PA 355 F5 1894 MAIN ^^^^KfUHmim /;v« UC-NRLF B M D33 SEE I THE DIFFERENCE BETWEEN THE GENITIVE AND DATIVE USED WITH enc TO DENOTE SUPERPOSITION LEWIS LEAMING FORMAN, A. M. A DISSERTATION ACCEPTED FOR THE DEGREE OF DOCTOR OF PHILOSOPHY IN THE JOHNS HOPKINS UNIVERSITY, FEBRUARY, 1 894 BALTIMORE 1894 PRESS OF THE FBIEDENWALD COMPANT BALTIMORE a ^s p r (6^54 " Vix quidquam tam lubricum est in syntaxi linguarum, quani hi loci, qui sunt de praepositionibus et de coniunctionibus." These words of G. Hermann/ though written many years ago (1831), must be recognized, however regretfully, as still true by any one who has attempted to answer even some less important question in prepositional usage. Classification he finds difficult, at times impossible, and the opinions of authorities widely diver- gent. On the general theory of prepositions, it is true, Delbriick announced in 1879 the following consensus of judgment : " Ueber die urspriingliche Anwendung dieser Prapositionen (dvd, eW, napd, nepl, irpos, npo, eV, e'*c, ^vv) ist man jetzt zu einer iibereinstimmenden Meinung gelangt. Man nimmt allgemein an, dass die Praposi- tionen ursprijnglich wie alle Worter Freiworier (sog. Adverbia) waren, und dann Begleiiworter wurden, und zwar von Anfang an in grosster Ausdehnung verbale Begleitworter, dagegen Anfangs seltener und erst im Laufe der Zeit haufiger werdend nominale Begleitworter. In der altesten Zeit war es die wesentliche Auf- gabe der Prapositionen, die Richtung der im Verbum ausge- driickten Handlung naher zu bestimmen, die Beziehung der Handlung aber auf einen Gegenstand driickte der Casus allein aus, ohne Beihiilfe der Prapositionen."- So essentially say Kiihner,'' Curtius,* Whitney^ and others before this date, and so Paul,** Brugmann,' Vogrinz** and others since. But it is only upon this general theory that a consensus can be obtained — so general indeed ihat it must ignore the question of the ultimate origin of ' Opuscula, vol. V, p. 50, quoted by Sobolewski, De Praepositionum Usu Aristophaneo. ' Syntaktiscke Forschungen, IV 126. ^ Grammatik der griech. Sprache (1870), 11, ^428, 3 and 4. * Erlauterungen (1875), P- i?^- ^Language and the Study 0/ Lang. (,1877), p. 276. ^ Principien der Sprachgeschichte (18S6), p. 316. ' Griechische Gramviatik (Miiller's Handbuch, II, i8go), §195. ** Gravimatik des komerischeu J)iaUktes (1889), p. 206. IGr,19B prepositions,' ;. c. whether or not they contain the dtems of B e griff sworter : whereas, if one enters into particulars even so slightly as to ask for a definition of the difference between the true and the "improper" preps., or what preps., if any, go with the true gen. case, he will obtain a great diversity of answers. Curtius, for instance, finds that the gen. depends upon di/W, t^poy dia, vntp "und vielen andern — gerade in der Weise wie von unserm Angesichts, laut, kraft."'' Delbriick takes issue with him, though admitting the Curtius-construction as a probability for dprl and a pos- sibility for Sid, because they may belong to the class of " unechten, d. h. aus Nominalstammen gebildeten Prapositionen."'' Vogrinz, Gram, des homerischen Dialekies, takes the gen. to be adnominal with drri (p. 21 1), 5id (p. 214), /card Sometimes (p. 215), vrrip (p. 216), ■napd "schwer zu entscheiden " (p. 222), Trpdy " allem Anschein nach " (p. 223). Delbriick now says: " Der echte Genitiv findet sich bei dm, vntp, fitd, fTTi, TTOTi, avd, dfitpi, nept, pfrd, Bci dvTi, xnrep Und Sid diirfte es der alte adnominale Genitiv sein, welcher uns bei den unechten Prap. begegnet."* It will be observed that as these lists are not co-extensive, the slipperiness of which Hermann complains is still present. For the purposes of the present essay, however, these larger questions need not be taken up and the general theory as above presented may be subscribed to. We proceed therefore to the proper subject of the essay — the difference between the gen. and dat. used with tnl to express superposition, or, to take a concrete case, What is the Attic Greek prose for ' with his hat on his head ' ? Is it fV( T^y K€cj)a\rjs, OT eVt ttj K((})aXrj ? Or if either, is there any shade of difference in the meaning? For the translation of so simple a phrase, one might expect 'See Grassmann, Ursp7ung der Prdpositiotten, Kuhn's Zeitschrift, XXIII (1877), p. 559. He maintains (p. 563) : " Keine achte Praposition ist aus einem Begriffswort entsprungen," as also : " Keine achte Prap. ist als Casus zu fassen." See on the contrary for Trapd, Osthoff, Morph. [Inters. IV 283, Anm., ♦' der alte Instrumental," and for irepi and ivi, Brugmann, Gr. Grafn., §194 (locat.). So too (5m ((^tni), npo, and others have been reckoned among the preps. " in quibus terminatio alicuius nominis latet, ex quo genetivus pendeat," J. A. Heilmann, De Genetivi Graeci maxima Homerici usu (1873), p. 25, note 2s ^ Erlduterungett, Y>. 177. ^ Synt. Forsch. IV 134. * Vergleichende Syntax der indogermanischen Sprachen (1893), p. 762. 5 clear rules and distinctions laid down even in the elementary books. The question is not one of origins. No matter what its derivation, affinities or ultimate meaning, fVi is certainly the proper preposition, while the case of the substantive should be settled by an examination of the remains of Greek literature ; and, if both cases prove to be allowed, the difference between them, if worth anything, should appear at the same time. Only in this last matter need one feel drawn beyond the Greek in search of the Indo-Germanic basis of distinction. Yet simple as the question seems, scholars are much at vari- ance about it. Stated in general terms the question is: i) Does Attic Greek prose employ «Vi with both genitive and dative to express concrete superposition of one body upon another? 2) If so, what is the difference, if any, between the two forms of expression ? The answers of the following authorities I quote at some length, that their text may be at hand for reference. i) Kiihner, Granimaiik der griech. Sprache (1870), II, §438: " eVt mit dem Dativ, i) raumlich zur Angabe des Verweilens nicht nur, wie beim Gen., auf, sondern, und zwar haufiger, in erweiterter Bedeutung an od. bei einem Orte od. Gegenstande." He then quotes among other instances of auf "Ken. An. VII 4, 4 0( QfjqKes ras aXtuTre/c/Saf ('ni rais K€(f)a\ais (f)opovv lirnatv, insofern die Pferde als thatig gedacht werden ; so Plat. Conv. 2i2 732, Eccl. 909. 5). Gildersleeve, America7i Journal of Philology, XI, p. 372, reviewing Sobolewski's book: "Under fVi c. gen. Sobolewski rejects Kriiger's distinction between fVi c. gen. and i-ni c. dat. in a local sense, a distinction which, it is true, might well be reversed theoretically as well as practically, for we should expect the natural position to be expressed by eVi c. gen., the unnatural by the dat. Fixity of position is in fact often denoted by eVi c. gen. (see my Justin Martyr, Apol. I 26, 15), and it is not impossible that there may be some such feeling as we have in regard to vit6 c. gen. and vn-o c. dat. In refutation of Kriiger, Sobolewski points triumphantly to Eq. 783 compared with 754, but he might have claimed here, not mere indifference, but, if one must refine, reversal. It would be easy to make Demos wriggle in the one ' See, however, to take the word Mr. Rutherford himself has chosen, Lycurg. Kara AeuKpaTovg §40 opdv 6' rjv inl fiev ruv Ovpuv yvvalnaq eAevdepag irepi6pov(: kta. passage and sit quiet in his ' fixed normal position ' in the other. At any rate, the gen. is much more common in Attic daily speech than the dat., as Sobolewski shows, though, as he also notes, Rutherford is wrong in denying fVt c. dat. in this sense to Attic (Babrius, p. 7)." 6). Transferring the question to Homeric Greek (Monro, Ho- meric Gram., §200) : " The gen. with tni is used in nearly the same sense as the dat., but usually with less definitely local force; in particular — i) with words expressing the great divisions of space, espec. when a contrast is involved (land and sea, etc.); as int Xepaov, (IT rjneipov, eV aypov ', Od. 12, 2y r) 6X69 17 «7ri yfj? aXyijo-ere — 2) where the local relation is a familiar one ; as (tt\ vtjos, «V' dnrjprjs, e(f) nrnwi', enl dpovov, eV ovboii, e'nl Trvpyov, err ayKcopos, tni p.eXir]S {(peiaOds). Thus inl vrjva-i means OH or beside ships, eVt vr\biv on board ships." (But for the Greeks before Troy was eVt vr\u>v a more familiar location than eVl vr]vai ?) Further quotation is needless to prove variety of opinion.' Mr. Rutherford is perhaps alone in denying to Attic Greek prose the use of the dat. in the sense of superposition. This point is natu- rally the first to be taken up, and could perhaps be determined by an appeal simply to Att. Greek prose. But it will be better to present at the same time and in historic order the whole material of the question. The following lists i) include only concrete substantial things, admitting abstrac- tions, metaphors or other unrealities only when they vividly sug- gest their originals, e. g. Soph. Ant. 189-90 (speaking of the TToXif^ ravTrjs em -irXeopTes ', Ar. Av. 39^4^ °* H-^" "Y^P ovu reTTiyes . . ■ . ('. Toiv KpaSwu adova , Adrjvaioi S' del e. rcov diKuv abovai ; Xen. An. II 5, 23 of the wearing of the tiara i. ttj Kf^akrj and also i. rrj KapBiq. 2) exclude on the contrary concrete objects where evidently the meaning is not purely local, e. g. z 423-24 navTas yap Kare- TTfcfivf .... 0ov(t\ €tt' u\nr68f(Tui (cf. vv. 209, 22i) ; Xen. Cyr. V 'I may, however, quote Kuemmell, De Praepositionis iiri Usit Thucydideo (1875), P- 30= " Structuris Genetivi et Dativi collatis demonstrabo saepe fere nihil interesse Genetivus an Dativus sit usurpatus." He then com- pares I 13, 5 with I 56, 2, £. Tov ladfiov and k. tu laOfxu ; II 93, 4 with VIII 106, 4 CLKpuTTipiov and -u ; IV 118, 4 with 105, 2 c. rfjq avruv fjiveiv and f. roZf iavTov . . . fziveiv ; IV 100, 4 with VIII 69 i, fTr' avToii (sc. tsixov^) and e. r£t>a ; III 102, 4 (and IV loi, 3) with II 80, 2 (and IV 10, 3) vewj' and vavci. 