tas deus, THY xpoay, wavO’ ds GAACTpta Stotxoueva emilyrel Kal... dpa wixerat kaetot, ddoyurret ppovel (7) yap poBetrar 4} rap’ 6Aiyov TéOvnkev) tva. K.T.X. This refers to the description just quoted in the famous lines of Sappho, daiveral poe xijves «.7.A. It can hardly be right that tas dyes should be cut off from ras dxods, which it would naturally follow or precede, by tiv yAdooay. Transpose the two. A more serious difficulty lies in the words, usually put into a parenthesis, H yop «7.4, Without dwelling on other proposed alter- ations I will suggest that 7 and xa/ have, as in many other places (cf. on 44. 9 below), been confused, and that we should read xai yap poPetra. xat—réOvyxev. rap’ ddLyov rébvyxev must be connected with dAoyiore?, and dofeirat with dpovei. dofeira: will refer to something in the later part of the ode which is lost and convey the notion of a reasoning apprehension, a state of mind so comparatively collected that it might seem strange in one who zap’ éALyov tOvyxe. A person almost dead might be thought incapable of rational, thinking, fear. ab. 4 ravi otae dpdov. The sense will be improved if we read wavri <8> ofpat. 13. 2 kai ye rovrou .. . éyadpeOa 70d cxorod. kal ye, kal ye, or something similar? Cf. 14. 1 odxoty Kat Has K.7.X. . > ‘\ ‘ - x , ~ n Ag, ab. 4 ei Ma wept mpwrelwy vi Mia mavrt Jun mpos Opnpov (TlAdrwv), ds dvraywnorys véos pds 75y TeOavpacpévor, tows x 4 Ne \ , 3 2 a“ uJ pev didrovixdrepoy kal oiuvel Siadoparifomevos, odk dvadedds 6 dpws, Sinpirredvero. Tlept “Trpous 251 I do not see how Sadopariééuevos can stand by itself parallel to giAovixdrepov. Possibly another adverb has been lost after xa, We might also get a fair sense by putting és... tebavpacpévoy before or after Sia8oparifdpevos. 15. 4 In the verses from the Phaethon éra 6& pante AtBuxdy aiep’ cicBarav, Kpacw yap typay obk éxwv dvida oh Karo Since what is kérw? It is hard to believe that kdrw Sujoer can really mean will meli it and make it fall. Has not KAT taken the place of K AIO (xaéwv)? cf. Aesch. Pers. 504 préywr yap adyais Naumpds Alou KiKhos pécov mépov Sune. 21. 2 domep ydp, et tis cvvdnoee TOV Ocdvrav Ta odpara, tiv popay attdv adypyrat, ovrws kal 7d 7dBos tard tov o- Sdopuv cat Tov dAAwv rpocOyKady éwrodilopevoy dyavaxrel. dyavaxrei is evidently unsuitable here, some word which will express diminution of energy being wanted. Several such expressions have been suggested, but perhaps ézrax- pale would satisfy the conditions better than any of them. Cf. 9. 15 % daraxpyy tod maGovs. 22. 1 The appearance of dyavaxra a few lines farther on is again not free from difficulty. The words are ds yap oi TO bvre dpyilopevor) PoBovpevor ) dyavakrodvres 7) bd Lydorv- wias i) Om’ dGAAov Twés ... Exdorore mapamirrovres K.T-A., and the difficulty is the occurrence of dyavaxrodyres in addition and as an alternative to dpy:fopevor. If we distinguish between them and say that dyavaxré expresses justifiable indignation, still (1) the distinction is hardly worth making here, and (2) at any rate the words should come side by side, not with oovmevor intervening. As we might expect grief to be mentioned among the emotions, dAyodvres seems not impossible. 24.2 A thing is said to be év TO wapaddyw. Read ev tov mapadoywy. Cf. on 31. 2. 30. 1 ered) pevrot 4 Tod Adyou vonots H Te Ppdors Ta AKL 88 éxarépou duerruKrat, 252 Tlept "Trpous For 3é the editors read 8 after Manutius, but 8’ éxarépov cannot, as seems to be thought, mean each through the other. Perhaps 8é should be ye, most at least of each subject. 31. 1 By an equally minute change I would turn éorw dp’ to gore yap. 1b, 2 raira yap éyyts wapagver Tov iubryv Gdn’ odk iSiwrever TO ONMAVTLKGS. For the last words, which are very obscure, read tév onpavrixav, depending on taira. Cf. the change above 24. 2. 32. 8 rots rovovrous eharTdpacw éemixetpOv duws adTd Kal 6 Kexidwos... defdponoey TO mavtt Avoiay dueivw TAdrwvos dropyvacOa. The Vahlen-Jahn edition gives half a dozen ways of rewriting duws airéd. Simpler than most of them would be dws airod or dAws adTg. If we took the second, the dative €Aatrwpacw would be causal, and this seems to me better. 33. 4 ofpa: Tas peiLovas dperds, ei Kal py ev mao diopa- Ailotey, THY TOD mpwreiov Woy del héperOau. He does not think that the verdict always is, but that it always ought to be, in their favour. Perhaps therefore AE] should be AEI (sciv). The corruption is quite fam- iliar. Or 8dciy may have dropped out after def. Cf. on 35. 4 below (p. 256). There is however also the possibility, supported by many passages, that oipa: dépecba can really mean oipat deiv dépec Oat. 34. 1 ei e dpiOpd, py Td didn det kpivouro Ta karop$wpuara, outws av Kal “Yrrepeibns TO Tavtl mpocxor Anpocbevors. Vahlen-Jahn pi 7G peyée after Pearce. Roberts follows a conjecture of Postgate’s, founded on confusion of dpeOds and dpos, and reads ei 8 dpw py 7d ddnOei, which is very plausible. When however we consider the general meaning, dAnGei so strongly suggests tAjOet (cf. 35. 1) that something like dpOpovpeva ro muvee or dpiOud pdvov Kal ro TAYOe May seem probable. Thuc. 8. 92. 9 the MSS, have wAnOe, the scholiast adyOei. Ilepi ” Yyrous 253 1b. 2 kat yop Aade (“Yrepe’Sys) per’ dbedrclas, 2vOa xpy, Kat ob wdvta EEGs Kal movotrdvas ws b Anuoobérys A€yer, 7d te HOixdv exer pera. yAvKUryTos Ob Autos epyduvduevor ddarot re wept airdy ciow doreiopot, puxtip ToAutiKéTaros, ebyéveta, TO Kata Tas cipwreias edrddaiocTpov, oKweppaTa od« ” a > r x x 3 x 2 ae 3 2 anovta ovd dvaywya Kata Tous AttTiKovs éxeivous agAXr é€mixeimeva, Swuovpyds te érideLios Kal TOAD Td KwpiKdV Kal pera. waidias eboroxou Kévtpov, dpipntoy d& eimety 76 év wact tovros éxadpdditov: oixticacbal te mpoogpvéatauros, eri 88 prbodro0yjoa KEeXvUméevos Kal ev bypG mvetpart SreEodedoar ert evKapTrys AKpus. So runs this difficult passage, following the MS. almost exactly, in the texts of Vahlen-Jahn, Hammer, Roberts, and in their critical notes will be found the attempts that have been made to remove some of the difficulties. There are several suggestions which I should like to make, not indeed with much confidence, but as possibly worth con- sideration. JI will first rewrite certain parts of the passage as I suggest they should stand, and then comment upon them. The changes are spaced so as to correspond with the spaced words as above given. : > 4 ten * , * c f # ob wavra é&js—Kkat wovotdvuas as 6 Anuoobervys A€yet, , 3 ‘ W . * a * , na > - TO TE HOiKoy exew werd yAvKiTyTos vy Aida divas epyduvopevov 1. TKOMpATA OdK dpovea odd’ dydywya Kata Tos Oea- XN > o 3 Xx , 2 om TpeKovs é€xeivous GAG eweKeipmeva, ..... dpipntoy 6 éxetvo To év maar TovTos éradppddirov’ oilKré- cacbai te mpoopvéctaros, eri dé puOodroyyjoa Kexupéevas kal év byp@ mvedpare SeLodedorat, ert edKapTHS akpus. (1) After é&s an adverb similar in meaning to povotovus seems to have been lost. éés itself is not such an adverb, for it means no more than i succession or one thing after another. Moreover ravra and é£js go closely together, form- ing a well-known Greek phrase, which occurs in two or three other places of this book (e.g. § 3 of this chapter), as Toup pointed out on 33. 5, where he was no doubt right in wishing to add some form of zavra. (2) #8y and v} Ava are sometimes confused (cf. Blass ad Dem: 4. 18 and the mistake 7b. 8. 7 Aéyovow idiws for Aéyovot vy A’ ds) and here I think that vy Ada has with the assistance of é@yduvouerov been turned into #bv. v7 254 Tlepi “Trpous Aia is quite suitable and in the writer’s manner. Cf. 13. 4 ei pH wept mpwreiwy vy Aia ... Simpiorevero: 43. 1 twa 88 vy Ala weptéxer Tis Ans a8osdrepa. (8) Itis at first sight tempting to adopt an old suggestion and simply put xara rods ’Artixovs éxeivous after dAAd, but on reflection this will hardly do. You cannot well say of an Attic writer or speaker that he spoke or wrote xard. rods *Arrixovs : xara implies a sort of assimilation or external conformity, not natural membership. Does not also the contrast with Demosthenes render the phrase unsuitable, as suggesting that Hyperides was Attic and Demosthenes not? Diog. Laert. uses @earpixds of a person in the very parallel passage 4. 52 jv 8& wat Oearpixds Kat odds ev TG yedoip (IyeAoiws or Ta TG yedoiw) Siadopycar, optixois dvopact... xpwpevos. Cf. D, Hal. de Vet. Script. 446 of a kind of oratory dd@opytos dvaideiag OeaTpiKky Kat dvdywyos, where the occurrence of dvdywyos as well deserves notice, and Ocatpixds of language repeatedly in the de Comp. Verb. (149, 154, 165, 172): Sidon. Apoll. Ep. 3. 13. 11 allorum quorum sermonibus prostitutis ac theatra- libus nullas habenas, nulla praemittst repagula pudor. (4) xdpw émixeiueva, would be like Eur. Suppl. 716 émixeiwevov kdpa kuveas: Ar. Pax 542 awédas... kvadous mpooxeimevat: Theocr. 23. 14 d8pw—repixeipevos: Plut. Pomp. 51 tiv otparwrikiy Sivamw mepixeiuevos: Appian B.C. 4. 124 rov airév of xivdvvov émixeipevor: Ep. Hebrews 5. 2 wepixerrar doOéveav. This construction of the com- pounds of xetoOa: is rather a favourite in late Greek, and Lucian has at least half a dozen examples. The construc- tion of cvvavamenmAcypéva 7a dotivdera in 20. 1 of this book must be similar, if the Greek is right, but possibly some- thing like éyoyra has been lost. That ydpw is the sort of word wanted appears both from the immediate context and from the contrasted statement farther on about Demo- sthenes, that he is very far from being érixapis. (5) duiuyrov 8 éxeivo is suggested by 28. 4 76 dyiuyrov éxelvo Tov ‘Hpoddrov. (6) xexupévos is probably not to be found elsewhere used quite in this way (cf. however ddidxv7os in 3), and the Tlept “Typrous 255 construction of an infinitive after it is questionable. The adverb on the other hand is right enough. The adverb &kpws which follows, is itself a similar and necessary correction of the MS. dxpos. (7) Finally I have put a comma after diefodetoar, so as to make ér: introduce, as it should, a new point, which he proceeds to illustrate. 1b. 3, 4. The passage about Demosthenes also presents difficulties. & 8 Anpoobérns dvnBorointos, adiayutos, yxiora typos 7} émrderkTLKos, amdvrwv ébqs TaY TpoEpnLevwy KaTd TO mA€ov d&morpos... GAN erevdymep, olwat, TA prev Oatepov KaAd, Kal ef moAAd, Spws dpeyeOn kapdia va govTos apya Kat Tov axpoarny jpepety evra (oddels yoty “Yrepetdny dvaytyveioxwv hoPetrar), 6 d¢ Wev EAdV Tod peyadopvectdrov Kal éx’ GKpov dperas ouvteTeAcopevas, bynyopias Tévov, éuuxa mdOy, aeptov- ciav, dyxivoay, taxos, eva by KUpLov, THy Taow ampdottov Sewédrynta Kat divapuy, éreidy Tatra, pypi, ds OedwepmTa Serva Swpypara (ob yap eiretv Oemitdv dvOpwrwa) dOpda és éavrov goracev, 5a TodTO ols exer Kadols dravras det vixg Kal bmwép Gv ovK éxeu. (1) émSetxrixds may be right. Demosthenes is not érideuxrixds, and émdexrix@s has been used just before of Hyperides. But with the other adjectives here émdeurixds does not consort very well, and I cannot but think émdéétos to be probable. Cf. Sacuppds émideftos in § 2. embeuxrixds might lead to the mistake, if mistake it is. (2) There have been some curious speculations about Kaposia. vypovtos, and appeal might be made to Plut. Mor. 503 F 76 év 7H Kapdia. Tod vypovtos ert Tis yAdrrys éoTt Tot peOvovros. My own conjecture is that it is simply a mistake for «at i8:a vjdovros, belonging to a sober mind. Cf. 9. 11 peyddyns Picews trodepoperys Hoy ididv éotw év yipa 76 pirduvbov : 30. 1 dds yap td dvre iScoy Tod vod Ta KaAa évépata : 32. 4 mdyBous wal téApys petagopor .. . ibid twa dAdcEipdppaxa. In the text of the "A@yvatwy Todcreia 40. 3 as first deciphered and printed by Kenyon xapdia cat xowg was given and was corrected by Bywater to xat i8ia xat xowj. The papyrus has however since been found really to + 256 Tlepi ’ypous give xatda. I imagine an old reader and copyist of this passage to have fallen into the same error as Kenyon. The 7 of the MS. is probably due to that in vjdovros. (3) évOev éAdv is perfectly right. This Homeric phrase (Od. 8. 500) is often used by late writers. To the examples quoted by Toup add Plut. Mor. 578, as corrected by Courier: Philostr. V7. Soph. 529: Lucian Asinus 6: Diog. L. 1. 102 (conj.) and 4. 63 (d6ev éXdv): Charito 1. 7. 6, 5. 7. 10, and 8. 7. 9: Heliodorus 5. 16 (end): Quintus Smyrnaeus 4. 148. It is used to express a man’s seizing an opportunity, starting from something and so plunging into a subject, and it is here intended to bring out how Demosthenes will manifest his great qualities at once. Before rod peyaAopvecrdrov insert rds. kai may only emphasise dxpov, but I think it is a conjunction. (4) It is very clear that xvpiov has no meaning here and that xadpeov should be written for it. Cf. 1. 4 twos xatpiws eevexOev : 43, 3 mapa xaipov éyxararatrépeva. In D. Hal. Lysias 462 dxvpodoylas is now corrected to dxatpodoyias : D. Chrys. 66. 19 xupéov to xacpéiov. (5) The author can hardly have written Oedmeumra Sewd dwpyyara. The word dewvds is inapplicable to a gift. Certain qualities in a man might be called Sed, but not (I think) dwpypara, especially when dewdrys in the characteristic sense has just been ascribed to him in the line preceding. A and A, a, and 6, are so liable to confusion (33. 4. above) that deiva (dévaa) inexhaustible at once suggests itself. Cf. such passages as Plut. Mor. 377 F dd tovrwy 8 tots... dwpov- pevors Hpuiy Kal mapéxovtas dévaa Kal Siapky Oeovs évopioaper : Xen. Cyrop. 4. 2. 44 rotro ds éyw Sod devadrepov piv divair’ dv tov 6ABov ... mapéxew: Plato Laws 966 & dévaov ovciav (being) émépicev: Com. Att. Fragm. Kock 3. 405 yAGrrav ... addy Adywv deivwv. (6) For the unmeaning kui irép dv odk exe I suggest kairep dv ovk éye. ‘ With the gifts he has he surpasses everyone, though there are gifts he lacks. y ec z 3 ‘ c a a 320) 2 x m” : 35. 2 dre fvors ob Tamevov Huds Ldov odd’ ayevves expwwe Tov avOpwrrov. Tlept “Tyrous 257 Perhaps éxpwe rév dvOpwrov determined to make, a use of xpivw which is common in Polybius, and quoted by L. and 8. from Diodorus and N.T. Pearce’s proposal to move yas into the next sentence would certainly relieve this of an awkward word. , tb. 5 éri ray rovovrwry drdvrwv éxeiy’ av etrromer, ds ebmd- pioroy pev avOpwrois Td xper@des 7) Kal dvayKatov, Ouvpacrov 8 Gpws del Td wapadogov. It is not true that the useful or necessary is always easy to provide, nor does such a statement contrast properly with the other, that the unusual or unexpected excites wonder and admiration. Read as 76 eirdpiorov pev dvOpdmois xpeo- des, Or xperddes prev GvOpwrois TO eddpicrov. What is easily obtained, the commonplace, may be useful or even necessary, but what fills us with admiration and wonder is the un- usual, though it may be of no use at all. Clem. Alex. Strom. 4. 149 (6381p) does indeed say tcpey ra Svomopiota od dvaykata, Ta 8 dvayKata cimdpiota yeyevnoOar piraydOus tape Tov Jeov, but such a statement does not make good sense here. 36. 1 7d 8 twos éyyds alpe peyahoppocivys Geod. Why has peyadodpoovvys no article? Perhaps peyado- ppoctvy, Geod depending on éyyis. 38. 1 Should péypr rod be péxpr Tod ? 1b. 4 épets you will ask, a late use. 39. 1 4 dd rOv Adywr air} rod otvOects ought I think to be 4 89 Tav Adywv aitav rod ctvOeors. ab. 4 dN aris ris Stavoias otk eAarrov TH dppovia. méepovyrat. Read éddrrov. Cp. p. 74. 40. 1 cwparoroovpeva 88 rH Kowwvia Kat ere Seopd Tis Gppovias mrepuxherdpeva. Probably 77 dpmovia, and for éri Seoua possibly émdopw. Cf. 41. 3 émiorvvdedeueva. 258 Ilept Trpous 1b. 2 § dpuws should perhaps be Sedyrws. Cf. od Sedvtws in 42.1, 41.2 domwep ra @ddpia Tos dkpoaras ard ToD mpdypaTos 3 ip ee ? zo - apérxet kai ép’ atta Bralerat. Does the writer mean to affirm this of all gSdpia, or has some specific qualifying expression been lost 4 43. 1 iStwrixdy . 4b. 2 od woddal pev kal rodutedcis oTpwpval Kat yavides, * 3 2 a A. a é 4 = J * XQ ae Ta pev dAoupyy, Ta 5 moiiATd, 7a O& AcvKd, ToAAAL S& oKyVal, KT. Can the neuter be justified? or is some word, e.g. iudria missing ? 1b. 4 rév dep Siadepet. rav dcamep Would be the regular expression and ought probably to be restored. Cf. 9.8 7a 60a: 16. 1 and this chapter § 6 rav dca. ab. elmep wavrws éBovdeTo aitdpxy otrws Oetvar. The question is of a man putting something in rather homely detail instead of wrapping it up in a vague dignified expression. adrdpxy is inappropriate. Perhaps the conjecture aira fyrads (or pyr&s ovrws) Ocivat may be hazarded. fyras is explicitly, in so many words. 44. 2 Oppo re yap dacw ixavy ra Ppovypara tov peyado- dpdvuv 4 edevOepia kai éreArrioa: kal dua SieAGeivy To TpdOupov THS Tpos GAARAOvS Epidos Kal... didoriulas. éreArioat may be right (in the sense of encouraging), but T rather suspect it should be éadciya. By a metaphor from the anointing of athletes ddcipw and éradeipu are used in this sense and seem to me slightly more proper here. deAGetv is obviously wrong. It has been altered by various scholars, but not convincingly. d:a/@ew occurs to me, aifw being inflame metaphorically as well as literally, but I do not mention it with any confidence. 4b. 3 rois abris eect. airis would be wrongly placed. Read ravrys. Tlepi Trrous 259 In the same § read probably d@yevoro. xaddéorov, though the article is not absolutely necessary. 4b. 5 of Tvypator, xadovpevor 82 vavo. Sense seems to require of IWvypator cadovpevor dy (or 34 kadovpevor) vavor, 64 being often added to xadovmevos, Neyd- pevos, etc. vavor is the more generic and ordinary name, IIvypato: the special one by which these particular vévor are called. In the Aristotelian Problems 10. 12 however the two words are used convertibly. 1b. 8 roovtwv év xvxAw should, I think, be rowvtrw pév KUKA®. 1b. 9 ededOepdv twa Kpityy Tov peyddwy 7 Sinkdvtwy mpds Tov aidva. Perhaps Singovrwv, ‘things that will endure.’ But he may mean old things, that have already endured, or gener- ically great things that do endure. # seems odd and prob- ably ought to be xaé: cf. on 10. 3. 4b. 11 dAws 8& Saravav epyv civar tov viv yevvwpéevev dvtocwv thy pabupiav. For daravév read rather dardvyy, literally a destruction, consumption, than Sdavov. This sense is established in the verb. DEMETRIUS epi éppuyveias. 4, THE cola of a sentence should not be very long, éreé tot ylyveras dperpos 7} ovvOecrs 7 SvorapaKxodovOyros. As duerpos here means only out of proportion, excessive in amount, we ought perhaps to read xai for 7. The two things go together and are almost one, not alternative. 15 rav 8& ras muKvas repiddous Acydvtwv ov8' af Kepadat fadiws éordow ... of Te dxovovres vavTiicn. There seems no point in ovdé Read ov6’, which con- nects with the re following, as in 19. Perpetual periods, he says, make both speakers dizzy and hearers sick. 25 (kdAa) wapdpo.a Tots éx’ dpyys (in respect of, by virtue of their beginning) .. . 7 as émt réAovs. Radermacher’s citations do not at all support this use of és, for they all refer to something in the mind, while és émi réXovs here is purely objective and matter of fact. Per- haps ds, like 7 in 4, is a mistake for «ai, both confusions being well known. 66 ds “Hpddoros ‘Spdxovres 5€ rov,’ dyoiv, ‘Roav ev TO Kavxdow péyefos, kal wéyeBos Kal rAnOos. The words (which D. gives as an example of repetition, dvadimAwors) are not to be found in Herodotus, and as they stand they are not intelligible. The second difficulty is however easy to surmount. Probably the phrase intended was something like péyefos , cal p. Kal zw. ‘astonishing in size, both in size and in number.’ 95 moet 8& pddiora peyadompéraay Sia 7d olov Wdodors éouxévat, kai pddiora TO Sévus. DEMETRIUS 261 Is not the first pdédiora an accidental and wrong anticipation of the second ? 103 ga yap py pybvta peiLove. daiverar cal trovonbevra paddov. After the downright py pybévra a waddov seems out of place. Perhaps we should read povov. In D. Chrys. 45. 10 the same change has been very plausibly proposed. Cf. p. 246. 116 drav SiOupapBodys ocvvreOy 4 Sitdwots Tod dvdparos. Perhaps SifvpapBwods. Cf. 91 dvpayBicds cvyxeipeva. Such a mistake is common enough. ; 121 éppnvetwy dporoy TH TyAc Rdg ToTapa. Probably zoraydv. ‘One like the river Teleboas’ is hardly possible for ‘a river like the Teleboas.’ 137 drav 76 abrd pnxuvdpevoy dxape yévyTat. As he means not ‘has become’ but ‘becomes,’ this should be yivyrat. In 216 editors correct yuwdpeva to yevopeva. 143 He quotes from some poet unnamed Séorora TAovrwy peAavorreptyur, touti Sewvdov mpo Trepvywv avTd Toincov. apo mrepvywv, which is of course metrically insufficient as well as difficult in meaning, may perhaps represent mpés rav mrepvywv, ‘by thy (or their) wings T adjure thee.’ For the apparently pointless aivé I would suggest the repetition of tobdro. touri shows the fragment tu be due to a comic poet, or at least to a poet of little elevation. 158 gorau... eudatve. Future and present together are unlikely. We might alter either. 164 76 88 yedoloy Kat <8’> dvopdtwv ebreAav? Just above we have had éxdéperar kai 8° évopdrwv KaAdv, and without dud the genitive lacks construction. Al lost after Al. 262 “DEMETRIUS 169 éa pév yap yédwros réxvou Kal xapirwv. ‘Sometimes humour and grace go together.’ This seems imperfectly expressed unless we add something, e.g. reading yéAwros, or . So in 7 I think xai has been omitted before ai Acrai. 226 daiveras seems a mistake for éfaivero. The sense needs that, and in the Platonic MSS. the word is éoéev. 237 kai <éri> rod Paddpidos tov tupavvov éby tis, aS in 218, 236, 285 (by Sauppe’s restoration for éme/), etc. The genitive as in 164, needs a preposition. NOTES ON THE PHILOSTRATI. For the following notes, which are mainly on the Life of Apollonius, I have taken Kayser’s Teubner text (1870) as the foundation, and my references are to the Olearius paging in his margin. 2 empoxyro. There is no reason for the pluperfect. The tense should be imperfect, érnoxeiro, like those before and after. 7 (end) qpoidy 8’ és FAtciav ev i ypdppara. Some verb, ¢.g. éudvOave, seems missing in the relative clause. 9 Birds can be trained to say xaipe etc. odre ciddres 6 Te Aeyouew otbre dtaxeipevor rpds dvOpadrovs. One would think that an adverb must have gone with Siaxeipevor. It might be something significant, like didixds, or something slight and neutral, like zws, which would easily fall out before zpés. 10 wovetrar and dumicyerar ought, I think, to be imper- fects. They have imperfects all about them, and such historical presents do not seem to occur in the book or to be suitable. 17 abrov ye py tov xpovoy ayjpw Te kal d0dvarov Tapa THs pewnpoodvys civat. Evidently rapa tiv pynpootvyy by reason of memory. 4b. av mpos dvdpa iSwow, épvOpidor. Not if they see a man, that is, any man; but only if he is a vir ptetate gravis. Some adjective is missing, or perhaps disguised in zpés, which looks wrong. rpecPurepov ? 264 THE PHILOSTRATI 27 wdda ydp ve AKovoper. Read dxovopev. 54 Some points of physical science may perhaps be studied best on mountains, but neither Athos nor Olympus will help you in moral and theological questions, ci py dtopey aita 7 Wuyxy, Hv ei Kabapa Kai dxypatos abt&v arrouro TOAAG petLov eywy’ dv dainv arrew Tovrovi rod Kavkdoov. The general meaning must be that the mind will go further, higher, than any mountain. drrev, used of rapid motion, the swiftness of thought, is no doubt right enough, though its implied application to the Caucasus rising into the air is a little strange. But ought not petfov to be pacoov, further or higher, not greater? I have suggested a similar change in Xen. Mem. 4. 7. 10, and the confusion of PGANov pacooy and péyoros pyxicros occurs, I think, elsewhere. 66 (end) gorkas .. . peraypade 76 iauPetor. Perhaps peraypdipev. 79 (end) éradav és rod Bactéws wrivwow. év for és? 81 wrdbev ottws exes huvis “EAAdS8os ; This may be right, but it looks as though an adverb governing the genitive, ¢.g. éye/pws, was lost. 83 (end) mAcovwr dé } eyo dpxet kal cidaiuwv 4 xopa Tapas mod THs évtaiba. eddaipovertépa or eddaipwv is required. apa rodv is only by much, and has not in itself any comparative force. 96 ovr’ voonoat... ovr’ av Tpwhels GAyjoa. 104 vomilew “EXAyvixois HOco. Read éeor. Soin 172 we ought perhaps to read adicdv mept 7a Oy, but that is less clear. a na a x ? 105 cidov... olxodvras éxt ris yas Kal ovK én’ airis Kal ie aA drexiotus Teretxirpevous Kal ovdey Kexrnévous 7) TA Tévruw. THE PHILOSTRATI 265 A remarkably clear case, though the editors strangely fail to see it, of 4 substituted by confusion for cai The contradictions must-of course be carried on by oidev kexTypevous kal Ta wavrwv. The same sentence, cidoy x.7.A. occurs with the same uncorrected blunder in 245. Yet in 106 we have the true sense and antithesis given in the phrase 76 8 pndty Kexryucvovs Ta wavrwy éxew dde 6 Aduis éényctrat «.t.A. Through similar corruption and want of perception we have in 194 76 terpyoecOar tov “IoOpsv 7 ob retpyoerOai, in spite of 163 73 rov I. rerpjoccOar Kai ov terunoecOar and of 162 obros 6 abxyiy Tijs ys TeTpyoerat, pardov 8 ov. With these examples before us we need not hesitate much in reading xai yap for 7 yap in 6. 110 riva Oavpaciuitepov Hyp tov émt Tpolay te Kal trip Tpolas eXOdvrwr ; éyd, ey, “AxiAdéa. We should read no doubt riva Gavpacuitarov. In 241, on the contrary, ore EuuBovtdAovs tyuds Biov woinodpevos AKw . mpecBitatos Te tydv . . . adiypevos airds av paddov cixotus EvveBovAcvoy tyiv, it is equally clear that we need mpeo Butepos. 4 ah, > ‘\ » 2 4 ae : 11 pdopa... mpoidvras airods nlavvev évrapaTromevoy TH Gpiry. There is no meaning in évraparropevov. évtarrdpevor, appearing in the ranks? In Plato Laws 797 rax0éra is a variant for rapaxdevra. : S 9 5 . 4b. 1d 8& wopbciv wédAas dotis edkAeeoTEpov HyelTaL TOU a dvouxilew moAw ovK éott. It is difficult to believe that, according to the writer, no one (ok éorw doris) thought this. It has always been the idea of the mass of men. Such separation of ot« éorw from doris is also strange. Is some predicate to ov« gor lost! A. dozen might be suggested. 115 xai Aapmrpivovres adrods éraivors olomep 7a dvOpdroda fnrwrods wéprete. Perhaps olowmep, or (lost after airots) éraivois olomep. 266 THE PHILOSTRATI 116 ods éBovdrAduny dv paddov Nipvyv aird wepiBrdoor véxrapos? meprBAvw is intransitive. 118 (end) éret 88 és dv3pas é€adrAdrres H8y (are growing a man), pedaipeOa Tov dvojrwy Kal edkddwy. edxéAwy is mere nonsense in this context. Probably eireAGv. Menander has éyo 9 dvdyros edrerAs irep- Bod (M. 4. 266: K. 3. 185). 119 (init.) guot 34, not 8é. 140 édoirwy . . , E€vov re adroy Hyovpevoe wat Biov EdpBovdrov Bupav re Spicews kal dyalparwv. By a mistake which is found elsewhere (cf. a suggestion of mine on Herod. 6. 52) #yovpevor has been written, I should say, for motovpevot. Of. 147 iarpoy wotodpevor airov Tod waOous: 171 Eévov re wapa 7G Adi érowwivto . .. Biov re vomoberyy: 178 Hyepovas adtovs rociobat. The word jyovpevor is even less appropriate to kal... dyaAudrwv than to Eévor. 142 dowriay xat rpuvpyv can hardly be predicates of éxetvov. An infinitive to govern them is missing. 145 He said a beautiful city was like the Zeus of Phidias, xajjobar yap airs (the statue)—otrws td Sypcovpyd Woke" rods 5é dvSpas ext mdvra Aeovras pydey daeouévar TOV ‘Opnpeiou Atds, ds x.7.X. The point is the fixity and immobility of the one, the freedom and movement of the other. But otrws. . . éofe gives poor sense, and the words are hardly grammatical, for %oge should be 8d€o1. Both faults are cured together, if we add és after the last letters of otrws and read cabjobat yap otTws 7G Snurovpy@ eBoéke. 146 (end) 4 epi atrois (not atrots) edBovdia. 148 (end) rév re Boivixa tpopéa kat draddv Kal Ta ToLadTa Tysov évopstev. Kayser rimav évouilev, which I have difficulty in trans- lating with the accusatives. It is pretty clear that ripdv dvopatev is what Philostratus wrote. The parts of vouifo and dvopdfw are, it is well known, often confused. THE PHILOSTRATI 267 163 (init.) Write 8€ for re after xopifowro. 169 xpivetoOar, not xpiverOat. 171 vai rw ov, ey. Is epy for épy vis (2) right ? 189 80x pos rév "AroAAdMOV érerxéepOat Td dv. pot is meaningless here, and should perhaps be pé. muotovrat 6¢ x.7.A. will then correspond to it. If, as appears to be the case, murotra. is the writer’s own addition (otherwise we should have mo7oitcOa. and droderev), EvuBaivor must be read for évypBaiver. 193 Nero performing in Greece deatvuy tiv govav Kat Sedtws Tov "HAciov H rov Aedddv, 7) py Seduds dv, xaxds Se ovtws troKpivdpevos THY éEavTOD TéxvyV os pa) pactrydcerOaL vouile pos TovTwv dy ards dpxev TéraKxTal, (Cf. the picture in Suetonius Nero 24 pavidus et metuens ne ob delictum certamine swmmoveretur.) The words as py p. v. seem to give the very opposite sense to what we naturally look for, ‘so badly as not to be scourged.’ Can a dcov have dropped out, as <écov> py p. v., §80 badly as just to escape a scourging’!? J have also thought of ds pepacriyaceo Oat vouiZev, but the sense would be odd. Immediately afterwards rots d¢ "EAAnot tiva, (= érepov) nym & Meurre; wérepa Béepénv xaramirpavta 7 Neépwva adovra; seems to lack a predicative word or phrase to go with tiva yp, ¢g. PoBepdrepov, or wAciw mapéxew mpdypata. 197 A word or words also lost with pets ‘Popaiwy aivtoxparopes, Which is again subject without predicate. So in 206 od rdyres is incomplete. 198 gore re pvOodroyia ; vy At’, etrev 6 Mévurmos, Av ye ot : Me 4 i mowtat éraivovot. Perhaps éracxovot. Of. 1 above and also 3 codias Hv & > a = q . eryoknoey. ématvotcr is senseless. 214 ws iwd yuvaiwy Arrnbets éweddbero rov dpxev adAG. cal rod Liv t 268 THE PHILOSTRATI 216 (end) BeAriov av jv should be B. av noGa. 241 é&v WvOayépov. Surely [vOayépq, unless anything is lost. 251 eet 8& miBavds Suiv Woke roipsv SiaBddrdrcw 7Oos. mifavas or Sia BdrArwv. 274 (end) ‘éyo 66... obra eyvey ovpavod Tpoyeverrépous dorépas,’ SiddoKxwv ore pnd’ dv yévoird te ToD év @ verar py évros. Read dvcera. 4b. kafiévros és adrots tropias, bf? dv Sueoryxecav éxkAyjotaLopery méoXus. The last words, as they stand, are unintelligible. But the last letters of the verb suggest a remedy. Should it not run deoryxecay é. dds 4 287 (init.) édidafav tuas gdyv Modo pymw és Sixas 7 SiaBodras ixaxOeicav. Read traxGeioa. 288 ds wydé should be 6 pydé. oN x \ 28 3 n - ‘ s 310 airs 88 76 Kew drip dvdpav KwduvetcovTa Kal Tovs Backaivovras aire mpdrepov érurydeious érotet. évSpav can hardly be right. Either add something to it or read dAAwv. In Plut. Mor. 603 B dddwv is a mistake for the Homeric dvépav. 323 Swxpdrnv . . . dre epuye THY ypadyv. Read edevye. épuye would mean was acquitted. So in Herotcus 660 read ypapyv exel dv tis, ofuat, pevyor (not vyor) py Tpupav. 324 Hdn pérpet, Bacired, Udwp: et yap Evyxwpyoeas aitg pijxos Aoywv, drayEeu Huds. Cobet 6éAfea for drdyée, but the words are very unlike. By the omission of one letter we can get dd fer, which seems suitable enough ; e.g. cf. Dem. 19. 242 eay ipds amayayn TO Adyo and rots dixacras dmrayayov dard THs trofécews. In Plut. Demetr. 5 TropGeiv Supiav kat ras mores dméyev Kal BidlecOa. Madvig’s dmayyev may be right. THE PHILOSTRATI 269 328 cis py 7d edOd exovons. Rather éfovoys. 359 (end) dxpoacacba. should be dxpodcco Bau. So too 312 (end). I add a very few notes on the other writings. Lives or Sopuists. 479 warépas 8 ob mpocéypaipa, pa Ai’, ob waow, dAAG Tots dx’ edSoxipwv. Punctuate pi Ai’ od, raow. Otherwise od could not well be repeated. 4b. of8a yap dy Kai Kpiriav tov coduotyy obk ex warepwv , GAN “Opnpov 8) pdvov ov Oo marpl éripvyobévra. | 500 Kat of ripavvor 8 aiperwrepor Tots dpxomevots dvetpevor PGANov 7 Euvteivovress ci yap dvycovow, Frrov pev dzro- Krevovow, yTtov Sé¢ Spdoovrai Te Kal dpmdcovrat. Spdcovrae is a vou nihili. Perhaps it represents Siacmdcovrat, p and « being exchanged, as often, and the repetition of aco leading to loss. But Iam not quite sure that diac7ay gives a good meaning here, for its sense is not plunder but rend in pieces, divide, etc. If we could run the two futures into one and read diaprdcovra, we should get a quite appropriate word. dveyevo. ought, I think, to be dveéyevor, matching Evyreivovres. The two forms are often confused. Thus in Clem. Alex. Paedagog. 3. 11. 74 it should be dvepéva (not dvigueva) Ta ToD odparos pedy, to match the perfects xabeevyn, etc. preceding. 576 (end) ‘EAAnomovrw yav ddlynv émBarov radryv ole oot pevelv. Obviously pevetv. 580 (init.) déotvra xatyyopeiy tod piv Anpocbévous Mydio pod, rod d& Aicxivov Pittrmir pod. 270 THE PHILOSTRATI Read Mydicpov .. . Brurmopdv. So in Letter 40 Katyyopet S& kal yipas Tov rpocwrov. 603 éxrjcaro Sé Kal oixias, Sto piv év doret, piav dé ev Tletpare? kai ddAny ’EXevoiv. Before the 8 of 8vo I suspect another 8 =7érrapas has been lost. Soin 510 Adyar & Aicyivov Kar’ éviovs pev Kat térapros it has been pointed out that we have to add a y' =tpeis (Adyou 8’ Aioyivou y: nar’ évious wey Kal Téraptos). x 3 an ae nn A a 605 px daetvas Tod tepod Tovs Oepamevovras. Read @eparedcovras, comparing 703 épnuov trav Gepa- mevodvtov, 116 dre py weracat tov Oeparevaovra, etc. The future is regular. This makes the fifth correction of present to future in these few notes. No error is commoner. Heroicvs, 662 Add dy to kat xdpva Soinv kat pipra Soiyv. Possibly cai kdpva 8 dv cat pw. 8., for the repetition of Sony seems to lack point. 682 (end) mepl pev yep TOU Tdpidos oid doveww afi ovddy . . . , mept O& Tou "Exropos . 4 OUK ay époiunyv yé te oS’ dv dxovoayut xaipwr, ei pn Siamndens ata pyd dpedas Agyous. ei wy) «.7.A. is quite absurd. ‘I should not care to listen, unless you went very fast and carelessly.’ Of course we want the opposite, not unless, but if. It would be awkward to double the uy (ei py Stamydeys), and therefore «i wy seems not unlikely, or «i py 
    
oo. Ee ae

It ie not a bad instance of the carelessness with which
editors follow one another that both Kayser (1870) and
Westermann (Didot 1878) give these words quite
erroneously as a question. The mistake was probably
made in some earlier text, which they reproduce.
THE PHILOSTRATI 271

705 (end) SiSwow tuiv Evpudxous éxardy wédes Kal os TH
Tpofav édciv waifovras. kai is unmeaning. It is some-
times confused with ds, and has perhaps here been repeated
from it by accident: that is, one compendium has been
interpreted twice over, once as xai and again as @s. The
alternative is to suppose an omission, ¢.g. kat 
as. Cf. a little below: rovotrod éopey .. . otot Tpolay pev
éarovdaxétes AaBelv, Kpyrny d& raiLovres.

