(pp. 228-241) e.s. roberts - the oracle inscriptions discovered at dodona

The Oracle Inscriptions Discovered at DodonaAuthor(s): E. S. RobertsSource: The Journal of Hellenic Studies, Vol. 1 (1880), pp. 228-241Published by: The Society for the Promotion of Hellenic StudiesStable URL: http://www.jstor.org/stable/623622 .

Accessed: 08/02/2015 11:11

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact [email protected].

.

The Society for the Promotion of Hellenic Studies is collaborating with JSTOR to digitize, preserve and extendaccess to The Journal of Hellenic Studies.

http://www.jstor.org

This content downloaded from 89.34.228.69 on Sun, 8 Feb 2015 11:11:39 AMAll use subject to JSTOR Terms and Conditions

http://www.jstor.org/action/showPublisher?publisherCode=hellenic

http://www.jstor.org/stable/623622?origin=JSTOR-pdf

http://www.jstor.org/page/info/about/policies/terms.jsp


228 THE ORACLE INSCRIPTIONS

TIHE ORACLE INSCRIPTIONS DISCOVERED AT DODONA.

STUDENTS of archaeology are now familiar with the splendid work in which Constantin Carapanos two years ago gave to the world the results of his discoveries at Dodona. The vexed question of the site of the ancient temple was finally set at rest, it will be remembered, by the discovery of a large number of inscriptions recording dedications to Zeus Naios and Dione. The immense quantity of relics and works of art brought to light in the course of the excavations has been exhaustively catalogued in the work, Dodone et ses auines, and they have been illustrated and described by various scholars and reviewers. The inscriptions, too, have, at least on the Continent, come in for some share of notice and criticism. A detailed account of these inscriptions-their contributions to the lexicon, to dialectology, to local and general history, and to topography-is still a desideratum. For, as was only to be expected, the interpreta- tions and criticisms of Carapanos himself are rather general than critical. His text, moreover, is frequently open to objection.

In a classification of these inscriptions our attention is at once drawn to an obviously new category; and it is with this alone that we propose to concern ourselves in the present article. The category comprises a quantity of more or less legible inscriptions engraved upon one or both sides of leaden plates often not exceeding a millimetre in thickness. These plates form a unique series of documents belonging to the archives of the famous oracle at Dodona, and contain the questions addressed, or prayers offered, to the deity by his votaries, who might be either communities or individuals. Whether, as Carapanos thinks, we have in some instances the actual answers vouchsafed by the, god, is more doubtful. We shall recur to this point below.



DISCOVERED AT DODONA. 229

The remainder of the catalogue may be tabulated thus:

(1) Ex voto inscriptions on bronze.

(2) Inscriptions on bronze or copper, recording (so far as they are legible)

(a) Decrees of citizenship, conferred respectively by the Epirote league, the assembly of the Epirotes, and by communities whose names have disappeared.

(b) Deeds of manumission. (c) Deeds of proxenia conferred on individuals, and in one

remarkable instance on a whole community. (d) A contract establishing right of intermarriage. (e) A gift of lands and other property, by a person or persons

not mentioned (or possibly by the town of Dodona). (f) Purchase of a slave.

(3) An inscription on an iron strigil. The reading is doubtful, but the subject is probably dedicatory.

(4) Two or three inscriptions on terra cotta.

(5) An inscription on a limestone tablet containing a decree of the Epirotes conferring the title of proxenos with certain other rights and dignities on Gaios Dazupos Rennios of Brundusium and his descendants.

The questions addressed to the oracle in many cases present great difficulties of interpretation and reading. The text is often rendered additionally obscure by the fact that the plates are inscribed on both sides. Possibly the first writing may have been in some instances partially erased by hammering, as in the Laconian inscription concerning a deposit made by Xuthias at Tegea (Cauer, Delectus, No. 2). How great was the task of decipherment may be judged from the photolithographic copies which Carapanos has given of four such tablets on Plate xl. of his work.

It would appear that the consultant was in every case required to put his question or prayer into writing. The discovery of these leaden plates, as Bursian remarks (Sitzungsber. d. k. Bayer. Ak. d. Wiss. z. MiAnchen, Philos.-Philol. Cl. 1878, pp. 1-28), confirms the explanation which had been previously given of the word 'sortes' in a passage of Cicero (De Div. i. 34, 76); ' Maximum vero illud portentum eisdem Spartiatis fuit quod




cum oraculum ab Iove Dodonaeo petivissent de victoria scisci- tantes legatique illud in quo inerant sortes collocavissent, simia quam rex Molossorum in deliciis habebat et sortes ipsas et cetera quae erant ad sortem parata disturbavit et aliud alio dissipavit. Tum ea quae praeposita erat oraculo sacerdos dixisse dicitur, de salute Lacedaemoniis esse non de victoria cogitandum.'