8 3j 34 *'. Toii vTTo^vyioii (cat 6}(T)fiaat KaTaXint'ip riva ] IsOC. IJ, 42 (oXitdoa) *'(/>' .V TToXXa xPVMa'-' ^v tyw 5eSa)/c(Ly, of a loan, not literally as Baiter translates it " naveni onerariam, cui ego multas merces imposu- eram.'" See p. ,'^3 3) include, but keep separate, all instances where fVi with either case is equivalent to ai or near? 4) include those instances of i-nl c. dat. after vv. of motion wherein a clear image is presented of superposition consequent upon the action of the verb.^ Hence n 579 t. veKpa KaTTTrftrfv and the like are included, but not such instances as en' a\\fjXoiai.v lovrts ' g<^i"g ^t each other.' The verb of motion is always given in the Hsts (sometimes also other words which may aid the memory in recalling the passage). 5) exclude all temporal, causal, or other developed uses of eVt c. gen. and dat. The bearing, however, of such uses on the ques- tion will be shown later. 6) include reference to a few works undoubtedly of the post- classic age, e. g. the Batrachomyomachia, the spurious Platonic dialogues and others. There is still a remnant of respect for these works, though hardly justifiable.^ 7) include, for convenience, under the department of History the dozen instances more or less to be found in Xenophon's philo- sophical works, since for him the department would hardly affect the style in tTiis particular. 8) Instances of the case-form -^i(f), which is neither gen. nor dat., have been omitted altogether. They can offer little aid in establishing a difference of the cases, since it is only by that dif- ference that their own usage is established. 9) Instances in the Att. Greek inscriptions down to 300 B. C. being few in number (12 of the dat. and 9 of the gen.) and not ' Attention will be occasionally drawn to the more striking or doubtful passages of this kind in the notes accompanying the lists, but a full account of the omissions is given in Appendix A. 2 Yet it must be borne in mind that the Greeks said j^pon {iT^i), and that this separation of the passages according to our translation or to our view of the facts of the case is only one of convenience in this special inquiry. The inquiry is, not the various meanings of 'mi in English, but the transla- tion of upon into Greek. * See Kiihner's Gram,, §447, for this so-called constructio praegnmis. * See, for the probable date of the Batrach., Van Herwerden in Mnetnos. New Series, X (1882), p. 171. alwciys of clear signification, are not included, but may be found in Append. B. lo) The lists are arranged according to the period and depart- ment of the authors, the words within the lists alphabetically. Substantives when given in the nominative do not stand in the passage cited, but some word of reference instead, which is always indicated. References in brackets, either by figures or abbrevia- tions, indicate those other lists or authors in which the same word will be found but i7i the other case. For example, the bracket [3. 4. Hd. X.] after the word opeo-t in list 7 indicates that in lists 3, 4, and in Herodotus and Xenophon the word will be found in ike gen. case c. inl denoting superposition. I. — Epic. a) Iliad (TTi c. Dat. = atyapos A 484 aKpordr^' aiixari X 82. aKTii B 395. CO 82. [Hm. 3. 6] anT]VD f 75 KOTtdijKfv. [Hm. Hs. 3] da-TTtr A 36 Trj. [2. 3. Ar. Hd. X.] /3^Xc3 ^ 202 intaTT}. ^Xe(f)dpoi(Ti K 26 ((f)l^avf. 3 1 65 XevT). a 364 ^dXf. ^ 398 cTrtTr- T(v. ( 2^ I (TTiKTev. p. ^2)^ fx.^vav. vJQ ennrref. 7r45I 0aXe, t 590 Xvddrj. 604 ^uX«. V 54 ex^^^"' (f> 358 ^ 1 18 KetTo. ^876 irdyr) (Nauck COnj. (Vt). k 165 KaTa 5^a€. Q 271 pvpo), KaTfBrjKav. (TCLKOS H 246 avTv Eupolis] TpoiTis »/Se (cat tOTOf /* 425 426 KfiTO. Sir y^l Kamreaov. a 196. C 153- n 67. 307- 6 222. t 89. X 461. fi 191. TT 439. o) 535 n'tTTTe. [l. 3] XiTcixn >!> 31. XPOI P 210 T]pfXO(Te, ylrajj-adois A 486 vrja epvaaav. ^ 853 '4(TTr}(Tav. y 38 18pvaei>. v 1 19 •^np.ad(o, Wf(Tav. 284 €Keifir]v. X 387 Kexvi/rat. «S/ioif o 61 jSaXeTo. [Hni. 3. Hd. 6. PL] Proper Names and References to Persons. l(pi8diJ.avTi A 261. KaXXtKoXwyr/ Y 53. "Oaarj X 315 6ip.(.v. OvXvujro) X 315 Gfpiv. YlnrpoKXco P 706. aXX7;Xot(rt X 3^9 pvT](XTfipfs . . . Kt- XVVTO. [P^'J avTif (of various persons) S 419. n 661. 12 666. A 470. p 236, where note contrast with xnro. ^ 381 (?). o) 525. Also in the phrase apa^rjat 8e reCxe' eV aira A 504. E (42). 58. 294. 540. e 260. N 187. P50. 311. a, 525. rw E 101 = 283 = 347. *■ 188. Q 445 Totat. Despite varying opinion I have here and throughout put the words /3co/:ioy and iaxapa under the head of superposition rather than of proximity. For whenever fire is mentioned, it is clearly a case of superpos., while elsewhere it must be borne in mind that the ^oifios regularly presented a foundation 7i/>0fi which the sacri- ficers stood, and that the iaxapa was in all likelihood surrounded with a paved space upon which, e. g. suppliants sat. Else why the mention of ashes in r^ 153 &>? etVcbz' /car' lip' efer' in ((Txdpj) ev KOVirjCriV ? In innouu of the above list Giseke in Ebeling's Lex. Homer. sees the genitive. And it is true that except once in Aesch. eTri always takes the genitive of this word. But ^aXXw in o 182 can hardly take the gen.,* nor would that construction be supported by the usage of fVi/SdXXa) (4 times in Hm.), for in the only passage where the gen. appears with it (z 68) the verb is in the middle voice. In B 89 TTtTOVTOi en' uv6eai, ini is ' tOwards.' So tOO in 'I' 82 1 en aiixevi Kvpe. In r/ I20— 21 oyxvt] in oyxvr) yrjpdaKei, fxijXov S' en\ p.Tj\a), airap e'nl (TracfivXr) (TTacJivXTj, avuov S' en\ avKca, the image of SUperpos. is almost as strong (yrjpuaKfi = ninrti) as in the phrases no'ipav (popov, ' See infra p. 56 12 oXyta, Kvdns, ovo^ia, Kpdroi) dflvai e'ni rivi (A 5O9. E 384. Z 357. ■*■ 4OO. 406. 6 554. X 560. T 592), or in eVt awfinn Kvpaas (r 23). But in accordance with the classification they usually receive, I have omitted them from the above list. en\ npodvpoKri (2 496. a 103) is also here and elsewhere regularly excised from the list of con- cretes, though I should prefer to follow Gerlach {^Philologus, XXX, p. 503) in understanding by npodvpa a portico or covered space before the gates of the court. See, however, Buchholz, Homerische Realieiis II 2, p. 96. fVi c. Dat = Proxiniiiy. (a)(aTifi /3 391 (rrfjcTf. i 1 82. k 96 (La Roche reads -rjs, citing I 280). [Hm. PI.] 6vpj]ai B 788. a 239 fjo-rai. x 25O. yjr 49. [Hd. 6] Kanrja-i 434 KUT(8i]aav. 8 40 Kore- 8qaav. KXr]l(Ti^ n 170. /3 419 KadiCov^^ ^ 579- ^ 37 Srja-diifvoi. t IO3 KaO'iCov =: 179 = 471 = 563 = X 638 = /a 146 = 221=549=: V 76. KpljVTj V 408. O 442. [HS.] Xi'/ifT^ Y 390. vrtvai A 559. B 4. A 513. E 79I. Z50. e l80yfi/ci)juat. 380. 531 fyelpofiev. I 425. K 306. 38 1. A 135. M 38. 90. 246. 403. N 107. 333 Vf((T(Ti. 381. 762. 832. ssiKfWo-t. 57. 65. 367. o 44. 248. 459. 494. 722 vUa-a-t.. n 18, 201. 547. 2 7 Kkovfovrni. 259. 294. 304. T 71. 135 vfe(TTToi(ri p 60. (iri c. Gen. =: dyKWfof K 80 6p6u)6eis. IT 702 ^fj (v. I. Irr'). ^ 494. dypov a 185 fa-rrjKfp. I90. 77330. 383. ;^ 47. 0)212.308. dKpijs {^vpoi) K 173 la-TaTai.. iiKprjs 6 508 epvanpTUi. [Th. X.J cLKTr^s Y 50 -(itoj/. (* 125 La Roche reads Kab b' Zip' d»cT^f with six MSS.) e 82 Kndrjpepos. I5I Kadfiptpos. K 140 K»Tr)yny6p(a-da, [i. 2. Eur.] Stiperposition. apa^a Q I90 avTijs, drjaai. 267 avT^s. [Th.] dirrjPTjs fi 275 pTjiop . . . (inoipa, 447. f 252 ri'(9«. [Hm.] ^utpoiv bases rj lOOearaaap. [l. 2. 3. An Hd. Th.] yairji N 565 (v. I. dat.). p 27 yi]s. ^85/3i//' 164. [Hm.] tTTTTcoj' chariot E 249. M 82. 2531 PaVT€S. Q 356. (In P 459 TOiO-l is perhaps the Trojans.) 249 u^a. T 238 tdi/Tt. 243 drrf- nffj.nop. 259 'iPX*''' • 339 *<*"'• 39 f px''M*'"'f • V' " 7^ ^<"''' 1 1 7 fUfordai. 301 (IX^Xvdas. [Hni. 3- 4- X. PI.] oudov threshold a 104 o-t^. 5718 rte. ^ 62 6'to/x€^'. p 339 rf€. <'" 33' X -^^3 i<\>ip e 455. [Hy. 3] nt'jprjs p 357 KOTfdqKep. Trvpyu>v 0519 Xf^acrdai. I 5^^ /3a«- voj/. M 265. n 700 TTvpyov eart], * 526 TTvpyov, f(TTt]Kfi. All at end of V. [Hm. 3. Hd. X.] aapihos

p A 3^ auroC (z*. /. avrw), [2. Ar.J fXeXiKTO SpuKap. Xapnrtjp r 63 avrap, pr]rj(Tap. rpdnf^a A 629 avTtjs (v. I. -ttj). fjLf\it)s X 225 (p(ia6eis. [4. Cf. 6. PI.] fiTjpos, -lov A i^6l avrap 8' o}po6eTr)(Tav Tponios t 2^8. [Hm.] =:B 424 = y 458 = /x 361. x^'p*^"!' S 284 /3^T>;i'. »c459=:X40I i/ijwv, -OS E 550 enfirdrjp. [e 528 =408 = 01 III. t 278 ex^aXe. ^opeovo-t.] N 665 e/3at;/e. 17 223 0)291. AUatendofv. drJK . 0171 dcfiiKfo. 260 (Z^ero. x^°'''^^ ^ 293 KaTf'drjKep. Y 345 )3 322 to>j/. 5 489 <5X€7-'. 817 Ketrai (Barocc. x^"''0- ■»■ 470 e/S;;. 6 5OO /3ai/r€j. t 535 eX^oi. X 115 I'etat. 5*-*^ rjynyov. 534 f^atpfp. p. 35^* »* 216 ot;^o)i'rat. ^ 188 dcfyiKfo. 295 ieacraro. 298 tTToprjv. 357 f^aipop. o 452 riyoi/x . 547 f/3;;. p 160 ^pfpos. ik^x^e\ [i. 2. 3] oipo)!/ A 46. (c 170 ojpou. [Hm. 3] Bot'TTpaaiov A 75^ ^rjaafiep (La Roche, towards). fVi fjjwi' m E 700 {irpoTptiToPTo pekuivi'iuyv en\ vj)a>p) means ' toward the ships.' fVt c. Gen = Proximity. (axariiis e 489 (z;. /. dat.). 454, X 45, though Giseke in Ebeling's Lexicon has them = supra, in. Too proble- matic to be included is a 278 =■ /3 197 U^va) oaa-a eoiK( (piXrjs sVi nai86s fueffdai. For the forms in -cf)iv, see p. 8. It will be observed that in this last list, as with the datives of prox., the image of superpos. is still clear, on answering almost as well as af even in English. b; Hesiod and the Homeric Hymns. eni c. Dat. = Superpositio7i. hKTois Sc. 213 r] 339- [Hm. 2. 3. 4. Th. X. 6. PI] •yXoxro-J? Theog. 83 ;^€toi;o-ti'. [2] yovvaai Epigr. 4. Batr. 3 drJKa. [X. 6] tpyoii fields Op. 549 ar]p . . TfTarai. ^avfiai Sc. 233 dnijapevvT' . TjXfKTpco Epigr. 15, 10. SeipeXia 2, II7 avrols, (drjKf. depedXois [Theog. 816]. Kaprjvois Sc. 236 fSovetTO . . , 0o/3oy. Kapnc^ wrist 2, 1 8. AroXoii/w 5, 272, 298. [Hd.] Kpari Sc. 136 Kvvf'rjv edijKe. 