722 7d dyakpa 7d év "IXiw véov tov “Exropo. kat pepaxiddy
gépet. Probably daive. Sexcenties im épew et aivew
scribae turbant et titubant, says Cobet (D. Hal. p. 112).

726 8cOjva1 8 od should probably be Se6jvar 8 dv, though
the mistake is oftener the other way, dv for ad. But cf.
for instance Thuc. 6. 91. 2 where M has dus 8’ ad for
duos 8’ av.

In the remarks on epistolary style appended to the
letters read at end of p. 364 ta roirw yodv 4 Bpaxvdocyia
Gpatfouro és GAAnY Hy wagay (for taca) orevty odera.
WESTERMANN’S BIOGRAPHI

Homer 1. 6 (p. 3. 61) rév MeAnovyévy seems a gloss on
pu.

Homer 5 (p. 29. 19) 7& Aowra sav cis abrév <éva>depo-
pevev Tompdatwvt The compound verb is always used.

Aratus 2 (p. 56. 7) cwy«pate... Atovvoiy 7d dirooddy
 eis pdovas perabenera |

Antimachus (p. 104. 25) ééxer Wydicpa smerornxévar
should clearly be Soxei. .

Aeschylus 1 (p. 122. 84) ei... XoylLoiro, datrov pev
 trodapBdvor.

Sophocles 1 (p. 127. 17) dcerovyby 8 ev waici cal sept
mudaicrpay. év madi is the the regular (late) expression for
as a boy.

ib. (p. 129. 46) tavrys yap (ris orepdvys) é& dxporddews
kAameions Kar’ dvap “Hpaxdjs yAwoe Sodoxrct, Néywv THV pH
oixotoay oixiay év defia ciovdyts épevvijcat, évOa exexpuTro.

The words rv pi oixodoay oixiay have been a puzzle. év
deEia ciowdvre suggests that some number must have been
given, and with this clue we may perhaps from the latter
syllables of oixotcay get x’ ( = Sexarnv) otcay, the tenth house
as you entered. Then py oc would stand for the name of
the place, and the o. makes one think of Meyapoi, a very
natural place for hiding anything taken from Athens.
Let us then read tiv Meyapot x’ oteay oixiay év dekia. eiovdvrt.

ab. (p. 129.51) gaiverar 8 cat rapa roddois % wpos Tov vidv
*lopavra yevouévy aire dixy roré.
BIOGRAPHI 273

The verbs ¢aivoyat and Pépopa are notoriously apt to get
interchanged (p. 271). We need here dépera, is reported,
ts found, ete.

Euripides (p. 135. 43) abrov pev <éy> ination gag?
vedrepov in 134. 28 should be vedrarov.

Ehinthon (p. 184. 30) 8pdpara 8 abrod Kwpixd TpayeKd
dy’. No doubt xwpixorpayiuxd. The word is not found, but
we know tragicocomoedia from the prologue of the Amphi-
truo, (Cf. Hamlet 2. 2. 415.) So in Frogs 207 it has
been thought that Barpdywy k«ikvwy should be Barpa-

XOKVKVOY,

Thucydides 1. 4 (p. 187. 24) ad odk av edror tis, TL
airG (Militiades) zpds @ovavdidyy ; eott yap otrws Tovrov
ovyyevys' Opaxes «.7.X.

ovk ay eloe tis; (if editors mean it, as I suppose, for a
question) is quite out of place, and is also awkward with
the undoubted question rix.7.A. Surely ddd’ ody, not &AN’
ovx, is to be read. I would also read éor: 8€ for gore ydp.
ydp is inappropriate and it often gets exchanged with de.

ab, 44 (p. 195. 48) &Oev kal r€yomey ds adobevéorepoy
méppactar ddcyov (or kal dd/yov). The subject of méppacrat
may be Book VIII of the History. Understanding it so,
some read xara Adyov in point of language for Kai ddALyov.
Perhaps we should read réppacro: 6 Adyos. Cf. the occas-
ional confusion of év éAcye, évi Adye.

ab. 49 (p. 196. 77) map’ “HpoSdtp cai 6 Sardis eorw 6
gidxKoos kal “Apiwv 6 kvBepvapevos poverty.

Tf this is right, cuBepydpevos is an unknown middle and
its object not expressed. Cannot we do better by a slight
change, 6 iAjKoas Kai “Apiovos kuBepvepevos povorky 4

Plato 2 (p. 391. 10) pera 88 rodro éhotryoe Swxpdrer cal c
érn map aire éroincey, NOuiy pirocodiay éxuabety Bovddpevos.

On éroinoey Westermann’s note is immo édoiryce ; I
should prefer to say tmmo émdvycev. Ta 7pd ToOTwY ToYy-
Gévra comes in the next sentence, but W. may be right
in suggesting roweévra.

<
HIPPOCRATES. [lept dpxatys tyrpurijs.

(Quoted by Ktihlwein’s sections and Littré’s pages.)

1. 570 L. éy woddoion pev Kal otor A€yovot Katapavees cit
M Y
dpapravovtes, wadiora, dé «.7.A.

For xat ofo. K. reads (but see Vol. IT. xvi) xatvotor ote
after Schoene. Surely it was kal  oicr

ibid. tiyy 8 av wavra .. . dtouxeiro.

As there is no dyin the parallel clause preceding, perhaps
rbxn & 84 would be right. Cf. doOeverrépou dé 84 in 5, 582.
In 16. 610 dy is a solecism, and é« rotovrov 84 tpd7ov should
be read.

3. 576 ws yap éracxov moAAd Te Kal Sed K.7.d.

és as an exclamation with a sentence of two or three
lines is such an unusual thing in prose, and also so out of
keeping with the style of this treatise, that we naturally
suspect it. As ws and xai are liable to confusion (see the
Index) read xat yép.

5. 582 (of sick men who could not take solid food or
even gruel) d@ikovro és wéuara, kal tadra, THOt Te KpHoEoL Kat
TO TAO Siadvddccovres ws perpiws Exor, pyre Trew ToY
dedvrwv pare dxpytécrepa mporpepdsuevor pide evdeérrepa..

Why pajre etc. and not otre? If we might read xzpoode-
pouévos, dependent on perpiws éxyou, the words would be
regular enough. If not, we must say that the construc-
tion goes on as though we had not perpiws éxo. but some
verb of which the men were the subject. I hardly think
perpiws éyorey likely. Instead of pjre...mjre...undé the
forms suitable to the sense would be pyre... wydé...prjre
HIPPOCRATES 275

9. 588 zoAAbv yap rod dagadcos dv Ser wepthapBdvovras
dyew eri 76 dcbevéctepov. One good MS. dodadéus-

It does not seem possible to make any sense of this.
We might think of something like rod\bv yap rod  dogaréus av eer repthapBdvovras x.t.X., provided
mepirapBdvovras can mean taking off, taking away, like
meptaipodvtas in Attic.

13. 598 After Enpdév there should be a comma only. In
16. 610 a comma has to be inserted after yévevrat.

14. 600 && Tovtwv was 6 Bios kat byaivovte Kai éx vovoov
avatpepomevy Te Kal KamVOVTL.

dvatpepduevos is supposed to mean convalescent. ‘ Well,
convalescent, and ill’ is an odd expression, and there is no
reason why convalescents should be mentioned at all, ‘in
health and sickness’ being all we want. Read therefore
dvaotpepouevm ‘a man upset, disordered by illness,’ Cf.
Thue, 2. 49. 2 émdre és rHv Kapdiay orypi~at, dvéorpepé re
abtyy.

16. 608 kai qv ye py mavrdnracw rayy TO cGpo.

py Should probably be omitted. But, if it is retained,
then xaf must go. They cannot be right together.

Just below in é& 76 aitG xupiw tiv Sarpy roretoOor
dorep Sceyvypevas read darep for dowep: ‘The same place in
which he stayed when thoroughly chilled.’ Idiom allows
and almost prescribes grep without év.

ab. 612 obdmep 7d fryos kal 4 Wikis veyvixwrdry Kat emt
mAclov évexpovicer.
Compare p. 314 and read wXciorov for metov.

18. 612 dfra 88 radra dre Hde exer ext THvdE TOY oNpELwv
mparov pev ert Ta havepwtepa, Gv K.7.r.

If we are not with one MS. to read Zor for the second
éxi, something like émi 7a avepwrepa  becomes
necessary.

19. 618 (end) ré yap airs dyooper civos Kpjowas aitav
dAAnV mpos GAAnKa. éxovcas Sivapuy ;
T 2
276 HIPPOCRATES

For these unintelligible words I suggest ri ydp airy
djoopey elvar  xpjotas abtav ddAAyv mpos dAAa éxovcas
Sivayww; cf. above in 616 def 8& Symov tatTa alria éxdorov
HryetoOar etvat, dv tapedvtav pty rowovrdrporov yiver Oat avayKn,
petaBadddvruv bé és adAnv kphow wavecOa.

4b. 620 dray récontar Kal év jovyig 7.

‘récoynrat M, waver re A, mut. in récoe re A” K. The
sense seems to require refOyj Te.

20. 620 A€youor S8€ ties intpol Kat codioral ds ovK etn
duvatés k.7.d.  ein?

22. 628 weprodiaOdvor te yap (av) Kal odk éxou eOpyy, eb’ js
weve.

As this is prospective and future, the present pévor is
hardly possible. Read pevot, or pevel.

4b. 632 drav 8 eyxupyoy... Kal... dvrimécy.

Read dyturaioy, as that word occurs a few lines below in
the form zpds 76 dvtimaioy, though dvrurirrev is also quite
legitimate.

The following need no explanation :—

6. 582 «idévar dre <éori> oior. Cf. 10. 591: 16. 610.

7. 584 Read dtvara: for Svvytar. Cf. dawv py edvvato
just before.

12. 596 Read déray for dri dv, and perhaps SeduvqcGat for
Svvac ba.

13. 600 wodAjy  amopinv. So in 17. 612 088 rot7’
 ein.

16. 608 Read xdvy (for kai) xéuvovow, and 20, 622 xéy
(for kai) otat ye.

ibid, Read é6éd« for 2OéAor, as five lines above.

1b, 610 7d8e  841

19. 618 daraddacodpevot S¢ rovTwy (for rovrov), and ovéé
(not od) waverat.
HIPPOCRATES 277

I cannot understand why in 23. 634 (end) K. accepts ay
éxuryoetos in preference to dverirySetos, which the antithesis
of the passage clearly requires, while an dy can easily be
added to ely pddtora, In 13.. 598 aapackevdcacbar is
distinctly preferable to the perfect.

Tlept dépov x.t.A,

7. 26 wept pev mvevpdrwy, d ré dor erirydera Kal dveriri-
Bea, Bde eet.
Read kat  dvemirndac. So just afterwards d ré éore

voowdea xal & byevdtara. This is therefore a stronger case
than Plat. Crat. 391 p w. v. Schanz.

8. 36 Read either éceveyxeiv . . . dvamerpetv, which is per-
haps best, or éveveyxav . . . dvapetpéwv.

11. 50 pare ddppaxoy diddvar exdvta 6 Te es Kowdiny pyre
Tapvew. és KotAinv  or some similar future tense.

20. 74 ob8 vouiLovor dia rHv immacinv 
dws av evedpor dou |

22. 82 tov wAciorov (for 7d wAciarov) Tod xpdvov ?

23. 86 tmép éwurdy Tods Kwdtvous aipedvras is not a Greek
expression. dvaipedvrat is most probable, but aipovra: also
would be right. Cf. p. 301 below.
PASSAGES FROM GREEK ELEGIAC AND LYRIC
POETS.

In the well-known lines of Tyrtaeus quoted by Lycurgus
and beginning reOvdyevos yap xaddv (10 in Bergk) 7-10
describe the condition of the craven who wanders into exile
rather than fight stoutly for his home:

a 7
ex Opbs pev yap Toto peréocerat, ovs Kev tkytat
xpyopocivy 7° elkwy Kal orvyeph tevin,
> ae ft x ee x ™ 2) 7
aicxvver Te yévos, xara 8’ dyAadv €idos eA€yxet,
mace. & drysin kal kaxorns émerau.

What business has peréooerat the future following upon
a present (éor dvinpdrarov) and accompanied by other
presents, alcxwver, éhéyye, éreror? We might turn aicyive
and éAéyxe into futures, but éreras is unmanageable. If
petécoera. is wrong, we might put jerépyerai in its place, in
spite of tyra: following. Cf. Od. 1. 134 strepdiddoioe
perehOdv: 6. 222 xotpyow eirdokdpowce pereAOav.

Though Bergk keeps the two lines at the end of this
piece, dAAd tis .. . Saxav, they seem to be out of place and
to spoil the ending. I do not know whether it has ever
been suggested to transfer them to the very beginning, so
that they should precede teOvdyevar yap kadov. If omitted
by accident, they might probably enough be appended by
the transcriber at the end.

In the ninth line of the poem following in Bergk (dA.’
“‘HpakAfjos k.7.A.) kal Tov devydvtwv te Stwxdvrwv 7° éyéverbe
would seem a probable reading: and in line 17

2 L . ait. sy a
dpyadéov yap dmiaGe petadpevov éore dalfew
dvipos pevdyovros Sniw év woken,
POETAE LYRICI 279

where dpyadéov makes nonsense and Bergk’s own fryadéov
is not very plausible, perhaps Aevyadéov may be suggested,
or dodadews... gore Satlew.

In 5. 4 dud’ airy 8 éudxovr’ I should prefer dud’ airy:
the confusion is a very common one. The MSS. of Strabo
seem to have dudw rds’. Pausanias, who quotes dud’ airy
x7.r., gives also (4. 16. 6) the Messenian distich 76 kai és
Heads ért addpevor,

és Te pecov rediov SrevuKAnpiov és 7’ dpos axpov
eter’ ’Aptoropervns Tots Aaxedatpoviots.

If the author or the people who sang it had any ear, the
second verse ran

rois Aaxedaipoviots eirer’ “Apiotopevys,
the subject being kept to the end as in the distich on the
heroes of Thermopylae,
pupidow tore TH78e Tprakociats edXovTo
éx TleAorovvacou xiAiddes Téropes.

Accuracy apart, would Simonides have written yudt0ddes
téropes Tov Naxedaroviwv? Cf. Arion 1. There need be
no hesitation about altering the order of words when so
well-known a line as xeiueOa Tois Kelvwr pyuact reibopevor

_appears also in the form n.7.«. meOdpevor vouipos (Bergk,
Simonides 92).

There is a quatrain too relating to an incident of the
Messenian wars, and quoted by Pausanias 4. 22. 7 (Bergk
Carm. Pop. 28 gives it in his note on the last), which must,
I think, be faulty, though Polybius 4. 33 has it in the
same form:

mdvtws 6 xpovos nupe Sikny adikw BacrAqi,
nope 6& Meconvys civ Aut rov rpoddrny
pyidiws: xademov Oe x.t.d.

The repetition of yipe in a really different sense is so awk-
ward that we may reasonably alter it in the second line to
elde (cf. Soph. Hil. 528 4 yap Aixy vw cthev and many other
passages), or better perhaps alter rév zpoddryy to 76 mpodédry.
There is of course no objection to 6 ypdvos nipe tov mpoddrny
in itself (cf. Soph. O. 7. 1213 épnipé o° dxov@ 6 wdv6’ dpav
xpdvos, Solon 4. 29, etc.): it is its. combination with ybpe
Skyy which makes it questionable.
280 POETAE LYRICI
Theognidea 95 :

ToLovrds Tot éraipos dvyp pidros ovte par’ écOdés,
ds kK ely yAwoon Ada, ppovy 8 Erepa.

Aga, besides its doubtful form, does not seem a very
proper word here. I conjecture A¢eia ‘smooth things.’ Cf.
852 ds Tov Eraipov padOaka kwridAuy é€ararav bere and 365
yAdoon 8 75 peidixoy aity éréoOw: Aesch. P. V. 647 zapy-
yopovy deiouor pois: Solon ap. [Ar.] AO. Mod. 12. 3 xwrid-
Aovra Aciws, ete.

In the couplet (151-2)
UBpw, Kupve, Oeds mpGrov xaxov dracev avOpi,
ov peAXrer xopyy pndeulav Oduevat,
perhaps we should restore dpyv ob pedAAa. If the words
got out of order, dpyv would be corrected to ydépyv. A
similar change has been suggested in 1066 rovrwy otd& rot
GAN’ env reprvétepov (oddé to. TovTwv), and seems pretty
certain: and in 831 miore. xpijpar’ ddeooa, dmiotin §
éodwoa I should suggest riore: dAcooa ypnpat’.

Line 424

ToAAdKt yop Td KaKoy KaTaKeievoy evoov djeLvoy,

éabdov 8 ééeAObv Adtov 7H TO KaKdv
is pronounced by Bergk ‘versus corruptus. Poeta videtur
dixisse: bonwm, quod divulgatum, plus nocwit quam
malum.’ That however would be a monstrously untrue
meaning. I suggest éoOAdov § éfeAOov Awtov H Kdxtov, ‘does
more good than harm.’ Kaxdy was written by mistake and
an article then put in to eke out the verse. Cf. Hippocr.
Lipid. 3. 4 fv 8 ratra poBepwrepa 7 Kaxiw, ‘more alarming
than serious.’

475 foll. The author tells us how much wine he has
drunk :
abrap éyd—pérpov yap éxw pedinddos olvor—
Uarvov AvotkdKkov pvyjcopat olkad’ iwy,
néo 8 ds olvos xapteoraros avSpt renda ba
ove TL yap vadw, ore Ainv peOiw.
But Athenaeus 428 p in his quotation has #xw, and that
is right. “Hxetv is used of being, or having come to be, in a
POETAE LYRICI 281

certain condition. This is familiar to everyone in the
phrases ed jxew, KadGs Hew Twos, etc.: but we find it also
in cases where its meaning is not always recognised. Thus
Soph. O.T. 1519 Oeois y’ &xOtoros Hew: OC. 1177 &xOiorov,
Gvak, bbéypa rod’ He marpi: 1b. 1266 Kal paprupO Kdkirros
avOpwruv tpopais tais catow grew (see Jebb). In Theognis
és olvos k.7.A. Shows that this is the meaning: he is just in
the state which (to use Hamlet’s word) is the most
‘ gracious.’

1007 foll.

Evvoy 8) avOparois trobjoopuat, ddpa tis HENS
dydabv dvOos éxwv kat ppeciv écOA4. voy,
TOV abTOD KTEdvwY cv TaTxEpEV.

Who ever used a genitive in this way after «b rdcyew?
According to Liddell and Scott Pindar did, for he wrote
(N. 1. 44) otk epapar wodty év peydpw mAotrov Kataxpuats
éxew, GAA’ édvrwv eb re Tabeiv Kai dxotoa dPidrov eapKéwv.
But there édvrwy is a genitive absolute, ‘if’ or ‘when’ I
have the money (xpyudrwy understood from zdodros
perhaps: cf. Ar. Plut. 503 évres tAovrotat wovypoi, | ddikws
aira €vdAdeEduevor). How then are we to deal with
Theognis? The answer is obvious. We are to write «
twice instead of once and read rév atrod ‘x Kredvev ed
TAC Xe[Lev,

MSS. of Theognis show just the same error in other
places. In 577 pyiov e& dyafod Ocivar kaxov 7 *k KaKod
éa6dov the ’x or éx is omitted by one MS., and in 431 doris
cudpov’ One tov dppova Kak xaKxod éoOAdv many have kal
kaxou or Kat kaxéy. Cf. above on Ar. Ach. 525.

In the thirteenth poem or fragment of Solon I have
three or four changes to propose. (a) He says

mAovrov 8) bv pev ddr Ocoi, wapaylyverat dvdpt
10 éumedos ék vedrov Tubpevos eis Kopupyv’
bv 8 dvépes Tysdow bh’ JBpios, od KaTa Kdapov
épxerat, GAA’ ddixous épypact meOdpevos
ovx éOéAwy ererau.

TeOéuevos:in 12 should I think be reopevors. Cf. 4. 11
mAourotow 8 ddixors épypact weOduevor and Theogn. 380
dvOpdruv adixors Epypace wefoxévev. Indeed it is not easy
282 POETAE LYRICI

to see how wealth could be said d8ikows gpypact reiPeoOat.
For one dative depending on another, épyaci on teBopeévors,
ef, 4, 22 doru | tpvxerat ev cvvddots Tois ddtKotor pidats, where
I would make no greater change than to read é8éxour as in
34.

(8) In 18 foll. he draws a simile from the wind which,
after laying the fields waste,

ie x. n oe 2% ef.
Sywoas kaha Epya, Gedy eos aimty ixdver
> # 2 if. 7 = mW 2 ~
odpavdv, aidpinv 8’ adbis eyxev iSetv:
Ade. 8 HeALoto pévos Kata tiova yatay
rs Lan ae , 2Qx Y 9 2 Vo ag A
Kaddv, atap vepewv ovder ér’ éoriv idety.
Idetv cannot be right atthe end of both pentameters. It
is not however of much use to suggest a specific alteration,
as we cannot say which idety is wrong.

(y) A little further on (43) he is speaking of the pursuit
of wealth :
6 pey Kata wévrov GAGrat
ev vyvolv, xpylwv otkade KépSos ayeww,
ixOvdevt’, dvémorrt hopedpevos dpyadéouow.

Scholars should have seen that, as the text stands, the
epithet ix@vdevra could not follow its substantive wévrov at
so great a distance. There is only one thing which would
render that possible, namely that ix6vdevra should not
stand alone, but be fortified by the addition of one or more
further epithets applied to révrov. When we have got as
far as this, it is easy to see that we should read dopevpevov
not qopeduevos. But can the sea be said ¢opeicbar?
Semonides of Amorgos thought so, for he wrote (7. 40)
moAAdKis 58 piaiverat | Bapvxtiroit Kipacw qopovpérn
(OdAacoa).

(8) In 65—6=Theogn. 585-6

mace d€ To Kivduvos ér epypacww, odd€ Tis ofdev
7 pAAG oxnoEY xpnpatos apxopévov,
is there not some awkwardness in cyyoew ( = reAevTHoeLWv)
referring to the man, while dpyeoa: refers to the business ?
The same subject should be said to begin and end, not the
business to begin and the man to end. Buchholz actually
POETAE LYRICI 283

understands ypjya as the subject of wéAdet cxjoev. But
Solon wrote dpydpevos.

There is an error in the poem on the ten ages of man
(Bergk 27). Dividing our life into ten periods of seven
years, Solon says of the sixth,

Ne ‘ , , , » ,
TM? $ €KTY) EPL TOAVTA KATAPTVETAL VOOS avipés,

but this is no more true of the years from 35 to 42 than of
those preceding. In them too a man’s mind is being
formed and moulded. The truth about the sixth age is
that the mind then is, not is being, formed: katjprurat,
not xataprverat. So Plato, as though to show us how to
restore Solon’s verse, says (Laws 808 D) mais exe. ryyiy rot
gpovely otrw .atyptupevnv. It is no objection to this
change the+ uccording to Solon a man is at his best voty
kal yAéooay from 42 to 56. Karjprvras does not neces-
sarily mean quite his best, only fully formed.

Some readers must have asked why Archtlochus (54
Bergk) should speak of a cloud standing straight or straight
up (6p66v) about a headland as a sign of bad weather :

Babds yap 78y Kipacw. tapdocerat
f z * ao” t 3 ‘ ¢ -
movros, appt & axpa Tupéwy éphdv totaras védos,
ORL0. XELLBVOS.

As applied to a cloud under these circumstances, ép@dv
seems devoid of meaning. The poet must have written
vobpév, dull heavy clouds, and v was lost after the y in
Tupéwv. ,

A more trifling error in the text of Archilochus is :

68 padyys 8¢ tis ofs, dore Supewy mreiv,
ds épéw.

Auéov to Archilochus would almost certainly be a
disyllable, as épéw is. Read perhaps dere ris dupewv.

Semonides of Amorgos in his first fragment paints a
gloomy picture of human life. We live without knowledge
of the future, like so many animals. “EAmis 6& qdyras
kdmureOetn tpéper ampyxtov dppaivovras: we go on blindly
trusting and struggling and failing. Our schemes are cut
short by age and illness and death, perhaps by suicide.
284 POETAE LYRICI

otTw Kakdy dm’ obdév' BANG pupiat
Bporotor kijpes xdverigpacrot Svat
\ - as ve 29) 2 s ,
kal mya’ éoriv: ei 8 éuot riBotaro,
> a ~ + io” 23 an
ov« dy Kaxov épGuev ob8' én’ adXyeow
Kaxots €xovres Gupov aixtLoiweba.

Kaxois and éxovres have been called in question, but no
one seems to have doubted aixfo/uefa, though Ahrens and
Nauck were perhaps on the way towards doing so when
they conjectured éxdvres. Yet aixiLoiuefa is by no means a
proper word. Shakespeare can say ‘As flies to wanton
boys, are we to the gods. They kill us for their sport.’
But would any Greek of Semonides’ time have said that
men aixi~ovras by heaven? It must be by heaven, if at
all, for no other meaning can be got out of the passive verb.
The saying that men are the playthings of the gods (Plato
Laws 644d, Plaut. Capt. prol. 23, etc.) is much less
strong. :

‘What is wanted is something that repeats the point of
xaxav épgpev. The poet is complaining of men’s folly in
clinging to life and making an ado about things, as though
anything mattered. The things on which we set our hearts
are only vanity and vexation of spirit. We know this, if
we would only consent to look facts in the face; but we
dissemble, we make believe that the things of this world
can certainly be had and are worth having. Now this
absurd make-believe, this affected ignorance of ours, can be
very well expressed in Greek by a word differing from
aixifoiueba. in one letter only, namely dxxtfoiuefa. So
Plato says in Gorgias 497 A oic6a GAN axxife. So Cicero
writes to Atticus (11. 19. 5) certi swmus perisse omnia:
quid enim axxilopeda tam diu? Reading dxxfoiueba, we
need not, I think, seek to alter the rest of the verse, unless
we think xaxots weak after xaxdv. “Eyovres Oupdy is
probably used in the sense of ‘setting our hearts upon’ a
thing: cf. Ouuds éore with an infinitive. Anything like
ovres Ovjdv (Meineke Fragm. Com. Graec, 4. 717) is seen
to be unnecessary and indeed inappropriate.

Semonides’ meaning is perfectly expressed in the fine
lines of Dryden (Aurengzebe tv. 1):
POETAE LYRICI 285

When I consider life, ’tis all a cheat ;
Yet, fool'd with hope, men favour the decett,
Trust on, and think tomorrow will repay :
Tomorrow’s falser than the former day.

* * * * * *
I'm tir’d with waiting for this chimick gold,
Which fools us young, and beggars us when old.

A minute change should be made in the second quotation
which Athenaeus 37 4 makes from Panyasis :

otvos Ovyrotct Oeav mapa Sapov dpioroy,
> / * a XX 2 , > #
dyAads, @ macat piv ebapydfovow dodal, «.7.X.
Read éyAadvy, remembering Homer’s dyAad Sepa, and
observing in Athenaeus’ first quotation just before :

otvos yap wupt too értxOovioow dveap,
éoOdov, aAcEiKakov, K.T.A,

I come to two of the epigrams ascribed to Plato. It is
remarkable that a very obvious blunder has not been
detected in the epigram on Archeanassa, which Bergk
numbers 30. It occurs in nearly the same form in
Athenaeus and in Diogenes Laertius, and Bergk writes
it thus:

"Apxedvaccay éxw THY ék Kodopivos éraipyy,
qs Kal ext puridwy mixpos ereoriv pws.

& Serol vedrytos dravrncavres exeivns
mpwtot)dov, . dans HAGerE wupKains.

Two things lead us to suspect dmavrycavres: first the
construction, for dravrav does not take a genitive, secondly
the comparative weakness and colourlessness of the word.
As soon as our suspicion is aroused, we see of course that
the author of the lines wrote déravOjoavres or perhaps dzav-
Giccavres ‘culled the flower.’ After thinking of this, I
found it most conclusively confirmed by the other and
quite different form in which the third line appears in the
Anthology (7. 217):

& véov HBys avOos drodpépavtes epactai.
’AmavOery seems not to occur elsewhere except in the

neuter sense of ‘ ceasing to flower,’ ‘ fading,’ and dav6i~w is
“to pluck a flower.’
286 POETAE LYRICI

But, it will be said, what a dreadful mixture of meta-
phors! a flower, a first voyage, and a conflagration, all in a
couplet! Can anyone have written so badly? As it happens,
mpwromAdov is the reading in Diogenes and it also appears
as a variant in the Anthology, while the first hand in the
Anthology gives rpwro8dAov, and Athenaeus zpwrorépov.
Now that we have got dzavOjoarres, there is no difficulty
in seeing that zpwro8ddov is the right word. Another
epigram in the Anthology (5. 123), ascribed to Philodemus,
speaks of Bérpus 6 rapOeviovs rpwroBoddy xdpitas (cf. ib, 61.
6), and zpwroPorciv of ‘ budding’ is also quoted from the
Septuagint. Thus dravOjoayres and zpwrofddov confirm one
another. It is a pleasure to save the unknown author from
the discredit of mixing three metaphors together, but I
fear we must still allow that he mixed two. The water
and fire of the common reading, dvres éyOicro. 15 mpiy,
were however more offensive.

There is another epigram ascribed to Plato(29), about which
I wish to say a word, though not to offer any emendation.
It is the well-known couplet on Aristophanes, of which
Bergk says dignum praeconium non minus eo qut laudatur
quam qui edrxit, showing that he, like some others, still
believes this epigram to be Plato’s, while giving up many of
the rest:

Ai Xdpures téuevds te haBeiv dep ody weceirat
Cyrodcat Wuxnv yipov ’Apirtopdvous.

Short as it is, it contains one thing, if not two, which
points to a late origin. First réuevos weceiras is a dubious
phrase.  {mos épya. Thucydides
seems not to use this double érepos in the same way
elsewhere, for expressions like 2. 51. 4 érepos ad’ érépov
Gepareias dvamipmAdpevor = a dx’ é\AyAwv are different ; but it
is by no means uncommon in Greek. Cf. Eur. Ale. 893
avpdopa 8 érépous érépa miefer: Ar. Pol. 1. 8.1256 a 35 of 8
dé Onpas CGou cal Onpas erepor érépas. In Pol. 4. 9. 1108b
34 mdrepov erepa Kal tadta Oeréov 7 Tots airois dmodoréov
dudw editors have assumed a similar omission and proposed
to read <€érépois> érepa.

3. 11. 4 & 7G atrG Se Kat Ta Kpdticta éri te Tods
trodeeorépous mparous Evverqyov Kal Ta TeAcvTaia AuTdvTes TOD
dAAov mepippypevov dabevéorepa euedAXov eetv,
298 VARIA

The article with reAevraia, which must be predicative,
seems grammatically indefensible. Kriiger conjectured
réde for rd. I should rather suggest ai-rd. av might
easily be lost after ac.

4.36. 3 Kal of Aaxedayudvice BadAduevol te dporépwhev
Hoy Kal yuyvopevor ev 7G adtG EvpTTdpart, ds pixpov peydrAg
cixdoat, TO ev Oepporidats——éxeivol te yap TH atparg
mepteADdvrwv Trav Tepody StepOdpyoay, ovroi Te dupiBoror qdy
vres obkért dvretyov, GANG ToAAOls te dALyou paxdpevor Kat
dobeveig cwpdruy bia THY oLTddetay brexdpovr.

So no doubt the passage should be arranged, if it is
what Thucydides wrote, and it is now supported by the
Oxyrhynchus papyrus. Thucydides, we must suppose, forgot
that the words oro re «.7.4. were part of a parenthesis,
and went on with adAAd... treywpovv as though they were
the main sentence. This may have been the case. Yet
in spite of the papyrus I would just suggest for con-
sideration a minute change which may save the author’s
credit. ’AAAd and dua are often confused: probably in
8. 69. 1. Read dua here, and the passage will run thus:
ylyvdpevot ev TG dro Evptrdpate TG ev OcpporvaAais (éxcivol
Te yop TH atpare repieAOdvTwy Tov Lleprav SuepOapyoay obrol
te dudiBoro. H5y bvres ovKérs dvTetxov) Gua Toddois Te dALyot
paxduevor kat dcbeveln cupdrov Sia tiv oirddetav trexepovv.
The word dua emphasises the combination of two causes
that made the Lacedaemonians give way. They were
much outnumbered and at the same time they were much
enfeebled.