Here the sortes are evidently the leaden plates on which the questions were written. The passage further shows that they were collected in a vessel which was brought into the sanctuary tc be examined by the priestess who delivered the oracular answer.1

The form in which the questions are presented varies con- siderably. The most complete of the petitions open with an invocation to the deity, corresponding nearly to our 'In God's name;' this is sometimes followed by an appeal, in the vocative, to Zeus Nalos and Dione. Then comes the name of the enquirer, with a verb, generally, but not always, in the third person, introducing the question; and lastly the substance of the question. One or both of the first two ingredients may however be absent, and all that remains is an abrupt question or a query as to the better of two alternative courses.

The invocatory formulae (with which we will include for con- venience' sake also those of the ex-voto and proxenia inscriptions) may be reduced to the following types: 1. 8e?6 7Trxa. 2. -eb'

~'-X JeyaOa. 3. d yaa ^

'Xa. 4 OE4 ' ^Vxa ayaa. 5. O•0v

T-•xav adya~dv. 6. Oe• '

xav dyaSdv. 7. [Oeoq 'r'xav a- y]aGcv.2 Of these, nos. 5 and 6 differ from any of the types collected by Franz (Elem. Epigr. p. 318). For, while 08e6s 7xCa arya9 may perhaps be explained as an ellipse for 8eos t6xa aryaa ̂ rrapelr), 'Deus bona fortunh adsit,' while OeO Tuvxa is said to require a similar supplement, while adyat9 7-aa, sc. dhi, seems to be a fair equivalent for 'quod bene vortat' and 8o01t is naturally supplied to 086'E TvdXav yaacdv; and while Oew -rvxa a7a08 finds its parallel in Oeo^v r7Xy (sc. e Xopas ?) of a late

1 Perhaps a reminiscence of this meaning of sortes lurks in the Vergilian use of the word. Compare 'Italiam Lyciae iussere capessere sortes.'

A en. iv. 346 'Hic ego namque tuas sortes arcanaque

fata,

Dicta meae genti, ponam.' Aen. vi. 72, 73.

And in the next line- ' Foliis tantum ne carmina manda.'

2 So Carapanos supplements P1. xxxvi. No. 4, reverse; but it may just as well be Ol6v or OEI.




Attic agonistic inscription of the age of the Antonines, C. I. G. 281 and the older 9eo^ d[ rrwtcopt'ot], i.e. d'rriKcovpLotq, C. I. G. 139-it is not so easy to see what was intended to be the syntax of

Be, r'6xav dryaOdv and O6eY 'Xav i iyaOdv. Perhaps

evXopaa 9evo (0e6v) vdxav aryaOav 8o0vat. The deities whose advice is sought are Zeus Nalos and

Dione conjointly. Though the fact of their joint supremacy as oracular deities is so amply attested by the constant union of their names on the leaden plates, there are, strangely enough, very few passages among those quoted by Carapanos in which Dione is mentioned in connexion with Dodona and the Dodonaean Zeus. Thus Apollodorus speaks of a Dione 7rapa A~w'oavat'osq, though he at the same time confounds her with Hera (ap. Schol. in Hom. Od. 7. 91, 'f2~ ical 2 "Hpa [Cvogdo-0r] Aultwv

'napa Aw•owvalotg, 6 'AtrowXX68opo,); and the Schol. in Hom. II. 2. 486 includes her in a list of several Dodonaean nymphs: Ze~ Ec 'TO) /f7 po flyevvV10rl'a ALo'vvoov -atL Aco•wviv t

vtv,- 'catL p'betwv

E"KtIcV, 'Apjpot'a, Kopovl&t, E 3app, A

tvn, Aloa5yI, I-oXv'o^. We find an allusion to -bv ala 7'Ov Awo w-

valov icat '27v AtcVrv in the Epistles of Demosthenes (?) (iv. 3); and in the speech against Meidias (? 53 = 531), in a perhaps later insertion with reference to oracular answers from Dodona, there is an enumeration of victims offered to Zeus Nalos and Dione. Similarly 6 ZevP,

e AtoSvY are coupled together in the speech

De Falsa Legatione (? 299 = 431). And in a passage of Strabo

(vii. 329) not quoted by Carapanos we read: drre~TB vcal a6vvao9

7I At'l 7rpoo-a.Treelt'XOl Kal 7 Atw'vrq. Bursian further suggests

that Servius (on Verg. Aen. iii. 466) had the same combination in mind when he spoke of a temple at Dodona consecrated 'Jovi et Veneri.' 1 Ovid, we know (Fasti, ii. 461; v. 309), used Dione as synonymous with Venus, and the step from the use of Dione = Venus to that of Venus = Dione as mother of Venus is easy. Doubtless, too, the epithet Dionaea helped the confusion.