6, 7 edrjKav. Kpordcfiois Sc. 137 dpapvlav, Batr. 131- [4] pfXeaa-i Theog. 152 = 673 ('netpv- /coi' = Sc. 76 = Op. 149. pifS Batr. 91 avTW. vu>T(o Op. 544 dp(f)i^u\r]. [Hd. Th.] owSet 3, 149, 284 Kadiacrai. ov8(o (yiypaos) Op. 33 1. °X^TI^> 'Tl^*'^ "''If Batr. 166 fartjaav. 223. 247. [X.] [oj^oto-t 5, 19 at end of v. "lam Vossius, postea Grashof, Sxf- acfiiv pro oxoiaiv scribendum esse iure censuerunt." Ebe- ling's Lex., s. v. oxos (2).] [Hm.] TreS/cp 2, 42 OT^y. Trerpr) Sc. 406. 375 dXXqXjjs. Hy. 3, 124,404. [2. 3. An] nXarapavi 3> 1 28 etpvcraro, irnvTcp 2, 216. Trpoo-wTTfi) 10, 2. [Hs. 4. 6] aTjpaTi Epigr. 3. [Hd.] Tvp^co Epigr. 3. i;8aTi Batr. 74. 99. (In v. 61 ev should be read with 3 MSS, Matthiae, and Franke. In v. 89 xi(f)', Bothe's correction, is necessary to the sense). [Hd.] xdovl Theog. 556. 564. Sc. 162. 462 KtiiS^aXe. Op. 90. 157. 252. Frg. 14, 4. [1.2.3] i6 •^afiadois 2,329 epvaavTO. 3, 79 dXXijXotf Sc. [379] TreVoi/, SC. ai/Spey. (ppi\}/(v. [7] , viaaoprni. Hy. 2, 316 /3a)/x5i/ Theog. [557]. [l. 2. 3. opovaa. "J, ID cio-ai/. 45. [Hm. Ar. Hd. Th.] 3. 4. X. PI.] Sicf)pav Sc. 306 ^(^aaiTfs. 321 Trpoaanov Sc. l/^J. [Hy. 2. 4. PI.] 8i 47' [3] rjndpov Op. 62^ f'pia-ai. Hy. 2,310 j^^oi/d? I, I33. 20,3. [l. 3] tpvaaadf. J, 22 d(f)wpfp. X^P^^ 3> ^^3 f^^T- Bpovov 4, 165 KaT(6rjKe. [Hm.] i. Pind. O. 6, 82 86^av 'ex^ '''"' ^- jX' aKovas Xiyvpas. [2] dfpBpioi Pind. Frg. 230 jSniveiv. [Hd.] ewaty Pind. P. 9, 12 jBdXev. [X.] Kopv(f>aLs Critias 7, 9 KadiCn- [Hm. X.] Kpr]p.vols Pind. O. 3, 22 'AXcjiiov. [Hd.] Xfx^fcrai. Pseud. -Phocyl. 189. [Eur.] 680J Pind. O. 10, 30. va-da Sappho 93. [Eur.] ocraois Sappho 29 apmeTaaov. 6(f)daXp.(H Pind. Is. 7 (6), 13 ea-Tuaas. [2] X(iXi(Ti Plato I, I. 32, 7 ^aivov. [Cf. Hd. Th. ;^ei'\€o? rac^pov] ;^5o»'i Theogn. 799. Archil. 56, 2. [Aristot.] 5, 12. [i. 3] "iSa Pind. N. 10, 71 TrXa^e /cepavj/w. aX6x(^ Pseud. -Phocyl. iS6 x^'^P°- ^aXrjai. dvepi Cleobulina I KoXXrja-avTa. TTcudi Pseud. -Phocyl. 210 rpecpeiv XclIttjv. Ion I, 7 uXXa, TTea-rj. Toin-^ecri (SC. irpaypaai) CritiaS 4, 3 a(t)pay\s 8 fjfifTeprjs yXaiaaTjs fVi TotffSeo-t Kfirai. (Tri c. Dat. = Proximity. SiV^o-t Anacr. i x-qQaiov. ^i^pot? Pind.N. i, i9€crrai'. [Hd.6] tfjxapri Semonid. Amorg. 7, 47 Kpava Pind. P. 4, 294. [Hs.] etr^iei. [X. 6] X[p.vr^ Theogn. 7. [Hd.] icTxmiaiai Alcaeus 69. [Hm.] pvXq Alcman 70. [Hs.] i8 noXeai Carm. Popul. 44. [3] TTpoxoTJai Solon 28. Aristot. 53 edea-av. Siuionid. 1 20. p(f0pois Piiid. 0. 13, 35. N. 9, 9. Is. 5 (4). 33' prjyix'ivi Find. N. 5, 13, Tvfx^a Theogn. 1203 olpaxdeU. Z8aTi Find. N. 3, 4. [Hd.] 'aX0€w Find. O. 8, 9. Evpina Find. P. II, 22. EvpwTo. Theogn. 1088 7rora/x&J. KaaraXla Find. F. 4, 1 63. Kr](f)i(Tav. Find. O. I, 77. [3] darrldos Tyrtae. II, 31 (peiaas. Find. P. 8, 46. [Hm. 3] arpaKTco Adesp. 52 (MS -ro)). ^adpidos Find. N. 5, I eVraor'. ^(opav Bacchyl. 13, 3. [i. 2. 3. An Hd. Th.] ytjs Tyrtae. 10, 31 a-rrjpLxOeU ^ 11,22. [I. 3.x. PL] ■yXoxrcrjjj Theogn. 815 jBovs . . . fiTi^aivav. Adesp. 87 -a?. [Hs. 2. 3] aiovoiv Sappho 30. BaKO) Alcman 87 r^ar'. So Bergk. [3] "imrov Corinna 14, i. [3] \ivK'nrna>v (SC. apparayv) Anacr. 12 B. vrjfbv Simonid. 105, 3. [Hm. 3. 4. X. PL] ^vkov Carm. Popul. 34, 3 Kad^piv-qv. 6XKd8os Find. N. 5, 2. TTiTpas Carm. Popul. 46, 25 KaOrj- pevos ("circa OI. 122, 3," Bergk). [I. 2. 3. Ar.] ttmXcov Hippon. 42, 1. See appd- Toiv supra. pivos Herodas 4 m'; • ■ • t?)" x'^^^i" . . . ex • MSS plvas. pvTiScov Plato 3O) 2 enecTTiv epa>s> v. [2. 3. Ar. Hd. X. PL] acnvibi Phoen. [1120]. 1 1 24 tW- (TTiiia. [2. 3. Ar. Hd. X.] fiofTTpvxois Bacch. 757 ''"^P 'ic^yepov. ^(OfjLols Aesch, Suppl. 6p4 d^iar. Alces. 133. [1.2. 3. Ar. X. 6] ya Ant. IJ4 Tvea-e. [Hm. 2. 3. 4. " Th. X. 6. PL] y\a>(T(rr] Ag. 36 (Bovs. Aesch. Frg. 316 kX^s. O. C. 10^2 kX^s . . . ^ejSaAce. [2] SeXro) Iph. A. 155. Sepvioia-i CycL 4pg. 8epr] Orest. 1653. dofioii Med. I2yo TVLTVOVT . . . a)(r]. [3] epKei besel of riyig Trach. 615 ofifin OijcreTai. f'axdpa Eum. 108. [X. 6] CvyM bench Ag-, 16 18. Phoen. 74 Ka&i^iT . yoke HeracL 854 crradevT. RheS. 766. doKois Iph. A. 795. [2] Superposition . iTTTToiy Aesch, Frg. 38 epnecpvp- ixivou [2. 3. Ar. Hd. X. 6. PL] v. Phoen. 1 131 (pf'pwv. Bacch. 755 edeaav. Eur. Frg. 863 (pepcov. [Hm. 3. Hd. 6. PL] dXX/yXoio-l Pers. 506 ttItttov. O. C. 1620 apcfyiKeipfvoi. [Ph] yvvT] Aj. 1295 f<^ l) ^a^a>i' • • • avbpa. 'iXioCTi Androm. JOI ^vyov fjXvde. p.oi Trach. gSi ^dpos. napOevoi O. C. 1611 aiiTois, Trrv^as . . . ;^et/3af. aoi Eur. EleC. I344 "lxvos ^dXXovai. Proximity. pr^ypiai Iph. T. 253. poat? Phoen. 574. Hel. 52. 124. Eur. Elec. 1273. Eur. Frg. 14. aKT^vah Aj. 3. Hec. 733. (Tuoha O. T. 21. a-recpdpoiai Phoen. ^86. aipayjj Hel. 1 582 ravpeia, aradeis. TfpdpLVOis Hipp. 5'J<5. Tvpi^M Eur. Elec. 1326. Xevpaai Phoen. /pj. KvKX(D77etotcri Here. Fur. 998. ^ipoevTi Troad. 810. Tpola Phil. 353. 611. It is not certain whether superpos. or proxim. should be under- stood in Choeph. 4 Tvp^ov S' fV' '6x6(0 rwSe K-qplaaa Trarpi, and in Soph. Frg. 342 fitSeis ivavtpiov \ipvas e'0' v\(/^i]\nis trTTtXdSf (rtn (of Posei- don). The readings are corrupt in Herc. Fur. 1003 e. Xd^w Ktap, Hipp. 1 195 e<^' app.aTi, while in Eur. Frg. 628 eV eaxdpms Ammo- nius' citation has (axdpas. To the words regularly excluded, as 21 e^odois, npodvpoii, repfiaa-i, roTrotr — for which SCe Append. A, 5 — mUSt be added Av8o7s Trach. 356, Xovrpolo-t. Soph. Elec. 445, ylrvxfi Ant. 317, pimai Soph. Frg. 511, (rTe(f)dvoi(Ti Phoen. 831, K(pKL(Ti Ion 506, in which there is either lack of concreteness, doubt in interpretation, or remoteness of metaphor. In Aesch. finally must be noted (nl. in the sense of towards ox for in Sept. 423, 714, 1059, Suppl. 1003, Frg. 69. ini c. Gen. == Superposition. dyKvpas Hel. 107 1. [^Cl. (rpiKpols O. C. 148] dypav O. T. 1049. Eur. Elec. 623. aKixris Hel. 897. aKpaav (sc. SaKTvXcop) Aj. I23O. aKTrjs, -as Pers. 449* 9^S deivovTus. Phil. 272. Hec. 778 (aKTals of the same fact in V.69S). Iph. T. 1 170. [i. 2.3] dm]vris O. T. 802 e'/i/3f^(as■ [Hm.] app.aros Aesch. Frg. 38. [3] See infra. ao-TTi'Soy Sept. 387. 400. 478. 510. 512. 520. 559. 661. [Hm.3] daTpd^T]s Adesp. 210. avxfvoiv Pers. 191 Tidriai. (Orest. 51 eV avxefos ^aXelv is SpurioUS. Seep. 57.) [Hy. 2. Hd.Pl.] /3«Xj3iSwi/ Ant. IJ2. ^Xecfidpcov Kur. Suppl. 284. [1.2] ^cojuoi; Androm. 1 123 eVrj?. Note Heracl. 238 Zevs) e'cp' ov av ^atpioi- daKf'is. [1.2. 3. Ar. Th.] yr]s Soph. Elec. 1 1 36. Aj. 235 yuias. O.T. 113. 416. O.C. 7705 ycii. i'ji2. Orest. 233 •yot'ns, «()/iocrat TrdSaj. Hipp. ^^^^ f^aaav. Alces. 869 n68a Tve^evav. Hel. ^25 TToba ;^pi/:i7rTo/iei'os'. Troad. 884. [i. 3. X. PI] ypapjjiij Eur. Frg. 382, 9 eV' avTJis rpeis Karea-Ttjpiypivai (z. 6. 'E'). So'/xwr Orest. 1574 I'lKpav. [3] hopoi ship Androm. /pj. edpas O. C. 85 • Farias Ag. I435. [6] tTTTTcoj/ Pers. 18. Eur. Frg. 675. [Aesch.] Kavopos Agathon Frg. 4 e'^' fpns. KKtpaKos Iph. T. 1382 dopav (Wecklein reads accus.). [X.] KprjTTidov Here. Fur. 1008 i'Kftro. Ion 38 Tldr]pi. [X.] Xdos O. C. I(?j. Xe'xovs Orest. 313 ptve. [2] vaav teinples Eur. Elec. 6 re'deiKe. v€a)s Pers. 18 vawv. Phil. ji6. 648. Iph. T. 102. 1000. RheS. 72 ^pwiTKcoi/. 97 veS)V. [Hm. 3. 4. X. PL] vrjuov Phil. 613 T)S vaiei. ^evrjs O. C. 184. 563. Androm. 136. ^vpov Aesch. Frg. 99, 22 e^rjp (so Wil.-Mollen.). Ant. 996. Here. Fur. 630 e/S^r'. o^coj' Bacch. I070t5pi^(ray. [Hm. 2] op.pdTO)v Ag. 1428. Eur. Suppl. 286 ^aXovaa. Phoen. 1452 rlOrjai. [Hm. 3] 6p6(f)(ov Aesch. Suppl. 6j/. 22 3pfa,vO.T.//q5. [Hm.Hd.X.PL] TTfpycifKov Iph. A. y62 (TTliaOVTai. nerpas Eur. Suppl. IO45. [l. 2. 3. An] TTfTpOV O. C 19' TToXeo? — -y^y Androm. /j/. TTpvfJLvrjs Iph. T. I 177 (Tra6eiiT€S. [Hm.] TTupSf Eur. Elec. 513. Ion 1258 (troch. tetram.) T^f. [3. 6. PL] TTvpyav Phoen. 109 1 «/c/Jco^. [Hm. 3. Hd. X.] TTwXou O. C. 312 fif^acrau. piTTos Eur. Frg. 397 TrXeoir. poTrijr Hipp. 1 163. [Cf. Th. and PI., also gen.] noonris Orest. 68 d)(ovp.f6n. (TKoKpuv Iph. T. 1347- awparos Ag. 1472. [3. 7] rpoTTis Hel. 412 f(/) ^r 6cr6}dr)y. [Hm.] xepa-ov Aesch. Suppl. 178. xBovoi O. C. 1256. Med. 781. Cycl. 543 6eU. [i. 2. 3] Xu>p.aTos HeC. 5-^4 '"xpov, e(TTt](T . Orest. 116 (TTas. [X.] Xtopas Trach. 300. [Hd.] cifxcov Soph. Frg. 344. Orest. 1532 (troch. tetram.). Eur. Elec. 813 ^pav. Rhes. 305. [Hm. 3] In Aesch. Frg. 38 cited above s. v. app-aros corruption seems almost a certainty. See Blaydes on Aristoph. Ran. 1403. Ant. 1 141 fx^'''"'' -favdapos TrdXt? eVi vocrov, even with the aid of FIdt. 6, 11 eVi ^vpov yap aKprjs fx^'''^'' W*" ''" irprjypara, doCS not COnvey tO all scholars the image of superpos., Jebb preferring to compare in (iprjVTjs. In Med. 135 eV apcj)nTv\ov yap eau) peXddpov yoov (kXvov Weil drops eni, by which " on retablit a la fois le sens et le m^tre," and Verrall asks why eVt may not mean ' in the direction of,' ' toward.' For Cycl. 384 Koppoiis nXarelas ^^axapas ^aXojj/ em SCe p. c6 • In Eur. Suppl. 272 yovvdrcov gocs of course With dvriaaov, and in Trach. 1275 the reading is uncertain. This is the sum of the exclusions. For fTTt c. gen. denoting proximity there are no examples. 4. — Comedy. a) Aristophayies. iTzi c. Dat. = Sziperposition. KVTOS dcopuKOs Pax 1 235 ^eKapfO). [X.6] apdeai Eq. 40J.^ yoj-ao-i Thesm. 1 182 (cXtiSeo-i Av. 2j8. KoiTius Vesp. 1040. Kopvcjia'is Nub. 270. [Hm.] KpoK(x>Ta/. = Proximity. oiKlaiai Vesp. 801. TTvXais Eq. 1246. 1247. 1398. poaia-i Thesm. 864. (XTopnTi Eccl. 1 107 Trjs eajSoXTJS' rpaneCn Pax 770. [Hm. 4. Hd.] Tvp^cp Ran. 1139. 'AprapiTico LyS. 12^1. Ar)vata Ach. 504 (Sobolewski, on). UaWabim Frg. 585. Sobolewski is sure that e'7r' ifrxapais Av. 1232 means at, not v) Av. 41- eXnis (as a ship) Eq. 1244 ?'(/)' ^? oxovpeda. Frg. 150, II f . XeTrraJf eX. ax^^O'd . inncov Lys. 6yp. [Aesch.] KaiOdpov Pax 81. KiXrjTcof Lys. 60. KfcfyaX^S Eccl. 222. Av. 487. 515. Plut. 1 198. [Hd.X. PI.] kXiVj?? Lys. 575. 