* *
*

Xen. Anab. 1. 8. 18 tats domion rps 7a Sdpara eéovrnoav.
Did they really strike their shields upon their spears,
and not rather their spears upon their shields, rots dépace
mpos tas doidas as in Plut. Humenes 14 (end) rds re
donidas aveihovro kal tats capicas émidoumypcavres HAdAaLav
See index under ‘Terminations interchanged’ and
Xenophon and Others, p. 303. A trifling example of the
same error is Simplicius on the Physics (Brandis 4. 404b
11) paprupodvros wepi trav mpatov Kal @codpdorov ypdipavros
VARIA 299

Ebony mept Twos a’tod Tov Suppaptypevwv dytrypdpuv ze

‘ irép év, oyciv, éréoretAas, where Evdnpov sone aite is
plainly required. Read also Ss ypdavros, or ypdyavros yap.
* *

*

Aristotle Bhet. 3. 6. 140849 oiov 76 pdvon tiv odAmeyya

elvat peAos aAvpov.

For civa: read tévac as in Plato Phil. 51D ras & 7 kaBapev
ietoas pédos: Laws 812D dda... edn tev xopdav ted.
A trumpet can be said to emit, but not to be, a strain of
music.

Rhetoric 3. 15. 1416 a 21 ddXos (tpdros) ei dAXot é, éurrapa-
AapBdvovrar, ods dporoyodar , wy évdxous elvar 7H SiaBodAH, otov
ei drt KaOdpros 6 jrorxds, Kal 6 deiva dpa.

Another way of rebutting a charge against a man is to
show that it rests upon some reasoning which, if admitted,
would warrant a similar charge against some other man
notoriously not liable to it. Instead therefore of the un-
meaning words in the text read «i dri xaddptos 6 
poxds, kal 6 Seiva dpa: ‘if, because A is a dandy, he is a
rake, then B must be so too.’

Xenophanes ap. Aristotle Rhet. 1377 a 20 otk ton mpé-
kAnots atty doeBet mpds cio Bij.

By reading dvoocBei for doeBet we get a good trochaic
tetrameter. So in Bacchae 263, where our MSS. give rijs
eioeBetas and the Christus Patiens rijs do alcOdverat. The loss
of yevera: will be due to homoeoteleuton, The use of
300 VARIA

yedous in the next sentence points to the probable occur-
rence of the word in this, and Aristotle has it several
times in the same connection with dys and doy. For a
somewhat similar omission cf. Magna Moralia 2. 7. 1204b 8,
dd rod dxotoat cal  dodpavOjvat, where iseiv kat
is Susemihl’s tolerably certain restoration from the words
that come a few lines further, éri dé ye rod iSeiy Kal dxotoat
kal éodpavOyjvar and from the fact that ideiy and dxodca: so
constantly go together.

Problem. 33. 7 1a. ri rov pey mrappov Oeov Tryotpeba elvat,
mH 5& Bixa 4 q THv Képutav ov ; H Store ex Tob Oeordrov Tav
wept yuas, THS Kepadrs, dOev 6 Aoyiopds eort, yiyveras ;

Did the Greeks regard sneezing as a god? They called
it ‘a bird,’ Aristophanes tells us, but to personify it as a god
was more in the Roman way. Read Oeiov, which is indeed
indicated by rod Geordrov. In 33.9 and in Athen. 66¢
the word iepds is used of a sneeze.

[Aristotle] *AO. TloA. 3. 3 ot évvéa, dpxovres dpuvvovow
domep ert 'Akdorou Ta dpkia Trommoety.

So the editors, but the letters of 7a dpxia appear to be
very uncertain in the papyrus. The phrase would be an
odd one and doubtful Greek. Can it be ra Sfxata ?

4b. 5. 2 Solon is quoted as writing about the civil strife
at Athens :

yryvdokw, Kal pot dpevos évoobev dAyea Ketrat,
mpeaBurdray & écopay yotav "Taovias
cawonerny (Blass: xaphouevyy Diels).

It is difficult to believe that xawouévy is a word that
could be applied to a country. If the other letters were
more rightly read by Blass than by Diels, is not pawoperyv
what Solon wrote? If Aeschylus could make y#, meaning
really its people, the subject of such words as orévw (Pers.
548), aidtw (1b. 922), wo6& (Ag. 545) and Sophocles «Alu
(0.7. 48), ériorapo: (O.C. 1006), there would seem no
objection to y7 pawopéry. O.T. 636 yijs otrw vocovons is
very similar, and still more so the metaphorical vdaos of
Ant. 1015 yore? rods. yi xatvouevy is different,

* *
*
VARIA 301

In the well-known fragment ascribed for a long time to
Dicaearchus (Miiller, #.H.G. 2. 258. § 14) the writer
describes the Thebans as Opaceis cal tBpiorai kat irepjgpavor
TAnKTal Te Kal Gdidpopot mpos wavra Edvov kal Syudrnv, where
ddiddopo. is understood, I imagine, in the unproved sense
of making no distinction. No doubt the original word
was evdidropor (from diadepecba) quarrelsome. It is need-
less to illustrate anything so common as this confusion of
ev- and d. I will only point out two cases in which it
does not appear to have been corrected hitherto.

Diodorus 13. 23. 4 od yap Suvardy rods dAdois dvnepws
Xpyocapevous adrois map’ érépwy Tuxelvy mote pidavOpwrias,
GAAG tpdéavras Seva tabeiv ebyvepova. This contresens has
been dealt with in various ways, but to my mind it is
clear that we should read dyvdpova.

Again among the fragments above mentioned, collected
by Mullach, we read (1. 489, 22) év pev ‘tats pébous mapor-
votow, év 5& tals dtuxias wapavootow of avéyro. Common
sense would suggest ciruxiais, even if we did not read two
pages before (487, 54) év wey rots cupmocios 6 pa) twapowdv
HdvUTEpos, ev SE rots dyaois 6 wy wapavoyiv. But the former
passage also gives us a correction of the latter, tapavodv
for zapavoydv. This seems better than reading zapavo-
povo. in the former. Foolish men are often intoxicated by
good luck, but there is no reason for saying that they
proceed to break the law.

* *
*

Theopompus (1) Oxyrhynchus Papyri v. 177. 10-16 a
Boeotian party try to get up a war with Sparta, reckoning
that Persia will find money, and that Corinth, Argos, and
Athens will take part: rovrovs yap éxOpots tois Aaxedai-
poviois évras avrois ovprapacKevdcev tods woditas. Read
rods SrAitas, 6wAtras TapacKevdlew corresponding closely to
xpyjpara mapéxewv. Similar confusion of the two words
occurs elsewhere, ¢.g. Xen. Hell. 6. 2.22: 7.2.7: 7. 2. 20
twice: Diod. 18. 38, 1 and 52. 1.

ab, 145. 17 e8i8ackov adrots dru peyav aipodvras Kivdvvoy is
a mistake for dvatpotvra: (ay lost after péy-av) or possibly
aipovrai, cf, p. 277.
302 VARIA

To fill gaps in the papyrus I suggest as a possibility in
col. 2. 8 Sqpaywyot i in its neutral sense: 3, 2-3  veds
and riv pev i vaiv: 13. 14 tiv wodw 814 rxov and 16-17 o . . . roduy, though the MS.
woAv is certainly casinet. this. In 2, 23 rév vicwy Tues
trav éx’' "A@nvaios obaGv ought not éx’ to be ix’? It is ure,
not éri, that expresses regular subordination, The editors’
yiveoOau in 2. 2 should surely be yevéoGat.

* *
*

Herondas 7. 102 xiv réocapds pot Aapetxovs trdoxyrau.

As a spondee in the fourth foot is not possible, may not
Aapexovs be a gloss which has ousted the real word,
rofdéras? We know from Plut. Ages. 15 that the Daric was

sometimes called rogérys from its stamp: rod Teporxod
vopicpatos xépayp.a Tokoray & EXOVTOS, dvaevyviwv | épn pupiows
Togdrais éEehadiverOat tis Actas. In 3.7 xal yop ovd’ azrap-
xedow al dorpayddat, Aaurpicke, I think it certain that, as
I and others suggested when the papyrus was first published,
dotpaydAa is a gloss on Sopxades, the name given them in
line 63 rijot Sopxdow wéprew dotpaBda (cf. SopxadiSes in 19),
though Blass and Wilamowitz actually thought that
Herondas wrote here a trochee (do7pa-).

4b. 4.12 tod adéxropos Todd’ Grr’ oixins Tofywy | Kipuca
Adu.

I still think there is a considerable chance that Herondas
wrote not toiywy, which has offended many scholars as
quite superfluous, but tpyyvv, a word constantly used of

voices.
* *

*
Diodorus 14. 62. 2 Staxdci pev yap Kxal revryKovra
paxpal vats eioérdeov..., pera 88 Tatra al gopryyol vais
> i = c + 7 / * Z nw
cicedpevar pev vrép tpioxiAias, pépovoat be mAciovs Tov
mevraxooiwy, al 6& waco oxeddv SioyiAta
I will not attempt to deal with the arithmetic of this

passage, but I think it is possible to correct cicOeduevat
with fair certainty, though Vogel’s critical notes offer
VARIA 303

nothing more satisfactory than ‘cicéfecov xevai Sintenis et
Madvig, cis moAcuov (vel Ayeva) Warm, oicduevar Stroth.’
What lurks under it is surely cis dpiOyov, a regular
expression for in number.

* *
*

Plutarch Mor, 2p oi 8 dpydrevot tpoxol rove KappOevres
BO” x ” f ‘ 2 2 fo’ bu; ro9 x n vO t
ovd’ ay ef te yevorto rHv e€ apxis OvvawT’ dvaAaBeiv cdOvwpiav.

Read tépvw, comparing Bacchae 1066 xuprés tpoxds répyve
ypapdpevos (though the meaning there is disputed), etc.

[Steele in Spectator 364, paragraph 7 This must certainly
be a most charming exercise to the mind that 1s rightly
turn'd for it. This may be right, but more than the turn
of mind is in question ; past reading is certainly necessary.
Did he write twn’d? There are other misprints still in the
Spectator. |

* ¥
*

Appian Celt. 8 zpociec be TG epyy.

Neither zpooieoOe nor mpocridecbe, which is suggested,
can mean apply yourselves to. Probably wpdcxeoOe was

the word.

* *
*

Athenaeus 5070 Suxpdrys... evirviov edyoev éwpaxévar
mAedvev mapdvtwy' Soxely yap épy tov TAdtrwva xopevyv
yevopevoy ent tiv Kepadyy pov dvarndyoayra 75 padakpdv pov
kaTackapipay Kal kpadlew mepiBrAcrovear.

Kaibel omits pov in both places as ab epttomatore
additum. A much safer and more obvious change is to
substitute éddxovv for doxeiv. It is the regular word in
telling a dream, e.g. Ar. Vesp. 15 éSdxovv derov karamTopevov
KT.

Posidonius ap. Athenaeus 234, after mentioning the
condemnation of Gylippus for embezzlement, adds tod be
dvariBepevou Ged kal ovyXwpoupevov Sijpou Kabdarep Koo p.y-
patos Kat KTNPATOS (t.€. gold) ov pasiov 7 nv vov Ovytov 6X
ywpor yevér Bat.
304 VARIA

In this passage djpov seems quite unmeaning, and, if it
were right, some word like idfa would have been introduced
later on to balance it as 6vnrév balances 6G. It is probably
just a mistake for Syrov or 57.

* *
*

Lucian Somn. 9 obdév yap ort py epydrys ever TG oopare
movav Kav tovTw tiv admacav éAmida rot Biov rteOepevos,
adavys piv abtés dv, dAlya kai dyevvi AapBavev K.7.dr.

For AapBdver read pav6dvev. The word AawPavw is used
but little of mental acquisitions ; chiefly of material gain,
on which so much stress ought not to be laid at this point.
Nor would éyevv7 AapBdvwv be a proper phrase in that
sense: a man AapBdve é& dyevrav, not dyew7 AapBave.
With dAcya xai ayevvq pavOdver is contrasted in 10 especially
Kal dAws druvra 6rdc0 éoti 7a Te Oca Ta 7 GvOpwrva ovK és
paxpdv oe dvddgopat.

Vit. Auct. 14

ON. ri yap 6 aldv éort;
HPAK. sais raiLwv, wecoctwv, Siapepdpevos.

In the words ais taifwv tecoevwy Heraclitus is of course
quoting himself: fragm. 52 Diels (79 Bywater) aidv zats
éort wailwv meaoevwv. When we notice how in the context
both before and after he is made to use terms in pairs,
usually antithetic (répyis dreppin, yraois dyvwodn, etc., and
again Oeol Ovyroi, dvOpwro. aOdvarot), and when we call to
mind fragm. 10 (Byw. 59) dda Kat ody dda, cupdepdpevov
Siadepdpevov, cvvadov Siadov, does it not seem likely that we
should read here in Lucian  dtadepdpevos 1
The MS. which Sommerbrodt calls y seems actually to
have cuppepdpevos written above diahepdpevos.

In a similar way I suspect an omission in § 5 xai
cewutov eva Soxéovra Kat GAAov Gpedpevov dAAov édvra eloeat.
The words dAAov .. . édvra are complete in themselves and
éva. Soxéovra only confuses them. Did not Lucian write
something like eva Soxéovtra ? So just
below dAAarropevov és popdas mrelovas. Cf. Theaet. 166 8B
(do you think he will admit) tov clvai rwa add’ odyi ToVs;
VARIA 305

Vera Hist. 1. 29 (end) padraxds évdiddvros tot rvevpatos
kal ovvitdvovtos emt THv Oddarray KareTeéOnpev.

‘Let down’ is what we want, not ‘laid down.’ Read
xabetOnpev from xaOinut, a form corrupted in consequence of
its rarity: indeed parts of tyuu (especially kainus, pebinun)
are often altered to parts of riOnui. Cf. § 9 KabijKev eis 76
aéAayos, where there is the v.1. xatéOyxev, and Book 2. 2 76
mAotov ... KabyKayev és tHv Oddarrav, which has escaped
corruption: 1b. 43 Cobet has already corrected xaraévres
tiv vaiv to Kabévres. (In Luc. Pisce. 47 and 51 I suspect
that the aorists xajxey and kaOjxa are a mistake for the
perfects xaeixey and xaGeixa, due again to comparative
rarity of form.)

Gallus 1. The cock says to its owner gpyv te xapteto bal
cou mpodapBdvav tis vuKTos Srdcov av Suvaiunv pOovav.

For the meaningless ¢fovév, omitted by Dindorf and
perhaps partly due to d@ovepdv preceding, read dwvav. CE.
éEvguvos above and ddwvdrepos below.

Ehet. Praec, 22 jv pév tis xadas etry... qv 88 perpius
evexOp, «.T-d.
Read dAcyOq for evex Oy.

Lucian (?) Amores 44, speaking of an ordinary boy’s life,
describes how he gets up and goes out in the morning rj
iepay xAapvoa Tals érwpuiors wepdvats ovppdpas. Why should
his chlamys be called iepd? What it no doubt really was
is éped, woollen, as Plato Crat. 389 speaks of a indriov
Awodv } épeoty and in Polzt. more than once of éoOijs épea.

* *
*

Aelian Var. Hist. 9. 3. Kyetros 8 efrore péAdow riot
xenuaritew, eri roppupov cipdruv BadiLwv rods Seopevovs
mpociero.

Certainly Clitus did not come to meet his petitioners,
nor did they find him walking about. For BadiZwv read
xabitwv. @ and «x are often confused, and in Pherecr. jr.
36 MSS. vary between éBadifopev and éxabifoper.

* *
*
306 VARIA

Clem. Al. Strom. 5. 3. 17 (654 P) dyvojoas tis efjrycev
kal Lytycas evpioxe: tov SiddoKadov eipav te éxiorevoev Kal
morevoas wAmioev dyamioas te évredOev eEopoitrar tw
Hyamnpevyy.

Stihlin in the notes to his recent text does not mention
that doubt has ever been thrown upon Amey, but I can
hardly believe that the very obvious jydryoev has not
been suggested. The trick of the sentence requires it, and
yAmicey seems deficient in meaning.

* *
*

Aspasius in Eth. Nicom. 4. 14. (Heylbut p. 125. 24),
Expounding the character of Aristotle’s PBwpoddxos as
adyTws TOD yeAotov oroxalduevos, Aspasius says ciot rives of
TO Taparay ovk diode. xpyoOa. TH watdia Kal ovr adrot dy
elrouey TrAov Tt br’ GANwV dxovceay ay ydews.

wAéov 71 makes no sense and is probably a mistake for
yéAody mu. y and w are liable to confusion, as in the
familiar case of tp@ypa and ypdppa.

* *
*

Heliodorus the paraphrast of the Nicomachean Ethics,
writing on Hth. 8. 5, says (Heylbut p. 169. 37) cuvdrres 88
Tovs didovs Td TOs adrots xaipew Kal rots adtovs direiv. Sad
rovro yap & Swxpdrys pidet tov TWAdtwva, ot. tov Swxpary
dpdwo dtrodor Kal trois Swxpdrovs ayadois dugw yxaipover.
Socrates loving Plato because both love Socrates is
nonsense, and it is odd that Heylbut did not see this or,
seeing it, could not put it right. For 6 Swxpdrys read
5 Iooxparns. Cobet in Novae Lectiones p. 677 mentions
two or three instances of the same blunder and _ has
corrected it also in Aelian V. H. 2. 36.

* *
+

Two Pythagorean comparisons (Mullach 1, p. 489), pre-
served to us by Stobaeus, admit of very obvious correction.

32 ovre ra Tod ’AxtAdEws drAa TH Oepoiry otre Ta Tis
Puxijs dyaha TE dppove dppdrre.
VARIA 307

ays Woxjs Should be ris t¥yys. How can a fool have the
goods of the mind? The confusion of yvx7%, rvxn is quite
well known, e.g. Isocr. 12. 9. In a fragment ascribed to
Democritus (Mullach 1, p. 341) dvonpoves purpodyra: rotor
THS Wuxns Képdeoe’ of be Tov Todvde Sanpoves Tolar THS coins
I think the correction riyys has been made.

35 otre wip ipariy mepioteiAar Suvatov otre aicypdv
dpdpTnpa xpovw.
~ The confusion of ypdvos and Adyos is equally well
established. Read Ady for xpdéve, comparing the saying
(Mullach 1, p. 501) 7a dpuoptypard cov repd pH Adyors
emixadvrrew aAAG Oepamevew éhéyxois.

* *
*

Platonius de Diff. Charact. 6 8 ’Apirropdvyns tov pécov
€Ajdaxe Tov dvdpGv xapaxrhpa, intermediate, that is, between
Cratinus and Eupolis. Though something like pécov or
péonv 6d6v éAatve would be intelligible, I am inclined to
think that pécov é\avvew xapaxripa is not right and that
for é\jAaxe we should read éfyAwxe, a word constantly

used in these writings.

* *
*

Vita Aeschylt rorAAG xarerdtepov Hv emt Odomid. Bpvixw
te kat Xoupidw eis roodvde peyOous THv Tpaywdiay mpoayayetv
H em AiocytAw eimdvta eis rHv Zodoxhéous eAOety TrederdTyTA.

In spite of one or two other passages (e.g. schol. Ar. Ach.
378) I cannot believe that ciety was thus used of
dramatists. Here émidvra is a very obvious conjecture,
coming after Aeschylus. Cf. Isocr. 4. 74 yaderov éorw
torarov érehOdvra A€ye wept Tpaypdaruv TéAGL TpoEAnPpEevev.
The two verbs occur together as variants Plato Rep. 4004
(Schneider) and Aristot. Post. An. 93 16.4 (Waitz). Twice
at least in Plutarch’s Moralia, 405Fr and 1108 Ff, éridy is
necessarily corrected to eizav.

* *
*

Stob. Flor. 4.107 Zrjvwv 8 &by yeAouov Exdorovs (éxdorov)
pev rots mpdypaow as Set Liv py mpoodyxew ds ovk eiddTwv, Tov
83 rapa wdvrwv eraivov Gavpdlew ds éxduevov xpicews.

x 2
308 VARIA

I do not know what sense readers have attached to
mpdypaow. Read zpoordypacw, the confusion of zpdrrw
and apoordrrw being well known (e.g. Ar. Hth. 5. 2. 1130b
23: Plato Rep. 407p: Lys. 14. 20 and 21, ete. 25. 13:
Isocr. 2, 34). We must adopt the variant éxdorov to
which ciéérwy refers loosely, unless éxdctwy was the

original.
* *
*

Suidas 8.v. Scdpwv' trois 8& xpdvois Fv Kara Bépénv cai
Eiperidyv.

Not only is Xerxes an odd person to date a Sicilian
author by, but he does not at all harmonise with Euripides.
The floruit of the two would be very different. Sophron
too was almost certainly later than Xerxes. Can we find
any Greek writer or thinker of about the same date as
Euripides, whose name might conceivably get distorted
into Xerxes? Anaxagoras seems to satisfy these condi-
tions. If xar’ ’Avagayopay (a followed by a consonant four
times running) became by accident xaragayépay, it would
be a fairly easy step to xara HépEnv. By a similar process
probably the absurd word ovveédpyew in Aaxiochus 368 D
(od yap edaiverd pot cepvov patvonery Siw cvvegdpxetv) was
produced out of cuveéapapreiv, ap having been lost
before ap.
SILVA CRITICA

Tue following is a very miscellaneous collection of
emendations which have accumulated in a notebook
during a good many years and which relate to all sorts of
books and periods from the Iliad to Eudocia’s Violarium.

. In the order of passages I have had no regard to chronology,
as that matters nothing to the present purpose. But by
some degree of system and arrangement I have tried to
make the silva rather less of a jungle than it was at first.
There are certain prose writers, Aristotle for instance,
Thucydides, Diodorus, Lucian, about whom I have many
suggestions as yet unpublished to make, and in whom
I could have found further examples of errors here
illustrated ; but I have reserved them for future discussion
and put hardly anything from them into the following
series.

One of the conimonest of mistakes is the substitution of

a present tense for a future.
Sometimes it is only a matter of accent, and of this

“pévo for pevd is a frequent example. Polyb. 19. 11. 2

meTecpevwy ... pevey Tors bmdrovs should be pevety: they
were persuaded they would remain. Arrian An. 5. 10. 1
TaUTHY THY @pav .. . mpoopevew . . . packer, ci ev TO TdTE

eipyo.ro: the optative in itself points clearly to rpocpeveiv.
Anacreontea 19.7 kav Avorn S€ tus abrdv, odk Eero, pever O€: it is
all future—if released, he will not depart but still stay. Cf.14.
19,20. In Aesch. Hum. 497 vead 7dGea mpoopevel, to harmonise
with two futures before it; in 894 nat d7 dédeypau ris S€ por
TYby jrevel; Seems very probable (as does the gfe conjectured
in 892): cf. Eur. Med. 386: Hel. 1059: Androm. 334.
310 SILVA CRITICA

The schol. Ar. Nub. 441 must be morevOjvar airév eupevety,
and érayyeANdpevos Sropeveiv: the paraphrast to Ar. Hth.
in Heylbut p. 194. 17 peveiy: Sext. Empir. adv. Math,
9. 62 Siapevet. In Eudocia 903 there are three futures
corrupted together, épxovs dre 8) Sexaeriav pévovor xal
orépyovat tods . . . vopous kai... 7d TyviKdSe Avovor, for we
must read pevotot.. . orépfover .. . AVoovcn.

Again where the o of the future has been lost—[Ar.]
Oecon. 2. 23. 1350 a 33 rév orparwrav ... mpds Tovs irevav-
tious dackévrwy daoropeverGar: Polyb. 4. 74, 6 otk ay
iyrépnoav gévuv Kal picboddpuv tov . . . TapepedpevdvTwv
(probably), and 29. 8. 7 wéupew ey... xdxet peorrever Oar :
Philo de Plant. Noé 2. 176 (Wendland) ovk dpa pebier 6
doreios (the future is constantly used of what the philosopher
etc. will do); or where the future y and é has become ¢
and x or y: Plut. Mor. 3503 Aioyivys cximrwv tov Anpo-
oben Aéyew dyotv Stu ypddera.. . diadixaciav, where we
have Aeschines’ own words 1n Ctes. 146 diadtxaciay ey
ypdwew. I suspect Alexander said (Plut. Alex. 31) ot
KAapw tHv vikyv, not ob kAérrw: cf. the version in Curt.
Ruf. 4. 13. 8 (47) meae gloriae ... furtum noctis obstare
non patiar. Read also in Eudocia 373 od xatadeipw tov
visv povev. D. Hal. A.B. 5. 29. 1 read ‘kal xaropOwoavre
xat Siapaprévee THs éAmidos drodavely bmrdpéa por, tt will be in
my power, and at the beginning of Anacreontea 16 ypade
pot Babvddov otrw tov Eraipov ds Siddéw (not dddoxw), as I
am about to tell you; cf. 15. 4 ypdde rHv enn éralpny
dxeotoav as av dw. Perhaps édcyéovrwy and éAdyée in
Plut. Mor. 4118 Seicba: pérpwv . . éAeyxdvrwv, where
éXéy£ovor follows, and Callim. 59. 3 oi8 eAaf?’ eEéracw 7H
@uxéos aris éhéyxer Tov pidov, and dvepéotoys in Porph. de
Abstin. 1. 7 obdev rpooedenfycay daddys aitias THs dveipyovoys
airots: cf. 1b. 9 éSdovTo Tov KwAdvcovros. Polyaenus
certainly wrote 2. 3. 1 ds airika dvaorpéyovom, and no
doubt Hac. 18. 6 ypyooper, not xpylouer, as logic and the
Anpopeba following require: the future ypyow is cited from
the so-called Timaeus of Locri, and Herodotus has ypjcat,
xpyoas. In Sext. Empir. adv. Math. 1. 196 write wepirpeovrar
and Pyrrh. Hyp. 1. 186 eiodéet.

Plut. Mor. 189 & éxéAeve ras olkias movciv aad mpiovos Kai
medéxews povov' aicytver Oar (read aicyvveirOat) yap «is oixias
SILVA CRITICA 311

Auras éxrdpata kal oTpdpata Kal tpamélas woduredcls ciope-
povras: Euseb, P.H. 14, 21. 2 érépou Set rod xpivovros (read
kpwodvros, comparing passages with ScicOa: and dzropelv
above): Sext, Empir. adv. Math. 10. 82 ef pi... héporro,
od oupBddrra (read cupBaret: indeed éora: immediately
follows), and 1. 192 «i 8 ob Svcyepaivovow (read dvcyxepa-
votow) AAG. . . cuprepipépowTo Tois Aeyopévols, Kal Hpets
atrots émyevotpev: Arrian Cyn. 19 égorw 6 dyyeAdv:
Athen. 4164 rudopévov... drt dard rotrwv épydforro, «ime
TlepoGv dvapibpyrovs ovykdpev we plainly need the
corresponding future épydcoiro and Anon. Vita Platonis
(Didot p. 5. 44) épyocev ravtas dvOpdrovs orovddoew .. «
pydéva d& Suvjoecba, aA’ Exacrov... roteloGar the other
futures call for zoujoeoOa.

D. Hal. A.B. 10. 4. 3 pérerrs should be peréora: and
Aelian V.H. 13. 3 7@ avoifavrt. . . odk eorat (for éorww)
dpevoy. There are various places in better Greek where
this correction has not yet been made, eg. Ar. Met. 7. 6.
1045 a 17 where éorat should be read, corresponding with
écovrat in the next clause.

Far less common in MSS. is the reverse substitution of
future for present. In D. Hal. A.B. 10. 17. 5 xwdvvetoopey
ody efew is not good Greek : we must read either xwdvvevouev
or otk exe. Plut. Mor. 8388 dapevos eis tiv émotoav Héew
read ye, for ddpevos here means bidding them.

Futures are written for aorists, eg. Arrian An. 6. 6. 5
as otk av diapvAdéovres, a very, common blunder: Euseb.
P.H. 7.1.3 rapaornoopey . . . éride(Eovres, where the sense
is I will prove that we may legitimately borrow from
barbarians by showing (or when I have shown) that the
best Greek philosophers borrowed from them too; and just
in the same way 15. 62. 17 émwoxemréov dd’ érépas Aoywv
apyis ... ryv SSacxadiav royoopevors. In Theopomp.(?) fr.
344 Seredoovras is presumably a misprint for dededoavras.

Aorists for futures. Arrian An. 4. 4. 3 od« edn
dmodeiéacOar: 15. 4 épacke . . . xataotpeacbar: 29. 1
Hyjoac0a: pdokovres: 5. 27. 2 mefoas pev afew ys, Tacbeis
8 ob Bidoacbar: Pollux 8. 86 duvvov.. . cvpprddéev Tors
vopous Kat pi SwpodoKycew 7 xpvoodv dvSpidvta dzorica.
Anthol, 9. 189. 3 &Oere wpds réuevos.. . &Oa, Kaddv orjoacbe
312 SILVA CRITICA

ef xopév should no doubt be orjoecbe; cf. the futures with
évéa Soph. Hi. 436, Aj. 659 and 1166.

Presents are sometimes written for imperfects, especially
Soxet for éSdxer: Philochorus fr. 79B apoxetporovet pev 6
Shuos ... eb Soxet 7d Gatpaxov ciodéeperv’ Gre Se Soxei,
edpdocero caviow 4 d:yopa Kal xatedeirovto x.7.r., Where a
number of past tenses shows what is plain enough even
without them, that we must write zpovxepordver. . . ei
eSdxer and Gre & edéxeu (or possibly Soxoty): Plut. Mor. 832F
(of Antiphon) zaparéraxey éws katadAtvoews THs SyuoKxparias
. . + w abrés Soxel ovyxatackevdoa, where éddxer was
thought seems probable: 7b. 868D ef éwiyeper... dp’ ovdx
dv aicyiora Kodaxevew ofe read émexeiper and schol. Ar.
Vesp. 1018 éexovpe: in otk éx Tod havepod, pyowy, émtxoupet 6
mourns Tos Ocatais iviv, where the text refers to the past.

Imperfect for present in Westermann Vit. Script. 104.
25 éddxee (read Soxel) 8’ "Avripaxos otros pydiopa, reronkévan,

Plut. Mor. 182, a slave having been enfranchised, ovx
av éBovdAduny (said his master) é&va ’A@yvaiov (-wv!) tn’ eéuod
pacrryodcba. Read pepacriyaoba, perfect for present, as
any flogging must surely have been then in the past.

Exchange of verbal terminations.

-ew -a. Euseb. P.H. 2. 6. 11 read ruyydver: TMept Kops.
8. 26 (Bergk) éxew for éyex: Hermogenes (Walz 3. 383. 13)
dpoupet.

-ev-e. Sext. Empir. adv. Math. 1. 203 read jxev for jee
and 7.77 et... cvpBeByxe for ci... cvumPéBynxev and again
el... cupBeByxe for ci... cvpBeByxe (cf. the confusion in
Laches 1834).

-ev -wv. Aesop 252 (Halm) ei cuwdpoveiv yap Svoxdbexros
ef Adwy, | THs viv pavels od peorda Spaces Saxpvwv; write
cudpovay answering to pavels.

-ev -v. Pind. Nem. 1. 27 apdooe yap épyw pev obevos,
Bovdaior 8¢ ppiv éoodpevov rpoidev. For the infinitive, which
is very awkward, read zpoidev.

-e-w. Anacreontea 17. 6 8dre 8 dvOéwv éxeivou | orepdvovs
olots muxdlo | ra pérwra py ‘mixalev | Td8e Kadpa.  olorts
wuxdtw cannot bear a final sense, with which to shade, and
its ordinary sense would be very flat. Read then wuaZer,
SILVA CRITICA 313

the subject being he (éxeivos), 7.e. Bacchus and uty rixafew 7..
depending on dére oreddvovs.

-ojnev -oupev -wpev, -dueba -dpeba. D. Hal. AR. 7, 52. 2
el 88 Pjoet, Se€drw cat pybev err Seducha Adyov. junbév shows
Sedpeba needed. Plut. Mor. 219 cvyywpotmey ’AdcEdvdpw
av Oédy Oeds yevérOar should be written cvyxwpdyey and
perhaps «dv or even 6G, for 6.y. seems to depend less
naturally on 6éAy. Hermogenes (Walz’ 3. 200. 15) read
Aéywper.

a. -av. Arrian An. 7. 29. 1 perayvaval ye éd’ vis
erAnmpednce pov olda tev réda Baciiéwy ’AdeEdvdpw trdp£ar
(read imdpéav) bd yevvaidrytos.

-at-et? ~=Plut. Mor.'2240¢ dcov yap ein rpaypa, Torotrov
(rocotros) Kat 6 Adyos @ xppoa. xpyoe? xpyoa’ dv?

-a-a. Sext. Empir. adv. Math. 1. 261 read dévicra for
aviord.

. -m-ean. Cebes 3 ci pev otv abtd ovvin tis, éowfero: «i 88
py ouvin, amwdero. Probably cvveiy in both places and
certainly drw#Advto.

-yre -erre. pay adyavaxrnre should not be imputed to
Polyaenetus (Hc. 31. 14) any more than py fyrqre to
Andoc. (2. 25) or wy ypvfns to Aristoph. (Plut. 598). Read
dyavaxteire OY d&yavaxTyionte.

-ov -wvra. Anon. in Ar. Eth, N. (Heylbut p. 408. 3)
6 Opacis exer Tov Oupdv drdkrws Spudv ois od Set «.7.A. Read
bppavra (like eye. thy émGupiay pérovcay in line 1) or possibly
épyavra (410. 9, 12, etc., but there of érufupia).

A very common mistake is confusion of the comparative
and superlative terminations of adjectives and adverbs.
Diogenes (Mullach 2. 302) épwrnbels rota ef Onpia xadera-
repo, ele x.7.d., and again (4b. 304) épwrndels ri yj Bapvrepov
Baordle ey ‘évOpwrov dzraisevrov.’ The form of question is
very familiar in such anecdotes and it is of course in the
superlative (yadetdtara, Bapvratov). Soin Plut. Mor. 1840,
asin Pyrrh.9, Wvppov of viol aides dvres Hpdtwv tive karadeiper
tiv Baorrelay: kat 6 Wuppos eirev ‘bs dv Spav dkurépay exy
Thy paxapav’: 217D épwrnbels ré Sewdrepov SéSpaxev ev 7G
Biw (ct. 229D 6 7 dvopdrarov épyov ev ro Bi wéxpaxras).
Almost equally clear in another way is 2b. 93 F rév Lowy 76
prdrexvoy tots povotdKots icxupdrepov éudverat. Polyaenus
314 SILVA CRITICA

Hac. 14. 22 sods é& trav aixyaddtrwv édaxlorovs Kal
dobevertépovs: Aesop 200> (Halm) tiv dpaidrepov mévrwv :
Didymus (1) ap. Stob. Eel. 2. '7 (Wachsm. 2. 48. 9) ratra
yap éore rév epi abris évvonudtwv dadovotepa: Strabo
5. 2. 5 the people of Pisa paxiudrepor Tuppyvov
trjpgav: Hermogenes (Walz 3. 383. 18) xaradhavécrepov
pevro. 6 A€yw yiyverat .7..; all these are to be corrected.

When Achilles (J. 9. 392), refusing to wed any daughter
of Agamemnon, says A. may choose someone else, ds tus of
7 éréoixe Kai 3s BactAetrepds éorwv, it does not really make
very good sense to understand god with Bactedrepos and
take it as said ironically. The sentence is rather of the
type of which we have two or three instances, all referring
to marriage: Od. 19. 528 érwpa ’"Ayadv ds tts dpioros |
pvarat; 20. 335 yypacd’ ds tis dpiotos dvip Kat wAciora
wopyow : 21. 162 yypac8’ os Ke wAciota mépou Kal pdpoyos
€\$or. Do not these parallels point to BactAevraros?

Not so frequent is the reverse corruption of -repos to
-tatos. But Euseb. P.H. 14. 3. 2 has it twice, for Moses
was certainly not wdyrwy ray “EAAjWvwy zpecBiratos, and
toratos ... Tav mpiv ‘EGpaiwy is probably a similar mistake.

Kindred corruptions are :

A. mAc(i)wv mreioros.  Aclian V.H. 2. 23 of po Fwy
ert rhelorov peuviprbat avrod: read éri mAéov say any more.
D. Chrys. 63. 5 daopav 82 él wAéov rédos darahyjoas KT... :
read émi mhelorov after long hesitation : 67. 6. woAddaus yap av
meious peraBohas ev Hypa pid peraBdddorro : read wAcioras
ever so many.