The epithet Ndioq (aqueous), the special attribute by which Zeus was known at Dodona, and by which he is qualified in the

majority of the inscriptions, is not altogether new. The follow-

ing passages abundantly illustrate the usage: 58 'wova&to?

1 Carapanos cites the passage in

support of his conjecture that a shrine to Aphrodite existed on the spot where

he discovered a small wheel of bronze with the dedication : 'f Xlovw 'Aqpoa8'rG vwe6y'ce (P1. xxvi. 1).




icat NdBo" vopXI a ydp rd Elcet Xopia (Schol. in Hoem. I. ii.

233); 7•

A d rT Nat( (Dem. Meid. 1. c.); Tv &U Awi8owvaZov xfeyov ica Nd-ov (Steph. Byz. in v. Aoscvq).

The most common formula of interrogation is drepero- (drrepoo- TrCvrT, C7repp6reT, cporoj, EE'eporT) rbv 6ia 'vT Ndvov icaA ~rdv Anrvav or

TwrcowviJraT (isrMowvraT, dtLcolyVCdvrat) 7de for T

Naio Icaal

Tr• A o~rva.

Thiswn, ascotva-Oat is only a dialectic form

for Att lcovwoa o-at, known, as Bursian points out, from

Attic usage (Plato, Prot. 313 b) in the sense 'to ask one his advice about anything.' Besides these, other formulae occur, as cpoirat, cheoi^rat (where the meaning is not quite clear), lrTope~-ticerevTet-atlTEt uctpa icalt icTreVes. The last two however introduce a prayer rather than a question, though perhaps a question, now lost, followed. The nature of the enquiries is, as might be expected, most various. The consultants are states, corporations or individuals; now it is man and wife together, now a would-be investor, now a suspicious husband, now a seeker after lost property, now a

shepherd. We shall give a better idea of the procedure if we subjoin a

more or less detailed account of some of the plates. The people of Tarentum seek information 7rep't ravrvxlaT

and certain other matters. 7ravTuxia is a new word; if Bursian is right in comparing it with 7ravwoXepla, we may assign to it the meaning 'general prosperity.' He is however wrong in

supposing that we have here a special Tarentine form of the article in [Ta 7r6]XtS "r Tapavrivov. The true reading is, as Blass shows (Rhein. Mus. 1879, p. 160), undoubtedly a; the rude cross before the second a stands for 1-, a form of the spiritus asper known from the Tabulae Heracleenses; and what more natural than that the metropolis Tarentum should use the same form as its daughter-town Heraclea ? 1 The resemblance in the character of the letters too is sufficiently close to entitle us to assign this to the same age as that generally assumed for the Tabulae Heracleenses, i.e. the period between 324 and 279 B.c.2 Blass acutely saw that the fragment, P1. xxxv. 4, of Carapanos formed the missing portion on the left hand of xxxiv. 4, and in

1 Mr. P. Gardner points out to me that this aspirate is not unusual on coins of Tarentum.

2 See however Meister, Curt. Stu- dien, &c., iv. 360.




this fragment the letters [-A are quite distinct. We subjoin the text as restored by Blass-

Oeoq r dvj arya9d [dwepcor?3 - a rro6 X' J- a rC^v Tapav ['rivtov

7'v a la r6v Ndtov Kat 7 [dT v A' dvav W7pt

•ravrvxta xcal 7r...

TaXeL eSp, ica 7rEpl ra0v...

He proposes in lines 5 and 6 icat T[C icd re ~ j 7-6aEe

Another inscription' presents a remarkable example of writing in cursive character; it would seem, however, from its contents to be anterior to the Roman conquest. A people whose name has disappeared, but who should be neighbours of the Molossi, seek to be shown how security may be guaranteed to them if they ally themselves with the Molossi.-A community whose name survives only in a defective form puts a question, the tenor of which cannot now be unriddled.2 The only decipherable words are, line 1, 'ETrutcotvarat Mov.&ta'rav rb icowyv 7rep Tov (= r20v ?), and line 3, re ical /3Xrtov

, (e~) KicXprPtEv. Line

2 seems to yield HlVpp6 (' ? not r) ra (?) Odt 7~ A'ia (?) dc T[WC]v dTart 'a Oept... In Mov.StaT&v one letter only appears to be missing and the 8 is incomplete. Above at the right- hand corner are the words Al

Nd, val Atd?va, but as these are

written with 12, which, as well as H, is absent from the remainder of the inscription (whence we may infer a considerable antiquity), it is likely that the four words form a later addition or are part of another inscription. Although the inscription contains a request made by a community, it is written on the reverse of a plate containing the prayer of individuals, Eubandros and his wife.-The Corcyreans3 ask to what god or hero they must offer sacrifice or prayer in order to secure the blessings of internal harmony. P1. xxxiv. 5, though much mutilated, seems to contain a similar request. The digamma is plainly marked in FOI COLtEv. On the reverse (5 bis) is the monogram A.