732. Eccl. po^?. [PI.]' 24 KopijfinTos Frg- 474* Kpndciiv Av. 40- Kpenddfias Nub. 218. \ldos Vesp. 332 f(f)' ov. [Hm.] Xo0&)i/ Av. 293 (troch. tetram.). [X.] \vxvi8iov Frg. 281. fifXias Av. y^2. fxripioop (as ^apLSiv) Thesm. 693. viuii Ran. 52. [Hm. 3. 4. X. PL] ^vkov Nub. 143 1 (iamb, tetram.). Vesp. 90. okiyov LyS. 31 euxero. Cl. ayKvpas. 6pSypAch.82. [Hm. Hd.X.P].] naTTaXov Vesp. 8o8. [3] nepiBpofxov balco7iy Frg. 133. TreVpn? Eq. 754. 956. Av. 836 -&)!'. [i. 2. 3. Ar.] nivaKos Plut. 996 firovrn. [Cf. TTlVaKLCTKOlS 4] Trpoo-coTTOuNub. II76. [1.2.4. PI.] TruyiSiwi/ Ach. 638 iiKpcop. pinos Pax 699 TrXt'ot. (TK^TTTpaV Av. 510. [3. Hd.] .6uv, etc., where the gen he says expresses " etwas der Person Zugehoriges, ihr Anhaftendes." See, too, Schmalz, Lai. Gram., §113, Richard Meister, Die griechischen Dialekte II, p. 297 ff. In the present case therefore (f. Wewv the prep, according to this theory goes immediately with the gen. Rather than that, 25 mvaKla-Kois Pherecr. Io8, 14 (I). rpanify Philem. 17 (II) Keinevov. Pollux reading -wf. [Cf. jriVa- Telecleid. I, 7 (I) -at?. [Hm. Kos Ar.] 4. Hd.] Trpoo-WTTW Eubul. 98, 7 (II). [Hs. ;^€tXe(ri Eupol. 94, 5 (I) eVfica^ifef. 4.6] a-ayfiaXta Cephisod. 4 (I) e'cp' oh Twfie (j'^. ww?) Cratin. 108 (l) . . . errecTTiv, tTrdo^ova . etri c. Dat. = Proximity. eCjmii Eubul, 53 (II). [Hd. 6] Tt^yuvois Pherecr. 127 (I). 'kaxavois vegeiable-inarket Cx^i\n. 49 (I). 2aypa Alexis ^^OJ (II). TToppfloicri Xenarch. 4, 4 (II). Proximity is perhaps also in Hermipp. 53 (I) &pa ndrTeiv e. tois ifpols. Menand. 1091 (III) seems corrupt. With Theopomp. 64 (I) KaraKeifievoi yuaXaKwrar' e. rpiKAm'w may be compared Eur. Phoen. 1533 ^'* ^^f'fiO'l-- eVi c. Gen. = Superposition. apa^av Menand. 396 (III). Adesp. 497 -rjs. [Th.] dvdpaKias Cratin. 143 (I) hexam. (Person's emend.). dvdpaKcav Ophel. I (I). Prjparos Menand. 1121 (III). yf,s Adesp. 352. [i. 3. X. PL] Si0pov Menand. 877 (III). [Hm.] fOTias Adesp. 463. [6] Kiopoiv Crates 15 (I). KOKpov Menand. 544, 5 (III). Koxaivccv Crat. 27 (I) Tcis rpixas Kadeipevai. Kpaviov Cratin. 7 I (I). [Cf. Kpavfi 3] KpoTciffiQiv Plat. 84 (I) ciKpav. [Hm. Hs. Batr.] padrjpaTos Amphis. 3 (II) ea-rrjKwi, as on a ship. ^iXov Alex. 222, lo (II). Her- mipp. 9(1) oiiTrirSj'^uXtBi/ (prob. not here official. See Append. A, 2). Trpoo-coTToi' Anaxandr. 58 (II). [2. 4. PI.] Teixiajf Eupol. 207 (I). [i-ftx" I. Th. X. 6] rriydvov Eubul. 76 (II). [4] rpdrreCn Alex. 26 1, 3 (H) e'cfy' tjs eV^eiT-'. [4. (6. PL)] rpoxoii Theophil. 7 (II). QpoKTis Adesp. 1 2 19. I should prefer the elliptic theory, or better 5'et, say that Oeuv, 6i6acKalov, etc., were new iudecliiiable substantives (in the gen. with hi, dat. with h\ ace. with fif), the neuter subst. didacKaT^ov, etc., presenting a notion as dis- tinct from that called up by J^Jdo/caP.of as e. g, cashmere the stuff is distinct from Cashmere. 26 fVi c. Ge?i. = Proximity. (f)dTPr]s Adesp. 719, from Photius. [Hm. X.] Menand. 202 (III) f. toZ o-apidlov is problematic, Epinic. i, i (III) is corrupt, but has been emended by Cobet, Mnemos. IV 322, who reads ewaXcjiLToiiTa for fV' dX^iVou TTivovTa. 5. — History. a) Herodotus. inl c. Dat. ■= Stiperposition. ciKpio 4, 195. 7> ^5' [X.J oiKrj/iara 2, I48 eKelvoiai. [X. 6j aKp(OTT]p!.a>(T0v opeos) J ,21 J fyevovTO. olKo8opT]fj.aTi 2, 121 it. [Th.]' 6>6art 2, 154. 4, 51, 53, 81, 87. 'ApTepiaico J, 183. 8, 21 (2), 42 (2), 43, 45, 46 (2), 66, 76. 9, ^ 98. 'Ao-cbtto; 9, 19, 30, 38, 43. QepfiudovTi 4, 110. 9, 43. loTpco 4i 80. AaKQij/a ;^;&)pr7 7, ^35 ^'^ ''^''JJ v^oos fTTiKeifievrj. Arjppio 7i 6 j/^croi fViKfi'/xei'at. 27 Ai^vr, 4, 153 v^aos, 156, 195. 'ne\onovv}ia(o 3) 59 frjaov. 2Kin% 8, 92. STpv/ioM 7, 25, 75. 8, 118. Soyn'cp 6, 87' TpioTTio) 7» 153 vi]aov. 'Yiravi 4, 53. Q/tfaj'w 4> 8. In 8, no e. Ti^ nXota and 123 ?'. tw 3co/xa) it is doubtful whether e? means £>?; or «/; in 5, 121 rfjv e. MyXdVotcri 686v it means / 16. Irnrov 2, 162. 3, 86 -coi/. 4' ^4 '<""> no roi^rcoj/, I16 -cov. 5» H2. 9, 44. [Aesch.] iirrov mast S, 122. [l] KaraiTTpaifiaTos 8, 1 18 fTTeoVrcof, II9« K€(f)aXris 2, 35 -emi/. 5, 12 (dat. in same cap.). [Hd. X. PL] AcAiV I. 182. [PI.] Kvd(fiov I, 92. KoXccvov 7> 44* C^^'J Kpriixvov 4, 103. [2. Th.] KpvcrrdXXov 4» 28. Kv/xaToiyfj? 9> 100. \6(l)ov 2, 124 (2), 127. [X.] tir)xav9is 2, I 25. /ia "''■ ('neareuis, opeav 7, III. [Hm. Hd. X. PL] ttoit; I, 132 ravrrjs edrjKe. TrXoicoj/ I, 205. TTpUipTJS 7' 180. TTup^j I, 86. 7, 167. [6. PL] pd;^iof (tov op(os) 3 1 54 enfovra. (TTjKOV 4) 62. (xtjfjLaTos r, 93 dVo). [Hm. & Ep. 4] crravpav 5> 1 6. o-rt/3d5of 4, 71- iTTOi;^ou 2, 125. (Tvp^oXrji (of fcocTT/jp) 4) 10 (tKprjs. rdcfiov 5) 47 l^pvcrdpivoi. [3. 6J Tu(f)pov 4, 201. [Hm. Th. X.] Tpdvrefa 6, I 29 airris Mp^'jcrnTO. [4. Cf. 6. 7] rifiajp 3> 23 nvTov (nnrXe'fLv. [Batr. 2] 28 v7T(op(r]t 9, 19. QpriUyjs 6, 33. Xfl^fos (^norafiov) 2, JO. 4, 141. XepcropTjaov 6, 39. [Hm. Cf. 2. 4] iraibes 2, IO7 (K^ivav eni^alvopTas. Xfipos 2, 141. [Hm. 4] <:;:/a>v 1,209(2). 2,35. [Hm.3] fVt ^. Ge?i. — ProxiviUy. yavitjs 1,51. 8, 122. BotTTTopo? 4, 87 nvroO. 6(i\(ip eTreunayTas 4, 84 See Append. A, 2. b) Thucydides. sVi c. Dat. = Superposition. alyinXols I 7, I. [X.] Ta>: VHI 69, I. [Th. X. Cf. I'lKpais Vn 34, 2. Vni 106, 4 -a. Tuxiav 4] [Hm.] apd^ri IV 67, 3. [l. 4. Hd. X.j KXtv-q III I16, I. ^(op6s VI 3, I f0' M dvovai. [l. 'ETTiTToXar? VI 97, 4. 102, I. 2. 3. 4. Hd. X. 6. Cf Twv 6eS>p AevKippj] I 30, I. 47, 2. I 126, 11] 'PiM II 84, 4. V 52, 2. top^ovv. [Th.] AlylvT) I 105, 2. ApTffiiaico III 54' 4" Aao-Kcoj/t VI 66, 2. Ar;X/« IV 101,5. V14, I. 15,2. 'Epcrpia VIII 60, I. K€Kpv(f)a\fia I 105, !• KepSv\la V 6, 3, 5. Kp^i/aiy III 106, 3. A^^5aXw VI 97, 5. AaKooviK;/ IV 54 > 4 vi] ^> 3 '°^' VII 62, I. 67, 2. Xo^coj/ III 97, 2. 105, I -01). IV 42, 2 avroi). 128,2. 129,4 -ou. 131, I -ov. V 7, 4 -ov. [X.] fierewpov IV 36, 2. V 6, 3. i>avs II 23, 2 avTu>v. 56, 2 -wi'. 57, I, -a>v. 92, 3 -oy. Ill 8, I -f)f. 102, 4 -wr. 115, 5 -av. IV lOI, 3 -wv. VI 37, I -mv. 91, 4 -w;'. VII 25, I civTcov. 71, 5 -S)V. VIII 74, I avT^y. [l. 3. 4. Th. X. Pi.] varov IV 4, 2. [Hs. 3. PI ] ^TJPOV I 109, 4. (TxeSiwi/ VI 2, 4. [X.] rct;^^;^ IV 32, 2. lOO, 4 avTov. V 7, 5. VII 28, 2. [Th. 6] X^i^ovs (rris Tci^pou} III 23, 2, 4. Xayplap III 97, 2. IV I02, 2 e^'" o5. [Th.] 'ETTtTToAwJ^ VII 43, 4. epaV I 56, 2. 57, 5. 59, T. 60, 3. 68, 4. II 9, 4. 29, 4, 5. 58, I. 67. 4. 79, I. 95, I, 2. IV 7, 3. 78, 1. 79, 2. 82. 102, I. 104, 4. 122, 2. V 2, I. 12, I. 21, I. 26, 2. 30, 2. 31, 6. 35. 3. 5- 67, I. 80, 2. 83, 4. VI 7, 3. 10, 4. VIII 64, 2. For I 126, II K(i6eCop(VOVS . . . . e. tS)1> afpvatv 6(a>v SeC SUpVCl p. 24. 30 fVi c. Gen. = Proximity. (f)v\aKrr]piov IV IIO, 2. AaKavm^s V 34, 2 K^lnevov. c) Xenopho7i. eVi c. Z?a/. = Superposition. alyioKa Hell. II 4i ^ KaTfcrTtjcrav. [x'] oKpa Hell. 1 6, 26 MaXf'a. De Venat. 4, 8 -atr rpixas opdas (sc. Kvvfs fxovai.). \_lj aKpa An. Ill 4, 49. IV 5, I e0 w. V 2, 16 -otr. VII 3, 44-01? Hell. V 4, 14. De Venat. 10 2 -01? (^SC. cLKpoKiviois). [3] aKpafila De Re Kq. 6, 7 KamTideTco yrj De Venat. 12, 6. [i. 2. 3. 4 " Th. X. 6. PI.] yovaai Cyr. VII 3, 5. [X. 6] «V^/x« Hell. V 2, 15. [Th.] /capSi'a An. II 5, 23. «6 10 TavTTj, irraKo- 86p.T]TO. [3. PI.] Xo'(^a, An. VI 3, 22. Hell. VI 2, 7! 4, 4. Vil 4, 26. [4. Hd. Th. X.] vairet An. VI 5, 22 yf(f)vpa ^v. Vivpais An. V 2, 12 f'lridi^X^a-dai. prjt Hell. IV 3, 12. [l. 2. 3. 4. Hd. Th. X. 6] 68v 4] (ppeaTia Hell. Ill I, 7 iTrea-Trjo-ei'. Xa>p.nTi Hell. II 3, 46. [3] v\j/r]\oTdTQi (tm) De Vect. 4, 44. [X.] ' 'Ofetw Hell. VI 5, 51. Tpaof a-Tr'jdei Hell. V 4, 50. OVrCp An. I 8, 27 CKUVTO. ini c. Dat. = Proximity. dpdpavL Conv. I, 13 o'rds. (ipKva-i Cyr. II 4, 25. dpxfioii Cyr. VII 7, 85. /3ao-tX«ioi? Cyr. VII 5, 26, dpvtpaKTOis Hell. II 3) S*-* emcTTrjvat,. 3>55- daXaTTT} An. I 4, I, 6. V 3, 2. 5,2'. VI 4, 4 (2). VII I, 28. 2, 36, 38. 3. 16. 6,43. Cyr. VII 4. 9. Hell. 14,3. IV 8,- 26. [Hd. PL] 31 6vpais An. I 9, 3. II 4, 4. 5' 31 (2). Ill I, 2. VI 5, 23. VII 3, 16. Cyr. I 3. 2. VI I, I. VIII I, 33, 34. 6, 10. 8, 13. Hell. Ill I, 28. [Hd. 6] Kpi'jvri An. I 2, 13 60' p. [Hs.] X./^m Hell. II I,. 23. VI 2, 7 ((t)u>pfJ.et.. vanrj Hell. V 4, 44 iyiyvovTO. vdirei An. VI 5) ^2 (yevovTo. niiyms An. I 2, 8. [Hm.] norap.^ An. I 3, 20. Cyr. VII 5, II. [X. 6] TriXms An. I 4, 5. VII I, 17. o-Kfji/aif Cyr. IV 2, 32. VII 5, 6 eyevovTO. (TTOfiaTi An. Ill 4, 43 ''""'' nXaicriov. Cyr. II 4> 25 -trt Toji' nopcov. Hell. Ill I, 23 aTpaTevp.aros, rdcfypa An. I J, I9. [Hd.] rjs An. IV 2, 20 eyevovTo. [2] Xt'^or De Re Eq. 4, 4 roi'rwi'. [l] Xo'c^ou An. IV 2, 13. VI 3, 1 r, 12. 5,28. Cyr. VII 3, 5! [X.] HvripdrcDP Hell. Ill 2, I4. [6. PL] vaav Hell. VII 4, 32. TT/jXoG OeC. IQj 14 "''<" Karndelrjs. TTvpapis A.n. ill ^,(^ravTT]s. [Hd.] 7riipyu>f Cyr. VII I. 39 ava^r/vm. [X.] panrd Hell. IV 1 , 30 av. o-ri/Sufioy Cyr. V 2, 15. rairibav Cyr. VIII 8, 1 6 TiOeaaiv. Tnx(ov An. VI 2, 8. Cyr. V 2, 2. Hell. IV 4, 12. VII 2, 8. [i. Th. 6] Tivxovs Hell. I 7, II. rpii]pa>v An. VI 2, 14. Hell. V 4, 5^ ■oiv. VI 4) 18 ft)*/. vfwj/ An. I 4, 3. Hell. I 6, 35 Tpo;^©? Conv. 7, 2 ou. 7, 3 -oi. avTcbv. 7, 32 -&)j. IV 8, 21. v'^T]\ov Hell. IV 5, 4. De Re [i.3.4.Th.X. PI.] V'jaoii' Hell. V T, 2 TToi d(f}iypei/os. ^ivrts De Rep. Lac. 14, 4. oiKTjpaTos Cyr. VI I, 53 -co;/. Hell. IV 5, 6. [Hd.] opeav An. IV I, II. 3, 7 -(OS. 7, 21 -eos, eye'vovTO. 8, 9 -eor. VII 4, II -eoy. [i. Hd. X. PL] oxriparos An. Ill 2, 19. Hell. Ill 4, 19 -cov. Ages. I, 28 -cov. oxdai An. IV 3, 5 S)j/. [i. 2. X. Cf. oxdos 3] Eq. 12, II. [X.] Xiopos An. IV 5, 19. upoTrXaTai De Venat. 5, 30 axnmv. epaicr]s An. VII 6, 25. Hell. I 3, 17. II 2, 5. V 2, 12, 24. Noriov HelL I 5) 14 ''■po''"'noj' aTi;(Tas. avTuv (^SC. a r] yrj (f)vei) De Venat. 5, 8 KaTaKXivovTai. erepov (^SC. di/Spof) Cyr. VII 5, 8 farrjKois. Proximity. ■norapov An. II 5, 18 hv. IV 3, 28. Hell. VII 4, 29. [i. Hd. Th. X. 6] 6. — Oratory. eVi c. Dat. = Superposition. a7VX^pn(Ti Din. I, 29 eniyeypappevov. yovacri LyS. 18, lO KUTfdrjKev. [X. 6] ypapparuov Dcm. 45) 1 8 ^1 yeypd- eXTTt'Si [Dem.] 17 r)K6fT](rav . . . dvpovs. eniypappart Isae. 5> 3^* earlais Dem. Frg. 17, [3. 