B. rpdrepos zpdros. Hermogenes (Walz 3. 214. 13) ras
dyribécas mpitas Tov Avoewy TiHdvar: read mporépas, and so
(tb. 211.17) py rd8e rovrou mp&rov (xpdrepov) etrety, ei xal Ty
rage. mpatov fv. Synesius 47 B rov "EAAyva rot BapBdpov
mparov (mpdrepov) ayw. 2

C. paddov pddiora. Stob. Flor. 98. 68 Bpaxis yap dvrws
& Bios, 4% 8% réyvy paxpd, Kal paddov (pddiora) Wy Tas THS
Wouxyjs vécous idcacbar Suvapevy: 1b. 24. 14 ris yap av Frrov
poBoird re} Oapooin padiora (uaAdov) } Gores K.T.d.

Confusion of case endings.

-os -ov. Schol. Ar. Hq. 320 daPBory tov "AOnvaiwy et «at
SotAo0s rap’ .adrois & Sijwos dmayyeAAerat: read dovAov... 6
SILVA CRITICA 315

Sijpos the deme of a slave, the slave having spoken of his
Oyporat.

-os -ov. Schol. Ar. Vesp. 1383 érel wapyvecey aitd 6 vids
mporepov Supynparixds (-Kév?) efvat.

-o -ov. Heliodorus ad Hih. N. (Heylbut 44. 5) a man
may not know 76 airod rpdcwrov Tod mpos Sv Tis ToLel 5 TroLel,
7d mpaypa b motel, eviore Sé Kal 7d Sia Tivos K.7.A. Read 1d
aitot tpécwmov, 7d mpds Sv «.7.A. In the abstract a man
may not know rév mpérrovra (Aristotle says), though in this
case himself (wis yap éavrov ye;). adrov certainly seems
wrong, and rod zpos dv questionable for Heliodorus: ro is
also suggested by the parallel clauses.

-o -w. Michael Ephesius ad Hth. N. (beginning) dv
éxdorn mpos TO (TG) Stxouocivy A€yeoOar Kat idiws 7H pev ppd-
vyois, 4 dé dvopeta, 4 Se cwppoovvy A€yerat: Teles (Hense
p.. 34: Stob. Flor. 95. 21) obxére otros rpds 76 (7a) rovely }
tnreiv ri éorw, a regular use of zpds, and so Plut. Mor.
6125 pds TotTo ‘yevdpevos Should be wpés rovTw: schol. Ar.
Ach. 398 Eipurisys évrad6a voetras 15 (7G) airds, is meant by
abrés: Alcid. Soph. 2 rods ém’ abro rotro (até tovTw?) Tov
Biov xaravadickovras.

-ov -ov. Plut. Mor. 8538 should not aidod mdvrpyrov
dvacrdaas be aidov wdvtpyrov? Ael. V.H. 3. 18 Kat ypdvov
Gv abrods obx doov ypels GANA Kal exetvov Surdodv. Read
éxeivou twice that. Eudocia 978 Philo Judaeus érdovryce
Aéyov wapdpo.ov IAdtwv: wdovreiy should have a genitive
Adyou and perhaps rapoposov.

Aesch. P.V. 109 vap6nkorAnpwrov 8& Onpdpat rupds | ryyiv
kAorafav. Should vapd. be genitive

-ov -wv. Cebes 22 xexpdryxey éavrod makes no sense.
aitavt mdvtwyt

-w -wv. Trag. édéorora 455 (Nauck) drav 8’ 6 daipwv dvdpi
mopatvy Kaxd, | Tov voov eBAape mpatov @ Bovdcierar. The
dative ¢ can hardly mean against whom. Should we read
the neuter dv 4

-ov -w. Solon ap. Plut. Sol. 16 and [Ar.] ’A@. Tod. 12. 3
viv S€ pou xoAovpevor | Aokdv SpOaApois SpGou wdvres ore
Syiov. The neuter dyiov (there would be no construction
for the masculine) is very awkward with dove, and dcre
Sjio. is often read. Perhaps dyip.

ov -ot. I suspect this change should be made in Pindar
316 SILVA CRITICA

Nem. 2. 16 ‘Axdpvor 8% wadaiparov | etdvopes and in
Aesch. Hum. 207 ob yap Sdpoot toicde tpécpopov poreiv.

-ov -nv. The first words of the Humenides, rpérov pév
ebyf rHde mpecBevw Ocdv | riv mpwrdpavrw Tatar ex 82 rijs
@duy «7A, might be the better for a change to rpdryv.
Aelian V.H.12. 64 aériy dudxey should be airév, t.e. Ptolemy.

-o -y. Plut. Mor. Bernardakis 7.173. 7 77 xpeia 10 mpos
npas Spifovres. Read rq for re.

-or -ov. Aristaen. Hp. 1. 20 Scoparys épvddrrero rap’ éuod :
read éoi.

-ovs -ots. Arrian An. 1. 17. 4 SYapSiavovs te kai robs dAAOvs
Avdots rots vdpors Tots waka Avddy xpioba edwxe. Should
not the persons be in the dative? and so again in Plut. Mor.
204.4 Sots peravojoat kat BeAriovas yevérOas Tovs rovypovs ?
Sext. Emp. adv. Math. 10. 68 ray wadatav tis Kuvixdv Tovs
Kata THS KWHTEWS Epwrdpevos Adyous amexpivato pev ovde EY,
dvactas 8& wepierdrei, St’ airys Ths évapyeias tiv Siavoray Tod
coduorou dvedifuv. AS Adyov épwray, Adyou Hpwrnuevat, etc.
are regular expressions (see Bekker’s Index to Sextus)
in the later dialectical sense of épwrav, read here rots...
épwrwpevoisAdyous. evapyeias should, I think, be évepyeias, the
act of walking, contrasted with Sidvorw. ([Ar.] ’AO. IToA.
7. 2 xaréxAnoev 8& rots vopous eis éxatov éry. We find
several times the expression xataxAefewv vouw (Rehdantz,
Index Dem. s.v. xara at end) and that is more accurate,
the proper object of xataxAclew being the person or persons
bound, not the laws which bind them: read then here rots
vopots.

-ovs -as. D. Chrys. 8. 31 adcicrovs Bods éxovra. Read
mieiotas, a8 Tas Bows follows.

-ovs -ws. Athenaeus 382B 7 érimerpiovs euavrdv ayw ob
peyadavxovpevos ex’ euavrg. This can hardly stand for
anything but éupérpws euavrdv dyw. Cf. 150 B piAavOpdrus
dyw, 153.4 dyduevos BaciAtxas, 393 E rarewviis Tye.

-ws -ois. Arrian An. 4. 4. 1 Qvcas rots Beois ds vopos abrd.
Comparison of 3. 28. 4: 5. 3. 6: 6. 3. 2 and 19. 4: 7.11.8
gives an overwhelming presumption in favour of reading ols.

-wv -n. Schol. Dionys. (Goetling Theodos. p. 58. 31,
quoted in Kaibel C.G.F. 1. 17 n.) éort kopodia pipnors
mpdéews Kadaptixav radyudtwv Kal rod Biov ovoratixy:
evidently xaaprixy, and the genitive due to waOyydrwv.
SILVA CRITICA 317

-a -qa. Appian B, C. 5.9 ’Avrwvin piv}... ermédeua...
neBrdtgvero, KAcorarpa 8 Gre mpoordgeaey eytyvero. The
nominative may be right, but KAcordrpa seems likely.

-a -o. Greg. Cypr. 1. 59 dv8pas xadds tpdccovros éyyiora
(éyyorot) pido.

-a -t. Ap. Rhod. 3. 1368 rov (rérpov) p’ ava xéipa AaBov . .
euBare. yxepi? Pind. Ol. 6. 104 city dé mAdov kapdrov
éxros éovra bidor. edvre or idyte !

-a-av. Plut. Mor. 404 4 dvaravopévy 8 abr@ wore pera
motov Kal xopeiav mpoomecovoay Sterpdgato. Read zpooze-
govea, on which airg depends.

-a-as. Arrian An. 5. 24, 6 read dpdcovra referring to rov
ypapparéa.

-as -g. Plut. Mor. 582 A oddemids réyvys Karappovelrat 70
puxpots peydda. kat 8. 6Aéywy moda Tpopnveecy. I think
ovdepia. Texvy in the eyes of no art. Possibly tz’ ovdemas.

“aus as. Antig. Hist. Mirab. 144 dvaBaivovtos piv jAiov
move, ert Svepais 8 idvros icrarat should probably be ézi
Svopas 8’ idvros, possibly émi Svopais 8 dvros.

Fes “as. Zenobius 5. 85 the Sardinians trois trrép 7a ‘éB8o-
pyKovta ery yeyoveras To Kpovy ZOvov yeAavres Kal domald-
pevot GAAHAOUS ai xpov yop jyovvro Saxpvew Kal Opyveiv.
Clearly yeAdvras kat domalopevovs, as indeed the parallel
passage from schol. Od, 23. 302 cited in the Paroemiographi
shows.

-n -qv. D. Chrys. 52. 2 4 tév Biroxrytou tégwv elite KKoTH
elre dpwayiy det A€yew. KAowyv would at once avoid hiatus
and be more symmetrical. Cf. Dem. 18. 20 and 270:
23. 156; Plato Theag. 121 c¢.

-n -ns. Od. 15. 393 ovdé ti oe xem | es pen porenouiae
Should we write zpiv dpys, as in Pind. P. 4, 4314

“n ~NS- Antig. Hist. Marab. 25 kat yap rots orehéxeou tov
dévdpuv kal rots pdXors Kal 7H yn Tov adrov Tpdrov dmravrt
Tore oupperaBddrer tiv xpoay. Read ris ys, dependent
on téry. cupperaBdrd\r\x—not changes along with but
changes so as to agree with—illustrates Aeschines in Ctes.
165 ’HAcio 8 adbrois uppereBddovro. Again 1b, 28 &
Opany TH KAyOeloy more Kedpurdder i is taken from Aristotle
HA. 9. 36 620 a 33 ev @pdxn TH KaAdoupéry wore KedpurdAc,
but in both @paxy is a mistake for @paxys, the so-called
‘ chorographic’ genitive.
318 SILVA CRITICA

-y -y. [Ar.]°A@. Tod. 23. 1 péxpe rovrou mpopAGev % mods
dpa 7H Sypoxparia ard puxpov adgavonevy. I proposed long
ago to write atfavopéry. Cf. Isocr. 8. 51 tiv Sypoxpariay
eiddres. . . ad€avouérqv: Ar. Pol. 2. 12. 1274 a 10 'avguy els
thy viv Snpoxpartiav.

-nv -ns. Theopomp. fr. 110. cvvaxodovbel . . . dvowa Kat pera
tairnv (ravrys) éxoAacia: Plut. Mor. 249 D mpés éxdorny ava
pépos BadiZovcat, where the sense plainly needs wpés éxdorys
to the house of each.

Add-os-tos etc. Schol. Ap. Rhod. 1. 517 ‘Epyijs Ad-yos (Adytos)
elvas wapadédora:: Arrian Tact. 2. 1 of wodéusot rodepodvrat
of piv év yf, of 8 év Oaddooy, read wédeuor as in Cyn. 24.5
moh¢hw 7 ava. kpdtos SiaroAcuovpevy : Aspas. ad Eth. N. vii
(end) & 32 A€youod twes ‘peraBoAy wdvrwy yAuKd’ wept THs
rovnpias (rovnpas) Kal evueraBdrou picews Néyouct.

-os -ysos. Alex. Aetol. ap. Gell. 15. 20 6 & 'Avagaydpov
tpddtpos apxaiov atpipvos piv euorye mpooemetv (anapaestic).
This is corrected by Valckenaer to yaiov and so printed
now. But xdios is a Doric word not very likely to be
used. Did not Alexander write zpodés dpxaiov! tpodds=
Opéupa is certified by Hesychius.

dpviOa, dpvw. The verses quoted by the scholiast on
Plato Rep. 479 c wept rod eivovyov ris Bods x.7.A. have the
equivocation or ambiguity expressed five times by te kai od
and once by xat od alone (dvyp re Koix ‘dvip | dpviOa Kod
Spud? iddv re Kod iSdv «.7.A.). Is it not probable that the
original words were dpvw re xoix dpvwv, altered when the t of
épviv came to be always short ?

Haplography, t.e. words or letters written once when
they ought to be written twice.

This sometimes leads to loss of the article, as in the
following cases: Plut. Mor. 825 ¢ ék trav xpnpdtwv 
évayikav mpocayopevdevrwy and 844 B KadXtortpdrov 
’Eurédov ’Adidvaiov: Arrian An. 4. 9. 7 adbrod 
Baciréws (not the Persian king): Appian B. C. 1. 1 és
KoAvow pddiota tov tmdtwv  ard THs Bovdifs
aipovpévov: Maximus Floril. 6 ddeAipous airois 
épyots ddA py Tois Adyous: schol. Ar. Ach. 961 civ dAyv
€opryv  Avovicw dyopéevny: Polyaenus 6. 18. 1 rrdpa
éyévovro  @erradidv dvdpes rerpaxioxtAco1, the 7, loss was
SILVA CRITICA 319

4000: Porph. Abst. 3. 20 éxetvo vy Ala  tod Xpvatmrov
mbavev jv: Demosth. ap. Macrob. 2. 2. 11 ov« dyopalw (-cw 4)
rocovTou <70> peravojoa. Like the latter instance in that
the haplography is not exact is Plut. Mor. 153 a ri faorov;
 450. In schol. Ar, Av. 1271 advres ASovrar of dvOpwror
<7@> dpvidas xpareiy the loss of rg was due to its similarity
to zo. Some of the other losses to be suggested presently
were also no doubt due to partial similarity of letters.

Anon. in Eth. N. Heylbut 225. 21 ryv trepBodnv 
or 
     imepeixer,

Arrian An. 6. 15. 1 xaraorpepdpevos ev mapddo 7d
*ABacravav e6vos abrovoyoy : cf. 21. 3. Plut. Mor.
504B dvow yap kal diAdvOpwrov  614 TotTo Kal
moAv@wvovt In various authors—notably Plutarch—it is
often a question whether dv has been lost after a neuter
adjective, e.g. Suvardv, where the full phrase would
certainly be duvarov dv, it being possible that ete. (e.g. Mor.
550B os ob duvarov év rots Sixalois xaropOoiv), but the
absense of dy is so frequent that loss seems unlikely.

Pollux 5. 41 pe6” jpépay <év> dpyerOar begin agam? or
should something else be inserted to give the sense needed 1

Plut. Mor. 148 © drws  qyepdtatos dv «.7.r. 1 249 D
jpwv : Polyaenus 1. 40. 9 "AdrKiBiddys wept Kvgixov
 Onpaperyy éxeupe: D. Hal. A. B. 10. 45, 2 read ov
88 <6>.. . dywriodpevos x.7.d. a8 in 11. 25. 2. In the same
way Anthol. 10.1.7 rai’ 6 UpinroséyavérrrédAopar 6 Aipeviras
.we should probably read 6 éAdtpeviras (obAAeviras), like 2b.
2. 8 6 Ipiynros | tapi évoppitas. In both cases the error is
due to the similarity of « and o.

In Aratus 942 (Dios. 210) aoddAdke Atpvatae 7 eivddrar
épvibes I suggest woAAdk., for the metre’s sake,
and so in Eudocia 960 eypapev AirdxOovas (éore 8 rept ris
’Arricns dpxatoroyias év BiBdious Séxa), wapawéres Ov erav
I would read Séka,  wapawéces: also perhaps Hera-
clitus fr. 60 Diels (69 Bywater) 686s dvw  xdérw pia
xal dur}. In all these cases similar letters occur before or
after.

Another fragment of Heraclitus (5 p: 130 8) is, I think,
a more striking case of haplography: xafaipovra: 8 é\Aws

11s there not a similar loss in Diomedes 491. 18 K mimus est
sermonis cuiuslibet  motus... imitatio?
320 SILVA CRITICA

aipart piawdpevot, olov et tis ets wydov EuBas wyAG daoviforro.
Both the imperfect sense and the repetition of wyAdv...
TAG Suggest Kabaipovrat 8 dAAws aipart  puaLtvopevot,
or 
      . So in Synesius 62 dda ovyypdéppara mpos
Gla wodv Kat reuaxiors mapaBadrdAdpuevos read 
Tepaxiots.

Dittography.

Athenaeus 260 © wAcious... ypet paAAov [ev] rais cvvoveias
tats Swpeais. ev is a dittograph of ov in wadAov. Schol.

Aesch. Pers. 110 ri 82 75 éuafov should probably be 7é 8
éuafov, but the history of this may not be quite the same.

Negative lost.

Polyaen. Hze. 39. 2 rovro yap roujoar obk duoca, but the
sense requires 
       oujoa, I did not swear not to do it,
or in good Attic yy roujcew : schol. Ar. Vesp. 635 weOdpuevos
ovrws mBavurara gueAAov épety should perhaps be 
meOopevos Ott bs mibaverara, not thinking I was going to
make out a very good case. In Plut. Mor. 1138 a woAA\Gv
dv tis POdvot Kal Tov viv KatayyvicKkwv dyvowy I cannot
make sense of @dvo. and suggest woAAGv  dv tis
$Odvo.. It is not always understood that oix dv POdvois or
any such phrase has two distinct meanings, (1) practically
recommending a course of action, the sooner you do so-
and-so the better, (2) stating what shall or will immediately
follow. The second sense appears in Plat. Symp. 2145
ook av pd. I will begin at once, Huthyd. 272 D otk dv ¢.
dxovwy you shall hear at once: Dem. 24. 143 ei py
TyswpynoeaGe TovTOUs, obK dv POdvo: 7d 7AHOOs Tovrots... SovAcvov
it will find itself at once enslaved. In this sense otk av ¢.
would make excellent sense in Plutarch, and it is so used
in another passage of the Moralia (Bernardakis, vol. 7. 173.
8) otk dv pbdvopev...cvyxwpoivres we should immediately
have to admit, In Origen c. Celsum 9. 63 read eéxxdévovor
To potxederv Sia TO 
       KowvwviKdv Kal mapa iow
elvat x.7.A., unless it should be cai <7é> rapa gpvow elvat.
Heliodorus ad Eth. N. Heylbut p. 44. 2 ot ciddva péyor
wéaov moévras én pefiew is senseless, until we read 
pebiew drink without getting drunk (uy, pe). A verse
inscription on stone at Amorgos (Kaibel Hipigr. Gr. 274. 5)
SILVA CRITICA 321

gives rpidxovta S€ yu (should probably be pe) AvedBavras
dvamAjoavTa... «ide wéAas Odvatos. As the « in tpidkovra
has no business to be long, whereas the u is sometimes
shortened, and as dyor-ov is the word preceding, it seems
possible that after ov an od has been lost,  tpidxovta.

We have the contrary error of a negative wrongly
inserted in Euseb. P.H. 5. 10. 4 riva [ody] tmepBodrry...
kataXeire: ; the sense being that the thing spoken of is as
absurd as anything can be. tiva imepBodnv... katadecrer; and
ovdcpiav trepBodyv...katadere have one and the same
meaning.

Loss of negative é-.

D. Chrys. 69. 8 xoddLovet Tovs vopouvs. Some MSS. have
tiWevrat for xoAdfovor: Wilamowitz deletes rovs vdpous.
Probably we should read tots dvépous. Cf. Xenophon and
Others p. 137 (Xen. Cyn. 5. 34).

But Plut. Mor. 5 eoedOe cis ropvetov, wat, tva paOys dre
Tav évakiwy Ta Timia obdev Suadéper (Diogenes) evidently calls
for d&iwv cheap opposed to ripia dear.

Loss of numeral, often written in letters, a’ B’ etc.

Polyb. 5. 1.11 rpiuyvov pic Podorjoat ri Sivapuy kat oirov
mpooOeivat pupiddas : Steph. Byz. “ABSnpa: tod ’ABS8npirys
péuvynta. Evdofos év rerdpty Tlepiddwv xal Tavoavias \
Tlepinyjoews cat “Hpddoros éB8ouy.

Loss of awword in enumerations ete.

Plut. Mor. 853 & dua tairéy dvdpt xal yuvarxt cab petpaxio
kal yépovre Kal oixdrpiBe  mpéwov’? 1b. 4390 a
third infinitive seems wanted in oivoyociy diddoKxovow,
épororety, or perhaps only 7 lost after v: Longinus Ars 554
the parallel of ddéas, ddogias mpdgéas, dpyias shows that we
must write , rAovTous OF tAOUTOUS, .

Loss of a preposition,

Arrian An. 7. 10. 1 tis dpi Tovijoas oidey  éuod
padXov 7 ey tép éxeivov ; (the first 7 seems a dittograph of
v-N H- in tyév): Simplic. ad Ar. Phys. 74 oix  rhv
airy éBddioe Tappevidy .. . ddéav. Sometimes in compound
words: Sext. Empir. 7. 255 ob mpoceixe 8 airy for ovdk tye

Y
322 SILVA CRITICA

8 airiv?t Synesius 57¢ of yip erociro codds elvat
Swxpdryst: verse quoted by Eustath. ad Il. 2. 698 rparov év
Tpola wecety | tov <ék>mpornSijoavra tis vnds (read veus) t

Loss of av.

Polyb. 18. 13. 8 ci wy... peréppupe..., pavepds 
dpSnv daokdder: 1b. 43. 10 otk  epy xkowwveiv:
38. 8. 10 brédaBov tots ‘Pwpalovs... trav  tbropevew:
Plut. Mor. 1808 darexpivaro Aapeiw pre thy yav jAlous dvo
pire rhv “Aciay Sto Bactreis  iwopevew: schol. Ar. Ach.
708 ot8  ris Ajyrpos fvécxero, as in the text:
Polyaen. Exc. 14. 7 od8tv  péya yévorro: Athen. 3748
bre yop py vixdy, AapBdvuv Buwxev  used to give.

Loss of pe.

Plut. Mor. 439B waddAcw  pavOavovow (nev, pay-):
163F paddov ecikds .. . ouprabety 7 SniGais  réga,
Avpat 8 "EAAnot kal addol cvprafotow ?

Loss of 8€ clause.

Polyb. 9. 8. 13 the clause answering to t@ pév tyepove
K7.A. is missing.

Miscellaneous words lost.

Many of the following suggestions are of course very
conjectural and at best approximate. Gorgias Palam. 10
gvdaxal 8 dv  od« eore daGeiv: [Ar.] ’AO. IIoA.
60. 2 6 pay vouos <éri> éoriv, 7 8% Kpiots Karadddvrat:
Anthol. 9. 826. 1 rév Bpopiov Sdrupdv  rexvyjcaro
Sadarén xeip . .. ciel Se «.7.A.: Plut. Mor, 48 8€o or xpein
to govern dvaBdvra dvaxpayeiv: 13F imperative after
ddeiAero: 1858 mapawov pr Sediévar tots "EAAnvas 
drodiSpdoxovras : 284. C elon tis avAs  or ? cf. 242p: 3348 aavtwv  oidpevos Setv
meptetvat, TOD Sixaiov 8 yrracOar: 335E adda. hépe mpds Gedy
 exredcoOqvar xal gavijvat towtrov épyov: 3455
Hevodav... ypdias & eorpariynoe Kal xatépOuce cal Oem
oroyeve Tepl TOVTwY  cuvTEeTaxPat TS Yupa-
kooiw: 413F domep , as e.g. 4168: 439D pH yuvat
, cf. E: 536 A  ordpara: 858 B
dxovricas 76 Sépu TotTo povov 76 xwpiov  jélwoe
SILVA CRITICA 323

doov eméaxey 9 aixyy: 1b. F pymote . . . dvopacrt Bojoal
 Tov atts dvdpa: Polyb. 4. 4. 5 <éuehhe> By olov 7
elvat: D, Chrys. 38. 18 of heralds ddicjoas odx eeorw avrav
ovdéva, as Tay Oeiv dmavras (or dvras) Tovs tmp dudéas
ayyédous, read perhaps  daravras : Polemo in
Callim. 27 & dperiis 8e 4 dyeov ev dvtimddy pev  Tijs Suvdpews
dorab ure de T@ THs TUS Kpiveras : Porph. vit. Pyth. 9 bpavra
Ti Tupavvida GvvTovwTEépay ovaay <7> wore K.T.r (v and y, cf.
above p. 200): schol. Ar. Ran, 218 +7 TPEPS n :
Stob. Hcl. 1. 8. 40 Zyvav ey... Kata rotrov yiyverbar 7a
ytyvopeva, kat  es Tepawopeva, aravta Kal Ta
évra evar: Antig. Hist. Mirab. 78 dy (worapav) rov pev
Aevxa. (rpdBara), roy d& peAava yewvay , sorely
coming again just after: 7b. 126 add ér: after xafdmep.

Loss of one syllable or more.

[Ar.] Oecon. 1347 a 6 ras Ovpas (Oupidas) Tas dvovyopévas,
as in ‘Ad. Hod. 50. 2: D. Hal. 5. 68. 1 obs pev dv dpdor TH
yaorpt Kal tais aicxloras élyxdras Hdovais is an absurd
expression, but, remembering Dem. 18. 296 77 yaorpt
perpodvtes Kal tois aicxiorors TH eddatpoviav we have no
difficulty in restoring 77 yoorpt kat Tois aicyicrots elytnKdras
qoovas. Plut. Mor. 347 F meroinka rH Kopepdiar: @KovopnTat
yop % Siddeors: Set 8 abry Ta orixidia éraoa.: the infinitive
is perhaps the remains of éruroujoat or émypaipau
(haplography: ae and yp. very similar). The plot is
Jintshed, I have only to add the words. Marcellinus in
his Life of Thucydides 51 says we may find in him
gpovnpa Tlepixdéovs cat Kdéwvos odk off’ 6 te ay eirou THs,
"AdxiBiddov vedTyTa, MemtctoKAcovs wavta, Nixiov ypyordryra
«7.4. Two or three attempts have been made at improving
adyra. Perhaps it stands by much compression for wiHavd-
tyra; Plutarch Nic. 3 says that Pericles (see Marcellinus’
words) needed no mi@avérys. The shorter Life of Aristo-
phanes has rpdrepov Kparivov xal EvrdAdos BAacdnpotvrur,
the longer muxpoTepov Te Kal aiaxpdtepov K. xat Ei. BAacdy-
povvrev 7) ee. Is apdrepov an accidental abbreviation of
amuxporepov ¢ So elsewhere ovre TiKpos diay éoriv dorep
Kparivos.

Diogenianus 2. 48 dyddves Aeoxair eyxabyperar ent
tov ddo\erXoUvTwV.

y¥ 2
324 SILVA CRITICA

Errors of spelling or division. ,
Theognis 401 pdtv dyav orevdew—zoddaxt 8 «is dper iy |
aomevder avip xépdos Sifjevos. The words els dperyv have
been variously altered by conjecture. They make no sense
as they stand. We might read to’ dperq and take the poet,
to mean that a man puts gain on a level with goodness:
lines 465-6 and 699-700 tend to confirm this. Theo-
phrastus fr. 81 otk, ef 89 wov..... , aiperdy éorat A€yw SE
otov od Kat xpuotov Tyudrepov xaAKod Kat THALKOY TOD xpuctou
mpos TO THALKOV xaAKoU peyeHos dytimapaBaddouevov m)héov
8d. It is not hard to see that od xai is really a second
ovx, ei, the apodosis beginning with the subsequent xai.
In a fragment of Chrysippus quoted by A. Gell. 7, 2. 3
Adam has anticipated me in pointing out that pera odd

pev ovy stands for peraroAAupevwr. ‘

Transposition.

Another Greek quotation in Gellius (16. 3. 7) contains
an error of a different sort, but hardly less obvious. The
words are éAoy:{éueba rapa tiv icyupav olprtwow Tis KoLAlas
elvat tiv odddpa dotiav, as though this affection of the
stomach produced excessive abstinence from food. Clearly
it is abstinence which produces the affection, and we should
read probably zapa rv odddpa douriav clvar tiv ioxvpav
ciprrwcw ris xoAlas. Arrian An. 7. 6. 2 evar yap odv
Mydtayv tiv ’Adedvdpov crodny GAyos ov cpixpov Maxeddow
épwpéevny : read ryv Mydicyy’A. orodjv. Cebes 1 tives xai
mote Hoav should probably be the usual tives wore xal joav.
In the shorter Life of Aristophanes ciAaPis 5: yerduevos
GAAws te kat ebpujs, in the longer evAaBys 8& odddpa
yevduevos tiv épxiv dAdws te kal edpvys, read as though
eiguys and ebAaBys should change places: else what does
dddws re kai mean? Hom. Hymn to Dionysus 24 py te
xorwbeis | dpon dpyaddovs 7’ dvéuovs cal AaiAama woAAjv :
perhaps épcy 7’ dpyadéous.

The difficulty of @vzot in Soph. O. T. 674 oruyvés piv
elxwv SHAos ef, Bapis 8 drav | Ovpod wepdons might be
removed by putting it after ¢l, as depending on cikwv.
There is no objection to Bapts 5 ending the line, as the
ordinary pause would then occur before the sixth foot. Cf.
Antig. 718 add’ eixe Guod (probably).
SILVA CRITICA 325

The particle wé& seems out of place in a fragment
(Mullach L. 495. 134) oKedy TH fey KOLO KpetTrova, piria dé
4 madatorépa (read oxedy pev 7a): Athen. 37 B ds vopilew pev
em Tptypous Theiv, xeyudLeo Gar 8& yaderds (read vopilew wAciv-
pev ert rt. or éml +. pey mAciv) ; schol. D. Thrax (Bekker’s
Anecd, 2. 749) rijs 8& wéons (kwpwdias) kat abriis pev oAdol
yeyovacw, érionjwos 8& Kparivos (read kai airijs toAAol pev).

A word substituted for some other by repetition or antict-
pation (see Xenophon and Others p. 307).

Plut. Mor. 82 c yapiey 76 tod Atoyevous mpos Twa veavioxov
épbevra piv ev karydelw, katapvydvra 8 eis 7d KamyActov.
The last word is clearly wrong, and from Diogenes’
comment may perhaps stand for érrdviov kitchen. Plut.
de Anima Bernardakis 7. 35. 17 woddd radla Srvdrrovra
ergs v trap 8 ovrw roANG 8 Urap (read dvap) épbeyéaro, dAAws
oizw dbeyyoueva. Pollux 8. 87 of Ocopobéras tpoypadovat
more det Sixdlew 7a Stukaorypia: I suspect that through
anticipation of dixaorypia an original xafifew was altered to
SuxdZew. Ar. Probl. 29. 14. 952b 4 of... KdAémrovtes Kata-
davels drat ylyvovrat ort pavepol ciow, Oore K.T.X. KAaTAPavas
otv Hn wovnpol SiateAodow dvres. Here gavepoi is an echo
of xaragaves and stands for the very similar wovypot. In
Antig. Hist. Mirab. 9 & 8 tals vicos...épduxes od
ylyvovrar, @dAQ Kav iby tis ardAAvvTarr Eror b& TovTov
TepaTwoeaTepov iatopodot Ort Kav idwow THY xwpav it is clear
that «av i8y is an unmeaning anticipation of kav idswow.
Read kav dyy, as in 10 kav eicaxOp reAevta. So in another
collection of wonderful things Apollonius 49 xatavAoupéevov
Tod ooparos KatavAnois yryvécbw ypepas mévre as eAdXioTa
the xaravAoupévov is due to the xatavAyois and has ousted
some such word as raAatrwpovpevov, Tovoupevov, KaTexopevov.
Cf. édv zt pépos Tod odparos év adyjpat. trdpyy just

receding. I have sometimes thought that Opéfacxov in
Il, 18. 602 dAdore 8’ ad Opegackov emi orixas éhAyAoiow Was
an erroneous repetition from 599 of & éré pev OpéEacxor, for
it does not seem quite appropriate to ért orixas.

Exchange of Terminations (see Xenophon and Others
p. 303).

Plut. Mor. 230 B NixavSpos eimdvtos Twos é7t aks abrov
Aéyovow “Apyeion ‘odxodv’ epy ‘Sixyv tivovar Tovs dyabous
326 SILVA CRITICA

kaxGs héyovres.’ This is most insipid. No doubt he really
suggested that their punishment for speaking ill of good men
was that they spoke well of bad ones: perhaps rots xaxods
dyaGGs A€yovres, or adyafovs. The adverb dyafds is
extremely rare, but it is found in Ar. Hhet. 2. 11. 1388b6
and Top. 5. 7. 136 b 28.

One word made into two.

Galen Hust. Phil. 24 of pev doadparov épacav (riv Wuyi),
of 5¢ odpara Kivetv: for the last two words read cuparcayy.
Hermog. (Walz 3. 325. 8) of ra orovdata yéAota ratra owte
Gévres : read orovdatoyéA ota.

Two words made into one.

Euseb. P.H. 2. 1. 51 édevOepiav trav rocotvrov Kaxav
ovderépus ebpduevor ) Sia povys THs... didacKadias, a mistake
for ody érépus, and 4. 14. 5 év rGde ye Ovew, a mistake for év 76
8€ ye Ove.

Various exchanges of words or parts of words, many
of them examples of errors that often occur.

é-, et-. Pollux 1. 186 ywpia dBara, tpaxéa, Siowopa...
eb8poza (read ddpopa).

am-, ér-, Arrian An. 7. 12. 3 read ényyéAAero promised
and 1b. 7 dq’ érov. Eudocia 300 dad (éri?) dvotvy BiBAlow
€avTov cuvetxe.

a6, tr6. Plut. Mor. 859F iBpicparos . . . bd (dz0)
Sapiwv mpdrepov tmdpéavtos. Antig. Hist. Mirab. 79 ia’
éprydvov kai Oeiov mepurarropevovs should we not read dx’?

av-, év-. Schol. Aesch. Pers. 28 (tAjpovi 8d€y) evuTootarw
Soxyoe: read dvuroordre.

dy-, ovv-. Dicaearch. 59. 23 dyopav oroats dverdnuperyy
Surrais: read ouv-, comparing 1b. 29.

dy, dv. Arrian 4. 3. 6 ef 59 7. Adyou dy (dv? or omit)
aétov vewrepilotro.

dv, ov. Plut. Mor. 1123p paddov dy ols od waparvy-
xévovor Siamitycavras avOpmrois kal mpdypact Kal Adyous
amAGs dracw 7 play éxeivoy ddnOy Kat trdpxovcay elvat
gpavraciay reoOevras. There is no sense in ols od wapatvy-
xdvover: read ols av wapatvyxdvwct.

dv, dy, D. Hal. A.B, 3, 30. 3 ef yap av wapaxweiv 7
SILVA CRITICA 327

toApnonte (read «i yap 5)... ToApyoere): 14. 9. 4 ri yap ay
(read 8) 8uvjcovra, comparing ri 84 in 15. 3. 7 and 8.
Arrian An. 2. 2. 3 apocdoxav éoeoOar av (84) odior...
apdAdcav (6. 6. 5 read aorist for future) and 6, 24. 6
et. . . éyxatahyndbciev, evraifa av (8)) éraAarradpouy, 84
emphasising the contrast. Philolaus ap. Stob. Hel. 1.
21. 7 orundv av tpdruv eyévero corrected now to wrmév
tpdrw, but read perhaps ormédv 84, or draw 84, Tpd7y.
Anon. Life Thucyd. 2 éuapripyoe Sé atta cal Oouxvdidys 6
pabyris ore dv, © ctpBovdos yévoito, prota. exeivos drphAarrev
eis Tas Sékas, evidently dz: dy. Eudocia 34 has the portent
of dv with present indicative, as dv (84) twés pacw.

ydp, 86 Arg. 2 Soph. 0.0. rerypwrar dé? (after ruyxetv
comma only): Hermog. (Walz 3. 190. 3) et dé jax.

dé, 54. Athenaeus 3518 pdve 84, not dé: Aelian V.A.
13. 24 d7dra 8.

duo-, ed-. Aesch. Hum. 266 Bookay depoinav mdparos
Svordrov, creating a metrical difficulty. Would not the
Furies find it edrorov? Cf. on Menander 203 above p. 95.

éy-, ex. Pollux 5. 73 rav dd 70d owparos éy(ex)xwpovyrav.

eis, @s. Schol. Ar. Vesp. 42 tov @éwpoy cis (ds) Kédaxa.
kopwdotct, So in the other scholium és KdAak cwpwdeiran.
Such a use as eis 7oAAG Kexwpwdyrar (schol. Plut. 665) is
quite different.