1 7repwTvTrL Tb I KO7bV TiV . . .

wv Ala Naov Kal Azcvar Ka['4r] ) ril ab'roTs

ovlt7ro•t'rTfovrt[v] e•rET& MoAooorov ao1-

9paX j..-Pl.

xxxiv. 2. ao4paxi is best explained as neuter plural.

2 P1. xxxiv. 3 bis.

3 OEbb T[6]Xav &ya0av. 'EV[I]crr- ,vp,"rai rol K[o]pxvpa[toiT 7r A lr 7i] Nd, Kal 7r A[t]6rva Triv Ka [OeC-v )] I ipocow O6op[v]esKAu f Ex[6E1evoi] I oIo- OSEv Il[j]j TcyvaOdY.-PL xxxiv. 4.




So far the demands of public bodies. We now come to those of individuals. We begin with a question put by Eubandros 1 and his wife, who would learn to what god, hero, or 8ai4ov

they must pray and sacrifice, that they and their household may prosper for all time ('iiov icaa c~LeLVov Kca rpao-ootev-a formula which frequently recurs). This inscription is especially in-

teresting from the evidence it affords of a distinctly local dialect. The orthography of Atel perhaps does not call for comment; the final et for t may be merely an indication of local pronunciation, just as in P1. xxxvi. 2 and 5 Aqcava for A twva may denote a pronunciation of q approaching the I-sound. But Evb'av8pol, 4e(Ov, 4cdov0re are new, and lq almost new, contributions to dialectology. The last may be compared, as regards the vocalism, with the iv of the Tegaetan inscription (Cauer, 117, iv T5 A'pya, cf. \lyVce XpIKoL), and with the Iq r0v'peyo (el rrVp(yov) and iE dpE'vt (elk pivtov) of two Pamphylian inscriptions (Cauer, 75, 76). The form does not occur in the Abou Symbul inscription, for the words XkOov 6\ K pEfoP IcaTi7repOeN II O [O] rora\v cdvl should undoubtedly be read, according to Blass's happy conjecture (Hermes, 1878, p. 381, seqq.), ia7v'r;epOE, v1 (= b' ro0) 6 woTapoaE d4Vi, ' as far as the river allowed.' With regard to

OeaS and netw G. Curtius, who is probably right in considering the inscription to be Epirotic, remarks (G. E. 5th ed. p. 485), that in these instances of labialism we must recognise a dialectic

process which at that period 2 presupposes the spirant instead of the aspirate pronunciation of the 0 from which the 0 came. He

aptly compares Modern Greek Fibae = ?Oe at, Russian Feodor = Oedopo. For the much older O'p = 0p he suggests another explanation. It is -remarkable that the ordinary form Oedo occurs in the initial invocation. Doubtless this was too

stereotyped by usage in the ritual of the temple to be displaced by mere local peculiarities. Ei/3avypo; will be readily identified with the familiar Vergilian Evander under another form. The

p which Rangab6 calls 'eingeschoben' is more accurately ex-

plained by Curtius (G. E. 5th ed. p. 586) to be a symbol, like

Oe - TtXaY &yaOvY. 'EVrLKOLw7iTcaL E6-

,av\8pos ical & 'yvv? hT ALE d T NdiaZ Kal 'r "l?lcy ri'S ca 4EW

- '4p6WV 6 4asd-

Iwv I fv'd eo'EO

L Kcdal q ovr' es Xc(OV K

ca

&ApEOIly pda'OOLEP Ko c al aCrol ical &

o'e'7nsTL Kad v O I cacL is rby aravra

Xpdvov.-P1. xxxiv. 3. 2 To judge from the character of the

letters, the second century B.C., according to Rangabe, Arch. Zeitfig, vol, xxxvi. p. 118.




the F in the Cyprian EbFaTypa9, for the w-sound (we should be

inclined to say th6 English w-sound) involuntarily developed out of the v before a vowel. We cannot acquiesce in Rangab6's statement that the /9 is an 'unwiderlegliche Beweis' in favour of the Modern Greek and against the Erasmian pronunciation of av, ev, rlv.1 The letters Evav, for the ordinary Evlavspov, are written on the reverse side, perpendicularly to the remainder of the inscription. Similarly are engraved the fragments ap and el4e[v]. It seems likely that in some cases these letters and isolated syllables found on several of the plates indicate the names of the persons who address the oracle.