4] ecTxnTia Dem. 42, 5 opos 'iiretjTiv. Cf. §§9, 26, 28 xP^<^s e. TJj eax. [Cf. I. PL] depeXlois Hyper. 4, 6 oiKoBopovai. 33 fiv^fiari Dem. 44, 30 ecpearrjKev. Alcid. Odys. 24 emypafinara. [X. PL] veveavievjj.fi'Ois (toIj) Dem. 21, 1 8 (TTfdrjKev. oSm (yj/pwf) Lyc. 40. Hyper, i , 20. opiois Tov ^lou (z. e. grave-stones) Lyc. 109. TTupo Aesch. I, 146 («£• (f)T)(nv 6 noitjTTjs). [3. Hd.] T(i(j)a> Dem. 44» ^^ e'^ecrTr/Ker. [Hd. 6] Tet;^€0-i Lyc. 47" L^^^* ■^* TfiX''^'' 4] cjiuiXais Dem. 22, 73 yeypanrai, gen. in same §. 24, 181. [6] yjrrjCpia-fia AcSch. 2, 68 (u eVsye- •ypaTTTo. 'Ep/Li^y statue of H. Aesch. 3, 184 fViy/ypaTrro. 'HSvXeiw Dem. 19, 148. yvvi) Aesch. I, 183 €(^' 17 at" dXto dcifjiapTi Lys. I, 30 fJ-0L)(6u Xa/Scov. Dem. 23, 55. Cf. also 59, §§41, 65, 67, 72, 85, 86. eV aKTa'is is found in an oracle supposed to be quoted by Aesch. 3, 112 and inserted in the text from Pausanias. Isoc. 17, 42 6\Kd8a (cf)' fi xpw^T rjv eym SeSwKO)/' is of coursc graphic (see p. 8), but if admitted to the list would make it difficult to exclude many other instances only slightly less so. Dem. 58, 55 rovs e. rots fioKpois irXoiois and many similar phrases, as also Din. i, 62 6 ?. rv. Isae. 5, I, 19, 25, 29. Frg. 4. Hyper. 2, 2. Dem. 29, 16, 18. 48, 50. 58, 32, 40 -0)V. 59,66. Aesch. I, 114. [6] epyaarripiiov IsOC. 7, 1 5. Hyper. 3. 33 -ov- dvpwv Lye. 40. [i. 2. 4. Hd. X. 6. PL] 'Ao-ms- IsOC. 12, 103. EvpaTTT]? IsOC. 4, 176. 5, 152. epqKrjs Isoc. 7, 9. 15, 108. Dem. 8, 64. 9, 26. 10, 65. 19, 219. Aesch. 2, 9. 3, 73. : Proximity. larpeiov AeSCh. I, 40, 4I, [50]. pvr'ipaTos Isae. 8, 27. [6. PI.] TTorapoi) Dem. 18, 2l6. [i. Hd. Th. X. 6] rd(f)ov [Dem.] 60, 30. [3. 6] Topiav Dem. 23, 68 a-rds. 'EfveaKpovvov Isoc. 15) 287 ^//■^J;(Oll- As doubtful may be reckoned Dem. 59, 67 eV' ipyacrrripiov Ka65>v- rni (see Append. A, 6), Aesch. I, 74 e. tS^v oLKrjpdrcov Ka6e(opivov, Din. I, 23 in olKrjparos earrjaiv, Dem. 34, 37 t. r^r (noas. Proble- matic is Dem. 19, 156 rdnlreixoyv. For official designations, as 6 e. rov opvyparos, See Append. A, 2. 7. — Philosophy {Plato), em c. Dat. = Stiperposition. e. Phaed. Z(^b. kuXois Timae. 76*? (nVpoir). avxevi Symp. I [3] _ yevos Timae. 75 ^ '''° '''^'' opOpamaiv ytvm crapKcoOJ] exov f

y. [Hd. X. PL] kKivZv Rep. 372 d KaTaKeiaBai. [PL] KoWonwv Rep. 531 b. Kopprjs Gorg. 486 c. 508 d (2). 527 a. Kpavlcof Symp. 195 e. KprjniSos Laws 736 e. [X.] kvkXcou Rep. 617 b (2) apwdfv, and paXdaKoo Symp. 195 e. peaov Parmen. 138 <:. d. prjxavrjs ClitOph. 407 a. ^epias CratyL 429 e. ^vXov Rep. 479 C ecji' ov. 6p(paXov Rep. 427 C Kadrjpevos. [3] oxr'ipaTOs Phaed. 85 ^. 1 13 fl^ tov- TCOV. nXevpas Rep. 388 a quoting II. 24, 10. [4] no86s Polit. 270 a. TTpouKecjiaXalov Rep. 328 C, ponas Locr. Timae. 97 pM, ii7r)7i'J7) , and itS reconcile- ment is not easy with the first distinction of the paragraph, viz. that the gen. is less definitely local. Giseke's theory I have reserved for the last, although histori- cally earlier than some of the others. Presented originally in his essay Die allniahliche Enistehung der Gesdnge der Ilias aus Unterschieden hn Ge branch der Prdpositionen nachgewiesen, Gottingen, 1853, it has appeared in a more enduring form in the article on inl in Ebeling's Homeric Lexicon, 1885. I therefore give the theory the space which its prominent position claims. Taking up the prep, in'i (p. 125) he says " dass in eVt vr)a (SaiVfii/ der Accus., seiner Natur nach, den Gegenstand bezeichnet, welcher das leidende Ziel einer Thatigkeit, der Endpunkt einer 40 Bewegung- ist ; f. i^rjos ^aiveiv hingegen heisst ' das Schiff betreten,' und der Gen. driickt aus, dass derjenige, welcher vom Festlande aufs Schiff steigt, in eineni andern Zustand iibergeht. Die Sprache fasst das Local Schiff als die Ursache, welche diese Veranderung in dem Zustande des bewegten Gegenstandes her- vorbringt und setzt, ohne Riicksicht auf die Richtung der Be- wegung, den Gen. vr]6s, weil derselbe andeutet, dass von dem Schiff eine Wirkung ausgehe." Then quoting B 351-52 vrjva-Xv in wKVTTopoKTiv (^aivov I 'ApyeZoi and translating the dative here by " auf, in Schiffen," he says : " Die Argeier und die Schiffe werden nicht als zwei getrennte Gegenstande angesehen, von denen der eine auf den andern wirkt (wei, beim Accus. eVl vrja ^aiveiv, die Vorstel- lung ' Schiff' sich andert durch den neu hinzukommenden beweg- ten Gegenstand, wie, bei eVi i>r]6s ^alvfiu, der bewegte Gegenstand durch das Schiff verandert wird), sondern es werden beim Dat. beide Gegenstande als zu einander gehorig und verbunden gedacht." Proceeding to examples he finds that the gen. is used in K 62 eV' oidov eCofxeada (of the Suppliant Ulysses and his com- panions), rj 169 eVl 6p6vov eio-e and Other passages, "weil das Local auf den Zustand der Person wirkend' gedacht wird." " Fehlt aber die Veranderung des Zustandes, od. soil auf dieselbe kein Gewicht gelegt werden, so steht der Dat., wie bei den Freiern, bei denen es sich selbst verstand dass sie nicht auf der Schwelle sassen (p 90 e. KXia-fioitn KadiCov), od. von den Ruderuden stets, z. B. jS 419, eVi kXjjjo-i addiCov, denn es versteht sich von selbst dass sie sich auf die Ruderbanke setzten, und nicht anderswohin, wenn sie entschlossen waren abzufahren." After other illustrations he concludes (p. 128) : " Daher kommt es dass der Gen. den Gegen- satz ausdriicken kann dass etwas sich nicht auf diesem sondern auf einem eindern Gegenstande befindet; cd. dass der Gen. den Ort mit besonderm Gewicht hervorhebt, wahrend beim Dat. das Gewicht auf der Handlung liegt e. KXia-pola-i Kddi^ov und eV' ovBov eC<'>fieda ; dass cndlicli der Gen. dasjenige hervorheben kann was zufallig und gleichsam eine Ausnahme ist, wahrend der Dat. die sich gleichbleibende Regel darstellt." The first point in this conclusion is remarkable. eVi c. gen. tells us that an object is upon one thing and so by implication not upon another. Thus if we are told that a man is eV dypov, the gen. bids us remember that he is not ev aarei. But does not the same ^But surely the time is past for finding tlie causal notion in all genitives. 41 implication exist in the case of the dat. ? If a man be said to be 6. vT)i, we infer, with quite as much certainty as in the case of eV' aypoi, that he is not on land. Spinoza discovered long ago in constructing his idea of an infinite God that any determination is a limitation and involves the negation of its opposite. Any desig- nation whatever therefore of the place discriminates against all other places. It may be noted how diverse the iinpressions are concerning the gen., for Giseke denoting the accidental and ex- ceptional places, for Monro and Gildersleeve the familiar and normal. Giseke then proceeds to cite and comment upon further con- firmatory passages, v. The dat. is not used. The phrase is in fact one word, as much so as emde^ios, though lacking declension. But there are only twenty instances in all. So a few other fixed phrases, as fVl dvpms (but with exceptions), eVl y^pws- 68<2, eVi ddfiapri., eVi divL (pr]y^7vi, xh^°^' rjTreipov'). But few of them are frequent, and most of them emerge only in certain departments or periods. Beyond this, phraseology does not disturb the question, ^^wj/, yrj, opos, oKTT], alyiaXo? and Others change with the picture to be presented. The second item, the habits of the language, may be instanced by the way in which the notion of a vehicle upon which is ex- pressed, viz. by the gen. This, I think, is the one grain of truth which Kiihner was struggling after above (p. 5) when he spoke of the 'Thatigkeit' of the horses, although the example to be sure was an unhappy one. In the above lists the following words occur literally or metaphorically as the names of vehicles, always in the gen. except the underscored words, which are sometimes dat., but in that case for the most part not as vehicles. Waggons : ciTrrjVT], ajj-a^a, appa, dppdpa^a, rjplovoi Viule-Cdr, ^ivyos, o)(os, o)(rjpa, KavaOpov. Add kKivt], (popflov, dirpdjit], ic^imnov, K€cj)a\r]. ShzpS .' pais, rpirjp-qs, irkolov., oKuds, 86pv, axeSia, ^KpOepa, rev)(os, pi-"^, poTrrj, poapr), eXms, pddrjpa, dXoyia, dvop-ara. BeCiSts of bufden .' Imros, iTTTrdpiov, KeXrjs, TToiKos, ovos, KdprjKos, arpovBos, KdvBapos, SeX0iy, rponis. On examining the passages where the underscored words appear in the dat., it will be seen that in the following cases at least the objects are not regarded as vehicles for the action expressed by the verb. f 75 iaBriTa . . . KaredrjKev . . . eV dnrjvji. Aesch. Prom. 710 TreSapo-ioi vaiova-' en' evKVKXois dxois (where the sur- prising nature of the habitation requires the dat.) Hm. Hy. 5, 19 dpTtd^as . . . eVl xP'"'^^^'-'^''^ oxoktiv is corrupt ; see the list. Plat. Protag. 315 d napeK.ddr]VTO de avrw eVl ra7s ttXtjitloi' kXivqis (a Defregger interior, where kXIpt] is of course not used as a litter, as in Andoc. I, 61. Dem. 17, 20 (rpitjpeis^ koI o-rpaTTjyov eV' avTois erd^aTe MfvsaOea — an official dat., see App. A, 2. Dem. 58, 55 rovs a-rparr^yovs K.a\ Tovs 67rt rois paKpdls ttXolois — official. '^ 3^^ ^'0 lirTToiiu pdartyas I'Uipav. o 182 f(j) Imvoilv pdariv (SaXey. Aesch. Frg. 38 tTTTroi 8 e4) "lttttois rjcrav tpTre(f)vppivoi. p. 425 e^dpevos S enl tois (^SC. TpomSi and Ictt^) 6eaav eVt dp-d^rj, nf.idovTe.s top apxoPTu, Sta rrjs Tu(f)pov {Jcge Kara ti)p Tdi' Tpane^av, iva vfiwv ttoXXoI dKTjKoacri, Koi aXXo6i, a solitary instance of eVi TpairtC^v in the sense of banks, for not only Lys., Isoc. and Dem. use the dat., but also Plato himself, Hipp. Min. 368 d ws eya TTore aov tJkovov /xeyaXavxovfievov, TToXXfjv aocjiiap Koi ^i)X(ort]v aavTov die^iopros ev dyopa fVt rat? rpane^ais — . And yet the difference is clear. The former passage is colloquial, to be sure, and circumstantial, yet melting into one thought, uncon- scious, making no point of the locality. The Hippias passage, on the contrary, is sarcastic, there are pauses between the clauses, each word tells, and a slight emphasis on the locality is not with- out significance in the case of the money-making sophist. In Lys., Isoc. and Dem. the dat. is of course in place, as in all busi- ness transactions dates and places must be carefully designated. 