éx, as. D. Chrys. 21. 3 otdels 7} ef tus ex TOY TpidKovTa
Read 7, ef tus, ets 7.7.

pets, buels. Cebes 33 6 pév pidos, & Eévor, ... ToLotros
huw (iptv) éorw: Eudocia 17 édevOepotv tas ’AOjvas Adyet
jytv (iptv) 6 Beds (the Pythia to Lacedaemonians).

te, et. Clem. Al. Pazdag. 3. 11. 7. 297 P. read dvemeva
tallying with the other participles.

xa, #. Plat Mor. 392d ravartia didotpev 7} picodpey Kat
Ooupdfopev kal Wéyouev: either 7 or the second «ai is wrong.
[Ar.] ’A9. Tod. 45. 1 dy rios ddicetv 7 Bovdt xarayvd 7
fnueoon : 4 or makes surely no sense. Polyaen. 1 prooem.
12 ratta pev 7 (kal) doa rotdra Siddoxnwv. Sext. Empir.
adv. Math. 8.12 dexrov dep aAnbés Te yiyverar 7 Wevdos: if
re is right, read cai for 7. Libanius 14. 4 aparos Kai pdvos
and 18. 217 pddiora Kal pdvwv can hardly be right. In
16. 8 he has ody duoov . . .) «.7.A. and Pausan. 7.16. 4 obdév
re yevopevos és ’Axatods dpowos 7) Kat KadAlotparos .. . mpos
328 SILVA CRITICA

*"AOnvaiovs and D. Chrys. 29. 15 duoov... 4%: but these
uses of 7 seem very strange.

cai, kav, Euseb. P.H. 9. 27. 36 they say 73 Moiow
Ociav poviv yeveoOa waragar tHv Oddaccay 7H pd BS kai
(xdy 1) diacrivar and a would divide : schol. Ar. Vesp. 508
ot 8 eddatpovotvres Kal (Kav) TodTo . . . Topicawro.

cai, ws. D. Hal. 10. 60. 1 word aapeorixer Séos...
drodwAexevat vopifovor tiv éAcvOepiay Kal déxa Bactrcis édo-
pévots av0 évos. Read as Sexa: the words must be subord-
inate, not parallel, to vopitovor. So again in 11. 1.3% a Sidvore.

+ Xetpaywyoupery dia. rey Adywv emt 7a épya kal _ (@s) pa) povov
dxovovoa TaY Aeyouevwy GAA Kal Ta TpaTTdépeva Spadoa: it did
not actually see them.

ovtos, adrds etc. Schol. Ar. Av. 1536 read airy for airf.
In Eudocia 927 (Flach p. 679. 17) drov wAetvcavras Tous
"EAAnvas 8: adrod Soxeiv Aiweva elvar SiapPapHvat for 8.’ abrod
read 64 16 Soxetv. Aesop 2 obros 8é.

ovrus, ovre. Sext. Emp. 9. 99 ovr’ dv should be odrws dv.

mapd, mepi. Eustath. ad Il. 3. 6 read xépara wepiriBevras:
Procop. de Aedif. 6 Séxnow rapexdpevov (not zept-).

mpos, mpd. Aesop 23 read mpocdydacw anticipated,
counted upon.

avv-, év. Cebes 10 Soxotow évetvar for Soxoter cvveivat
(perhaps with oo. after Soxoidct).

&, 9. Carcinus 10 & wodAd riodros SvaTuyéoTaTos Kupay |
opus pwéyirtov CHAov év Bporois éxe. Surely 7 toAAd. For 7
strengthening modvs cf. Herod. 3. 72: Plato Huthyd.
2714: Ar. Hq. 1290: Pind. P. 9. 24.

os, écot. Democritus (199 Diels) dvojpoves 76 Liv ds (Guws
con}. Diels : read dco.) otvyéovres ov eBeAover Seipart atiew.

as, otoa. Aristaenetus 2. 9 read émeirep oto’ éadovatarn.

dyw, exw. Anthol. App. Plan, 4. 244. 2 4 tf wapaxAivas
obas dyets (exes?) Kadduy ;

ddeaav, ddyv. Schol. Ar. Nub, 386 all the allies sending Botv
tvOnodpevov, cvveBawey ddeav elvar Tov kpedv. What could
ddeav mean 1? Have editors supposed it = dpBoviav q

aizuspevos, (k)extypevos. Aristaen, 1. 13 odparos pev
aavij tratTopevos dAyndova, Woxiis 8& rats dAyecars € epwrixny
airudpevos (read (x)extnpévos) vocov. airievos gives just
the wrong sense.
SILVA CRITICA 329

dxpws, axpiBas. Athen. 21 F dkpws tais yepol 7a Neyopeva
dexvis. dxpy8Gs would seem more proper. Cf. Ar. Hth.
2. 9. 1109 a 34 érei oty Tod péoou Tuxely dxpws yademor,
where I have proposed the same change. In Plato Theaet.
148 c the MSS. vary between the two words. But in Athen.
248 F dkpws yé pow éreriunoas we should probably read
TLiKpOs.

dAAot, roddot. Heraclitus 1 Diels (2 Bywater) rots dé
dAXous dvOpwrovs AavOdver seems too strong and too vain.
Read zrodAovs, as in fragments 2, 17, and 29. The confusion
is well known.

évr(Soo1s, dvéravots. Appian B.C. 1. 4 émt Bpaxd e&dygav
at otdces ért SvAda, Kal xaxov dvridocts jv avy 6 BvAdAas
eipydfero. dyridocts is unmeaning.

dé, adéw. Arrian Cyn. 9. 2 xives ... dturep movnpdy Tod
xpwrds TH wAnodlew Te Kal exOepuaivery éragovow. Both
verb and tense being unsatisfactory, read éravgovow.

GAs, dAdws. Olympiodorus Vit. Plat. (end) xai ddAAws
moAXaxas.

droxpivopat, drodaivoua. Appian B.C. 1.104 drexpivato
has no propriety. See Xenophon and Others p. 1.

doxav, Soxav. With the correction of [Ar.] ’A6. Toa.
23. 2 compare the apparent imitation in D. Hal. 10. 27. 2.

yev- yeyv-. Plut. Mor. 392 Oeiperar 6 dxpdlwv yerouévov
yépovros : rather yryvopévov, for as one comes the other goes.
Ch. E rotr’ aird peraBdddXe yeyvduevos Eerepos e& Erépov. So
in arg. 1 to Soph. O.C. Ismene should be said to announce
Thy yeyvoueryy, not yevoueryy, dpiw rod Kpéovros. In 397
she speaks of him as #éovra Basot xpévov. Porphyry Abst.
1. 46 the other futures show yiyveoOa. to be an error
for yevyoer Oa.

yev- Aéy-. Aen. Tact. 4. 7 perhaps rapayévyra, but the
sentence is wrong otherwise too.

yepnopia, icpadopia. D, Hal. 2. 10. 2 é dpxais Kai
yepnpopias xai rats dAdats tats eis Ta Kowa Samdvass.
redehopios is suggested for yepypopias. Should it not be
iepagopious or iepoavrias? Cf. 2b. 73. 3.

Sevrepos, vo. Euseb. P.E. 2. 6. 12 and 14. 4. 14 eis kat
Sevrepos should of course be eis xai dvo. The mistake arose
from f’ which stands for either.

SiaxetoOat, Savoeiobar. Plut. Mor. Bernardakis 7, 152. 21
- 330 SILVA CRITICA

"Avrimatpov . . . A€yovow év ebruxnmatos péper SvaxetoOar
(read dtavocioOat) riv ... eis "AOjvas edrAcav. He thought
‘it lucky. Cf. Plut. Adem. Paul. 27 pydey as povipov...
Stavoeio Bar.

dteABeiv, Stedetv. Alex. Polyhist. ap. Syneell. 7. 4
(F.H.G. 2. 501) read &eAdvra (not deAOdvra) trav rod mhotov
Mepav fados tt Kal iddvra x.7.A. Cf. Herod. 5. 33 dueAdvras
THs vews kata TovTO, Where d1éAKovras is a quite needless change.

ddou, det~w. Athen. 1288 dwcouey should I think be
defEouev, or perhaps dyAdcopev, though syAoctvres comes just
before.

eldov, Fyov. Plut. Mor. 603 B év 7 (xépa) Hevodav pera
TH oTpareiav Aurapoy Eide (Hye 1) yipas.

eldov, 45y. Himerius Or. 14. 23 ob rHv pev (dirdocodiav)
cide (yder 1), rHs de eueAnoe: Callim. Ep. 18. 2? see p. 69
Aesop 274° 6 8 dpvés ey ‘ dxpyv tpopyy ob eldov’: is eidov a
mistake for oida 1

ein, etvat. Schol. Ar. Vesp. 1302 read eivar (av ein just
before).

elvat, Seiv. Proclus in Plat. Crat. 37 ’Avribevyns edeye
py detv dvriAéyew. This should certainly be efvar: indeed Sety
is nonsense. Cf. Plat. Huthyd. 285": Ar. Top. 1. 11.
104 b 20 and Mez. 1024 b 34.

elyov, £w. Anon. Vit. Plat. p.5 Didot xat odrw 8& mpociyev
should apparently be mpodée: or zpoééery with no stop before
kai ovrw dé.

évdpyeia, evépyeia. Sext. Emp. 10. 69: see p. 316.

émtBovdy, émiBory. Aesop 4° the words are not very
clear, but read ém:Body.

éxinpo, émixepa. Ap. Rhod. 4. 375 cut my throat édp’
érinpa pépuiat éouxdta papyootvyow. Surely émiyepa. See
Jebb on Soph. Ant. 820.

etp-, aip-, dy-. Schol. Ar. Av. 1021 Sov eipebe’s should,
I think, be aipefeis, and Anthol. 10. 84. 2 Sdxpvae 8’ &v
moAXois Tov Biov nipov ddov pretty certainly Fyov (cf. under
eldov above). Aristaen. 1. 13 rév 88 voty perdpctov adywv : as
ay and aip are confused elsewhere, read aipwy, a much more
suitable word.

exovres, Livres. Polyaenus 7. 1 wodw ody elyov, éxovres
dvev vouov kal dixns, dyovres kat hépovres GAAHAOvs. Roth’s
dyovres for éxovres, adopted by Melber (Teubner 1887), is
SILVA CRITICA 331

rather Latin than Greek and in any case could hardly
immediately precede dyovres kal qépovres. Read évtes,
Zyovres being either an ordinary corruption of it or a
careless repetition from ¢fyov.

tnretv, aittov. Plut. Mor. 3850 émei b€ rod didocodetv,
éfy, 76 Lyreiv 76 Oavpalew Kat amopety, cixétws Td TOAAG K.T.A.
Here 16 ¢yreiv is pointless and the real connection of the
infinitives with rod didocodetv is not brought out. Read
something like rod ¢irocodeiv, épy, eotiv airiov 76 Oavpdley,
comparing Plato Theaet. 155 p and Ar. Met. 1. 2. 982 b12
88. 76 Oavpdlew...ypgavro pirocopeiv.

HOos, 20s. Arrian An. 7.9. 2 vouors kal HOecr (ec ?)
xpyorois éxdopycev.

fv, qe. Plut. Mor. 349 F dre Kapxivos ’Acpory ouviy 7
"Exropt “Acrvdduas. I donot see what ovvjv can mean.
cuvjes would be competed.

Geios, Sotos, a well-known confusion. Plut. Mor. 488 F
ri hedyeas 'ApraBavov, mat, Oetov dvra kal TlepoGy dpiotov ;
Is not dotov more likely 1

iepds, Netos. Plut. Mor. 854 af Mevavipov kwpwdiar
abOdvav dddv Kal tepdv peréxovcr. Actos is sometimes
applied to a taste and would be suitable here.

daB-, Bar-. Strabo 6. 1. 12 mpocédaBe SE ry S0ky x.7.A.:
Arrian Ind. 20. 10 wodXdév cuvertAaBérbar és edOvpinv TH
otparty. In both cases the other verb should be substituted.
The confusion is quite familiar.

Aéyw, 2yw. The corrupt and obscure line of Hedylus in
Athen. 345 B should perhaps be not 74 8 épav poy pdvov od
déyopev but 76 8 Spav Keiwevov odk exomev (again a familiar
mistake),

Aeur-, Aur. Paroemiogr. 2. 154 Aavaidwy widos: ert trav
det éxdurdvrev (read ékAewr-) kal py wAypoupévuv. D, Hal.
A.R, 20. 13. 3 read éxAevrovcas.

vépos, Adyos. Phalaris Hp. 4. A man can hardly be
dxdAaoros év vopors ; read év Adyous.

olov, rotov. Cebes 3 certainly rotdv tt ;

épovod, suoroyd. Plut. Mor. 8E pévos trois whovotous
Spovoeis (Sporoyets ?) tas iojxas Sddvat. 3

rou, 6o@ or dréoy. Correct Arrian An. 5. 4. 2 rod’Ivdo0
petovés eiou al wodd 8 pedoves, drov Kai adrds 6 "Ivdds rod

Tayyov.
332 SILVA CRITICA

dopa, _ Tepyopat. Theognis 570 reibw e éparéy dos
jedriovo, | gumns 8 écOdds ey dopat ovdev ert. Somat is
weak and 567 78 tepropevos wailw strongly suggests
Tépop.a.

mapadacw, trnperoow. Plut. Mor. 440 4 éxetvor piv ydp,
as pyow “Apddoros, rods oixéras éxtupAotow Srws Trapaddow
airots. The right sense would be well expressed b
tanperoow. Of. tanpérior réxvais immediately following.
In Phaedo 82 p I have suggested iayperotvres for tAdrrovtes.

mapeis, mayeis. Aesop 8 eurapertwy (eurayévtwv) aitod
TeV évixuY ToIs padAois.

mapovoa, mpéerovoa. Aristaenetus 2. 1 reOivijécrar... cf wy
Ti wapotoay (mpérovoay t rpooyjKovoay ?) Oepameiay éxwetoets.

mAciwy, 7diwv (Cobet’s Misc. Crit. p. xiii). D. Chrys.
18. 6 wAciwy (read 7diwv) 4 alcOynots dmraddayévte THs wept TO
dvayryvioxew doxoXias, 1.€. it is pleasanter to hear than to
read. 7Aefwy seems unmeaning.

moXurixés, touxiAos. Plut. Mor. 839 ¥ Isaeus pros «
oxnpatifey ypEato Kal tpemew emt TO ToAttiKdy THY Sidvolay.
In connection with cynpatiZew what can 76 wodutixéy Mean
here? Perhaps 76 zorxiAov, as D. Hal. Isaeus 3 says of him
h 6€ "Ioatov (Ades) rexvixwrégpa ss... Kal oxnpario pots
dtecAnppevyn moxido.s, and several times subsequently uses
motxiAos and zrouiAXw with regard to him. Plutarch Nic.
“1 calls Thucydides zoiAwraros. When Zosimus (West-
ermann Vit. Script. 298. 45) says of Demosthenes
eLnruwoe . . . Oovvd(Syv cai Ta wAciota ex TOV éxetvou Adywv
eri 75 moditixdtepov petéppace, he is probably using
modurixds a8 Isocrates does when he speaks (9. 10) of
modTiKa dvopata, ordinary, everyday expressions. This
sense will not suit the statement about Isaeus.

mpoa6e(v), mpovorn. D. Hal. 1. 53. 1 mpdoGe(v) yap kara. TV
dginow "Edvpos 4 ard Tod BactAtKod yévous dv. mpoabev yep iy,
mpoereOn, Tpoetxe have been conjectured. I suggest xpotory.

aporepos, erepos (see my Xenophon and Others P. 226).
Athen, 4618 odr’ & “Ypapais ovr’ émt Trav “Tporepov cipijoopev
motypiov eipéyebes cipyacuéevov. There is no sense in the
antithesis of év ypapais and émi rév mpérepov. Read éxi rav
érépwv in the other arts.

cowry, oroAy. D. Chrys. 30. 4 xal rq cowry (crodq?) Kat
7G Bodiopatt.. . oe eysetro.
SILVA CRITICA 333

témos, xpévos. D. Hal. 4. 7. 3 Tarquinius lived to the
age of 110: rodro 8 7d pjKos tov Blov od dépovow of Kal?
Hpas toro. Surely xpdvor.

tpépw, pépw. Schol. Ar. Av. 1404 éxdorn dvd} Atovicov
(-ow!) tpépet SOvpapBorodv. dépw is a technical word for
tribal appointments, eg. ’A@. IloA. 56. 3, while rpédw
maintain would be odd here.

tropia, drowia. Arg. Ar. Vesp. 14 read 6 8& mats éretparo
ddatpely Tas aromias tod mpdyparos, vouberav Tov ‘yépovra.

xpdvos, Adyos. Stob. Hel. 3. 6. 14 (6. 31 Mein.) ascribes
to Socrates the saying otre wip ipariw mepiorretAar Suvardv
obre alcxpov dudprnua xpdve. ‘Time can only veil things
long after. Read Ady, for words can cloak a thing at the
time. MSS. vary between the two words in Aesch. P.V.
875 and Pers. 713.
ON THE USE OF THE WORDS spaywdds
AND xapwdds.

1. Lippett anp Scorr in their Lexicon tell us that
tpaywods is first ‘a tragic poet or singer, these characters
being originally one,’ but ‘later, when the poets ceased to
act, the term rpaywdds, tragedian, was for the most part
confined to the tragic actor ..., the tragic poet being called
tpaywooroids Or tpaywdodiddoKados (but tpaywdds continued
also to be used in its old sense... ).’ They mention separ-
ately tpaywdoi ‘members of the tragic chorus’ and the
plural tpaydof ‘often used =rpaywdia.’ They therefore
hold that in Attic Greek the word has at least three
distinct meanings, poet, singer, actor. For xwywdds they
give only two, poet and actor, thus leaving out, perhaps by
inadvertence, the very first sense to which the form of the
word points.

These ideas are probably universal among scholars who
have not looked into the facts. In a way they have come
down to us from Ammonius or earlier writers whom he
followed, for he says xwuwdds Kal rpaywdds A€yerar 6
xopevTis Kal broxpirys’ Kwpwdorods b& Kal tparywSorotds of
motntai’ éviore d€ cvyxéover THY Stadopdv. But in the schol.
ad Dem. 5. 6, dtmoxpiras éxadovy of dpxaior robs viv
Tpaywoovs Aeyouevovs Tos Tointds, olov tov Eipuridyy xat
*Apioropavyy, Tods dt viv broxperds’ otro: b¢ joav Svo* Tov pev
deuvrepaywrioTny, Tov 6& tpiraywrioTyv" aitovs St rovs mountas
tov Spapdrwv tpaywdovs Kal tpaywdodidackddovs, in spite of
the obvious corruption (which Lambinus, Sauppe and
others have tried to put right),! it seems affirmed or implied

1 See an article in the Amer. J. of Phil. 29. 206 (1908) by Capps.
Tpaypdos AND Kxapmdds 335

that ‘the ancients’ did not use rpaywdds of an actor, but
did use it of a poet. The object of this paper is to
question the tradition and to ask what grounds there are
for saying that in good Attic the words were ever used
of either poet. or actor as such.

2. It is clear from the very form of the words that
tpayyooi and xwywdoéi originally and properly meant singers
of a certain kind, that is members of the tragic or comic
chorus, who before the introduction of actors were the sole
performers.1 This sense of singers appears throughout
Greek literature, and there can be no doubt about it.?
But in certain phrases and contexts the words came to
signify the performance rather than the performers ; not
the men in themselves, but the men as doing something.

The common form of this is a phrase which occurs over
and over again in inscriptions and is found also in speeches
or the documents, genuine or spurious, which they contain
and in a few other places. The phrase directs a proclam-
ation to be made, or speaks of something as done, rpaywSois
or tpaywdots xawots. See C.I.A. ii. Index s.v. tpaywdoi :
Aesch. in Cites. 36, 45,176: Dem. De Cor. 84, 115, 116.
Cf. Plut. Mor. 603c: Luc. 5.51: Ael. Vi. 2. 13,

The dative is half local, half temporal, like Atovucios,

’ From these words are formed directly nwpwdodiddcxnados and
-adla, rpaypdodiidoxados and -arla (the last abstract substantive not
found, but no doubt existing). Xopdy diddonev, rpaypdiav biddonev
are both idiomatic (cf. 5:@upayBodiddenados), just as in English we
can ‘teach’ both a class and a lesson. The spelling of these words
and of xwupdoronrhs is established by the metre in Ar. Hq. 507,
516: Pax 734, 737, and so with rpvypdoropoverny in a fragment of
the other Thesmophoriazusae (333). But metre is not decisive as to
tpvypsorods in Thesm. 30 or rwumdoraxav in Vesp. 1318, and in
good Plato MSS. the spelling of both tpaye3(i)omoids and Kap@d(t)o-
motds varies. The forms in co are favoured by analogy, for the first
part of words in -roids usually gives the thing made, ¢.g. éromoids,
abAomotés; but the grammarian Moeris says Kwy@domoio) ’Arrixds:
Koppdiomoot ‘EAAnviKas, and kwu@doypdpos, as against the usual
Kwppdioypdpos and tpayydioypdpos, is guaranteed by metre in Anth.
P. 7. 708, ascribed to Dioscorides.

2 Compare xibappdds, abrAwdds, medwdds, fpappdés and any other
such words. A foaq5és might compose his own verses, a xOappdds
his own verses and music accompanying, but this is in no way con-
veyed by the name, which only connotes the performance. So with
aoidds itself.
336 ON THE USE OF THE WORDS

Tlavafyvaio. It means ‘at the performance of the
tpaywbdoi,’ ‘at the tragic performance,’ ‘at the new tragedies.’
Tpaywdois cannot signify simply the men. It must mean
the time and place at which they appear, their appearance,
their performance, just as in Latin a thing is said to
happen gladiatoribus, that is at the appearance of the
gladiators, at the gladiatorial shows (Cic. Phil. 1. 15. 36,
ad Att. 1, 16. 11 parallel with ludzs, and elsewhere).
When xawvots is added, this is still clearer. To whom or

1 In Greek we may compare the datives (euplois, votfots, Bopetors
(Ar. Hist. An, 9. 28. 2: 6. 19.4: 8. 12. 10), meaning at the time of
such and such a wind, when it is blowing, and the genitives amn-
Aidérov, Bopéov in Thue, 3. 23. 5: in Latin Virg. Aen. 9. 668 pluvia-
libus Haedis, Juv. 9. 68 aquilone Decembri, Tac. Ann. 3. 28. 3 pace
et principe. Perhaps we may add Ar. Eg. 410 4 wh wor’ ayopalou Aids
omhdyxvoist mapoyevoluny, for 4 man cannot in strictness of speech
be present at the orAdyxva, but only at the doing of something to
them. Aaunds, cuvwpls, etc. are used as names of contests, e.g.
Aaurdda tpéxewv, Th Aaumdd: trav walSwv, éré0n ovvwpls. Closel:
akin is that extremely common Latin and occasional Greek
construction, by which a substantive or pronoun with a participle
stands for something done to or by the person or thing. Thus
mwapicay mapowourrt avrg (Antiphon 4. 1. 7) is perhaps in essence the
same as omAdyxvots mapayevécdat, Aesch. Pers. 728 vaurinds orpards
kakwOels meCby Sreoe orpardy ; Herod 8, 131 robs 8& “EAAnvas Td te
ap yryvdpevoy fyeipe xal Mapidmos ev @eacadln eév are good examples
of a large class. The best English parallel that occurs to me is the
university slang by which we put e.g. ‘the Eights’ for the races in
which boats so named take part ; ‘the Eights are over,’ ‘I saw him
at last year’s Eights.’ Like rpaypSof, the boats stand for the per-
formance, and the expression is at once local and temporal. College
boards sometimes bear the notice ‘there will be no river to-day’ :
those who read it understand not that there will be no water in the
river, but that the college boat will not go out. ‘River’ stands
for ‘practice on the river.’ So ‘the Old Masters’ are (an exhibition
of pictures by) the Old Masters: ‘is there to be an Old Masters
this year?’ ‘The Christy Minstrels’ is at least half a place or
performance. ‘Punch and Judy,’ ‘a Punch and Judy,’ is a perform-
mance ; sometimes perhaps the men and things necessary for it.
Often places are put for events that occur or have occurred there :
‘Henley’ for ‘ Henley Regatta,’ ‘ Waterloo,’ and ‘ Trafalgar’ for the
battles. So Plutarch Comp. Aristid. et Cat. 5 6 Mapaédv, % Sadauls,
ai MAaraal, and Mor. 4968 rhy Sadauiva ... Thy @emiorondéous,
and elsewhere. For persons cf. Lys. 25. 25 werd robs tpiaxoctous :
Lycurg. 124 werd robs rpidovra: Thue. 3. 68. 1 wera tov MFSov : ete.
na are very many such instances to be collected, were it worth
while,
tpaypdes AND cwoppdos 337

what in a tragic performance was the word new properly
applicable ? Not to any of the performers, nor to the poets,
for poets, actors, and many of the choreutae would
probably be old hands. What was new was the perform-
ance itself, the plays they performed, and the phrase means
really when new tragedies are (were) played. So xwpbois,
which is much rarer, in the decree of a deme (C.I.A. i. 585)
ordering a crown to be proclaimed xwywdois. Compare the
terms of the law given in the Mezdias 10 drav 4 ropa}
q+ ++ Kal of kwp@dol Kat of tpaywdol «.7.d., In which of rpaywdoi
and of xwuwdot signify like 4 roury not men but some-
thing that men do. So in Aeschines in Tim. 157 év rots
kar’ éypovs Atovuciows Kwpwddv dvrwv év Koddure : tn Cles.
41 yeyvopéver tov év dore: tpaywdav and 154 peddovrwv domep
vovi Trav Tpaywdav yiyverOar: itis not the men that yiyvovra.
but their performance. Add in Cies. 34 rpaywddy dyov-
fopévoy xawév (so Plut. Mor. 710 ¥F), where as before
xaivav is not applicable to the men themselves. But 7b.
204 peddAsvrwv tpaywddv  , ¥
Tous yap Tpaywoods mpGrov, ei BovAe, oxdret
ds dheAodou wavras.
342 ON THE USE OF THE WORDS

Menand. Sam. 245 otk dxjxoas Neyovtwv Tév Tpaywdav ;
Philemon (M. 4. 44: K. 2. 512)

7a 8 dpyvpwpar’ éeotiv 7 re Toppipa
> . ‘ eb eT, > > . ,
eis Tos Tpaywoors —,_” » ovK els TOV Biov.
XPHoUW’s
Diphilus (M. 4. 388: K. 2. 549)
@s of tpaywdo! acu ols égovcia
éorw déyew adrrayra Kal Totety pdvots.

Hibeh Papyri i. p. 47 (Hippias of Elis?) padAov ray
Tpaywoar o' ovTas dvdpefous Tov 61d. TavTos seneney ed’ dppovias

ade.

Plato Rep. 395 a ddd’ ovd€ Tot drroxperat Kwppdois Te Kat
tpaywoois ot atrot. Phaedr. 236c tva 8€ py 76 TY Kopwdav
optixdy mpaypa avayxaldeba. rovety dvramrobidvres GAAGAots.
Leges 935D tiv trav kwpmddv tpobvpiay Tod yeAota eis rovs
avOpwrous héyetv.

Aristot. Poet. 3. 1448a 37 ds Kapupdovs ovK dad Tov
oped ew AexGévras. 5. 1449 b 1 kat yap xopov kwpwdav oye.
mote 6 dpxwv euwxev. 22. 1458 b 34 ’Apippddys Tous
Tpaypdovs éxwpmdet, Ott & ovdsts av elon ev TH Stad€xT@ TovToLs
xpavrat.

[Aristot. ] Occ. 1. 4, 1344 a 20 4 88 ta THs Koopyoews
(6mtAla) oder Stadepoved eori THs TOV Tpaywdav ev TH oKEUT
mpos GAARAovs Sutras.

Diodor. 4. 56. 1 xaddov Sé dia thy THv Tpaywday Teparetay
a * - e f x * 2 ,
mouxiAn Tis Kal Sidopos toropia wept Mydetas efevyvexra.

Dionys. Hal. de vet. script. 2. 11 trav 8% Koppdav pipetrar
(i.e. Euripides) ras Aexrixas dppovias: 7b. 9 tape émt rors
Tpaywoous.

Plut. Cimon 8 tiv trav tpaywdiv Kkpicw: Mor. 870¢
irepraiovra Tovs Tpaywoors ddafoveia.

Dio Chrys. 13. 224 rzpaywdots éxdorore spare ois
Atovecios. 21. 272 7d wh mdvy pirciv rods tpaywdots pyde
fprotv. Cf. 11.7: 15. 10: 38. 39: 66. 6 and 29.
tpayesos AND copmdds 343

Lucian Herm. 86 76 rév tpaywdav Todro, Geds ex pyyavns
> ’ € t oe. € * < ’
émiavets. Tox. 9 éaorav tyiv of tpay@dot tas Tovavras
gidias ext rhv oKyviv avaBiBaoavres Setxviwow. Cf. Lupp.
Trag. 41. Anach. 22 rots 8€ ye kwymdois kat drooKkdmrenv

~ a 3 y 2 5 iP x 7 s
kat AoWopeiaGar epiepev és Tovs ToXiras. Leds tpaywdds is
the name of a well-known dialogue.

Pollux 5. 97 ékadetro 8? rapa rots kwpwdois kat éyxAactpioia

«7A. But cf. 7b. 101 and 6. 18.

Polyaen. 1 prooem. 11 orpariynua Odvaecéews of tpayydot
adovety, and tpaywddv 7 oKyV?.

Philostr. 197 of rév tpaywdav ripavvot.
Sext. Empir. adv. M. 1. 281 as ei kwpwddv jxpoaro.
Heliod. Aethiop. 1. 8 rotro 8) 76 r&v Tpaywdav.

When Ar. Av. 512 speaks of Priam appearing éy rots
tpaywSois, the editors rightly explain this (with one scholiast,
while another says something about rév xopdv ray Tpayixav)
as = év rats tpaywdiars. We must not however suppose that
it means literally ‘in the tragic poets,’ as we might say
that Richard ILfI appears ‘in Shakspere.’ “Ev is never
used thus in good Greek: even the use of wapd with a
dative of the author’s name seems to be later (rapa rots
xopmdois Pollux 10, 26: Ar. Lys. 188 é& AicxvAw has
variants and is very improbable). Tpayqdoé are still the per-
formers standing for what they perform: ‘on the tragic
stage.’ Equally clear, perhaps clearer, is the use in
Philemon. If rovs tpaywdovs meant the persons to whom
the things were serviceable, we should have had ois
tpaywdots, not els rods tpaywdovs. This passage deserves
special notice, as it lends itself very easily to misconception
and yet certainly cannot’ mean even ‘for the purposes of
tragic poets,’ but only ‘in application to tragic performers,’
‘for tragedy.’ No one will contend that eis o& xpyjoima
could mean ‘useful to you’ or ‘for your purposes’: it could
only be ‘useful to somebody else as regards you.’ The
344 ON THE USE OF THE WORDS

antithesis of eis rov Biov further points the meaning. In
the Hibeh papyrus the reference to rpaywdot as singing is
explicit.

After what has been said we can have no difficulty in
dealing with most of the above quoted passages: there is
no reason at all in most of them for thinking that zpaywdot
and xwyu@doi mean specifically poets ; it is still the performers
standing for their plays, their tragic or comic stage.1 But
a few places may be mentioned separately.

Commenting on Plato Rep. 395.4, Ast in his Lexicon
Plaionicum takes cwpwdol and rpaywdoi by an odd oversight,
perhaps only a clerical error, for actors. In such a context
this is manifestly impossible, and the words are usually (I
imagine) understood to mean that tragic and comic poets
employ different actors. But we can now see that the
words may quite well mean (1) that tragedy and comedy
have different actors, or (2) that there are different actors
at tragic and comic performances. Our choice of interpre-
tations is like that we have with regard to yopyyeiv
tpaywdois.

In Ar. Vesp. 650 rpvywdois is simply comedy: in. the
fragment of the Goeyeades Tév Tpvy~dav are either again
comedy, the comic stage, or strictly parallel with the yopav
that follows and used of the singers, the chorus.

The Jines Vesp. 1537, Pax 806, Av. 787, though of a
different kind, have been misunderstood in the same way.
In all three, certainly in the first and third, the words have
been understood of poets, but there is not the slightest

1 The use of odrupo: for a satyric play is partly parallel. See Ar.
Thesm. 157 Srav caripous rolvuv wots, nadciv eve: C.I.G. i. 1584
mromths catépwy along with roinths tpaypdidv and momrThs Kopwdiav
(for other cases in inscriptions cf. A. Miiller, Griech. Biihnenalt.
p. 391, n. 4); Strabo 60 *lwy év "Oupdayn cardpos: Athen. 407¥F év
"Ixaplots carupos : 1b. 420 4 ypdas carbpovs Mevédnuoy with Diog,
L. 2. 140: arg. Medea @epiorais cartpois: Suidas sv. Mpartvas:
mparos typaye caripovs: Horace A.P. 235 Satyrorum scriptor.
Hence garupoypdgos. Like rpaypdol, cdrvpo: is the name of the
chorus (from whom a Greek play so often took its name), but
tpayydol connotes the performers, cdrupo: the characters performed.
Once or twice it is the singular odrupos which is used: eg. C.I.G.
2758 iv and Demetr. 7. épu. 169 ev cartpw Kal év kwppdlars and
odrupov ypdwer avr) rpaypolas,
tpayp~dés AND xwopodds 345

occasion for this in any of them. In all three the words
are closely connected with yopés and describe the persons
of whom the xopdés consisted, as in yopds waidwv, xopos
maidikos above cited and probably in the zaidwy, etc., of the
inscriptions (Oivyis raiSwv). To these three passages join
that from Poetics 5, which presents no ditliculty. The
xopyyos tpaywdav of Dem. 21. 59 may be recalled in this
connection. It cannot mean choregus of tragic poets,
because each choregus was associated with one poet only,
nor is the relation of choregus to poet ever expressed (as
far as I know) by sucha genitive. The tpaywdo/ are almost
certainly the chorus, whose choregus the man was, though
it is just possible that the word may mean here of tragedy.
In the Parian Marble 54 xwuw[dav xolp[os yip|éOy is only
conjectural.

There remain a few passages still to be considered. In
Ar. Thesm. 391, if the text is right, tpaywdot and xopoi
seem to be distinguished from each other. But the
scholium to Plato Theages 127 ¢ quotes it as Oeatal xai
tpaywsixol xopoi, and this reading has been adopted by
several editors, including Velsen. Cf. Ach. 886 zofewy pev
tpvywoicois yopois: Lysias 21. 4 wading xopd. In three
passages of the Vespae (1480, 1498, 1505) tpaywdds is used
with special reference to dancing. Although Carcinus and
one at least of his sons here introduced were writers of
tragedy, it is clear that tpaywéds does not refer to that, but
is still the name of a choreutes, no small part of whose
professional business was dancing. There is no reason why
Philocleon should challenge tragic poets to dance. He
challenges the trained dancers. Photius 598. 4 records for
us that Aristophanes used the verb tpaywdciv = xopeve.
Cf. the glosses in Hesychius: tpaywdia: xopefa, kwuwdia and
tpaywdds' xopevtys, kwuwdds; in which we ought perhaps
to read tpvywéia and tpvywdds. Xopevew is seldom, if ever,
used with regard to choral song. In Pax 530 the sense of
tpaywdav is indeterminate, but aiAdv and peday help to
indicate it.

Let us take the last passage of the pseudo-Aristotelian
Oeconomics. Here, if in any place discussed so far,
tpaywdot seems so used as to apply to actors, not indeed
expressly, but taken along with the singers under a general
346 ON THE USE OF THE WORDS

expression. The other words hardly admit of our making
it here the performance. spay8oi must be the performers,
and when we consider the sense we cannot very well
exclude the actors, who have most of the épAia apds
dAAjdous. It seems to me that in this place the word is
used confusedly and applies to actors and chorus together.
I say ‘confusedly,’ because I think the writer had first in
his mind the idea of of tpaywdoi=the tragic stage, with
which sense the words zpos dAAyAovs SmiAia are not in
strict logic consistent. Be it remembered that we do not
know when this sentence was written, or whether the
writer was an Athenian and fully master of the delicacies
of Attic speech.