To continue: An Ambraciot enquires concerning his health, fortune and general prosperity, and would know what gods he must propitiate ([i]Xao-c0'eYvo9) to gain his end. In Carapanos's text, given below,2 ia should probably be added after Oec^ov.- A man named Socrates wants to know3 how he may trade most profitably for himself and his family. On the reverse of this plate is the beginning of another query (['Errtcotv]iwvrat Alt' xaL A[Cvqja]), and a fragment in larger and deeper characters (...101 Kaa

a 7a 1 n O Xl'oe j ...[d]]9 YV]EV 7VA6V To Xa9El), which betrays a Doric origin.-Several consultants ask advice in their hesitation between three courses: whether they shall prosper best by going to Elina (a place not otherwise known) or to Anactorium, or by effecting a certain sale; and a woman seeks relief from a disease. These last two queries are on one side of a plate containing on the other three fragmentary lines in Attic, apparently an enquiry about offspring.4

Another fragment, also inscribed on both sides, appears to contain a query about the advisability of applying for citizenship in some town or other. The words are: "H alrcopLat 'rdv i...

1 Whatever sound is here assumed for the 8 (whether a B- sound or a V- or a W- sound) it is perfectly com- patible with a genuine diphthongal pronunciation of a preceding av, ev, s7v.

2 ..... r 'AUBpaclf~T7rs] Adl Nal icl A[qrOv6] I repI vy7elas abroO [Kal]

I TOtV brapXO'VTWV [Kal PVP] Kal eCIS Tb grflra [XP] I d'Oov, rlvas Oew^v [1] Aacr- K4AEVOS A65LrOI icalcd &LeLov/ wrpd[ao-oi].-- P1. xxxvi. 5.

3 [Tc Ail 7T Natere na rP1 As c.Sva

IWKPcaT71S drLKOll [va^ai TI Ka] ipyaCo-

pepos Adrnov Kal &pEtvovI [w7pdaao& ab'r]bs Kal abri Kcal yeveci.-Pl. xxxv. 2.

4 (a) "H Els "EAc arpI prepAo. .. . j 4 Es 'AVaKrdpIOV .. . . i-wAo^.CVres TbY

(b) ['EepCer. . . .] a Tlu/ Ka eLw^v Ov'ova I [Cal ebXooe'la /A.LEpYO ] wpdt-rtot Kal Tas vo'ov I [&7raAAaXetlZq?]

(c) ['bV] Ala cal TiP AcIyI f . ..

rbz BVby drepwra, I ... dIC 70S yvviacKdS. -P1. xxxv. 1.




wroXtrelav 4E rav'r[at?]... I i nTO elOLoVTro1 (P1. xxxv. 3). No sense can be extracted from the inscriptioh on the reverse, originally ten lines long. In an inscription evidently complete in itself the enquirer plunges in medias res without ceremony of invocation or other introductory formula, and asks: "H abCos w7Traed/.ovo T dv EdrXL olcta ic Ka b XopIlov, 8AE'X-to6' kotL I

e-y Kqal 'roXvwx6Xe[i]T6[p]ov (P1. xxxvii. 1): 'Whether if I myself have acquired possession of the house in the town and the farm, it will be better and more profitable for me.' The form WrewraCLvov is sufficiently familiar;

7roXv•o•eXcrrepov is due

to Blass (Rhein. Mus. 1879, Heft i. p. 160). It is not so certain that we should follow him in assuming a mistake of the engraver in dr6"Xt for Ed' 7rdXt. We can hardly infer carelessness in all such cases, as e.g.

d•avrtveta for 4d1 MavrTveL'a in the Praxiteles-inscrip-

tion of Olympia (Arch. Zeitung, 1876, p. 48; Cauer, 32), 6d'-T•Xr for dv orj~X of the Athenian decrees, and the Boeotian erao~Lv for g~praotv = ey/K"~ryov (C. I. G. 1564 = Cauer, 118). The

tendency to complete assimilation of consonants is shown also in /3dXrtd, got. The construction is summarily changed in the apodosis, as in the following inscription (P1. xxxvii. 3) ; ...'I rrcw- urparo v rt 2 ' Ndc ical t [(a t)Aava t'va T ica Oedv

Y w]v

Xa'o/y itat ap/Letvo/i 7Trpado-aootL, to the beginning of which

perhaps d'nLEoW a"rat should be supplied.-We next find ' Agis asking about the mattresses and pillows, which have been lost; whether some foreigner (or, some one outside the household ?) might have stolen them. Carapanos restores drancXoX[ev a'Tr] ]; but adr7hoXoXev, doubtless a mistake for daro'Xwhev, is surely not transitive in sense. 'E?rPeprTE is a new form; so is 'porAi7 in an inscription in which Lysanias3 enquires concerning the

paternity of the child of which Annyla is pregnant-a 'AvvtXa Kete, as Bursian reads for the ungrammatical 8a Qv NriXa IVes

1 'EwrEpwre? 'A-ys MAa Naov [Kal Atw- wav] bIrEp TCv opwQdrOWV IC[alc i' 7rpoto]

KElpaxatCWV, T& d7r3SAoA[e], 4 TrcV

'rwel rs T &v K[A E 4%EV].-PL xxxvi. 1. On the reverse, Al and B.