2) dX\j]Xa)v. — Plat. Theaet. 195 ^ ^^^ ^^ irpos ndcn TovTois eV dXXr)Xu>v (TVfjiTrenTcoKOTa § vno has a gen. with it, all points are itemized. In our own language we draw hundreds of just such distinctions, unaware of their existence until some unlucky foreigner fails to observe them. 4) avxfvav. — The next case of single instance is Aesch. Pers. 191, where Atossa is relating her dream. Xerxes yokes the two women apfiaaiv 8 vno ^(vyvvaiv aiiTo) koi XeTradi' eV avxtvav | ridrjai. Would not the ivapyua of the dream be better served here by the dat. ? Perhaps so ; only there would then be too much (vapyua. The women already fill the picture, with Xerxes, the yoking and the collar. Mention of the part of the body is incidental, em- phasis upon it would be ludicrous, evoking the question : Where else, pray, if not on the neck ? 5) Kprivdooi>. — Hes. Op. 757-8 /xjjSe TTOT iv npo^pj) ttoto/xcov oKabe Trpopfovrav, fjLT]8 eVi KpvjvcKav ovpeiv. Here eVl KpTjvdcov must mean hnmediately over, just as occasionally eVi TTOTOfjiov, eVt daXdaarjs are uscd, for which see p. 53. 6) nrjyai. — X 1 53 of the fountains evda 8 eV ai/rdcov TrXv^oi evptes eyyvs eacriv. The fountains had been already described. The nXwoi are now taken up, the fountains serving as a mere point of reference. We too, we accent ^near them,' not 'near ihem.^ Compare this with the very different effect of the datives in the Catalogue, where the homes of the various contingents are most carefully set forth, e, g. B 523 ol T€ h'CKa\.ov fXOi>, nrjyfjs ewi Kr]v ^ Thuc. uses hn' avrC) twice, Herod, once, of proxim. to a river, Herod. £7r' avTi) once of proxim. to an island. 51 yovdrav avet\tyfj.evov ^i^Xlov. Xen. Conv. g, 4 tnei yt fjLrjv Kart'idfi/ avrfiv (^SC, ApidSvTip) 6 Ai6vvv yovuTav. In the former case Pindar has two things, a wreath and an open book, the place of the book being desig- nated incidentally, but furnishing no third item in Pindar's make- up. Dionysus in the Xenophon passage dances up to Ariadne and seats himself upon her lap — not a remote or surprising place for the lover, as the dat. would have represented it. In the instances of eVt yovacn. either the verbs more naturally take a dat., as Karidr]K€v (Lysias), dfivm, KaOiaa-as (Hm.), Ka6i(opivr] (Ar. Tliesm. 1 182 — where it is not for love, however, as with Dionysus and Ariadne), or there is pathos and the etching is deep. X 500 Andromache lamenting Hector from the walls and telling of Astyanax, 6s Trp\v p^v ioZ enl yovvacrt narpos \ pveXov olov ebicrKf. Xen. Cyr. VII 3> 5 '''^^ ^^ ywaiKa Xf'yovcrii' cos KudtjTai x^M"' KfKoapijKvla otr fi\€ TOP avbpa (her slam husband), Tj)f Kf(f>aXf]v atroi f)(ovija fTTi To'ls yovaai. 8) Qvpeoav, -wv. — In placc of fVi 6vpais, the regular phrase in frequent use from Hm. down, em dvpiwv (-wv) occurs three times : Hdt. 3, 120. 5, 92 y. Lycurg. contra Leocr. 40. I confess here to seeing no reason whatever for the departure from usage. In the Herodotean passages one is almost led to suspect corruption of the text, some one of Herodotus' late admirers substituting by slip the phrase of his own times. For that the use of eVi c. gen. to denote proximity became more common — though never exten- sive — in later times is seen by reference to Polybius (see Krebs, Die Praepos. bei Polyb., 1882) and to the New Testament. Ly- curgus' break with the habitual phrase is the more possible, or rather probable, as already in his time iitl c. gen. had lost its earlier stability of signification and showed various metaphoric uses and connections with abstracts. Why not, therefore, also in connection with concretes show development ? Besides, in his desire to increase the tensity of his expression (already through- out quite tense enough), he may have seized on the less emphatic gen. because unusual, and by this contravention of the usual attained the emphasis aimed at.' Or can it literally mean {6pav 'This method of obtaining an effect (viz. by reversal of the natural means) may be seen in any art which has reached its full growth. In modern music, for example, love scenes and the aiidaute movement of the symphony are often given fortissimo, the finale on the contrary pianissivto. 52 8 ^v eVi fJLfV Twv 6vpa>v yvvaiKai' (Xfv6fpas 7r€pi(f)6^ovs KaTenTr])(yias) that the timid women pressed cowering upon the doors ? 9) Ke(f)a\r].— ln Plato's Sympos. 212 e (too long for quotation) stands first a description of Alcibiades standing at the door raiplas exofTu en\ Ttjs Kei^akrjs navv noWds, and within the same paragraph he says vi/v 8e tJk(o em t^ Ke(f)a\^ €Xer]v los el^ov apicTTa, iir v8a>p enepnov ayyos eVt tt/ Ke(f)a\Tj '4)(ovaav Koi sk tov ^pa)(iovos lttttov (TjekKovdav Koi Kkadovaav \ivov, while in the last sentence of the same chapter the same woman appears ^epovo-a to vbap eVt r^? KecjyaX^s koI eVe'XKOvo-a e'/c TOV ^paxiovos tov imrov /cat crrptcfiovcra top arpaKTOv. " Here again nO difference," say some. " The gen. and dat. are as undistinguish- able here in function as in the dual they are in form." But the true explanation is this : Herodotus having once painted the picture in detail, has no further need of the itemizing dat. Like a good artist, he chooses for the repetition the more summary and incidental gen., at the same time shifting it, as Plato did, into the less prominent position after the verb.^ In his minutely- detailed account of the dress of the various tribes under Xerxes, Herodotus invariably uses, when describing their head-gear, eVt T^ai Kf(f)a\TJ(Tt, as might be expected (VII 70, 72, 74, 75, 76, 79, 84), the dat. preceding the verb in every instance but once — one other ^ This same lightening of the touch on repetition is seen in Eur. Hec, where in v. 698 the slave-woman tells of the dead Polydorus err' anTalq vtv Kvpu 6a7iaaaiaig full of the horror of the scene, while in v. 778 the grim Hecuba, still self-restrained before Agamemnon, responds to his question as to the finder f/6', kvrvxovaa Tvovriag aKTfjg ekl. See also Herod. Ill 28 k. pev T(I> pET^KCj) . . . e, de tov vutov. 53 instance lacking^ its verb. Xenophon also uses the dat. in describ- ing head-gear (An. V 4, 13. VII 4, 4); further, in an instructive instance of the attributive position, where the attributes are them- selves contrasted, and do not, as is usually the case, stand sub- ordinate and in the gen., viz. An. II 5, 23 rfiv fih yap fVi rfi K((pa\f/ Tidpav ^aaiKei iJ.6v 6(f}crdpfvos bi Kol TOV Boanopov (TTrjXas earrjaf 8vo en avToii Xidov Xeviiov (IV 87), while in the other (III 5) «V6 TavTr]s (sc. KaSCnos TToXtos) TO. ipnopia ra enl daKdacrris pexP*' ^li'^'o-ov noXios fWt tov 'Apa^lov, Herodotus is mapping out the land and uses the sea as an adjec- tive (the sea-ports). The gen., consequently, is the only proper case. We turn next to those cases in Aristophanes which to Sobo- lewski prove the indifference of gen. and dat. He compares Eq. 752 ff. 6 yap yipu)V o'lKoi pei> uvdpcov fan df^iutTarot, OTav 8 eVt Tovrrjai KadfJTai Trjs ntTpas, Kixfivev ktX 54 with Eq. 783 o-e yap .... fVt Tn7s (Te Kadrjfievov ovtcos ovx (oanfp e'yw pa'^afj.fvos s ktX. From all that has been said, the distinction here must strike every one. In the first passage there is a contrast of Demos at home with Demos at the pnyx. The places are used attributively (//^w Demos and ikai Demos), and in themselves are of no importance. But in the second passage how pathetic are those hard rocks — real rocks that hurt you when you sit upon them, not the pnyx rock of the other occasion where men voted and clamored. *' I have bought a cushion which I made for you. Now you can sit softly on those hard rocks." Evidently the sausage-seller makes the Trerpai all-prominent. The difference between the two passages so far from being 7n/, is enormous. Sobolewski again holds up Vesp. 1040 TjniaXoi kqi vrvpeTol) 01 Tovs TTanpas t Tjy)(ov vvKrcop Kai rovs Tramrovs airenviyov, KaroKKivopevoi r eTTi rais Koirais erri roiaiv anpaypocnv vpa>v avTapocrias Koi Trpo &anep ttokop iv ^aXavfla eKTrkvvavTas Tr)v olcrwaTtjv, eK Tr]s noXecos enl kXivtjs eKpa^di^eiP tovs poxdrjpuvs ktX. and Lys. 732 aXX rj^oi Taxeois vtj to) dea,- ocrov 8ian€Tdaa() presents the image of two objects, not one, and the gen. is impossible (see p. 59). So in the official phrases (p. 60) where the wagons, camels, machines, ships claim our thought more than the inan in charge of them, while the official gen. on the contrary sinks these things in the personality of the officer, ovrri Tov opvyfiaros, for example, was perhaps as concrete an indi- viduality as the modern Beefeater. So the various actions and occasions at which one may be present (eV' aySxri, a-vuSoXaiots , diadrjKT], deiTTva, etc.) So, temporally, to do one thing after another, and in the phrases enl tovtco, eVt rw rptVa) a-rjiitiM, cf)6vos enl (jiQvcp, ('ixdos eV axdd, etc. So to post One next to another rarreij/ (^KaTaa-Tfjo-ai) riva inl Tivi, tO foUow on another (TTeadai eVi Tin, the rearguard ol enl nacn. So Eur. Ale. 373 nrj yafxeiv aW-qv nua yvvalK e'0' rj^lv. So the hostile tnl m ixrjxafaadai \TiKTaivt(y6au, avvofivvvai, TaTTdv, fidx^crdai) irri rivi. So the eVi of price, calling up two objects and their exchange, while the gen. subordinates the price to the thing purchased. So the further extension of tni to denote con- dition, cause, purpose. Throughout the whole list two objects are presented. em c. gen. on the contrary becomes phraseological, fuses, and presents a composite, whose elements are not easy to disengage, by themselves not making obvious sense. From the time of Homer's evx^ade . . . aiyjj i e^' eavTov and yeZ-av etp' eavrc^. 