It ought to be pointed out here that the verbs rpaywda,
copwda are used without the limitations of the nouns.
Thus in Ar. Thesm. 85 Euripides the poet is made to say
é7iy TpaywdO Kal Kkak@s aitras Aéyw: and in Nub. 1091 ri
dai; tpaypdode’ éx tivwv; the parallel of cvvyyopotow and
dnunyopotor suggests that it is poets who tpaywdotow. It is
Euripides, not his choreutae or actors, that Aristophanes is
pretending to attack. The scholiast however, followed by
Meineke 1. 149, refers it to choreutae. So too Aristo-
phanes uses xwuwdd of himself, the poet: Ach. 631 ds
Kwpwdel THY TOAW Hyav Kal Tov Sypov KabvBpifer: 1b. 655
kupwdyoe Ta Sikora: Pas 751 otk idudras dvOpwricKous
Kwpw@dayv odd yuvaixas. For other uses of the two verbs,
sometimes in a derivative and not dramatic sense, see
Plutus 557: [Xen.) R.A. 2. 18: Plato Rep. 395= and
452p: Ar. Poet, 22. 1468 b 34; Dem. 18, 13 and 19, 189.

6. Our result so far is this: The evidence that in Attic
of this date the words were ever used of poets is absolutely
nil, and the evidence for their having been used distinctly
of actors is practically nil too. Thesm. 391 and Oecon.
1. 4 are the only passages that lend any colour to the latter
theory.

If from other Attic evidence we knew with certainty
that the words sometimes bore the meanings in question, we
might consider the propriety of taking them so in a very
few of the passages above given. The utmost which can
really be said is that a few passages admit of such a
meaning, supposing the possibililty of it to be established
tpayodds AND xawpwdos 347

otherwise. For instance, in Crates, Menander, Diphilus.
Timocles, Plato, Aristotle Poetics 22, of tpaywSot might be
poets, if there were independent and conclusive evidence
that the word ever bore that meaning. But it has been
shown that other well-established usages fully and naturally
explain the passages in question, and that no such meaning
is in any degree needed for them. To establish the sense of
poet or actor, what we want is some passage where no
other explanation is plausible or possible, whereas we have
seen that in all those above cited another sense is quite
satisfactory. Are there any passages, such as will be
quoted shortly from different sources, where a specific
author, e.g. Euripides, or a specific actor, e.g. Theodorus, is
spoken of as tpaywdds or xwumdds? If any such can be
found, it will need the most careful consideration, but no
number of passages proves anything in which the words
can quite well be ranged under another use. I should be
very sorry to say confidently that I have not overlooked
some such use, our lexica and indexes being as yet imperfect.
I will only say that I know of none.

7. It is very significant in this matter. that in literature
the words hardly ever occur in the singular. They are
almost always in the plural. If, as I have argued, they
mean (1) singers, in most cases forming a chorus, acting
and spoken of together, (2) a performance, (3) tragedy or the
tragic stage, comedy or the comic stage, (2) and (3) being
derived directly from (1), it is natural that they should
generally appear in the plural. If however a single actor
or poet could be so called, why do we not find the singular
occurring in that sense? An actor, a poet, is mentioned
often enough, but never by this designation. In good
Attic Greek I can cite for the singular tpaywéds only the
two passages Vesp. 1498 and 1505; and a fragment of
Menander (M. 4.300: K. 3.231) tpaywdds jv ayov Atovicra,
where tpaywdés would be an adjective and rpaywday is
conjectured. Cf. the adjectival use in Callim. fr. 8c
tpaywdds potoa: he has also (fr. 84) the obscure od zpov
pev ypiv 6 tpaywdds iyetpev, Which Schneider interprets of
the cock.

8. We have now to consider some evidence which may
very probably be thought to show that owtsede Attica the
348 ON THE USE OF THE WORDS

words could be used in the sense of tragic and comic actors.
It consists mainly in inscriptions relating to festivals at
which, though they were not Dionysiac, performances of a
dramatic nature took place along with others. We will
begin with two that appear in Boéckh’s Corp. Inser.
Graec. No. 1584 (Vol. 1), relating to the Charitesia at
Orchomenus and dated by Béckh about 200 B.c., gives a
list of victors by their own names and professional desig-
nations. The names may be omitted here as immaterial.
The professional designations are as follows: oiSe évixwv rév
éyava tév Xapiryoiwy: cadmorys, kppvé, paywdds, roynris
érv, addAntys, adtdAwdds, KiOapioTys, KiPapwdds, Tpaywdds
Kopwdds, ToyTys caripwy, broKpirys, wowTys Tpaywdidy,
troxpiTys, TownTyS Kwpwdidv, broKpirys.

No 1585, relating to another Boeotian festival and
belonging to imperial times, enumerates in like manner
(after some others that it is unnecessary to give here)
paywdds, wdavrys, KiOapiorys, tpayydds wakads Tpaywolas,
ToTHs Kawys Kwumdlas, troKpiTys Kawihs Kwpwdias, TommTys
Kawys tpayedias, troKpiTys Kawvas tpaywdias, yopavA7s,
veapwdds, carupoypados, d4 rdvrwv. In this Béckh plausibly
supposed that between the third and fourth victors a
Kwpwdos tadatas kwuwdias has been accidentally omitted.
In the first of these records the rpaywdds and xwpwdds (each
in the singular) are clearly distinct from the various
tmoxptrai mentioned subsequently, as they are from the
poets. In the second the tpaywdds madaias tpaywdias
is distinguished from the troxpurns Kawis tpaywdias, and
if we adopt Bockh’s suggestion, there would also be a
distinction between cwymdds radatds kwywdias and Sroxprriys
Kaus Kwpwdias.

We may place here a third inscription (965 in Rangabé’s
Antiq. Heill.), relating to the Amphiaraea at Oropus and
referred by Rangabé to almost Roman times. In this
after some other competitors we find zomrys carvpuv,
Tpayydds, Kwpwdds, romgTys tpayydias, TomnTHsS Kwpwdias,
émwixwov, etc. but, written small at the side so as to be
associated with ourys tpaywdias and rors Kopwdias
respectively, dmoxpirys with a name is twice added.
Tpaywdds and xwuwdds are therefore again distinguished from
troxpirys.
tpayedos AND copmdds 349

It is well known that Bockh founded on inscriptions such
as these his theory of ‘lyrical tragedy,’ supposing tpaywdds
in the first inscription to be the same as tpaywdds mwadatas
tpoywoias in the second, and taking rado.a tpaywdic. to be
an older type of tragedy, lyrical in character, in which new
plays continued to be composed. The records since found
at Athens (C.I.A. i. 973), in which waded tpayedia indis-
putably means only a tragedy that might be called old at
the date in question, e.g. the Orestes of Euripides in 340 B.c.,
have}thrown very great doubts on this, and it had been
much questioned even before. Cf. also the statement
preserved in Bekker’s Anecdota 1. 309 rév rpaywddr oi
pev qoov mada, of mada Spduara eioayaydvres, of Se
Kawvol of Kowa Kal pydémrore cicevexOevta. It is indeed
obvious that the very phrase xa.vol rpaywSoi above examined
implies an antithesis to something that might be called
madadv. Liiders (Dionysische Kiinstler p. 129), A. Miiller
(Griech. Biihnenalt. p. 387), and Haigh (Tragic Drama
p. 447, n. 7) agree that in these inscriptions tpaywdds is a
protagonist who made himself responsible for the production
of an old play. They do not however explain why in these
cases the actor is called not émoxpitjs but tpaywdds or
Kwpwods. ie a probably 2nd century inscription (J. of
Hell. Stud. 7. 150) Saoxpirys wadads tpaywoias occurs,
and subsequently rowraé and éroxpiral ray Kawav Tpaywdav. |
Even if the latter words were in use for actors, there must
be some reason why in the same inscriptions certain per-
formers are formally styled izoxpirai and certain others
tpaywSoi or kwpwdoi. It is no explanation to say that the
actor who brought out an old play (or part of it) was
called rpaywdds. Why was he called sot As we have seen
that in good Attic the words appear not to have been used
of actors, it becomes still more probable that the difference
of name followed some difference of kind in the per-
formance. ,

May not the words, as used in these and some other
places, signify a performance by one or two people in which
singing was paramount? We know how much there was
of this even for actors from early times. The commoi
properly so called and other commatic scenes are a marked
feature in all three tragic poets. Monodies, though they
350 ON THE USE OF THE WORDS

did not begin with Euripides, are prominent in his plays
and form the subject of an amusing parody in the F'rogs.
Cf. Paz 1012: Thesm. 1077. It is plain that these
monodies became very important in the later tragedy.
Why should not solos, duets, trios have been the per-
formances to which the names tpaywdéds and xwpwdds
sometimes refer? What had once been the part of the
chorus, when a play, e.g. of Euripides was first brought out,
might often in single scenes be quite well taken by one
voice. Monodies apart, we can find plenty of lyrical or
semi-lyrical passages in extant plays that might be picked
out for performance in this way, just as in our theatres on
benefit nights and others bits of many plays are often given.
Such would be the great Cassandra scene in the Agamemnon,
the commos in the Antigone, Phaedra and her nurse in the
Hippolytus. As actors sometimes recited famous speeches,
so songs and whole lyrical scenes from old plays may some-
times have been given, and of course new solos and duets
could also be composed. Some stories that have come down
to us, e.g. that in Plutarch’s Lysander of the song from the
Electra, and what he tells us in Nicias about the lyrics of
Euripides which the Athenians in Sicily were able to turn
to account, may support this conjecture—I mean only as
showing how popular such songs were. Songs from comedy
might not be so easy to find, but some would certainly
be forthcoming, and perhaps even anapaestic or trochaic
passages might be included. The tragic lyrics or lyrical
tragedies, which were so popular at Rome and elsewhere
under the Empire and which will have to be mentioned
again presently, would afford a good parallel for such a
practice of reviving old or producing new monodies. Nero
(Suet. 21 and 24) tragoedsas cantavit and had a hypocrita
with him, that is, he was himself a tpaywdds. Caligula how-
ever is said (Calig. 54) tragoedo pronuntiants concinere,
where tragoedus appears to be an actor. Philostratus (Vit.
Apoll. 5, 7-9) uses tpaywdss, tpaywdia, ddev, and pedwdia
of Nero. Cf. Plut. Galba 14 rpaywdiay rot aitoxpdropos. A
passage of Philostratus (Vit. Apoll. 4. 21) seems clearly
to imply that such performances were at a later date
familiar in Greek cities. Apollonius thought the Athenians
went to the theatre povwdias dxpoaropuevous Kai peAoroiias
tpaywdos AND xopuwdds 351

mapaBdacedv Te kal prduav, Srdcot kwpwdlas Te Kal Tpaywolas
cioiyv and was disappointed to find they only went to see
dancing. In much earlier times what are we to make of
the lyrical dialogue between Aegeus and another (1) person
which forms the 18th (or 17th) poem of Bacchylides? It
is exactly such a composition as I imagine two rpaywdoi
may have sung, and for some such purpose it must have
been composed.

If this practice existed, a tpaywddés would not necessarily
be a different person altogether from an actor. The same
man might be called tpaywdds and troxpirys according to
circumstances. The actor in an ordinary tragedy would
sometimes have to sing monodies or bear a part in lyrical
scenes. Perhaps even by this he became a rpaywdds: still
more, when he had little or nothing else than singing to do.
We should not therefore be surprised, if the same man
were spoken of by both names, and this would not prove
that the names meant just the same thing. With us the
same actor and actress may appear, or at any rate may
have appeared, in both comedy and comic opera. Yet the
two things are by no means the same. Singing does not
cease to be the characteristic feature of the one, though
the same performer may make his appearance in the other.

We have to consider very carefully four interesting
records of performances which took place at the Delphic
festival known as Soteria. They were first published in
Wescher and Foucart’s Inscriptions de Delphes (Nos. 3-6);
but may also be found elsewhere, e.g. in Liiders’ Dionysische
Kiinstler, p. 187, with a discussion of them beginning at
p. 112, and the first of them is given in Dittenberger’s
Sylloge (404 in ed. 1: 691 in ed. 2, 1900), who now dates
it not much after 270B.c. They vary slightly in details
but are very uniform, and it will be enough to describe the
first of them. I leave out, as before, the personal names,
which are given all through, and only put down the style
under which the various performers are enumerated. oid«
hywvicavro Tov dySva Tov Swrypiov—fawwdoi 2, xfapirraé 2,
xBapwoot 2, waides xopevrat 5, avdpes xopevtai 5, abdnrat 2,
SidacKador 2, tpaywdot 3, addyris, duddoKados, 3 More names
(apparently zpaywdoi as no designation is given: so regu-
larly) with atAnrys and éiddcxados, 3 more names with
352 ON THE USE OF THE WORDS

aidyrys and diSdoKxados, kwpwdoi, 3 with abdnrys (a didd-
cxados has probably dropped out by accident), 3 more
names with aAnrjs and di3dcxados, three more with the
same, three more with the same, yopevral kwpixot 7, iporio-
picbo. 3. In two of the inscriptions the number of the
maides and dy3pes is much larger, 12 and 15. In the
fourth an aidAyrys and Si8acxados are only recorded with 3
out of 9 kwpwdoi. All four have 7 yopevral xwpixoi: none
any xopevtat tpaytKoi.

At first sight we are disposed to say, as Liiders, Ditten-
berger, A. Miiller, etc., do say, that the three rpaywdoi and
the three kwu@doi are obviously the three actors of ordinary
tragedy and comedy. Perhaps they are. But, when we
examine the lists, we are struck by the fact that, with one
or two exceptions, each set of three performers, tragic or
comic, has its pipe-player and its teacher. The former is
less noticeable, because an actor would probably need him
now and then: it was however only for singing that he
was wanted. The remarkable thing is that there was a
teacher with each set of performers. As far as our inform-
ation goes, I believe teachers are never mentioned in
connection with actors properly so called. We hear of
them usually as teaching the chorus, that is they taught
singing and dancing. The teachers of tragedy are such
people as Sannio in the Metdias, 6 robs tpayixovs xopovs
biddocxwv (58: cf. 59 rdvra tov pera tadra ypovoy diddoe
Tovs xopovs). Xopodiddrxados is a more explicit word for
the same thing. There is no doubt that the dramatic
sense of diddoxev came from the original sense of teaching
the chorus. It is never used, so far as I know (though
Foucart de Colleg. Scaen. Artif. apud Graecos, p. 75,
quietly assumes this to be the meaning of the word), of
teaching actors who had simply to speak their parts. When
then we find teacher and musician going along with every
set of performers, does not the idea present itself that the
performers were rather vocalists than actors proper?
Perhaps it may be thought that, although only the actors
are specified, there must have been dramatic choruses too,
and that the teachers were for them. It is indeed difficult
to understand how a complete play of the type known to
us can have been performed without a chorus. Thechorus
tpaywdos AND xapwdos 353

could not be simply dropped out of any tragedy we have.
But with regard to these Soteria records (1) the lists of
performers appear complete, the names for instance of all
those in the dithyrambic choruses being enumerated ; (2)
in each of them there actually is one set of seven xopevrat
xwpixot mentioned with their names, though no xopevrai
tpaytkoi are mentioned at all ; (3) the critics above named
agree in thinking no ordinary dramatic choruses to have
appeared and the yopevrai xwpxoi to have been dancers.

‘With these Delphic records it is natural to associate the
Amphictyonic decree found at Athens (C.I.A. ii. 551)
guaranteeing safe-conduct, etc. to the ‘artists,’ i.e. of mept
Acévucov texvirat, of Athens. There are mentioned in it
mpéoBas: ’Aotvddpas momris tpaywdias, —pos tpaywdds.
Astydamas, if it is the famous tragic poet of that name,
carries us back well into the fourth century, and the
tpayyxdés has been plausibly thought to be Aristodemus,
the Athenian actor contemporary with Demosthenes.
This is the earliest example I have found of zpayw8ds used
in the singular of one performer, except the two lines of
the Wasps above quoted in which it evidently refers to a
XopevTys.

Other inscriptions in Liiders, pp. 183-186, afford a few
examples of tpaywSds and xwpwdds in the singular without
(I think) throwing any light on our difficulty. C.IG.
1845, referred by Béckh to the second or third century
B.¢., arranges for 3 aiAnrai, 3 rpaywdoi, and 3 Kwpwdot to be
supplied to Corcyra, and another in Le Bas (Asie Min. 281;
also Liiders, p. 181) for two of each kind to be sent to
Tasos, Liiders and A. Miller suppose that in these cases
the tpaywdot and kwywdot were three protagonists, accom-
panied by other actors and necessary persons like the 3
aiAnraé expressly mentioned, to appear in competition with
each other. Each tpaywdds and xwywdds therefore would
virtually stand for a company of performers. This is
somewhat improbable in itself and necessitates a quite
different interpretation from that of the Delphic decrees,
where three tpay doi are certainly not understood to be
three protagonists. How are we to explain the record in
an Egyptian inscription (about 250 B.c.: Bull. de Corr.
Hell. ix. 131) of one rtpaywdds, six xwpwdoi, and four

AA
354 ON THE USE OF THE WORDS

avvaywvicral tpayxo!? Some Delian inscriptions (Bull. de
Corr, Hell. vii. 105: Dittenberger Syll.? 692) seem to
give usually two xwpwdot and three tpaywdoi: once one
xopwdds. For a few more references see A. Miiller’s Griech.
Biihnenalt. pp. 384-388.

If the view here suggested be correct, that rpaywdds did
not for a long time lose its proper sense of singer, but that
an actor might be called zpaywéés on particular occasions or
with particular reference to any singing that came into his
part, it may explain why the well-known actor Neo-
ptolemus is called NeowrdAcuos 6 tpaywdds in Diodorus 16.
92. So late a writer very possibly meant by the word only
actor. But it is at least a curious coincidence that,
when the word is used, Diodorus is about to quote those
lyrical verses of sinister significance which the great actor
sang before Philip of Macedon. Perhaps the passage that
tells most strongly on the other side is one in which
Athenaeus 538 r, following Chares of Mytilene who was a
sort of chamberlain to Alexander and wrote memoirs,
describes a great festivity given by the king. Many
eminent performers of various kinds appeared at it:
tmexpiOnoay 8& tpaywdot ev Oeaoadds xat "AOnvddwpos kal
*Apioréxpiros, Kkwpwdot d¢ Avkwy Kat Poppiwv Kat ’Apiorwy.
It may no doubt be that these performances were rather a
matter of singing than speaking, but some of the per-
formers are known as actors, imexpiOyoav is used, and
the presumption is the other way, especially when we com-
pare Plut. Alex. 29, where the same well-known actors
(éroxpirac) are mentioned in connection with Alexander:
though Mor. 334 D rpaywdoi is used. A good deal turns
on the question whether we have here the very words of
Chares himself, or only the substance of them in the lan-
guage of Athenaeus, whose use of rpaywdds for actor would
not be noticeable. It must be allowed that he seems to be
quoting, but the point is not entirely clear, and elsewhere,
when he seems to be quoting, his words are not always
identical with the words of our texts. In the parallel
place 584 p there is no reason for thinking that 'Avépovixou
zov Tpaywoou is quoted from Lynceus: Athenaeus seems to
be summarising throughout. In any case the passage proves
nothing as to Attic usage, Chares not being Athenian.
tpayeses AND xopwdos 355

The last is the point on which I wish to lay most stress.
Many readers may think that this passage from Athenaeus
along with the Amphictyonic decree, and perhaps with
some of the other inscriptions, either amounts to proof or
at least raises a stronger presumption than anything I have
said can rebut. Nor am I prepared very strongly to deny
this, though I call attention again to the remarkable combin-
ation of 8ddcxado. with tpaywdoi at Delphi and to the
necessity of explaining the words tpaywdds and xwpwdds
where, as in Boeotian inscriptions, they are clearly distin-
guished from fézoxpirjs. But all the passages which it may
seem natural to understand of actors are at any rate non-
Attic. This may point to the actor’s functions being
different, song more than speech. It may however point
only to a difference in the use of words. Though we talk
a good deal about Attic purity, it is probable that in
practice we often fail to realise the many minute differ-
ences between the Greek of Attica and the various forms
of Greek current in other states of Greece. In studying
the promiscuous vocabulary of Xenophon, often curiously
unattic, I have had occasion to illustrate this point very fre-
quently. There would therefore be nothing surprising in
the fact, if it were established, that outside Attica zpaywSds
and xwpwdds sometimes bore a meaning never given to them
by a careful speaker or writer of Attic. There would
indeed be nothing surprising if under such circumstances
a Xenophon or an Aristotle or some other writer who
(unlike Aristotle) was an Athenian, and (unlike Xenophon)
lived mainly at Athens, was now and then betrayed into
the less Attic use of the terms in question.

With regard therefore to the use of tpaywdds and Kwpwdds
for actor, although I do not believe it to be Attic of the
fifth and fourth centuries, I hesitate to affirm more.
Others perhaps will be able to throw further light on the
question. The new inscriptions of Delphi and the other
finds which we are now constantly making may help us.
If it be proved, as it may, that I am quite wrong, this
argument will still not have been wasted, should it lead to
a thorough mustering and examining of the evidence.
Many scholars have had something to say about the use of
the words in the above-cited inscriptions, but it has always

AA
356 ON THE USE OF THE WORDS

been taken for granted that the meaning actor was a vera
causa, a fact known otherwise for certain. No one, I
think, has tried to show that, at any rate in the best
Attic, the words never bore that meaning.

9. When Attic and later writers have occasion to speak
definitely of tragic and comic poets, by what names do
they call them? Aristophanes refers to himself eight or
ten times as (6) rowyryjs, three times as 6 diddoxados. He
also uses the words xwpwSoromrys (Pax 734), xupwdod:5d-
oxaros (Eq. 507), tpayedorows (Thesm. 30), rpaywSoddd-
cxados. xwpwdds does not so much as occur in him or in
any of the comic fragments collected by Meineke. Yet
he would have been very likely to use it, if admissible, of
himself or some other poet. In other writers besides
TounTyS, Tpaywdias OF Kwywdias rorys, Which seems the
regular and perhaps technical expression in the formal
language of inscriptions, zouris tpaywdiav (Apol. 22 a),
and 6 rouoas tpaywdiay or kwpwdiav, I have found without
much seeking :—

tpaynds Sannyrion, Alexis, Plato, Aristotle, Theo-
phrastus, Strabo, Plutarch, Hephaestion, Athenaeus. tpa-
yexds mornrys Aeschin. 3. 231 and xwpuxds rougrys 1. 98,
dvyp tpaytxos Phaedo 115 a. Aelian has zp. and «x.
mors 8 or 10 times, never tpaywdds. kwpixds Polybius,
Plutarch often, Lucian, M. Aurelius, Hephaestion, Athe-
naeus, Pollux, and much later Photius and Eustathius.

tpaywo(t)orouds Plato, Heraclides Ponticus zepi rév rpudv
tpaywoorooy, Aristoxenus ep! tpaywSoroav, Hephaestion.

xwpwoorotds Plato, Aristotle, Plutarch, Athenaeus.

Kwpwdorountys Pollux.

tpaywdodiSdoxados Aristotle, Isocrates, Dionysius H.,
Lucian, Athenaeus.

Kwumdodiddoxados Lysias, Aristotle, Dionysius H., Pollux.

tpayw0(t)oypdgos Polybius, Diodorus.

xwpwdoypados Dioscorides: xwpwd(t)oypddos Polybius,
Diodorus.

In the Poetics the word rouris occurs from thirty-five to
forty times. In many of these places the sense is general,
but in many of them, though certainly a minority, a tragic
poet is meant. Yet Aristotle never designates a poet as
tpaywdds, unless it be in 22. 1458b 34 rods Tpaywoors
Tparypdds AND xoppdés 357

éxopwde, which I have explained above quite differently.
On Athenaeus’ incessant use of xwuwdoroids and kwuixds
see below.

If an actor is spoken of, he is baroKpitys, TpayiKos or K@MUKOS
troxpirys (Xen. Symp. 3. 11: 6. 3: Dem. 5. 6: 19. 193:
57. 18: [59. 26]: Aesch. 1. 119 and 158), troxperis ris

-rpaywdias (Ar. Pol. 4. 17. 1336 b 28: Alciphron 3. 48. 1).
Of kopexoi ix. a line of Alexis (Meineke 3. 423. 13: Kock
2. 329, 13) must cover actors, even if it cover the chorus
too (cf. schol. Nub. 538, 539) but I have not found «capuxds
or tpaytxés used distinctly elsewhere in good Greek of an
actor, though perhaps they are. (If Alciphron 3. 71. 1
wrote 76 yop@ Tov KwyiKkdv ovAdapBdver (pe), TOY KopiKov
must mean either choreutae or poets; but perhaps we
should read 7G xwpixg.) Actors are also texvirat, mept
Aiévucov rexvirat etc., but like artifices this is a more
general word and covers all persons, sometimes even poets,
concerned with dramatic performances.

10. Whatever may be the true state of the case with
regard to inscriptions, in literature proper for a long time
after the Attic era it is difficult to find clear or even
apparent instances of rpaywdds or kwywdds in the sense of

_actor. Passages, far apart in time, of Chares and Diodorus
have been cited above. I cannot adduce any others for
the centuries covered by these names, but that may be due
rather to scanty knowledge on my part and to the imper-
fection of our lexicographical aids than to the fact that
they do not exist. Our Greek literature of these times is
also itself so scanty that no dearth of examples can warrant
us in alleging the words not to have been freely used in
this particular sense. The Latin use, which I will come
to presently, would be likely to react on the Greek, even if
it was not itself derived from a Greek, though not a good
Attic, usage. This doubt may be cleared up by those more
familiar than myself with the literature and inscriptions
of these centuries, or the discovery of fresh texts may
remove it at any moment.

We do not know very well how to date the treatise epi
Yyous, though the trend of opinion now is towards ascribing
it to the first century. The sense of zpayqdoé in ch. 15 is
not clear. In § 2 the author says, referring to the Orestes,
358 ON THE USE OF THE WORDS

évradé’ 6 romrys abrés eldev ’"Epwvas, and then in § 8 of
pyropes xabdrep of tpaywdot BArACrovow “Epuas. At first
sight we take tpaywdoi like rowrys in § 2, but he goes
on kai ob éxeivo padeiv of yevvato. Svvavrat dre & A€yov
"Opeorys ‘uebes x.7.d.” havrdteras rad’ ore patverae which
looks rather as though rpaywdod were illustrated by Orestes
and therefore meant men in plays, actors. The latter
sense seems for other reasons much more likely than the
former, but on the whole I believe this to be a rather
confused instance (like that, perhaps, above in Oecon. 1. 4)
of of tpaywdoi for tragedy. :

In Plutarch it is pretty clear that such passages as the
following exhibit the sense we are seeking: Cicero 5
‘Paokiy 7G xopedd...Aicdto 7G tpaywdg: Phocion 19
kai Tore Oewpévwr kawwods tpaywdots "AOnvaiwy 6 pev tpaywdds
eioévan pedAAwy BactASos rpdcwrov x.7..; Sulla 36: Pelopi-
das 29: Demetrius 34(%): Moralia 334D-F yeyovacr 88
wept airov (Alexander) rpayedol pév of wept @erraday xat 6
"AOnvddwpas . . . kwpwdoi 8 Hoav ot rept AvKwva Tov Sxappéa'
tovTw 8 els Twa Kwpwdiav épBaddvre ortxov x.7.d. 4b. 785 B
TldAov 5& Tov tpaywdév «7.A. But at times, eg. Moralia
63.4 of tpaywdol xopot Séovrar Pitwy ovvaddvruy, the use is
not clear and rpaywdoi might be tragedy.

A few later passages may be added both for their own
sake and as illustrating the sort of evidence which is so
lacking for the alleged Attic use in early centuries. (1)
Arrian Epict. Diss. 1. 24. 18 tpayw8d mpocépyy, ob 7a
troxpity, GAN aitg rg Oidirods: 1b. 1. 29 tpaywdds occurs
several times, referring perhaps rather to a singer, and so
in 4. 7. 37; but 3. 4. 1 xop~wde is pretty certainly an actor.
(2) M. Anton. 3. 8 ds dy ris elo Tov tpaywdov mpd Tod
tedéoat cat Siadpaparioa: dradAdcoerOau and 12. 36 ofov «
kopmdoy drolver tis oxnvis 6 wapakaBov orparnyds. (3)
Lucian Navig. 46 of rots Bacrrcis Sroxpwopevor tpayedoi
eEeAOovres dd tod Oedrpov x.t.r.: Anacharsts 23 describes
the dress, including the foot-coverings, of tpaywdot and
kwpwdoi in a way that clearly refers, at least mainly, to
actors (cf. Hpist. Sat. 19), and so in de Hist. Conser. 22
the high shoe of a zpaywdés is mentioned. In the Zeis
tpaywdds (lovem tragoedum, Suet. August. 57) there is
nothing to connect the word specially with song. Cf. the
tpaywdos AND xapdds 359

Tpaywooroddypa. Some passages again are ambiguous
e.g. Pseudol. 10 the proverbial 'IAeds dv tpaywdots euro buow.
(4) Phrynichus p. 163 (Lobeck) ob pévros évOa pev xoppdol
Kal tpaywdol dywviLovrar Aoyelov épeis, eva dé of atAnral
kai of xopol épxnotpav.

When Diodorus 4. 5. 3 writes kal Sarvpous dé paow
abrév wepidyerOar Kal rovrovs év tais dpxjoect Kal Tats
tpaywdiais réppw . . . wapéxerOa: 7G GeG, he must mean
songs by zpaywdiats. Later again the words are sometimes
used with a distinct reference to singing, e.g. Arrian Hpict.
Diss. 3. 14. 1 cs of kaxol tpaywdol pdvor dorat ob Suvdmevot GAAG
pera woAAGv: Lucian, Pisc, 38 rpaywddv twa. . . KexivypKopev
dodpevov Tas Ppvyay cuudopds. Sometimes the noun and
the verb appear to refer to the ‘lyrical tragedy,’ if so it may
be called, which was in imperial times so much in favour
at Rome and elsewhere, and on which we may refer to
Friedlinder’s Sittengesch. 2. 404, and sometimes it is
difficult to say how far plain acting is meant as distin-
guished trom the acting of a singer. See for instance Dio
Chrys. 33, 396 c rpaywddv twa émidypijoavta yvdxdouv
éridei~acbo. KkeXevovres: Lucian De Hist. Conser. 1 ’Apye
Aaos 6 Tpaywdds .. . Tpaywdyoas THY “Avépouéday, and two or
three passages in Dio Cassius in which Nero éxifapwdyce
kat érpayddnoev, etc. Cf. the cantare Orestem, cantare
tragoedias of Suetonius.

How rpaywd6 lasted on in the sense of singing appears
from the scholia on Theocritus (cited in the Thesawrus)
1,19 ddev S& 7d Kowds tpaywoety: 3. 38 docipar Awpixds
jyow doopat, 73 kawds tpaywdjow. Cf. Sophocles’ Lexicon
under this and the kindred words.

11. We come now to the use of tpaywdds and xwpmdds
definitely for a poet. The following passages must, I
think, be admitted to be fair evidence for the second
century of our era, even if there should be error in one or
two of them. Phrynichus Hcl. 258 (p. 291 Lobeck) & ru
kopwdia dpyaia mpooribepevy Tydrexdeidy TG xopyde: 326
(p. 344) 76 88 ‘épyoSoreiv’ rapa tut TOV vewrépwy Kwpwdar :
and in an extract from Phrynichus’ “Zo¢icr. Hapack.
(Bekker’s Anecd. 45. 33) d4Aov rowotvros Tod Kwywdod dre
«7.A.: Celsus in Origen contra Celsum 329 6 mapa ro
copwdd Zevs, and again 330 6 kwppdds yeAwroroidy tov Aia,
360 ON THE USE OF THE WORDS

meroinxe «.7.d. (probably Celsus’ own words): Athenaeus
35D Xaipjpwv 6 tpaywdds, though this must be unique or
almost so in Athenaeus, who uses xwpwdorouds at a rough
estimate 50 or 60 times, xwpixds very often, occasionally
tpayixds and tpaywdorods: kwuwdds I have not noticed.
Chaeremon is 6 rpaywdorods in 679F, 6 tpaytds 43.0,
562 z, 608 a. But probably there are other second century
instances to be added.

For the first century Plut. Mor. 88D 7d rod tpaywdod
dAXwv iarpos x.7.A., if right, is evidence, and there may well
be more. [In Philodemus de Musica p. 99 (Kemke)
Tpaywoav Tis 7) Kwuwdoroimv ... ody doov 7} KOapwdav the
words seem to be participles.] I say ‘if right,’ because the
three forms in -txds, -wdds, and -wdorédids easily get confused,
and this must be borne in mind even as to the second
century examples. Thus in Athenaeus 3c, 430, 101 F,
there is the variant xkwyixds for xwp@dorods, and in Lucian
Iupp. Trag. 32 some of the less good MSS. have cwpwdds for
kwpixds (Os 6 kwutxds épy), while 7b. 41 one has tpaywdds
for tpaywdorotds.!

12. It is worth while to add a few words on the use of
comoedus and tragoedus in Latin and on a Greek use which
was perhaps due to Latin. No one, as far as I know,
maintains that the Latin words were ever used of poets,
and this is an important confirmation of the view that the
Greek words too were not used in that sense. On the
other hand they were regularly applied to actors, though
this is far from proving that the use was admissible
in good Attic. See for instance Cic. pro Rosc. Comoedo
10. 30 ea pessimo histrione bonum comoedum fiert posse:
Quint. 11. 3. 91 cum mihi comoedt quoque pessime facere
videantur, qui etiamst iuvenem agant, etc.: cf. 2b. 181: 1.
11. 1 and 3: Juv. 3. 100 natio comoeda est: Mart. 6. 6. 1
comoedi tres sunt. No one will doubt that tragoedi are
actors in Plaut. Poen. 3. 2. 4 condoctior sum quam tragoedt

‘ It is curious that in quoting the Greek Life of Aeschines A
Miiller (Griech. Buhnenalt. p. 197, n. 4) has inadvertently substi-
tuted rpayydod for tpayehiowowd, and that Haigh (Attic Theatre,
P. 242, ». 2) has written tpaypdol for rpayySorool in quoting Plato
Prat, 425 v.
tpayedos AND xap@bdos 361

aut comict or in Cic. de Or. 1. 28. 128 vox tragoedorwm with
Quint. 12.5.5 vow quidem non ut Cicero desiderat paene
tragoedorum sed super omnes quos ego quidem audierim
tragoedos. So, too, we may-take Hor. Hp. 2. 2. 129 qua se
credebat miros audire tragoedos, though perhaps Horace
was thinking of the idiomatic use of tpaywdot for the per-
formance. In Plaut. Pers. 4. 2. 4 tragict et comici are
actors, like comict in Poen. 3. 2. 4 above, but as a rule
they are poets. Scaenicus is another word for a performer,
not an actor only: but see Vitruv. 5. 7. 2.

I do not remember to have seen it pointed out that in
the above passages and others comoedus seems to have the
generic sense of actor rather than the specific one of comic
actor. Quint. 11.3. 91 certainly illustrates from Menander,
but any one can see that there is no stress on the comic side
in any of the places cited. In Juv. 6. 73 comoedi ig an
actor in general, with some reference to singing, and ib. 3.
100 natio comoeda est means that they are all born actors,
not comic actors. The point is that they can play a part,
not that they can raise a laugh. In Mart. 6. 6. 1 tres
seems to indicate tragedy, but this is uncertain. But of
course comoedus and tragoedus can also be distinguished.
Thus Cic. Orator xxxi. 109 et comoedum in tragoedis et
tragoedum im .comoedits admodum placere vidimus.

The words of M. Aurelius (12. 36) above quoted, ofov «i
Kwopwooy drodve THS oKnVAS 6 TapaAaBov otpaTyyds, seem to
mean an actor of any kind, and so probably Arrian Epict.
Diss. 3. 4. 1 (cf. 10) crovddcarros kwpwdd ti. The text is
uncertain in Lucian 44. 1, but, whether we read kwpwdetv
or xwpodiay, it refers to the use of poetical diction and metre
and therefore clearly means acting, not comic acting.
Lucian 25. 4 xwpixdv Sopuddpyya and Bachmann Anecdot.
2. 329. 7 (quoted by Kock 3. 416) 8o0pudopypara . .
kod. mpdowra, & ocvvedépyerar piv Tois Kwpm@dodow, ara dé
obdty Siadéyerar suggest tragedy much more than comedy.
When we read in Athenaeus 620 D doxpivacbat ‘Hyyoiav
tov Kopwddv 7a “Hoiddov, ‘Epuddartov d¢ 7a ‘Ounpov, we can
hardly think that Hesiod’s lines were delivered by a comic
actor.