2 For the interchange between the terminations -aw, -ow, -ew in the so- called contracted verbs, it may suffice to refer to Curtius, Das Verbum, &c. (Eng. ed. p. 244, sqq.)

3 'EpwT'i ,Avoa-~var AMa Ndi'ov I al

Agrc6Yav 4 ob) EuiC C'(TI L E' ro I 7 b rat~c'- ptov 'Avvha K6EL.-P1. xxxvi. 2.

The omission of the iota in EpwiTr finds its parallel in inscriptions of a somewhat late date, as this may well be; and even in so old a document as the Tabulae Heraclienses, e.g. doluia- OwO?, i. 112.




of Carapanos. 'Avv6Xa may be a diminutive of the foreign name "Avva: an Illyrian female slave "Ava is mentioned on a Delphian inscription (Wescher and Foucart, Inscriptions re- cueillies d Delphes n. 439).1-Another inscription, interesting also from a dialectical point of view, records the question, expressed in the first person, of a man who wanted to learn 'whether he should succeed in trading in such way as might seem to him expedient, living (?) where he pleased, and carrying on at the same time his own craft.'2 The dialect is difficult to determine; g'tetv for elvat reminds us of the Rhodian inscriptions, which however offer no instance of Tyj

for -, nor any exact parallel to the vocalism in Xpewtevoq. The v again -in ob'rv9, though it would ordinarily bespeak a Boeotic or Aeolic origin, may perhaps be compared with the v -- ot in the Abou- Symbul inscription previously alluded to, which is presumably Rhodian and Doric." The writer may well be a Geloan or Agrigentine; for that the Rhodian colony of Agrigentum had dealings with the Molossians is shown by a decree 4 in Carapanos's collection, in which the Molossians grant the Agrigentines en masse the title of Proxeni. The characteristic S1pewv appears also in this decree.-In an inscription beginning with the remarkable formula Cpoi3rat, KX6eoi0Tat, the enquirer asks5 'whether it is a good investment and advantageous to him to take up sheep-farming.' On the reverse, in larger letters,

1 The name "Avva occurs in C. I. G. 4003 b (Phrygian), 4315 c (Lycian), 4379 g (Pisidian), and in several Chris- tian inscriptions.

2 T'Xa d7ya0d. 'H 'uVyxdJvoLIr

Ka eL-

7ropevuevo. 3Irvs Ka c ocK o6b1popov FeLy I Kal &ywv 7r i KC 8o0c I &94a T_

TCX•q XPEVLEjos.-Pl. xxxvii. 4.

The plate containing this inscription is literally ' scribbled over.' On the first side can be recognised the words Tb~ Ala brbPW AW•o aov and some letters of two other inscriptions in large character. On the reverse there are no fewer than four inscriptions: the few decipherable words of one (. . . .

dyarOv Tas yvvaoc s v 'ras 7rapa/h[4p]fov .-. .) seem to be part of some enquiry touching conjugal fidelity.

3 Compare also the v Ka So'Awvral of a Cretan (Doric) inscription (a treaty between the Hierapytnians and Mag- netes, Cauer, No. 461. It is on the same stone with two others, one of which is Rhodian).

4 P1. xxviii. 2, '8o?e -roTs MoXooAoo~s 7rpolEYiav 8d~UEJV roTr 'ApayaV•~roir. This document is interesting as being without example in history. Perhaps what is meant is merely something like the honorary title of fratres accorded by the Romans to the Aedui (Caes. B. G. i. 43; Tac. Ann. xi. 25; Cic. A4tt. i. 19). The decree is engraved 'au pointille' on a plate of bronze.

5 'Epo~'rat IKCov-rat rby Afa tal rc&v

SAtdrav aT ~Orrt ab'rol7 7po3arE•Eo•r• &vawoL ical pe'XYou.v.-P1. xxxviii. 1.




are the words 7rep 7rpoaaTE?'aV, doubtless indicating the subject of the other side. d vatov is new to the Lexicon; abTroZ for ab ^ is also to be noticed.' 'Epohrat may be a collateral form from 'pota&, but cXeoirat is harder to explain. May it be

cognate to KaXkB(, ieXev'&W, /cXopat, with the sense ' calls upon,' 'appeals to'? See however Curtius, G. E., No. 29, b.--Ai Athenian, Diognetos, son of Aristomedes, 'begs and beseeches you, Lord and Master, Zeus Nalos and Dione, and you Dodo- naeans, to grant (a certain favour) to himself and all his well-wishers and his mother Clearete (TeFr p jrp& KXeap'Tre&).' 2

The inscription is incomplete. The dialect is Attic; the form et for ,p in the dative feminine of A-stems is not un- common in Athenian inscriptions, e.g. in the expression e0SXOa t