56 the military phrases eVl cjidXayyos (^Ktpas, ivds, rpiaKOvra, oXi'ycoj/) ayeiv, TUTTfiv, yiyveadni, etc. So even in the eVi of direction (eVt ^apov nXelv), where the gen. is the " characteristic of the motion " and goes as immediately with its verb as the word wesi in our phrase ' to go west." In all cases the gen. presents an object which melts at once into the chief object of the thought or serves as a mark by which it may be recognized. The developed uses there- fore of eVt c. gen. and eV/ c. dat. favor the distinction maintained. A second support may be found in the behavior of the sister preposition vn6. Why does it prefer the dat. in its local sense ? (That it does so may be seen by reference to Appendix C.) Evidently for this reason. If one thing be upoti another, we see it without effort, there is nothing to call forth remark, and, on ordinary occasions, we would use the gen. eVt nvos. But the being under a thing does not strike us as a natural (normal) position. Things under other things are apt to be hidden and to be passed over, as proved by eVi 1455 times against vno 345 times (not in- cluding Plato entire or the lyric fragments). But if we do see things under other things, their situation cannot but impress us. And what case can better draw attention to locality than the dative? Hence v-no-nvi, just as tVi nvi, to emphasize the place. A third point in favor of the theory that the distinction between the gen, and dat. with eVi must be based in imagery and not in logic is the remarkable difference of construction shown by cer- tain verbs, e. g. by ^dXXeiv and Uvai, The latter prefers the gen., ^aXXe^l' and eVi^aXXeij/ the dat. or accus., never perhaps in all classic Greek taking the gen. They are not found with the gen. ^ Definition of the cases may as well be abandoned. If the nearest defi- nition of the accus. is : '• eine Erganzung oder nahere Bestimmung des Verbalbegriffs," then countless genitives are accusatives. For, as said above, the dispute as to whether it is the verbal element in a noun or the nominal element in a verb which attracts the gen. seems sheer logomachy, particularly supererogatory in the case of the Greeks, who cudgelled their brains many a century before discovering (or fancying they had discovered) the difference between a noun and a verb. Query: did the Greeks lack dis- cernment here or is it we, warped by early training into seeing distinctions where none exist? Yet at last we may be floating back to the definition- less open sea of the Greeks. See Professor Gildersleeve,_/^/?«j Hopkins U7i. Circulars for 1883, p. 67 : "The adjec. is a ptc. at rest, the ptc. is an adjec. in motion. A similar difference is seen between the abstract noun and the infinitive." But if rest and motion are relative, then verb and noun are one. 57 in Homer/ Hesiod, Pindar," Aeschylus, Sophocles, Aristophanes Herodotus, Thucydides, Xenophon, Plato,^ or Theocritus. All editors are dissatisfied with Sappho 102 rjp' tn irapdevlas e'nii3u\\ofiai, and Eur. Orest. 51 has been rejected as an interpolation, while yovmTCiv in Eur. Suppl. 272 goes of course with dprlacTov, and (o-xiipa^ in Cycl. 384 is accus., thus leaving as the solitary example in classic Greek* of (SaXXetj/ with the gen. Eur. Suppl. 286 ri KXaitLs XeVr' eV' onfidrcov (jidpt] \ ^aXovaa tS>v (tS)v ', ^ Why this difference of construction in verbs of the same signification? Clearly because of the mental image evoked by them. With jSuWa our eyes follow the missile to its goal, while levai. simply lets it fly." teVai therefore chooses the subjective case, /SdXXw the objective. But if verbs elect their construction by reference to the imagery of the cases, why may not also prepositions ? It would be an additional acceptable support to the theory if the use of modifiers and of the definite article gave evidence for it. For on first thought we might expect that the case which tended to phraseological formations would seldom show modifi- cation, or, if any, that it too would be of the stereotyped sort; also that the article would be more frequently lacking than with the livelier dative. Now, though the figures do not run counter to this surmise, yet the difference between the number of modified genitives and modified datives is not great (28 per cent, of the genitives to 2fi per cent, of the datives). And on second thought we could not expect it to be otherwise; for stereotyped phrases and brevity of phrase are just as necessary when speaking of two 1 Z 68 tvfipwy £7r/^aAAo//£vof is the nearest approach, but the verb is here in the middle voice, the thrower is the viissile. ^01. I 58 neipaAdi ftaldv is ' from the head.' 'Sisyphus [Plato] 391 a does indeed show rhv TrAeicraiag ^d}.7.ovrn rov OKOTTov, but Sisyphus is not Plato's. *The following examples of ^uA7.£tv c. gen. are quoted in N. T. lexicons : jiejiArjjihov, -rjv, ett\ KMvTjg Mt. 9, 2, Mk. 7, 30 (which hardly count, as fid/./.u is in the perfect), idv iivOpunoq (id?iri rhv anopov t~l rf/q yi/g Mk. 4, 26. (ia/iovaa yap avrrj to fi'vpov tovto inl rou auuaTog uov Mt. 20, 12. *If thegen. be allowed here, however, to the introspective and erratic Euripides, jajopof may be gen. in Supp. 272. ^See Schmidt, Synionymik der Griechischeti Sprache III, §104, p. 151 : " jSdA/lEiv unterscheidet sich so von ^yliTTeiv und livai, dass die Erreichung des Orts-Objekts dabei direkt ins Auge gefasst wird." Page 155: "Endlich ist auch hervorzuheben, dass mit Uvac nicht einmal notwendig die Tren- nung von dem Subjekte angegeben wird. Od. 6, 231. II. 19, 383. 22, 316." 58 objects as when speaking of one, if not more so. For example, we speak of the man fV' airocfiapco (without the article) briefly, yet clearly see culprit a?td crime, while the phrase ovn\ rod opvyixaros (with the article) calls up but one image — the hangman's. Here, therefore, there is no evidence either for or against the distinction maintained. I am well aware that the principle oirepraesentatio here claimed must, if true, be of much greater extension than the present essay has set forth. A picturesque use of the cases would hardly con- fine its exhibition to a connection with two prepositions (fVt and ijrd). But certainly there is nothing in the nature of the two cases to render such a distinction improbable. On the contrary many points have been presented in its favor.^ If repraesentatio has been found so potent in the moods of indirect discourse, why not also in the cases ? Its images may sometimes appear illogi- cal, and logic be compelled to retire baffled. But if analogy is admitted to have played many pranks with logic, why may not repraesentatio do so ? Reason is not yet dominant in language. Thanks to imagery, the sun still ' rises.' Without absolute revolt, therefore, from logic and statistics, we may do well at times (especially if we have statistics with us) to go back to Dionysius of Halicarnassus and judge the phenomena of language by aXoyos ato-^T^o-tr, or at any rate to avoid the Charybdis of dj/at'o-^)?rof \6yoi. ^ Other small points maybe pointed out, ^.^. that vv. of action prefer Tspi c. ace., vv. of thought and speech Trept c. gen. For the difference between Jid c. gen. and ^lo. c. ace. I quote Gildersleeve, Introductory Essay to Pindar, p. 98 : " With the gen. the passage is already made, or as good as made. In Pyth. 9, 133 wapdevov ayev cKnevrdv 'Nofidduv 61' b/ii'Xov, we may imagine elbowing, but it may be imagination." But it is more than that. 59 APPENDICES. A. Excluded histances of tni. A necessary complement to the lists presented in this essay is a list of exclusions, the heads of which will be here given. The question what to omit and what to include has been by no means always an easy one, and the decision will often perhaps seem arbitrary. For while in this special inquiry the guiding principle is apparently simple, viz. that of the concreteness of an object, or, in the case of metaphors, the vividness of the image presented, yet just what the vivid image is and just what metaphor has become so remote as to be dead are questions which can be cor- rectly answered only as one approximates to Greek thought and feeling. Neither logic nor the analogies of our own language should be allowed to influence the decision. Yet I think such has been the case in the first of the following heads : I. It seems to be the common opinion that in such phrases as eadietv enl r«3 ctito) oyf^ov (Xen. Mem. Ill I4» 2), Kapbafiov e^^"' *^* '''V aiTco (Cyrop. I 2, ii), eni c. dat. is not purely local but means 'in addition to' or 'with.' In deference to this opinion I have omitted such phrases from the lists. Yet if our own phrases had been ' bread on meat ' and ' butter on bread ' instead of what they are, perhaps the purely local notion would have been more readily allowed to the Greek. This error of classifying the phenomena of a language by the translations made from it into one's own is an old one, against which Rumpel raised a warning voice in his Casuslehre^ p. 8o. The following is a complete list of such exclu- sions : dXi Ar, Ach. 835 ■na'uiv . . . fxciSBav. . . . e^'^o'"'' ^^ -' ^7 »'■««'"«' vfi&Jp. aX^tVoif Ar. Plut. 628 fiefxvtTTiXr]- Mem. Ill 14, 2 f'crdiovai . . . fievoi. oy^ov. afi^poaia PL Phaedr. 247 e in Taplxit. Ar. Ach. 967 \(>(fiovs Kpa- nvT^ veKTap ewnTKrev. daive'rco. Ar. P rg. 63O yfXcora ^uXXavTia Ar. Eq. 707 (f)dyoti . . . /caWfio/Ltni. Chiondes 6 (I) ay, KOTTTtTOP (?). KoWvpa Ar. Pax 123 e^erf . . . o-^of rplyXr] Autiph. 26, II (II) k(itkt- (Tt' avrrj. 0i-fi' yovr en\ pia rfjv ovaiav. Xdx^iva Aristophon 13, 8 (II) >|,(u;ia. Xen. Mem. Ill 14, 5 ox/^wv ttIvovitiv (TTi TovTOis v8 J-5' f'""' '''? ""I'pi Anab. IV 2, I4« eVl rois vno^vylois /cat 6\fj^a(Ti. Cyrop. V 3, 34. in\ Tois vava-'i Hellen. I 5) 1 1* alpelad ai em rots Traicrt (^e(f>TJ^ois, reXeiois dv8pa(n) Cyrop. I 2, 5- Tarreiv em TpiTjpea-i Dem. IJ, 20. em rais ripcopiais IsOC. 5> 1 17* e'm TTJ Koivrj cj)v\aK^ Dem. 17, 15. See also Dein. i, 112. Dem. 60, 22. Aesch. 