It appears likely that this use established itself in Latin
362 ON THE USE OF THE WORDS

first. Bentley suggested ‘as a guess’ that xwuwdia was
originally used of both comedy and tragedy. There is no
evidence of this, but in late times it may have been the
case with both xwywdds and kwywdia. Every one knows
that it is often so in modern languages, Littré in his
French Dictionary defines comédie as representing ‘incidents
ridicules, plaisants, ou intéressants’ and draws no sort of
distinction like ours between comedy, comedian on the one
hand and play, actor on the other.! Balzac’s Comédte
Humaine is far from being always comic, nor did he mean
it so. Though we should hardly use the English words in
that way now, there are many instances in literature, e.g.
Shakspere Twelfth Night 1. 5. 194 ‘Are you a comedian ?’:
Hamlet 3. 2. 304 ‘if the king like not the comedy,’ unless
Hamlet is quoting: Lear 1. 2. 147 ‘pat he comes like the
catastrophe of the old comedy’ (?): Colley Cibber’s Apology
(end of Ch. xi.) ‘her Majesty’s only company of comedians’

and often: Burney’s Evelina (letter 23) ‘not a village but
has its barns and comedians’: Burke’s Reflections on the
Rev. in France ‘They act like the comedians of a fair’:
Hazlitt (Works 8. 356) ‘fellow-comedian’ of a frags actor :

Lamb’s Complaint of the Decay of Beggars (end) ‘you pay
your money to see a comedian feign these things.’

On the other hand foreign languages are less ready than
English to apply the words tragedian, comedian to an author.
Even in English limitations may be noticed.

13. It may be well in conclusion to summarise the main
contents of this paper. I have suggested (1) that there is
no evidence for the sense of either actor or poet in
good Attic of the fifth and fourth centuries: (2) that
outside Attica towards the end of that time and onwards
there is evidence for actor, but not of a quite clear and
conclusive kind, and that conclusive evidence does not

1 A passage in Heine’s seventh letter to Lewald illustrates at
once the usage of two modern languages and the words of Juvenal
above quoted: ‘. .. alle Franzosen geborene Komédianten sind... Die
Franzosen sind die Hofschauspieler des lieben Gottes, les comédiens
ordinaires du bon Dieu, eine auserlesene Truppe, und die ganze
franzésische Geschichte kommt mir manchmal vor wie eine grosse
Komidie, die aber zum Besten der Menschheit aufgefithrt wird.’
tpaypdés AND xoppdds 363

appear till the first century after Christ, though we need
not doubt that the sense existed earlier: (3) that the sense
of poet is not found till the first or second century of our
era. But these points have been made (let me repeat) in a
tentative and interrogative rather than a positive and
confident spirit.

Postscript,

Mr. O’Connor in his Chapters in the History of Actors
and Acting in Ancient Greece (Chicago, 1908) adduces
against me along with other evidence and arguments two
fragments discovered later but apparently belonging to the
well-known and unluckily very imperfect inscription C.I.A.
ii. 971, which records dramatic victories at Athens. They
will be found in Wilhelm’s Urkunden dram. Auffiithr. in
Athen pp. 23, 28, and refer to the years 386 and 339 B.c.
respectively. The first, as restored, uses the phrase raAauov
Spaua mpdrov rapedidagay of tpaywdot, and the second waAatov
Spapa mpaerov apedidagay of kwuwdot. Mr. O’Connor, and
with him as I understand Professor Capps, do not of course
contend that the words here refer to poets. They take
tpaywoot and kwywdol to be the leading actors, one of whom
they, like A. Miiller, think to have brought out the old play
on each occasion after the practice came in. But, if this
were the meaning, we ought not to have the plural, as only
one play and one chief actor is in question in each case,
and we could not therefore say that the leading actors did it
for the first time. How are we to understand the words?
There is no need to take them of the actors in the
play as distinct from the chorus nor even of chorus and
actors together. They may mean the chorus only, for the
chorus were always regarded formally as the chief per-
formers (e.g. Plato Symp. 173.4 of Agathon rdmwixia ébvev
airds Te Kal of xopevtai, where the actors are ignored), and
the main competition was between choruses. The victory
of a play was the victory of the chorus, not of the actors.
I should take the phrase partly thus, partly as an exten-
sion of the old application of tpaywdoi and kwywdoi to a
364 tpaypdos AND Kxopmdds

performance. In truth the two senses, chorus and per-
formance, run into one another, as we have seen.

For the application of xwu@dds to a comic poet in the
second century a.D. Pollux 7. 201 rd 8& dvosa rotro etpyné
Tis Tov véwv Kkwuwdav Evdofos is further evidence, as Mr.
O’Connor points out. For the first century he cites Plut.
Mor. 88D, which I had in the meantime observed.
APPENDIX

ORATIO PROCURATORIA

360
: ORATIO PROCURATORIA
MENSE APRILI A.D. MDCCCLXXXVII IN DOMO
CONVOCATIONIS OXONIAE HABITA.

ProcuRaTOREs egressi sumus e laribus nostris ego et
collega meus: domum revertemur artium magistri. Utra
sit condicio potior, nolim ex hoc loco testimonium dare:
tantum dico, me hunc honorem nec mihi confidentem sump-
sisse cupide nec prorsus invitum deponere. Sed, antequam
potestatis nostrae insignia egregiis viris successoribus nostris
tradamus, quod more receptum est mihi quoque faciendum
esse intellego, ut quaecumque hoc anno digna quae memo-
rentur usu venerunt breviter attingam ad animosque vestros,
academici, si vacatis, revocem, et, si quid fortuna in alteru-
tram partem attulit, aut gratuler vobis aut dolore nos, sicut
aequum erat, commotos fuisse ostendam. De quibus dicere
aggrediar, si prius de eis quae ad nos procuratores praecipue
pertinent pauca dixero.

Iam ab initio anni nostri sperabam atque confidebam
animo omnia nobis tranquilla et iucunda fore; tantam
iuventutis eius quae nobis quodam modo commissa esset
sciebam esse modestiam. Nec me fefellit, sed, quod ab aliis
acceperam, id sum usu quoque expertus. Delicta, ut in hac
multitudine iuvenum, pauca fuerunt et ea maximam partem
perquam levia. Sunt sane graviora quaedam quae magis
occultantur, ut interdum haud facile dicas lateantne an nulla
sint, sed ne eam quidem contagionem latius manare arbitror.
Paene dicam nihil eos nobis exhibuisse negoti. Illud quo-
que saepe sum admiratus, cum ad me iussi venissent et
poenae eis nonnihil irrogari oporteret, quanta humanitate
etiam quod displiceret audirent. Concurritur ad procura-
torem a turba ‘mane salutantum.’ Ipse sedet, exspectat
quem quinque solidos, quem decem, quem etiam plures
pendere, quem domi post cenam coerceri iubeat. Adsunt
iuvenes, admittuntur, comes, urbani, etiam subridentes :
discedunt multati, non mutati.
368 ORATIO PROCURATORIA

Nonae illae Novembres ut non sine rixis, ita sine magno
tumultu transierunt. Cuncti vellem nossent quod plerisque
tandem est persuasum, eius modi rixis immisceri et etiam
quaerere quicum pugnet non tam animosi iuvenis quam
parum esse sapientis. Sub ipsum finem anni necopinantibus
nobis et vice-cancellario et invitis accidit ut severius in
quosdam animadvertendum esset, propterea quod ludis publi-
cis ob magnas pecunias ipsi equitantes certaverant. Neque
omnes sunt comperti. Sed, quot nescio quo pacto cogno-
veramus, horum commissa ut par erat vindicavimus.

Admonet me res ut hoc loco mentionem faciam statuti
quod nuper perlatum est de procuratoribus eligendis. Ea
de re diu atque acriter est disputatum. Una erat sententia
eorum qui omnes condiciones aequandas esse censebant
collegiaque omnia una aequitate continenda ; altera eorum
quibus alia causa maiorum collegiorum esse videbatur, alia
eorum quae minora essent. Sed quae tandem ea magnitudo
esset. parum inter ipsos convenit, neque aliis persuadere
poterant ut arbitrarentur aequi quicquam aut utilitatis in
his differentiis inesse. Accepta est lex quae omnia collegia
uno et eodem iure complectitur.

Magna est aliis de rebus statutorum multitudo, quorum
maxima pars in examinationibus versatur. Periit vetus
illa ‘in rudimentis fidei et religionis’ examinatio, quae si
verum confiteri libet adulescentibus nostris odio simul et
ludibrio fuit, fabularum fons, verae pietatis opprobrium
offensio periculum: exstitit nova quaedam quam voluntate
subeat si quis eam utilem fore crediderit. Reliquum est
ut primae quoque examinationi eandem adhibeamus
medicinam. Cautum est etiam eis qui minus litterati erunt
ut ex hoc tempore in prima examinatione ne necesse sit
Graece et Latine scire. Quo onere sublato homines id
genus et plures apud nos exstare oportebit et suarum
rerum scientiores: non quo dubitem quin etiam nunc
plurimum habeant scientiae, sed magnum est levamentum
litterarum detractio. Omnia autem enumerare longum est,
neque ignoro esse qui dicant ‘corruptissima academia
plurima statuta,’ quibus ego nullo modo assentior : immo
illud viventis esse et vigentis academiae puto rationesque
suas temporum mutationi accommodantis. Sed ad alia
transeo.

\
ORATIO PROCURATORIA 369

Aedificiorum quae hoc anno coepta sunt fieri vel
absoluta haud scio an par sit statutis copia. Insignissimum
illud! quod museo nostro adstructum est, non ob suam sane
pulchritudinem sed propter rerum quae ostenduntur prae-
claram seriem. Spectantur hominum, qui disiunctissimis
in locis diversisque temporibus fuerunt, vestitus orna-
menta, arma navigia, instrumentum supellex, operae artes
eodem undique collectae et tam sollerter dispositae ut nemo
non videat quibus ab initiis ortae quam sint ad has formas
pedetemptim provectae. Et summa beneficentiae oppor-
tunitate haec eo tempore a viro docto donata accepimus,
cum musei custodem eum habemus qui se praecipue his
studiis dedit.2 Quo adiutore et magistro constat acade-
miam quam maximos inde fructus esse percepturam.
Neque illud mihi praetermittendum esse videtur, quod
collegium Trinitatis nova quadam apud se officina?
ubi utilissimis artibus iuvenes erudiantur instituenda
benigne eis consultum voluit qui non habeant in vita quod
sequantur. Scilicet alumnorum nostrorum extrema haec
est spes et ratio. Iam dudum omnia conferta videmus
iureconsultis, clericis, ludorum magistris, ludorum inspec-
toribus, qui ita sunt multi ut vix cuiquam quicquam loci
superesse videatur. In his angustiis ad novas artes decurri
necesse est. Manu et fabrili opera enitendum est nobis,
academici, ut victum quaeramus.

Quod ad praelectiones attinet, instituit nuper artem
poeticam tractare professor novus,‘ thesauri ille aurei
conditor, qui sicut multos abhinc annos subtili eligendi
iudicio in poematis scriptorum nostrorum ordinandis usus
est, ita nune in ipsa arte exquirenda et illustranda dili-
gentiam et sollertiam adhibet. Quid quod Roscium®
summum artificem audivimus de scaenica arte pauca illa
quidem sed scienter disserentem : pictorem nobilem Poly-
gnotum,® dum munus professorium obit, etiam pingentem
vidimus?

1 The Pitt-Rivers Museum.

2 Dr. E. B. Tylor.

3 The Millard Laboratory.

4 F. T. Palgrave, editor of The Golden Treasury of Songs and

TCS.

6 (Sir) Henry Irving. :

* (Sir) Hubert von Herkomer.

BB
370 ORATIO PROCURATORIA

Encaeniis civem magnum, Ioannem Bright, hominem
honeste popularem et nullo artificio eloquentem, principem
nisi fallor huius memoriae oratorum eundemque in tempo-
ribus reipublicae plus fere quam alios videntem, qui praeclara
beneficia et saepe alias in civitatem contulit et cum maxime
confert, cum aliorum audaciae, aliorum ignaviae et timiditati
iam senex pro virili adversatur, hunc virum serius omnino
quam oportebat sed frequentissimo theatro plaudentibus
qui aderant cunctis gradu honoris causa donavimus. Alteri
quoque seni idem est honos habitus prope modum nostro,
nostra certe lingua utenti nostrisque hominibus non secus ac
suis noto, Olivero Wendell Holmes, qui nulla contentione
dicendi sed sermone lepido legentium animos oblectat,
acceptissimus mensae matutinae arbiter.}

Postremo nonnulla sunt quae de viris academicis cum
mortuis tum vivis dicenda esse existimavi.

Duas nobis mortes, quae quidem ex hoc loco commemo-
rari debeant, annus attulit. Morbo absumptus est is quem
eram primum sortitus huius muneris sustinendi socium.
Erat in Ricardo Shute ardor animi, ingeni vis, disputandi
subtilitas, morum summa mansuetudo, Delectabant eum
maxime ea studia quibus continentur universae scientiae
fundamenta et ipsa ratio cogitandi. Aristotelis scripta non
primoribus modo labris attigerat neque tantum inde exhau-
serat quantum ad iuvenes instituendos opus est, sed ea
quoque quae rarius leguntur penitus cognoverat et prae-
stantissime interpretabatur. Veritatem et amabat magno
opere et librum de ea investiganda scripsit. Idem communis
vitae negotia minime contemnebat neque erat ab eis alienus,
sed ad res agendas iudicium firmum adhibebat. Denique is
erat cuius memoriam non modo amici sed collegium suum
cuius causa multum laboravit et tota academia magno
dolore prosequendam esse duceret. Senectute est oppressus,
postquam collegio Exoniensi plus triginta annos praefuit,
Ioannes Prideaux Lightfoot, ruris amans, negoti bene
gerens, ingenio praeditus bono atque molli, qui iustitiam ita
colebat ut eam ipse ut qui maxime observaret, erga alios
plus clementiae tribueret.

Sed ad eos redeo qui vivunt. In vicem Ioannis Percival,
viri impigri adeoque laborum non dicam patientis verum

1 The Autocrat of the Breakfast Table.
ORATIO PROCURATORIA 371

etiam avidi ut sescentos pueros eorumque magistros regere
quam collegio non ita magno praeesse optabilius ducat,
electus est qui succedat Henricus Georgius Woods, cui
collegi Trinitatis praesidi bona optant omnes omnia. Summo
apud nos honore, quo iam quattuor annos industria paene
singulari functus erat, abiit tandem vir illustris, Beniamin
Towett, collegi de Balliolo magister. Per hos annos experti
sumus, pace eius dixerim, aliquantum, quid eventurum esset si
Platonis sui sive votum illud sive somnium est appellandum
re consequeremur, qui rerum humanarum ita optimum fore
statum confirmavit si aut reges philosophi facti essent aut
philosophi reges. Ille vero quo consilio, qua constantia, qua
diligentia, qua dignitate personam academiae sustinebat !
Te, insignissime vice-cancellarie,! qui in locum talis viri
successisti, laudare proinde ac mereri te omnes intellegunt
coram non audeo: habemus certe exploratum, dum rerum
nostrarum caput eris, numquam te commissurum ut commoda
nostra ulla ex parte minui patiaris. Utinam nec tibi nec
vobis, optimi viri, qui iam estis in his sellis consessuri ubi
nos nostra collegia aliquamdiu sedere voluerunt, plus
contingat laborum atque molestiae quam in tanta dignitate
forsitan putetis etiam convenire.

1 Dr, Bellamy, President of St. John’s College.

“BBQ
&Bovdos 208
GBpoeluwy 171
&Bvocos 130

ayabas 326

aydAAw 162 163 166
dydvvipos 164
ayaves 180 167
aydrnua 167

GyéAn 154 166
ayépwxos 171
Gyhparos 220
Gyhpws 154

ayndaAn 149 166
dyaads 149 164
GyAevehs 164
ayvebw 210

dyopedw 130 168
typurros 162
byxicra 206 207
adeApds metaph. 220
&Soros 134 168 206 210
aeliw 161
deluynoros 220
delvws dévaos 130 171
delorros 164
deawros 152
a0éui(c)ros 210
*A@nvala 130 140
&0pa 130

&Ouppa 161

at aiat 130
aiytaAds 131
aidotua 31

aidhp -éptos 131 170
aidés 181

at@pa 166 167
al@pios 161

atdwy 162 166
aindAAw 131 164
aivara 131

INDEX I

aiviyués 131 166
aiddos 149 164
airiacts 208
aixuave 170

alow 162

aidy 160 1638
dxdAuwros 168
axdparos 170
&kacna 161
axeatés 208
axrehs 131 208
akovgiws 210
deris 165
erAaad(o 171
arAdorwp 167
dayive 162
&aexros 162
GAéxrwp -opfs 154 161 163 164
GAcilnaxos 152
datos 166

&ars_ 181 161 167
dAurhpios 206
éArtpla 131
&Animos 149 164 168
GAAoTplwy, ev 43
GAAdpuA0S 162

“BAAvEis §=162

BAdrws te nal 233
GApupés 164
&aroxos 154
arom 131

das 164
aavond{w 161
&aAto 167 169
dagdva 163
dpa w.dat. 131 154 169
apardivw 131
Gudprow - 161
Gpaprwrla 131
Guérpnros 154
374

auhxavos 152
tpotpos 167
aumpebw 152
dpbve 131
auperw 164
aupt 181 167 171
aupiradas 171
ay 30 188
avd 181 164
avaBadnv 4
avayiyvécna 205
dvayvos 206 210
avadtoua: 131
dvaxetoOa: 154
avandralo 206
dvarplyw 206
avanés 163
avarshs 154
avdpeoros -rG 154 162 167
&vat &vaoon, 131 161 163 164
168 170
dvaraiderw 132
dvanvéw 192
évaronplrws 206
dvanrep@ 132
avantw 170
avaruvOdvouat 1382
avaplOunros 154
avapphyvuyt 132
avacxerés 132 165 167
dvavdns 166
avapoBa 154
avaxwp& 206
avexds 132 161
aveAenudvws 210
aveuduns 154
&vOeuov 161 164
avOeudins 152
&vOpwros 91
ave’ dv 132
aula -u@ 152162171 ,
avouticrws 210
avrdéios 163
dvravya 166
aytiAoy@ 154
dvrivmos 132
avttos 164
avritiyzwpodpar 132
avripeplCw 154
avupévaios 169
tvw verb 149

INDEX I

kia: 208 209
aka think 220
aradAn -nua 182
amapive 154
umapvos 208
amacrla 149
amavds@ 132
dreds 206
amelpwy 154
Greddyxw 208
amredoyhOny 205
ameumoA@ 182
annvhs 154
amrodindlw 208
aroid (eo 208
&moxaralo 154
amokéyw 154
amordtatos 206
aropepunpl(a 182
amoudpyvuse 149
andmruaros 149
amocrata® 149
amérporos 132
amdpevtis 208
aurqv 154
&rwbev 132
anwrépw -drw 132
dpa 132

apap 164
&pape 169
apyadéos 182 163 168
appyw 1382
api(hawros 154
apx@ 206 207
&poros 154
apphetws 182
dpxéaas 132
apxnyés 168
dpwyds 165
doeBotuar 220
&onuos 206
aonnOhs 164
doralpw 206
dooov 149
&araros 172
darevaxrt 182
aorparnpopa 132
aratpwros 132
dre 182
aréxerros 208
&repduwy 182
INDEX I 375

&rn 164 Bpéras 183

arnpds 182 Bpépos 166

arpeulCw 206 Bpovrnotxépavvos 155

abdadla, 149 Bpdretos 167

abdévrns 210 Bporés 133 165 twice

abdiryerhs 166 Bpbew 133

aban 154 Bptw 183 161 166 167

ative 226 Bvdds 149 164 165

avos 152 170 Bwodrw 149

avpa 182 161 Bworpé 138

aire 32 154

avrodaé 149 yaia 163

abrémpenvos 1:49 yaatrn -ds -i(w 183 167 172

(6) abrés oftos 384 yauhaAwos 149 169

aire 132 yava -vupat 1 26 133 155

avxnua 171 yatpos -i@ 133 161

abxye 160 yéynda 155 164

&paros 132 166 yeved 155

&pOrros 155 yévero 159

puntos 27 yevndS 168

agunvifw 161 yevvatomperhs 155

*Axapynlins 149 yepatds ynp- 183 162 206 207

axnvia 133 220

&xpis 170 yepalpw 149

&wpos -£ 206 yepovraywye 133
ii (x@év) person. 300

Balvw 183 160 167 169 ynpadéos 161

Bads 1 133 170 ynpoBookds 149

Barxeia 155 yAvndbdupos 155

BaotAela 40 yaadrra 117n.

Bdows 149 yropicr}s 208

Baonw 155 yoval 133

Baord{w 162twice 169 171 yévos 149

Bdcrduxpés 164 yons 133

Brua 133

Bla periphr. 170 daivupa: 166

BiBpdcnw 149 ddios 155

Bios 114 dals 161 165 171

Bloros 155 Sayatw 166

BaAdBos 183 208 ddpap 162 164

Baaoravw 183 168 Sdmwedov 155

Barérw 34 61 183 168 twice dapddrrw 133

Bopé 152 déinexclam. 193

BopBopdéupos 155 55H -re ©1388

Bordy 152 161 Sedirropa: 155

BovduTd 133 SeSdenuat 155

Bovah 188 Sénua 138

Bevanddépws 168 SefAacos 133 168

Bovaurés 188 Seina 149

Bpadivw 133 Bermatyw 155

Bpepw 154 Sexware 134
376 INDEX I

deftoduar 134 eyyiorra §=206

dedre 32 éyyin 59

dnrAadh 149 eyipynps 184. Cf. 171

Ané 140 165 167
Siayvduwv 206
Biadatporoua 167
bid wesOovs Exw 302
Siawréew 39
diamdvrios 167
BiaotiABw 149
Biaxpomar 155
diSdonw -kados dram.
Siepevva +166
3len =184
diknv w. gen.
Siofyouar 134
Sidwrns 134
Bios 164
dixa w. gen.
edidnaboy 134
Soxnoddttos +162

Sora 29 134 164
Sonhaw edéxnoa 149
ddaos 134 twice

Séuos 166 167 twice 171
dove 184 167
Sotarrhs 208

bépe. 149

5épn 168

Spouaios 155

dpa 6-206 207
dvaBovala 149
dvodalpwy 171
duondbapros 134
duoKAehs -ea §=155
SvoxoAdnapmros 155
SuonoAdnorros 155
Suocpevhs -eca 134 206
déouopos 134 169 206
dvomdAaoros 164
dtomoruos 168
dbornvos 134 165 168
dbopnuos 169
dvepops 134 162 169
bvoxeluepos 184
Bveddys 164

Saoua 134 166

Sapov 184

Swpotpa: 164 165 169

352

134 166 171 twice

134 166

eyyerfs 149

eyxove 134
éyxunAotpai 155
eyxerpa 134

dyb ydp... 3 28
weno 135 “iss 206 207
eliés w. gen. 209
eixoBorA@ 155
eludrepov 205 206
elxdy -d(m 49

elua 164

eluappévos -y 168 208
elvera, 168
eloaxovtl(a 170
eloGe impers. 55

ex w. pass, 206 208
éxate 185

exyevva 162
exdtidoxeo 185 155
éxxad¢w 171
éxvduios 135
éxovotws 210
éxmetpGuat 135
éxmepS 166
expapiad 167
exrlOnut 47

terowos 149
exprcyw 149
éaaclBpovros 135
éaacadvws 206 208
éaappés 172 206 208
éAcalpw 155

éAwvdw 135

&uBa 152

éupavhs 168 170
eumal(w 185

funas 164

zumvovs 206
éuwopevouat 135 164
éumupebw 149
éupephs 1385 162 168 171
zuppav 206
év=%veor: 168
évavta 152
evaonidoiua: 185
evdéxouar 135
&ouxos 149

ev800 170

évérw 167
INDEX I

%0a -Gev 149 168twice 164 166
evOev Edrav 256
tio = 179

€vowos 171
@vomAos 171

évrds 221
éfdxeots 155
ekaxovri(a 165
étakodw 135
éEaumperw 152
ekavaréd\Aw 162
ekavdpotmar 135
ekavéxoua 152
ékavoiyw 135
étamiyns 185 169 170
eteralpw 150
eEedpnua 152 162
einwepomedw 135
étijs mavres 94 253
éiind = 150
eEorxodoue 162
ekdAAuue 1385 162
eoupara 135
éLoudpyvup: 149
etdmiabe 185
étopu@® 149
étuBpl(w 220
érayratfw 149 155 161 166
émalw 135 :
erapgievvups 168
éravaBow 152
érdvw 163 166
emrdétos 166
érapea 169
émavpécda: 171
éref of time 186 166 170
émelyw 155

émjy 186

éxt 9

émvyava 167
emryAwTTamat 136
émlinros 136
émCety 136
émiardrrios 164
émixoupa -pos 155
éntvoia 186 163 171
émiobaa, ) 67
emippéemw 209
émtxeipa 155 208
éropa: 186 169
érouBpla 149

377

émomta 172

dros 186 161 163
émoupl{a 136
énrdmvaos 166
émwpeag 152
Epapar 152
épydCouce 162
Epdw 136

épelio 136
épérroune 161
épevva 208

Epis -i(w 136 220
éptérn 155

éprw 136 164 169
épuOporolxtAos 164
épxduevos 95 205 223.
ép@ will ask 257
épwrd 316

rw 164 170
érepos 80 89
érepos erepov etc. 297
evavOhs 155
ebdpoua 169
e¥dw 186 162 164
e¥epos 161
evqaAws 149
evxAehs -era 181 155
eturés 220
eVxuxros 149
evAoya@ -la 186
ebpaphs 164
evmevns 136 169 170
ebunxavos 153
evyn 155

evod@ §=163
ebradhs -era 162
eimdAapos 161
ebrarépera 169
ebmetas 206
eUntepos 136
ebpdrws 208
etpvduos 136
eboeBotuac 206
eVoxios 162
evotou@ 136
evowpaTra 136
edtpdme(os 167
edppalyw 136
evddns 162
epetw 149
éphpepos -tos 136

Cf. 229
378

epidrdAw 187
eplepa 158
épluepos 166 170
epornds 153
éxGalpw 137

Exe eidévar 295
éws 31

€ws ob 230

(d0eos 155

(dopos 168

Zdv Znvés-{ 140 155 168
(qr w. infin. 210

Copepds 171

(wrupo 150

dé 162 167
duAdyos -@ 168
jdupeAts 156
ovowos 164
qduradjs 165
neds 137

eovta eis, Td 208 209
qaArdQouae 187
HAtBaros 170
FA 153 206 208
HAbyn 150
huevos 170
Tupay = 67
Tuépios 156
hulOeos 156
qvioxd 156

mos 137 167 170

aniw 187

Aa ge gn3

 

 

Oaros 156

OddAapos 167
Oardaotos 164
Carla 156

OdArw 137

Oaud 137

Oapivd 158
Oavdornos 206
Odoow 156

ded 150 156 166 167
Oelyw 187

Oetopavhs 167
OeAnthpios 169
6éAw 185 137 168 206
beompdmos 171
Oedoenros 156

INDEX I

Bepuds 206

Ocopds 161 167

Oeomigda 137

Onrtppwv 187

Ohy 164

Onp& -Gyar -arhs 150 168 169
171

Ovryydvo 162

ols 156

Ovyanw 150
Ovnréds 156 162
Sotvn -Guar 161 164
Bobpios 153
Opéupa 153
Buyarépos 33 187
ObeAAa 156
Ounay 187
Ovpatyw 137 168
Ovpordwy 156
bupds 137
Oupotpar 137
Oupodpuevov, 6 207
Osos 162

Owrd 164

Owpuds 153

teuae 138 162
fepomperfs 168

Uw

ixvodua: 150 163 165
Ymepos -elpw 138
ivddAAopat 138
innnddy 188

imG 149 161
ix@udppovs 167

id 138

«ad éva 70
xablornut 210
xaBecia 138
KabuBpl(o 156
kawol Tpaypdoi 335
kaxoddkimos 165
Kakds kanas 55
xdxoopos 138
waives 171
KaAAterhs 156
KarhaAluoppos 165
KahArpatrefos 163
xkadbmroua: 138
kovaxe 161
INDEX I 379

képa 161 163 164 166 171
xapadonad 138
kdpminos 167
wapropépos 156
ndpra 150:
naprepés 138 169
Kapxapddovs 156
nard = 254
xarayAaltw 171
karédeoa 156
xaradorodpar 205
xatabvyjoKw 165
xarayi(w 171
naradovpa: 163
katatdarA@ 138
xaraléw 138 150
Karakpara 169
karadkapBdvw 205
xatdAeemros 156
xarévacdev 156
xaraoroda §=6153
karagréAAw 138
karapbelpw 165 168
Karappiyw 156
karaxhvyn 156
xarelBouat 138
Kareptxw 156
«droie 166 twice 169
ndrotvs 150
katémy 138
xarwxdpa 163
Kavxaouar 161 166
kéap 1 138
kexpagiddpas 156
xéAados -36 156 163
knarts 207
Kunraqonw 163
xwduvebw 199 228
Kivdpouar 138
xlpynpt 138 171
kAayyh 167
KAddos 167

KAadua 138
xaewvés 188 162 165 166
naéos 156 165 twice 170 220
kata 138 156
KAdvos -& 153
navsov 166 171
kaAtw 1388 162
wrdv 171

xvépas -aios 188 171

aynuls 156

kvwdarov 138 161
koadpa: 188 160 163 167
xoirn 156 161 171
nodoouprés 156
noutd(w -acpa 158 171
komi@ 139

«dros 158 166 171
xépyn -os 189 166
kopuds 153
Kopvooona 172
rouptdios 139
xpadatvw 139
xpavads 139 140
xpdros 162
kpeccardvws 207 208
xplvw 169 257
Kpoviins 140

xpovyéds 156

xreivw 156 208
xromos -@ 189 171
xuBepy@ 165 169 208
xvddCw 165

xudomuds 139

Kimpis 140
Kumpoyevera 156
kuptés 170

ropmdds 334f.
Koppdotpayydta 341
Kopgda 346

AdCuuae 189

AdOpq 157

Aaids 150

Adivos 162

AapBavoua: 220

Aaunds 8386 338

Adumw 189 169

Aads -of 189

Adoxw 189

Aagpuypds 139

(Aéyw) Zeta etc. 189 162 twice
163 166 three times 209

AelBouat 150

Aelmw 160 166

Aelavoy 150

Aexrés 158

A€oxn 166

Aevedaogpos 157

AeveowAnO@ns 1389

Aéxos 139 150 169
380

Aeds 139

Ahyw 150 167 170 220
Afioy 162

Aja 139 150 165
Anvatrns 157
Auralyw 166
Auraphs 150
Aimapéds 160
Aloooua: 140
Aixpopuos 157
Adytoy 140
adyxn 150
Aoflas 140

Aoxa 162
AvorreA@ 208
AwBnrhs 140
Agaros 140 162

pdnap 157 163
paxapirns 157
paddrrw 140
padrCards 140
payreios 157
parebw 150
péya adv. 157
peyareios 165
peyaroxdopov 165
peydaws 140
pedéwy -ovoa 140
pebhna 153
petdia@ 140
pelwy 165
peAavénrepos 157
perddnua 167
péAeos 165 170
perera 167
perlCona 163
HuerrAovy 159
BATH 150 166
pevowe 150
pevos 151 171
pépyva §=158
péoaros 140 168
perdBovdos 157
perayevhs 168
MeradAdoow 165
perdpowos 140
petaxwpa 157
perewpoxon® 157
pérexos 207
pérpa 249

INDEX I

wh 140 161 165 twice 214 274
288

pAsopat 157

pinos 140

pnvipa 207

pams 169
pntpddev 140
penxavh etc. 140 208
pialyw 207 220
puapla 207
pivbpouas 141
puvupds 163
puodmoAts 151
pvela 153
ponudouvey 158
pvhpov 141

poreity 141 161 163 165 171
wet be 157 167
Of 14
ies -la 151 157
povoxoira® 157
povodpayris 151
poveota® 157
povootpar 141
udxGos-& 153 157 162 169
poxAeuThs 153
pubetw 171

publfw 141

podos 141 162
puxdpar 157
pupiorAnOjs 166
puoapds 157

puxds 157
popopa 141

v, &long before 83
vaérns 166

vatw 165

vapa 171
vavotoA@ 141
vautiros 166
vavpaperos 141
veadns 141
veapds 157

vetkos 151

véuw 208

veorala 141
veorrotpopa 157
veoxuds 153 161
vepédn vépos 141 161
vy Ala 174
INDEX I 381

vimios 141 207
vywitios 142

vida 165
vurdBovaos 142
vindpatw. gen. 208
vipders 15

vénua 142

véoros -& 142 162 171
voudecta 157
vurtepelotos 142
vuoraxths 142
vitro 142

vixios 157

OyedAAopae 157

éyea 151

bd 6 avhp 208 209
ddoimAavG 142
ddoumépos -@ 151 162 169 170
636s 88

édévn ete 142 168 twice
of of 170 208

ola how 151

ofyw 171

ofSanev 207 208
otxnors 142

oixhtwp 165

olxos 153 169

oixtipw 142

oixrpoxod 157

oluat think right 252
dis 142

olorp@ 168

olws 142

BABos -tcos 142 169 171
GAryodpavhs 157
6AAbw 171

GrAopuppds 157
bAdpwvos 161

Suamos 163

BuBpos -tos 151
bulxan 142

Supa 142 163 167 twice
Sudmrepos 142
dudsomopos 165

duo 151

bynots 168 twice 169
évnoipdpos 171

dvivnur 142 169 207
SFdOvpos -obpat 142 165 169
dtuxdpSios 151

ban 142

brAtois 157
émrebw 157
émrhp 208

brama 166

brows 148 212
épyh 148

bpyia = 148

épy@ 157

6pexOG 158
GpOotpar 208
éptBdrns 151
éplyw 166
Spxwpora 143
Gpobw 143

Gppwid 143
dpGya: middle 157 161
domep 22 194 287
Boris 51

briph 214
érorb(w 157 171
ov after dore 181 201
obAos 157
ovpdvios 157 161
ovpavouhkns 157
odptos 153 171

maryyAukepés 157
mdyxpnores 153
mdyxu 165

maudt, ev 272
mada tpaypila 349
Twarayevhs 167
madaoyeris 157
madraids 151
Tadatod, é 207
maddpn etc. 148 164 171
Twahapvaios 163
maAlyxotos 151
Tlaaads 140
mdAAw 143
nopBactrea 143
Tauptoapos 157
napmnota 148
mwavhpepos 153
mavoupya@ 209
mavoAeOpos 143
marat 143

mapd 209 343
Twapacrar® 153 165
mapapépw véuov 207
382

mapevoaredw 1538
mapolxoua: 161
mépos 140
mdoxw Te 95
mdrayos -@ 143
mdtpa 143

matrw 143
mavpot 157
mavoTnp 167
meddpotos 140
médsov 143 166 twice
meCouaxa@ 157
wetOapx@ 161 163 171
(weidw) miOdv 148
meip@pnat w. partic. 205
meAdyios 166 twice
mwéAas 167 169
meAdpios 151
mévOos -& 207 220
menxpds 143 167
mémeipos 151
wempwuevos 166
mewTapevos 157
mep 151

nept 13

meptiiw 207
mepielAw 157
mepicarAts 148
mepleAvoros 166
meptvoor® 142
weptopyhs 170
meploeuvos 157
mepioxeAhs 168
mepiTeAAduevos 157
mepiprta 157
mépynut 143
merenvds 165
méepuna 243

my 142

whant 157

nijua 166
anpalvw 158
muvutés 166
mlovvos 157
mitvAebw 157
mAdtn 166
maaris 143
mrékw 170
waéws 143 161
mwAnoidxwpos 158
mwrovbuyleia 158

INDEX I

avoh 148 165

nodhvenos 162

modevds 148 165twice 168
mola 151

.rowh 207

mom 181

moreul(w 158
morguiorhpios 144
méAtopa 144 170
moAtocovxos 171
moditikés 382
mwohAov adv. 158 162
mododuar 144
modvmepos 158
moAbmAoxos 151
moAtppodos 154
moautiunros 144 165
moduTAtpov 144
moAvupvos 158
moAvpvaros 162
moAbxpucos 170
mévros -tos 161 162 165
mova 289

mop0@ 144

méptuos 144

aépos 144

méats 169

motavés 165
aétvia 144
morapar 144
mpayos 144
mpdxtwp 207
mpdovws 144
mpeuvoy 151
mpéeoBus 154
mpevpevhs 144
mply 102 151 317
mpivddns 158
mpéBnua 144
apéBAnua 158
mpoBovajs, éx 209
mpopavres 161
mpodéAuuvos 144
tmpoarclrw 144
mpdmodos 144
mpéppi(os 144

mpés 68 77 166 315
mpogeupephs 144
Tpoonyopeb@ny 166 168
wpoonvhs 169
mpdobe 144
INDEX I 383

mpocbryydvew 165
TmpochapBdvouet 144
mpocduotos 144 158 168 172
Tpoomenatrtadevpévos 169
mportpdmatos 207
tmpoopephs 144 166
mpoopidhs 168 170
mpéapopos 144 171
mpocpiw 158
mpoopdds 171
mpdowbey 144
apétepov # w. subj. 207 208
mpobvonoduny 209
mpovoedoduev 151
mpoxoh 158
mpdxous 158
mpara adv. 144
mpotns, ard 209
mpatioros 144
mpwrémeipos 167
arepoddyntos 145
mrepders 158
arepwrds 145
ntép0os 162
Tlvypatoe 259
muxadwm 161
muxivds 166

mupy@ 158 169
aiprvoos 170
mupmoA@ 145
mupsdns 158

pélw 165

pevoas 20

phyvums poviv 158
pimrd(o 145
pd0coy 158

poiBdos 158
polfnpa 145

pdun 145

piopa 169

caxerpdpos 163

aodkos 166

odrupos -ot 344 n.
capnvhs -ea 165 209
oeBiCw 151

oéBw -ona 154 158 207
gedayas 145

ceAnvaly 151

ceuvds 158
oBévos -w 145
oratoupya 158
oxiaypapeé -ia 292
oxémedos 158
oKoTtiwpodpat 158
opepdardéos 158
omaviCw 145
onapdttw 151
omeppatoAdyos 165
omAdyxva 145
onopd 169
otéyn 165 167
orelxw 165 167
aorTéAAw 220
arevagw 169
orevaxw 145
arépvov 167
ateppés 158
orépos 161
orlABw 149 166
oripos 151
ortxes 145
aroAh 145 169
ordaos 145
orpatios 158
orupedAryuds 158
cuyyvwords 145
ovyypapw Slkas 209
ouyKerpapar 145
abyxorros 171
obfuyos 145
avAAhrtwp 207
oupBovdetw, ob 241
cuunapacrdrns 145
cuumpdetwp 207
ouupéepopar 209
cuupépw 97 145
cippowvos 158
otv 145 160 162 164 165 167
207

cuvavta 145 163
cuvéumopos 172
ouvéropat 165
ouvinut 151
otvvouos 165
ouvvod 151
ouvtvxla 158
cuvpdds 154
cuvwpts 336 n.
awalroris 145
384

oéooua passive? 236

tayds 146

Taras 146 162

raradpwos 146

ravaés 170

tavitrepos 162

tapatindpdios 151

tappts 166

Tavpopdyos 158

TabTa...rauta 188

Taxéwy, 3d 220

tdxos, eis 151

a» » wsete. 146

taxbBovros 158

taxtve 158

te 146 207

réyyw 158

teipw 158

réxvoy etc. 146 162 twice, 163
168 169 170

téxos 154

rextalvouar: 146

tTéuevos 286
Tépua 158
réprw etc.
71
terpeyalyw 158
rertvyopdpas 158

tebxw 166

Texvalw -acua 146
rexvapar 146

rnaavyhs 158

rTnrepavhs 168

Tyros 146

tnpodua 146

rlOnut 168 171

TiywpS 210

tlyw 167

TAhuev 146

rajva etc. 146 163

rotos 21 69 165

toxebs 220

téApnua 147

7vév trois. pronouns 161 166

rotétns 802

rerophaw 147

técos 21 167

rovro pév... TovTe 8é +209

tpaypilay rod 6 14

Tpaypdds 3384f.