Tet 3ovXe.--Another plate seems to contain a question relating to the inscribing, sealing, and dedicating of a writing-tablet

(rtvrdlctov) at Dodona.3 Among the dialectic peculiarities the terminations of ypaoOi3pevv,

aapav•t.ev are well enough known

from inscriptions (Ahrens, Dor. p. 315). o-a'owv for oy~ket'WV would supply Ahrens with at least one certain example from a sufficiently old inscription of Doric 1 for et when followed by a vowel. According to him (Dor. p. 164) instances like 7rXiovav,

xp av, dra'd87o , daapawrov, are to be found only 'in titulis infimae aetatis,' e.g. C. I. G. 2060 (= Cauer, 35), written in the time of Tiberius or Caligula.-In a fragment in which Amyntac appears to enquire concerning the prosperity of his son,4 the letters ey Xto are lightly traced under the name 'A/wtk a[v]. These may be the remnant of another inscription, or may mark

1 It is doubtless a locative form used for the dative, as is plainly the case in the Elean Damocrates-inscription lines 21 and 28 (Cauer, No. 116).

2 [Oebs Tv'xnl &yOai-1 AE'oTroTar &vac Zei Ndz' eal AtcL IicLal Awova•loL

(sic), acre? L as I ical IKEE'veL A1lyVrTOS'Apwo- ToU7t] ov 'AOWaQos BovvtaL arrIj ical

-TOL &avTov eUvoLs &7raorv I Cal 'TE7 pr.Tpl

Kx.ap~cre ical ... .-P1. xxxviii. 3.

3 ... os ica Olca -ra•6I oAW

v .ry oApv'-

S. . . [e]cq/apvaro of8' wacE'a0eTo at

a [ Swuepov WeZv eI]s Awahvav 7rep TOV;

7rwvaldov . ... . a 6v vKicejaio ob5cs

rexpa .. .. e ypdPOeqCP KCal aCCdiPOt-.

/Aev.-P1. xxxviii. 2.

mdvaicia in C. I. G. 76, are according to Boeckh 'syngraphae mutui, diptychis s. codicillis inscriptae,' and on the word in C. I. G. 50, which is a list of donaria (Q 33,

tAaCnXOSo Xp'6o[S abPy ?]L ?va- KCqp tv . . . . dve'le 'AplTE1lM Bpav- pcWOI), he remarks, 'exiguam puto tabellam operis tessellati esse, quales nos quoque in anulis gestamus.'

4 'Ewepwre7 'AjA6v'a ... I A[la Nd]c'ov cal AtcSvav ( xrov Cal &.ELLr [vov Pir'kp ro]iv; Tabs j .. ayda'raa•• (?).-P1. xxxviii. 4.




the nationality of Amyntas.-Finally, Heracleidas prays for the blessings of good fortune, and asks whether he will have other offspring than his child

A'T'yX•l. The dialect, to judge from

Vj7gY ayarav, ryeWL&Lj, A'yvXrg, is Ionic in origin. On the reverse an enquiry is made (orTopel) by one Nicocrates.

If we add further that on P1. xxxvi. 4, Carapanos makes out a word wop.tnrara (?); that P1. xxxvi. 6 appears to run: "H cvlp/7eFiov[rt] I

auTr7 O rw-p 4T[o]05 7rpa/ytaro•os [v] ITVa Ica

7rp[ov] J...tv....arata... Ictpov I3i-Lo[v] ica. Tam..evov I IJ-

o'-aitcov (?) lalet7Ta w (?); we may return to the question, Do

any of the fragments contain answers given by the oracle ? That these, if they survived, should have survived in much smaller numbers, was a priori to be expected. It was only natural that, if indeed they were committed to writing at all, the enquirers should carry away home with them the precious tablets. The fragments in which Carapanos recognizes answers or portions of answers are the following :-

(a) [7- c7rt]ore6ove Y e Tt dXaOe[9]. (P1. xxxv. 6.) (b) del [b]Xa TM. (P1. xxxvi. 1 bis.) (c) [1d]8E 6To PaP771tOV Yry c Xpqol ice dXc'ie. (P1. xxxviii. 5.) (d) dcXXav pao•rTeled. (P1. xxxviii. 6.)

And further the plate (xxxviii. 7) containing a fragment of a prayer analogous to that of the Athenian Diognetos,2 was concealed in its central portion by another fragment of lead adherent to it, and wrapped in the larger plate. On the smaller can be read

aav- ^c... I abrr T ... O 0

aor'oo yl[ovei o ?]. Carapanos, whose reading is slightly different, says that three other small plates are found in a similar way to be wrapped in larger plates, all containing inscriptions. The most probable explanation, he thinks, is that the larger contain the question put to the oracle, and the smaller the answer returned. A photograph of such a combination is given in no. 3 of P1. xl.