2, 73. ecf)e(rTdvai eiri — Isoc. 3, 48, and many other verbs. Also with the article 6 (of) im rais pr^xavals Xen. Cyrop. VI 3, 27. em Tn7s KUfirjXoii Cyrop. VI 3, 33. em toIs aXKois dppaai Cyrop. VI 3, 36, em ra opvypari Dein. I, 62. toIs Trpdypatn Dem. 8, 76. 9, 2 ovres. em rots ^evois . • . eyevero Dem. I, 74* Here must be catalogued, and not elsewhere, such instances as the following: dXX', el ^ovXei, pev em rw (tt par ev pari Xen. Anab. Ill 4, 4^' ^o'^^ (TT paTTjyovs Koi rovs em rot? puKpois irXoiois Dem. 58, 55. naprjcrav ai . . . vrjes . . . Koi en avrais vavap^os Uvdayopas Xen. Anab. I 4, 2. Ki'rjpov pev vavap^op en ovra (Aral tovs OTrXtras) em vava\v oXiyais evdi's mptvova-i Thuc. II 80, 2. This last passage gives trouble to Kuemmell, but the fact is that the local has vanished before the official usage, and Captain Knemos is sent with or z>z command of a few ships. As a test, let one try to satisfy himself with the purely local notion, and he will see that the image presented is uncalled for by the context, and absurd. On the other hand, if the notion of conveyance had been intended, the whole usage of the language, where no emphasis is present, would have demanded the genitive. Kuemmell may as well revert to the purely local notion of VTTo m ihuc. IV 44> 4 °^ KardhriiXos T] pd)(T) rjv vno Toii opovs. Examples of the official genftive are : ol em rcov oTrXirav Lys. 32, 5- 6 em Tov opvypaTos Lycurg. 121. tovs em rrjs noXiTelas e(j)e(TTr) kotos 6i Dem. 19, 298. Tovs eVt 70)1/ TrpayiiaTmu Dem. 1 8, 247. Deiii. Proem. 30, I. eVt rfov npci^eu>p Dcm. Proem. 55, 3. t6i> (7t\ rwv viTr]pniKU)V Aeschin. 2, 73. Tovy eVi rovroav firfarfwras Herod. 4, 84 ; etc. For the difference between these and the dat. see p. 55. 3. A third more difficult and indeed impossible line to draw was that which should separate the local en-i and the more vivid of the temporal eVi's from the gradually less vivid temporal use which finally loses all imagery in e. g. {ju.) eVl tcvjoi^ (ro-de). Beginning, therefore, with r; 120 oyx^T] tn oyxpi) and S 130 ev) ovs fVi Ilarpd/cXo) ne'cfyvtv . . . 'A;^iXX€i^f, where enl IS often translated 'in honor of Compare also Xen. Anab. I 8, 29 01 p.ev (fiaai ^aaiKea KeXfvaal tip' f'Trtcr(f)d^ai airop Kvpco (said tO be local) with Cyrop. VII 3, 7, where the same phrase fma cfidTT fip npi is said to mean 'in his honor.' In accordance with custom, I have there- fore ruled out the following and like instances, though in many the local sense seems sufficient: Xen. Hellen. Ill 2, 5 ^(i\//-acT€f rois iavTwp Koi 7To\vp olpop eKTTiopres in alrois. Acsch. Ag. 1547 '"'^ " *""•" Tvp^ios alvop in dvbpX 6ilo) . . . noPTjcrei/ Aesch. Eum. 329 in\ Be t(^ TiOvpfPcp Tode fifXos. Eur. Alces. 148 ovkovp in' avrrj npaaatrai ra np6 24. Kipa (of army, either rrg-ki or /(?/"/) Hd. 9, 102. Thuc. I 49, 6. II 90, 2. IV 43, 4 6^' «. 93. 4- 94. I- V 67, I (2). VI 67, 2. loi, 4. Xen. An. I 8, 20. VI 5,11. Hell. Ill 2,15 (2). IV 4, 9. V 2, 40, 41. Oec. 4, 19. AuSois Soph. Trach. 356. XooTpolai Soph. Elec. 445. /Lte'ffto (of army, like Kepas) Xen. An. VI 5, II. ottXois camp Com. Frg. Adesp. 663 (III). Thuc. VII 28, 2. VIII 69, 2. Xen. Cyr. VII 2. 8. op'iois Thuc. II 12, 3. Xen. An. V 4, 2. Cyr. II 4, 31. VIII 5, 21. Hell. VII 2, I. 4, 39. Andoc. I, 45. Lycurg. 47. Dem. 18, 174, 230. opp.r) Plat. Timae. 27 c. opuco Thuc. Ill 76, I. opois Aesch. Prom. 666. Eur. Med. 540. Hd. (ovpoia-i) 3, 91. 5, 52. Xen. Hell. VII 2, 20. ovpa (of column of men) Xen. Hell. IV 3, 4. 7rpo6vpoi.cr(i) 2 496. a IO3. Plat. Com. 4, 2 (II) (Bergk eVi, legebatur hi). Eur. Alces. lOi. Plat. Phileb. 64 c. irpvpivois (ayopai) Pind. Py th. V 93. aKonnis Xen. Cyr. VI 3, 6. orei/w Xen. Hell. VI 4, 3, 27. arparevfiaTi Isae. 4i 26- (TTpuTOTTibco Xen. An. VII 3, i. TiKevTrj Xen. Mem. I 5, 2 roG jSi'ot;. Aeschin. 3, 205 rijy dTroXoyta?. Plat. Gorg. 516 a ToC /Si'oi;. re'Xci Plat. Euthyd. 291 b, of a discussion. Menex 234 a. Leg. 730 c. 818 a. Polit. 268 d. Rep. 506 d. 532 b tV' ovrw ■ytyj/erai rw tov vorjTov reXei axTTVfp eKeivos tots tiri rw tov oparov. Ttpp-ari, -v Thuc. Ill i8, 5. aKOTTTjs Xen. Cyr. VI 3, 12. Keparos Hd. 9, 47. Xen. An. I aTpuTonibov Xen. An. VI 5, 4. 8,9. Hell. 114,13. VII 5, 25. Plat. Leg. 674 a. on\a)v caynp Xen. De Rep. Lac. reXfuTi";? (jov \6yov) Aeschin. 3, 12, 7. 257. TrXeupwi/ (of column of men) Xen. roVoi; Dem. 10, 23. An. Ill 2, 36. ({)povpas Xen. De Rep. Lac. 13, npoaareiov Thuc. II 34, 5. I, II. Here too, it appears, is there an oscillation between gen. and dat., but it is not a vacillation. The context will show that the dat. is deictic, the gen. adjectival — a difference strikingly shown by the two following passages : Dein. 3, 8 eVi /ueV twv ciWcov aSiKr]- p.uTa>v (rKfyj/apivovs aKpi^oiS Set fifd ijarvxias ical raXrjdis e^eraaavras, ovTcas TTiTidevai Tols TjdiKrjKoai rrjv Tip.(opiav, eVi de Tois (pavepals (cat napa ndvTuii> wixoXoyijpfvais npodoaiais kt\. IsOC. 15, 20 koi yap alaxpov enl pttv rasv aXXcoj/ irpuypLaTdiv eXerjpoveaTaTovs opoKoyelcrdai . • . , eTrt oe Totv ayuxri Tins ivddde yiyvop-tvois Tavavrla ttj 86^tj Tavrrj (fiaivfadai npuTTovTai. It Will be seen that secondary matters (aXXa) are disposed of with the geni- tive, important matters emphasized by the dative (note toIs e'vOude yiyvofievois in the last example), 6. Finally, the phrase eV oiKripams (^epyadrrjpiov, reyovs) has been excluded, when signifying places of prostitution. The passages are; Hd. 2, 121 e, 126. Dem. 59, 67 epyao-ri^piov, by euphemism. Aeschin. i, 74. Dein. i, 23. Plat. Charm. 163 d. In view of the fact that Grecian houses were built low — perhaps especially the case with cheap houses of prostitution, mere slaves' quarters — Professor Gildersleeve has suggested as somewhat more than probable that the women literally sat upon them, just as other wares would be exposed to view.' See his note on v Ka\ Toi ^^paros — a passage to which Lutz fails to draw atten- tion. 1 See the scholiast, however, on Plat. Charm. 163/': t. tov deafiuTT/plov, wf Avcr/af, ?/ e, nopveiov, wf 'Attikoi, 64 B. Examples of 1-kI in the Attic Inscriptions. The following are the instances of eVi c. gen. and dat. in local sense in the Attic Inscriptions down to 300 B. C. : eVi c. Dat. Vol. I I, 40 /3a)/i(B. 273, 14 SoDI/tW. 273, 22 HaXXaSiw, also in 1. 5. 321, 20 roix<^, also in 1. 43. = a/. 322, 9 yovia ad angulum Boeckh. 322, 83 Trpoorao-ei ad poHicum B. Vol. I 322, 90 enia-TuXlois in epi- styliis B. 324 a I 44 Kuyotarioi/ . , . to ini (sic) rw emarvXia. So 324 f II 12.' 432 a, 32 SiSciM ("is locus, ubi terrarum situs t'uerit igno- ramus," Kirchh.). Vol. II 163, 19 ^w/xw (twice). eVi c. Gen. Vol. I 31, 17 A ra)^ em OpdKTjs. So 181, 3. 446, 46. 157, 6 KOpt] XP^'^V ^'"''^ O'T^Xjyj. So 170, II. 173, 6. 319, 19 KXifMOKe) ((ji' cov 01 Xidoi eVeko/xi^oi/TO. Vol. I 322, 44 Toixov ad parietem Boeckh. So lines 51, 67. 322, 86 Kopu^v supra puellas B. 324^ I 18 TQV apdpa TOP eVt ttjs ^aKTT]pias elarrjKOTa. Vol. II 167, 63 Toixov. Table showing the Local Use of In 6. Gen. = sub = U7rc'K Total Gens. Dative, Total Dats. Accus. Total Aces. Horn, and Hym. 28 17 (18) 133(1 34) 109 212 70 72 Hesiod 12 2 32 14 25 2 2 Pindar 3 4 12 12 26 5 6 Aesch. 7 48 15 22 14 14 Soph. 10 60 10 10 5 6 Eurip. Aristoph. Herodot. Thucyd. 12(13) i(?) 5 1 I (4) (?) (?) 13s 157 457 366 41 6 14 4 45 9 40 13 29 9 22 6 30 12 45 43 Xenoph. Orators 702 1294 28 8 52 63 24 See note. 33 26 Plato [6 .? 5 16 13 15] 65 In preparing this table I have relied in the case of Homer, Pindar, Aesch., Soph., Aristoph., Thucyd., the Orators, and Plato, upon the lexicons of Ebeling, Rumpel, Dindorf, Ellendt, Sobo- lewski (dissertation). Van Essen, Lutz, and Ast respectively. The last-named is of course incomplete, and I have bracketed the figures. For the other authors the count is my own. The fol- lowing notes may be added as explanatory, or of interest : The phrases vn6 ctkotov, -ov, -&> {(n(f)ov, crKids, aiiyiis, (vBiav) have been included in the list as local, the metaphor being a vivid one, if indeed they are to be reckoned as metaphors. But the phrases Sa/X7yi/at (Te^^ai/at, etC.) vno ;^fpat (^naXafirja-i, etC.) rivos, OV {jiro rivi, have been excluded from the purely local list, as also vno fjniXTjs. This last phrase has certainly lost all literal significance, as shown by the word itself, which occurs only in the genitive and only in this phrase {naa-xdXr} being the literal word). These are almost certain marks of " adverbiale Erstarrung,"' rendered doubly certain here by the fact that the sense of occidte is necessary in some and admissible in all of the passages where it is found, viz. Aristoph. Lys. 985. Xen. Hell. II 3, 23. Lysias Frg. 54. Dem. 29, 12. Plat. Gorg. 469 d, Leg. VII 789 c. Sobolewski is inclined to admit this : " nescio an hie quoque (Lys. 985) haec significatio {clanaihivi vel occulte) praeferenda sit," but Lutz quotes the Lysias fragment as " das einzige lokale Beispiel fiir 'vn6 c. gen. bei den attischen Rednern." In the usage of the individual authors the following points may be noted : Homer shows but two cases where vno c. ace. may be taken as temporal, n 202 vno \i.r]viQ\i6v, X io2 viixB' vno. The sub iviperio sense of vn6 c. dat., which is steady and frequent throughout, begins in Homer with nine instances, at the head of which may stand vno aK7]nrpa — the source perhaps of all the others. Hesiod, according to Rzach's constitution of the text, shows vno c. ace. only twice, one of these being Ini x^*'"" Theog. 304, the other ovpas 8' vn6, Op. 512. Pindar uses vn6 c. dat. seven times to express agent, against vno c. gen. five times. His preference here therefore, as well as in the local use, should decide for the dat. in Ol. VI 40 \6xfiais vtto Kvnvims, altogether aside from the question of picturesqueness (as one item of which note the imperf. tense of (tikt(). Note also the difference between In' AiTvar 01. XIII iii and vn' 6