146 163 166 167

INDEX I

tpaygds 346
Tpaxudépuwy 165
Tpéuw 207
(rpéxw) Ope—- 159
Tpéxw Tmeph puxiis 3
tpidnovta 321
tptBh 147

TplBw 163

tplBwy 147
Tpiroyervhs -ea 140
rpvyydds etc. 147 841f.
tpvalBios 158
tTpupdrcca 158
tpupepatvw 158
tpbxw 147

tbuBos 147
Tuxnpas 147

tytela 38

bypd 158

bral 147
brévepbe 158
brepadyG 158 .
Srepnvopéwy 147
bréppeyas 147
brépraros 147
brepxord 151
baloxoua: 147
invd 147
iménrepos 152
broromotpa: 147
Smovpy@ 152
Sroxwpo 48

pudpds 158
géyyos 147 158
pepe 176
gepéoBios 165
ohun 165 220
épnoa 166
peeypa 147
pbelpw 161 168
pelvw 167
gpirdumedos 152
ptrdoxoviwy 165
prdtevos 160
paéyw 158 171
pase 147

¢68n 161
tmepdBnuat 159
PoiBos 140
INDEX I

golvios 147

ophy 147 165 167 168 169 171
pitta 147
gpovrts 207
gpdvrioua 147
gpoddos 147 209
guh 172

puadAds 159
@uAdroBoas 159
pidrov 159 207 208
pbs 165

guredw 147

gud 147 168

xatpnddv 1 147
éxdpnv etc. 159
xarérrouat 171
Xapace 152
xavidvw 152
Xdpy W. gen.
Xapua 165
xaoredCeo 159
xetua 165 171
XetmdCoua: 159 168
xelpioros 220

xepds etc. 32 169

xeés 152

xOdvios 159

xodv 162 166 170 171
XAG 152

147 220

385

xors 151

xopela 152

xpedy 148

xprteo 148

Xpiua 148 163 twice 165 twice
168 171

xpnorés 104 «

Xpuooxddivos 159

xpo wish 161

xpos 148

Xwpopirs 209

xwpG 148 162 twice 163 167

watw 162 207 208
Wevdnyopd 168
Wevdopxa 159
Wibupl(w 159
Wuxiis, tpéexw wept 3

Gide hither 165

copurds 148

as and os &y final 72 148 161
168 210

asuhen 99 148

», in wishes 58 148 169

=@onep 159

>, =dore 7 152 209

Sore ov and infin. 181 291

apeanua 154

wpedhomos 152
INDEX II

In most of the passages referred to it 13 proposed to alter the first word or

expression of the following pairs to the second. The very many pro-

posals for altering terminations in verbs, substantives, &c., for inserting

something apparently lost, and for making other miscellaneous changes,
are not indexed.

&- (neg.) Suve- 300
93 lost 321

byavaxtd drarud(w 251

- ary 251

byto Bkw 268

byw exw 328

K5emay ddnv 328

dw dw 30

dei Sely 252

+, Gmodes 86

ainlCopar diexlCouoa 284

aipoduat dvatp- atpouat 277 301

aisxpés yAloxpos 236

altia tka 47

»,  @ikla 105

airtouar (x)éxrnuat 328

a&xovaréovy yevotéov 52

&xpws d&upiBGs 329

GAnOhs wAROos 252

GAAd dua 298

BAAeoOat &ArAEcBa 24

Ada moAAol 187 213 329

karo abrol 213

dpeivov pelCwy 233 292

&y a@y 2138: &p’ 11: ad 271:
84 217 239 274 326: e& 46:
bv 326: ob 826: &yvlost 61
87 185 196 212 214 223 241
248 264 270 272 276 twice
322

&v- auv- 326

dvayeioat yavioa 25

dvarate: dvdnavow 108

avdpav tArAwy 268

dv éxdy Unaros by 86
dvingt tyw innt 64
avridoois dvdmavots 329
&vwbev dvabdpev 44

kiwo attw 329
anavrhoavres amavOho- 285
drep tcamep 258

amAGs 4AAws 329

aré brd 91 191 326
droxpivonat aropatvoua: 329

darecOar emibécOar 56
tpa Ady 244
» «yap =—-252
&pyarkéos devyaddos 279

Bpicros xpyords 103
ond 80x 329
&rowos &movos 95

’Arrixés Oearpucds 254

ab 84 «6381

atte done 177 190 226
aitdpxn ottws abtapytas(ofrws)
258

abrol etc mdyresetc 388 41 60
174 315

abtés abOts abtis 80

55 vis 181

abrés etc obros ete 43 216

247 258 261 328
airav patrdAws 223

Badi(wo xabl(w 305
Bhoopa: mdcouo: 294
Bip Bla 107
Bpaxts taxts 290
INDEX II

yép 8 93 173 218 240 273
327

ye TE 56 113 246

yeypaupévos mwempaypevos 118

yev- y(y)v- 228 244 261 329

yevérOat -hoecOar 182. Cf 329

vepnpopla iepnpopla 329

yapdoreo yap Béoxw 97

yoov vy &v 83

Aaperxéds rotédtys 302
bye 252
5h 48 259 266 327
»» lost 21°
Bewds aelvws 256
Sedpevos pedduevos 111
Seamdrns Seamdris 294
Bedrepos Bio 329
SynAotrar SedhAwrar 244
dijuos Byudrns 246
Shuou Shrov 8h 304
bid 8H )=—257
Siaxeio@ar SiavoeicOar 329.
Sioxeto Oar § 66
Brea Bety SieActy 330
Sucd(w KabiCw 3825
Boxetete éddxer etc 35 303 312
Spdooua diacmdcoua 269
Svvacbar SeduvijcOa 276
34 Sécw 199
ddcw Sel—w InrAdow 330

”

éautg én ait@ 9

ey- ex. 64 327

E(no’ ets 94

ec te 269 299 327
eldov Sev Fdn 68 330
>»  tyov 330

eldds eimdy 108
eiddres SteAdvtes 219
etm elvat 330

elvat iéva: 299

1, elv 830

eimetvy éxeivo 254
einévy émidv 307
eipnucvos dpicpévos 291
els tis 2382

eis ws 327

» to’ 324

e&i)s ev 264

387

elaOeduerat els apiOudy 302

eixov eaxov 222

ais Exe 181

$5 éw 3830

éx eis 5 84 91 327

», lost before « 6 281

ev én 244 245

éxet émef 194

€AeO aidoduar 31

éAfhAana é(hawka 307

Ow €béAw 114

eAdeivy afew 258

ay fev 52

eumoptCopar eee 100

év aby 35 106

» ey 251

ey él 74

evdpyeia evépyera 316 330

evdarba evavcacba 218

évex@@ AcxXOG 305

mm eR 74

évrijta: éeumpta: 41

éleAdoas éAndcas 46

étdy ety 230

Zoixa elwOa 112

éravG éracka 267

énGoat emimojoa 323

emeAmion emadrenpa: 258

émt (ér) aad (amr) 81 112 115

émiBovdAn emiBodh 330

eninpa eémtxepa 330

emiderinds emidétios 255

épnuoduat juepoduar 249

épduevos eaduevos 21

Frooua Epxouat 278

éorl @rra: 311

» 76 293

Erapixds épwrikds 246

ért eal =257

ev- &- 104 228 293 301 326

», Suve- 95 327

eBroros ebtedhs 266

eUmopos &movos 228

ebp- aip- ady- 330

eipov (nipov) «lAov
33 . €ltrov

236 279

79 105 186

224

éxparo elpyro 74

zexav (av 330

Qnteiy attiov 331
co 2
388 INDEX IT

H Hh 48 Aemm- Aimw- 331.
>» BA 98 Aerrés Ards 112
», os 35 Alav Ala 174
Hyotpa mowodpa 266 Avyupds trAapds 228

Hiiorov eOrardy 105
Hd6 érdo8 «106

» vy OC 253

ROos os 105 264 331
Hréqnoa eica 223
fama jydanon 306
jets tuets 175 186 327
-nuévos -@uevos 21

hv fe 381

Oavety Odvaros 289
Oefos datos 331
Oeds Ocios 299
Opéupa Odapa 225
Odpas Ouplias 323

eldévar 99
Aads 191
épya 38

»  epea 805
iepds Agios 331
tet and comp.
inavés adds

idety
Bios
iepd

rlOnu: 3805
107 twice

Kabjna Kadetxa 305

xablnu: rlOnus 3805

kal nan 62: by

1» Kalrot 99

Gs 658 175 176 191 260
271 274 328

» # 189 213 222 235 250
259 260 265 327

kalvopat palvoua: 300

kanéds Kadds 100 114

kanav évexa tov 201

naral mdaa 11

Kapdiqg Kal Sia 255

xatd «at 229

KataAtoes Katavicas 98

ndtw xalwy 251

dun Keparh 224

kbptos Katpds 256

31 276 328

AaB- Badr- 64 203 331°
AapBdvw parOdvw 3804
Aey- yev- 329

Adyw exw 381

Aurapés 110

Adyos xpdvos 111 248 307 333
>,  wWéyos ByAos 181

Adan bAyos 290

Aga Aeia 280

HGAAov pddwra 246 314
33 pévoy 246 261
a pérAre 223
petGov paocoy 264
pev misplaced 62 248 325 etc:
lost 322
péve perd 309
petd xard 290
beta ToAY pev obv
pevay 324
ph 3h 56
uh Alay vh Ata 174
4h olxodoay Meyapor 5! otcay 272
bende phre 36 225 274
picOds udxOos 107
bot pév 267
cor 90

MeTamoAdu-

”

vy y 40 72 252

», 7 200

véwy wo(t)ov 73
ving Sdever 108
voul(w dvoudtw 266
véuos Adyos 216 331
Hépins ‘Avataydpas 308
B ob 196

8 dé rdéde dé 214

of of 192

olde der 222
oinréov moinréov 240
olxos dyxos 98

»  €ixds 291

ofov doy 69

ofos motos 195 331

1 8aos 180

bAlyot woAdol 2 222
GAlyos dS Adyos 273
Bupa vinua 79
duovod duoroye 331
INDEX II

duovootcav dpuolav otcay 219
Suws Saws 252

> «OvTws §=258
dv lost after -ov
6 vots dudvous 118
bvras dé vovs 109
brov da érécm 331
brws mas 41

bpGs dptoas 79

Gp0ds vwOpds 283
Spria Sica 300
bpGuar ddropdbpoua: 232
bs ofos 87 265

»» Ore 104

5 268

dotwy Ovary 215
bri by bray 276

» 7 ef bre 221

ov e@ 104

ovdéy obdéva 229
ovdty by ebAoyay 72
ovdemore ovderdmoTe
ove oby 273

,, 088’ 46 192
ovnal ovx ei 324
obv &y 88. Of. 88
ovala oixta 246
otte ovdé 191 260
>»  obTrws 328
oUTws eidas 22

» Ovrws 7

53 étt ds 320
Poua: Tépwouae 332

a mt 197

wavTa midavdtnta 323
mavraxod mwavtaxdbev 65
mapa mept 1 50 228 328
>» mwavTa 231
mupaddvres mapaBdyres 35
mapad@oww brnperaow 332
Tapavoua mapavod 301
mapels moryels 332
wapirye mwapixe 288
mapodoa mpérovaa 332
mapovalay émtotcay 67
mathp Bsorjp 81

matpls mpanls 43
melOecOa ribecOar 242
meldw melrw 21 218
mendtnuat memolnuct

245 258 319

204 230

's

389

wénpaya mwémpaxa 288
mep wept omép 22 256
néoy maton 276
mupds xbptos 2138
mtAoduat elAov- €Axo- 244
midpevos miOdueros 83
ainrnot mlrrovts 64
midy mivwy 289
waatloratos mAatiorouos 295
mwAe(t)wy mAetotos 275 314
y> | Bleoy B32
mre(i)ov mAc(ijon 74
5 yérowv 306.
mo(i)nua mveipa 249
mo(t)hoacba Sencacba 197
molt) move 273
i axoTme 180
motxtAos muxwds 290
méAts madat 238
mwoAirns moAditinds 240
on énAltys 301
moditixés motklAos 332
wécos sopds 80
mworé tte 217
» Thott 82
mod ToD 257
Tmpayua ypduua 106.
e mpboraypa 308
mpds mpd 828
mpdce mpotorn 332

Cf. 118

ampoolerOe mpdoxetcbe 303
mpérepos mupétepos 213 323
“is trepos 332
mp@tos mpdérepos 98 201 314
‘3 mpsopopos 101
sy Tmprrov 87
arHkat mera 295

wip eis Sapeis 109
TWA® mow 238

clwmh atoAh 332

oxnvi oxevh 200

omovdaia yéAoiw omrovdaioy- 326
orjoa arivat 202

opadrepds aoparys 107

o@ (ov hxev tows avijcev 115
Swxpdrns “Iooxpdrns 306
capa oXijpa 239

cépata kweiy cwparixny 3826
cwrnola ebnpepia 110
390

Td aitd 298
thyahpa torkyorua 26
Tapdttw Ttartw 265
tadTa tabrd 193 197

3 «Toatra 180

»  wavra 226

te 8€ 32 267

» Te 87
vi 76 41 71. Cf. 63 108
tipjs idow’ 111
twa thy 100

to 768° 73
Totovtos togovTos 80 197
tolxwy tpenxiy 302
tévos tépvos 303
témos xpdvos 333
rotvoua Td véonua 68
robs mis 71

Tovto obrw 91. Cf, 221
tobrav y' otrws 57
tpépo atpépw 275

55 gépw 333

Tpépar ypdyar 292

INDEX II

imdpxw mapéxw 178
bmoyla aromia 333

gdvar pdoxew 81
gavepds movnpds 325
pépw galyw 271 273
gebyer mwépevye 182
gevy- puy- 268
Pbovs pwvd 305
popotpat pbepotpar 42
ppovs pera 293

xpévos Adyos 111 248 307 333
Xpopevos Kexpapévos 228
xépn Spy 280

wuxh tbxn 307
& 4 328

@s Kal see kat
1» oot 84 328
1» otoa 328

» @ 229
INDEX III

To the authors and many of the passages emended. Aristophanes, the

Orators, and the few books comprised in pp. 240-277 are net given ;

the names of the comic pocts ore given, but the fragments are not
specified. Italics indicate the page.

Aelian V.H, 2. 23 314: 3. 18
315: 9.3 805: 12. 64 816: 18.
3 811: 18. 24 327
Aeneas Tacticus 4. 7 829
Aeschylus Ag. 964, 1828 291:
Hum. 1, 207 316: 266 827:
497, 894 309: P.V.109 315:
1031 291
Aesop 2 328: 4b 880: 8 332:
23 328: 200b 314: 252 312:
274a 330
Alcidamas Soph. 2 3
Alexander Aetolus (Gull 15. 20)
3818
Alexis 61 66 83
Ameipsias 75
Amphis 81
Anacreontea 7. 15 288: 16 310:
17.6 312: 19 309: 31. 10, 32.
8, 83. 14, 84. 4 and 13, 35.
11, 36. 9, 39. 7, 48. 27, 50. 12,
58. 9 288-290
Anaxandrides 79
Anaxilas 82
Anonymous :—
Comic Fragments 174
Tragic Fragment 455 315
Vitae: Aesch, 307: Aristoph.
823 324: Plato 311 380:
Thuc. 327. See also 272
273 ;
Paraphrast to Aristotle’s Ethics
810 818 319
Philosophorum Fragmenta :
Mullach 1. 495 325: 2, 302
304 313

Anthology 9. 189. 3 311: 9, 826.
1 3822: 10.1. 7 819: 10. 84.2
330: Append. Pl. 4, 244. 2
828

Antigonus Hist. Mirab. 9 325:
25 and 28 317: 78 3823: 79
3826: 126 323: 144 317

Antiphanes 65 76

Apollodorus 112

Apollonius Rhodius 3. 1368 317:
4. 375 380

Apollonius Hist, Mirab. 49 325

Aratus 942 319

Archilochus 54 and 68 283

Aristaevetus 1. 138 328 330: 1.
20 316: 2.1 382: 2,9 328

Aristomenes 65

Aristophanes (Fragments) 61 74:
see Anonymous and Scholia

Aristophon 82

Aristotle [’A@. IoA.] 3. 3 and 5. 2
300: 7.2316: 12.3 315: 23.1
818: 28, 2 829: 45. 1 327: 60.
2322. Eth. 2.9 829. H. A.
9. 86 317. Met. 7. 6 311.
Ocecon. 2. 2. 4 823: 2. 23 310.
Probl. 29. 14 825: 33.7 299.
Rhet. 3. 6 299: 1.15. and 3. 15
3800. Somn. 2 299

Arrian Anad. 1. 17. 4 316: 2. 2.
3 827: 4. 3.6 826: 4.4.1
$16: 4.4. 3311: 4.9. 7 318:
4.15. 4 311: 4.29.1 311: 5.4.
2 331: 5.10. 1 809: 5, 24.6
817: 5. 27, 2 811: 6. 6. 5 311
827: 6.15. 1819: 6. 24. 6 827:
392

7. 6.23824: 7. 9. 2 831: 7.10.

1 821: 7.12. 8 and 7 826:
7. 29. 1 313

Aspasius in Eth. 4. 14 806: 7.15
(end) 318

Athenaeus 21¥F 329: 374 286:
87B 325: 128B 830: 2844
308: 248¥ 329: 2600 320:
845B 331: 8513 827: 374B
822: 382B 316: 416A 311:
4618 332: 507 c 303

Axionicus 87

Callias 75

Callimachus 18. 2 69 3380: 59. 8
310

Carcinus 10 328

Carmina Popularia (Bergk) 28.
279: 46 287

Cebes 1 324: 8 313 881: 10 828:
22 315: 33 327

Chrysippus 324

Clearchus 87

Clement Alex. Paedag. 8. 11. 7
3827: Strom. 5. 8. 17 306

Crates 72 73

Cratinus 72

ss Junior 66

Demetrius 76

Democritus 199 328

[Dicaearchus 59.14 801: 59. 23
826]

Dio Chrysostom 8. 31 316: 18. 6
832; 21. 3 827: 29. 15 328:
30. 4 9382: 38. 18 3823: 52. 2
317: 68. 5 and 67. 6 314: 69,
8 321

Diodorus 14. 62. 2 302

Diogenianus 2. 48 323

Diomedes 491. 13 319 n.

Dionysius Com. 87

Dionysius Halicarn. Ant. Rom.
1. 53. 1 832: 2. 10. 2 829: 3.
80. 8 826: 4.7.3 3833: 5. 29.1
810: 5. 68. 1 323: 7. 52. 2
818: 10. 4. 3 311: 10. 17. 5
811: 10. 45. 2319: 10. 60. 1
828: 11. 1. 8 828: 14. 9. 4
827: 20. 13. 8 831

Diphilus 64 92

INDEX III

Ephippus 81

Epicharmus 69 70

Epigrammata Graeca( Kaibel) 274.
5 320

Epicrates 83

Eubulus 80

Eudocia 17 327: 34 827: 300
826: 373 810: 903 310: 927
828; 960 319: 978 315

Euphron 113

Eupolis 73

Euripides Bacch. 440, Hipp. 115
298: Ion 511 294

Eusebius Praep. Evang. 2. 1, 51
826: 2. 6. 11 312: 2. 6. 12
829: 4. 14. 5 826; 5. 10. 4
821: 7. 1. 8 311: 9. 27. 36
328: 14. 3. 2 814: 14. 4. 14
829: 14. 21.2 311: 15. 62. 17
S11

Eustathius ad Ji, 2. 698 322: 3.
6 328

Galen Hist. Philos, 24 326

Gellius 7. 2.3324: 15. 20. 8318:
16, 3. 7 824

Gorgias Palam. 10 322

Gregorius Cypr. 1. 59 317

Heliodorus ad Eth, 44. 2
(Heylbut) 320: 44.5 316: 169.
37 306

Heniochus 88

Heraclitus 1 329: 5 60 319

Hermogenes (Walz Rhetores
Graect 3) p. 190. 8327: 200. 15
$13: 211. 17 and 214. 13 314:
825. 8 826: 383. 18 312: 388.
18 814

Herondas 4. 12, 7. 102 302

Hesiod Shield 148 294

Himerius Orat. 14. 23 330

Homer Jl. 9. 892 314: 18. 602
825: Od. 15. 393 317: H.
Dionys. 24. 324

Kwpydlas, wept 8, 26 312. See
Platonius

Libanius 14. 4: 16, 8: 18. 217
327
Longiuus Ars 554 321
INDEX III

Lucian Am. 44: Gall. 1: Rhet.
Praec. 22: Somn. 9: Vera H.
1. 29: Vit. Auct. 5 and 14
3804-6

Macrobius 2. 2. 11 319
Marcellinus Vit, Thucyd. 51 323
Maximus Floril. 6 318
Menander 63 93-112 (Fewpyés 93:
Lefebvre Frr. 101: Tvduat 103)
Michael Ephesius ad Eth. 315

Nicomachus 1174
Nicostratus 81

Olympiodorus Vit. Plat. 329
Origen c. Cels. 9. 63 320

Panyasis 285

Paraphrastes : see Anonymous

Paroemiographi 2. 154 331

Pausanias 7. 16. 4 327

Phalaris 4 331

Pherecrates 64 73

Philemon 62 69 90 115

Philippides 712

Philo de Plant. Nog 2. 176 310

Philochorus 79 B 312

Philosophi: see Anonymous

Pindar Ol. 6.104 317: Nem. 1.
27 312: 2.16 315

Plato Gorg. 527p 58: Phaedo
82D 332: Phaedr. 2348 292:
2440 294: Rep. 484d 196:
607B 174: Eryx. 398A 74:
Epigr. 29 286: 30. 285

Plato Com. 60 64 74

Platonius 307

Plutarch Alex. 31 310: Aristides 8
182: Demetr. 30 175: Moralia
2n 8038: 48 322: 50321: 8E
331: 138 F 392: 820 3825: 93F
318: 148c 1584 319: 163¥F
1808 822: 1821 312: 184¢
3138: 1858 822: 1898 310:
204A 816: 217D 2198 224¢
813: 2808 325: 234c 322:
249p 318 819: 334E 3355
3458 329: 847F 823: 349F
831: 3508 310: 385c 331:
3892p 827 829: 418¥F 439D

393
822: 440A 882: 488F 331:
504B 319: 586A 322: 582A
317: 603B 830: 6128 315:
786K 23838: $25¢ 318: 832F
312: 8385 311: 839F 332:
8448 318: 8538" 316 321:

854c 331: 858B and F 322:
859 F 326: 868D 312: 1128p
326: 1188 a 320: (Bernardakis
Vol. 7) p..85. 17 325: 152. 21
329: 173. 7 316

Polemo in Callim. 27 823

Poliochus 1274

Pollux 1. 186 326: 5. 41 319:
5. 73 827: 8. 86 311: 8. 87
826

Polyaenus 1 proocm. 12 327: 1.
40.9319: 2. 3.1310: 6. 18.
1 318: 7. 1 330: Eue. 14. 7
822: 14, 22 313: 18. 6 310:
81. 14 313: 39. 2 320

Polybius 4. 4. 5 323: 4. 74. 6
810: 5. 1. 11 822: 9. 8. 18
322: 10. 82 12 182: 18, 13. 8
and 43. 10 322: 19. 11. 2 309:
29.8 7 310: 38. 8. 10 322

Polyzelus 76

Porphyry de Abst. 1. 7 310: 1.
46 329: 3. 20 319: Vit. Pyth.
9 823

Posidippus 113

Posidonius 303

Proclus in Crat. 37 330

Procopius de Aedif. 6 328

Scholia :
Aesch. Pers. 28 326: 110 320
Apoll. Rhod. 1. 517 318
Aristoph. Ach. 898 315: 708
822: 961 318: Eg. 320 314:
Nub. 886 328: 441 310:
Vesp. arg. 8383: 14 3383: 42
827: 508 328: 685 320:
1018 3125 1802 330: 1388
315: Av. 1021 330: 1271
319: 1404 333: 1536 328:
Ran. 218 323
Dionys. 316
Dionys. Thrax 325
Plato Rep. 479 c 318
Semonides 1 283
394

Sextus Empiricus Adv. Math. 1.
192 311: 1. 196 310: 1, 203
812: 1, 261 313: 7. 77 312:
7. 255 321: 8. 12 827: 9. 62
$10: 9. 99. 828: 10. 68 316:
10. 69 830: 10. 82 311: Pyrrh.
Hyp. 1. 186 310

Simonides, Sayings of 296

Simplicius 298 321

Solon 4, 22: 18, 12 ete.: 27. 11
281-3: in ’A@ Toa. 5. 2 800
12. 3 318

Sophocles 0.7. 674 324: 772
292: O.C. 266 292: Phil. 300
176: [0.C. arg. 1 829: arg. 2
827: Aj. arg. 85]

Sosipater 113

Tacitus Ann, 4. 52. 6, Sueton.
Tib. 58 114
Teleclides 60

INDEX III

Teles 315

Theognidea 96, 152, 424, 477,
1009 280-281: 185, 1015 294:
401 324: 570 332

Theophilus 68 89

Theopompus Chius 110 318: 344
811: Oxyr. Pap. 301 302

Theopompus Com. 75

Theophrastus fr. 81 824

Thucydides 2. 40. 3: 3. 11. 4:
4. 86. 8 297 298. See Marcel-
linus and Anonymous

Timocles 88 ;

Timon 295.

Tyrtaeus 5. 4 279: 10.7, 11.9
and 17 278

Xenarchus 89
Xenophon Anab, 1. 8, 18 298

Zenobius 5, 85 317
INDEX IV

Accusative with portitw 22,
mpooépxouat 65, of time 67,
217, reflexive 189, with éni
(wept) repos 254

Adscripts 77, 100, 193,
227. See Glosses

Agrippina 114

Alexander 310

Amphiaraea 348

Amphictyonic decree 353

Anapaestic verse (comic): spon-
daic endings 72, diction 119,
124

Anaxagoras 308

Anticipation, see Repetition

Antiphon: tetralogies 204-212,
vocabulary 206, Nicole papyrus
209

Antisthenes 330

Aorist subjunctive or optative of
past or completed action 53,
participle with eiul (ylyvoua:)
14-20, corrupted to future
311

Aristomenes 279

Aristotle’s use of yAGrra 117 n

Attic Greek 121, 355

Attraction 184

201,

Bacchylides

Basileia 40

Bétant’s Lexicon to Thucydides
128

Bright, John 3870

Burke 96, 102, 362

Burlesque 122, 125, 160

18, 351

Cantare tragoediam 350, 359
Caravella’s Index 128
Case, see Terminations

Chares 354

Chariteria 348

Clouds 25, 27, 73

Comédie etc, comedy 362

Comoedus -ia 360, 361

Comparative adjectives 37, ad-
verbs in -ws 208: comparative
and superlative interchanged in
Mss. 179, 229, 240, 246, 247,
265, 275, 313, 314: comparative
and positive 280

Comparisons (eixdves, eixd¢w) 49

Compound words 126

Daric 302

Dative with passives 82, edruxla
235, another dative 282, eiud
mpds 68, 315, xopnya 3838:
Tpaywbdois Katvors etc. 336

Dittography 320

Division, errors of 324

Dryden 284

Dunbar’s Concordance 128

‘Eights,’ the 336

Electra 27

Ellipse 3, 21, 88, 227, 336 n

Enumerations, word lost 321

Epicharmus, vocabulary 164

Euripides 116

Events- persons, things, places
put for 336 ~

Forms, poetical 124

Future after wépuxe 248: cor-
rupted to aorist 227, 269, 311:
to present 22, 46, 63, 112, 176,
190, 198, 202, 212, 218, 224,
228, 236, 241 twice, 242, 244,
396

247, 259, 261, 204, 267, 268,
269 twice, 270, 276, 309-11.
Genitive; time  (dmndrdrov,
Bopéov 336), duration of time
81, for, in return for 102, with
avyxalpw 187, xivduvedw 199,
yupvacdels (2) 202, dvemirt-
bntos 208, vir@yor 208, mapa
(not person) 209, «8 mdoxw
281, piv 317, mpés to house
of 818: ovyyvéuns eivt 200,
*chorographic’ 317, Olvnts
maldwv etc. 339

Gibbon 14

Gladiatoribus 386

Glosses 31, 33, 181, 223, 272,
290, 302

Gods, poetical or religious names
of 140

Haplography 6, 61, 188, 193,
237, 245 twice, 248, 261, 266,
268, 270, 273, 281, 283, 297,
818-320, and elsewhere.

Holmes, O. W. 370

Homoeoteleuton 299

Hope, E. W. 127

Horace Ars P. 67, 344n

Hyperbolus 74

Imperfect corrupted to present
35, 72, 76, 178, 237, 262, 263,
312: to pluperfect 263

Infinitive used as genitive 2

Interrupted sentences 22, 34, 36

Isaeus 332

Jowett, B. 871

diction 117,

Latin, poetical
imperfect sub-

present and
junctive 18n
Liddell and Scott’s Lexicon 128
Loss of words 183-5, 320-3, and
passim, of syllables 328
‘Lyrical tragedy’ 349, 359
Lyrical diction in comedy 120,
180
Lysias, genuineness of Or, 2 220

Megara 272

INDEX IV

Negative lost 8, 186, 189, 198,
222, 245, 247, 820 (a- 321):
intruded 275

Neil, R. A. 128

Nero 267

Nominative, irregular 183, 189

Numeral lost 223, 270, 272, 321

Odyssey 249

One word corrupted into two
269, 273: two into one 221,
234 twice

Optative after present tense 178,
237, with &» of past time 15,
combined with past indicative
17, in -ow for -out 161

Orators, diction 121-2

Paphlagon 12

Paratragic 125

Participle with unusual verbs
205, with cvupépe: etc. 97, with
substantive or pronoun= event
as in Latin 336: one subor-
dinate to another 68, 217

Perfect corrupted to present 182,
283, 312

Persae 50

Plato’s name 295, epigram on
Aristophanes 286

Pluperfect corrupted to perfect
46

Preposition understood 92, lost
321 and often

Present corrupted to future 8,
229 twice, 280, to imperfect
264, 272, 312

Pronoun, uses of 189, 199

Protagoras on the Gods 295

Question, indignant (éy ydp':;)
28, punctuation wrong 37, 80,
88, 180

Quintilian, the Maricas 14

Quotation, an unobserved? 248

Repetition and anticipation of
words by mistake 8, 10, 20,
26, 40, 42, 48, 79, 88, 107, 187,
190, 196, 197, 204, 219, 2380,
INDEX IV

232, 246, 261, 282, 289, 297,
325, 3381. See Xenophon and
Others, p. 307.

Religious language 123

Rutherford, W. G. 127

Satyrus -i 344n

Shakespeare 48, 125, 293, 362

Shute, R. 370

Singular and plural confused 10

Sneezing 299

Socrates 75

Soteria 351

Steele, a passage in the Spectotor
emended 303

Subject understood 32, changed
295

Subjunctive with ews 31, rpérepov

i
Syllables lost 323
Syrus, Publilius 108, 111

Terminations corrupted, sub-
stantival 314-318 and passim,

397

verbal 312-3 and passim, by
assimilation or accommodation
63, 247, 261, 316, by inter-
change(see Xenuphon and Others
p. 303) 63, 89, 112, 288, 298,
32

5
Themistocles, tomb 65
Tmesis 124
Tragedy, diction 116
Tragoedus 360
Transposition of words passim
Tribrach in 6th foot 52
Trochaic verse 60, 119, 160
Trochilus 37

Van Leeuwen, J. 128
Veitch 128

Wordsworth 116

Xenophon 122, 211, 220
Xerxes 308
RicuaRD CLay aND Sons, LImMiTvED
BREAD STREET HILL, &.C., AND
BUNGAY, SUFFOLK.
i
faa
URE

t,
ahs: ; tre iy
Hi aN AN PIAS ae oN mr
MR Aa Mie Sa Aha
basta
Gi

inca:
Mey
a! ee

i

4
Ce eh)
pabiet Ss:

 

c

el
is ty
fe
ae

ek