With regard to the first of the examples quoted (a), Cara- panos remarks, ' La ressemblance de formes avec les lames de

1 [eeb] 'HpaicA[e]l8as alT? c br f [AlaC

'b Nroval ro Ia Yr V Aid6n 8oIva ab-] I [C,7] 7Txny dyaO, ,ical

r[o^&S abr-o Kial ?•rEpw'c] I . ... . repl yEVe ' 4 'r,-a[ actfr &i'pal. . . I A[i]yAys 'iijs vPUv

exeL.-P1. xxxviii. 4. 2 The Syntax is quaint: Ze' Ndie

S. . . cereE r .. . [icK] al ov'a ar?[?CV . a... I r ci /k c.ov uo ie c[ ala]

S. .[vyapl Kap I-] rab'.]



.240 THE ORACLE INSCRIPTIONS

la planche xxxviii. Nos. 5 et 6, et le sens de ce qui reste de l'inscription peuvent la faire classer parmi les rdponses de

l'oracle.' If by 'formes' he means the forms of the letters engraved, it must be said, on the other hand, that on many of the plates containing queries the resemblance of forms is sufficiently close to those of the three plates containing supposed answers, so that no stress can be laid on the resemblance of these latter to one another. Of the sense of the words which remain (7~T WTT1re0oV VT1i& aEXaO9p), the most that can be said is that it is not against the conclusion of Carapanos. The same remark applies to the words 6r? [a']ka TL (b), of which, in spite of his restoration, he suggests no interpretation. The fourth fragment on the list (d) has been more satisfactorily explained otherwise by Bursian, who refers to another plate (xxxvi. 4), on the reverse of which we read aXXav olceo I[-as] i.e. i) 'XXav oicrj?[av or -avTe'] or

OlK7.70LV. S So q "XXav itao-retee (for aoTrej6Lt)

should mean 'or does he seek another ?' In that case the e for 1, arguing a respectable antiquity, furnishes one more instance of the use of the pre-Ionic Alphabet in these inscriptions. We

pointed out one such case on p. 233. The remaining example (c), 0TOi azvrg[t]ov deyw Xp" , 'I give this response,' has certainly all the appearance of being an oracular answer. Mr. Gardner's

objection (Academy, August 17, 1878), that the ie dXae' which follows does not seem well fitted for an oracular reply, loses its force from the fact that on the plate these letters are added by a different hand. The facsimile has MANTHKON in the first line, but the K is close to a fracture, and the two shorter strokes

may be only rays in the lead. The Ionism in Xp'co and ptav- Trtov, as Bursian remarks, seems out of place at Dodona; but it is conceivable that the deity through his sage Peleias may have answered the enquirers each in his own tongue; or a simpler explanation is that the oracular answers at Dodona, like those of other oracles, were expressed in a dialect of strongly epic colouring, as at Delphi, where the local dialect was certainly not epic.

We have now examined in detail all the inscriptions in

Carapanos's collection out of which it is possible to extract or infer any meaning. There remain a few fragments repre-

1 [Oebs -r'xav dty]aOhv. ndrepa r.•.-

x[dvo1] I *I. . . . ea &aE xav olieo . . .

Can n'rav be for 14 •dY--'

either this or another' ?




sented on Plates xxxvii. and xxxix. In these the individual letters are for the most part plain, but only here and there can a word be made out. In xxxix. 4, the name

['A]Xepi,uaXoo is clear. The examination has established a new proof of the con-

siderable part which the consultation of the oracle played in the public and private life of the ancient Greeks. In circum- stances of importance or embarrassment the confiding Greek had recourse to the omniscient deity, and sought from him the means of succeeding in an enterprise or grappling with a

difficulty in very much the same way, to use Mr. Carapanos's simile, as we now consult a lawyer or a physician. The sur- viving inscriptions doubtless form but a very small portion of the mass which must have accumulated at Dodona during a series of centuries; but those which we possess range over a long period of time, from perhaps the fifth century B.c. to the date of the final destruction of the temple in 88 B.c., or even later. Aptly confirming the testimony of authors to the wide-

spread fame of the oracle, these haphazard survivals depict for us as seeking the help of the god, not only the simple peasant of the

neighbourhood, but members of other Greek communities far and near-Ambraciots, Corcyreans, Tarentines, Ionians, Athenians. We may perhaps confess to a little disappointment that on the whole neither in the oracle inscriptions nor in the others of the Dodonean collection is the gain to Greek dialectology as great as might have been expected. Nor is our previous knowledge of the history of the Greek Alphabet greatly increased or confirmed. None the less welcome, however, is the munificent contribution of Mr. Carapanos to the history of the social and political life of Ancient Greece.

E. S. ROBERTS.

11. S.-YOL, I,



(pp. 228-241) e.s. roberts - the oracle inscriptions discovered at dodona

Documents