May, 1873.] HINDU WORDS IN GREEK. 119 OPAA0PO—aria or ardha (P) + athra, OPAArNO— ardha + agni or vereihraghna * * § p see Benfey in Z. d. D. M. G. VIII., 450 seqq., 460 seqq. Hence the following results are to be drawn. Firstly, as to the vowels:— a stands so regularly for a, d, that there is no need of addacing examples;—foro in Uarrdka—potahi, —for ava in Kov&oxarqs, Pandae,—for ago, in Bu^ai'- TIOV. [Indian a is further represented + by o, as in ’Avov- poypappov,' ApiTpoxa-rqs, 'Epawofioas, ’Orropo-, K ip- iadoot, Kavoyi(t), Kokko-, Kovboxarqs, Kopptvaargt, Mo&ovpn, M.o(f>is, Sav&poKvirros, Jomanes,—by t in Epewcats, Acp3.it, Kopptvaagr, KrjpcfioOprjs, MfOopa, Scpvot;—by t in Ctyyt^cpt, Aom/3ape, 2i<tikv7Ttos ;—by v in Powrvpov X (bhutdri), Avpfatot, (Kapvo)(f>vovX, NcXicvv3.i (nilakantha P) ; by ov in o-tpowdov, Av8ov- (i.ipiosy sulphur (Suludri);—by a> in M<o(f>is, Miu- <ra>Xo?. The representation by a is however by far the more prevalent]. c stands for a (as above), for ay a in Kt;«ccoi, for ava in Ov£fvror, for i in Epavvofioa? NcXkuv3o, for u in IloXcptor (Pulumdi), for e in apycWtov (?), (tyytpcpi. i stands for i, i, of which examples are not re¬ quired ; also for a (as above), for u in KavoyiCg. [Indian i is moreover represented by c (as above); by ct in ASriaaOpos, Elptvov, Kianetpatot, ToxrxXft: by at in EUXai-* ; by v in ’Yapcortr, 'Ybatr- nrs, *Y</>acm, Bapuya£a (P), Bvfamov, ^vXXtrat; by o in 0£upanr.] o stands for a, i (as above), for u in Odombo* rae, ’OCgvrj, ’OrropoKoppat, on*<iXXtor, B i(obrfo, -fioQpa, -fiodpgs, -(3opa, K^pffiopn, KoX^ot, Kopapin, kootos, M(Oop-i, poaxos> IIoxXafr, Sodpaxat, (with Ev-), sa- con (?), Stvdos, 2oavor, Soaoroff, Sobpai ;—for o in Kokko-, FovbaXot; for au § in Odomboerae, Bolingae, Colubao, IIopos;—for ava in Ilupof, M.iXXot. v stands for a, i (as above), for u in firjXvppos, Ktipvo-, Karra-, Katnrairvpor, Kvtv8ptvrj, ’Ogvbpaicat, 'O^vparis, lavbpoKvirros, iiaiKirnrosy 2v8pot, Supatr- rprjvrj. [Indian u is represented also by t, o (as above), by ov (Latin a) in Ovapaa, Avovpo-, BovSvas, Bov88j, (3oxrrvpov, -fiovpa, (Karrv)^ovptvrf, Ataopova, KanovOts, Nayayovva, Mobovpa, Movcrucavos, -trovpa, Sovpaarjvot,
- Comp. /Sao-tXeuy, fdaatkuov ucyap Opdayvqs on a
coin of Quduphara- in Thomas’ Catalogue of Hart nun Coins, No. 28 (Lon<l. 1857). Tlie other words of this kind, such a3 A0PO, MriPO, MAO, MAN AO BAIX) are by Lassen also referred to the Zend, more particularly to Persia; whence it would indeed be difficult to separate the above words from it! With OPAH0POY moreover, after the prefix OPA = ardha an entirely hybrid formation would have to be assumed. But of course the explanation of OPA through arta is also hazardous, because (1) art a sounds in Zand asha (the form arta seems to belong only to the B estern Persians), and (2) because the same falls entirely away for OPAArNO Opdayvrjs : if it bo conceived as verethraghna; but to seek in it perhaps arta + agni (!) is precluded by sulphur, Srovpa;—by fu in IlfvKfXa ; by « in Ku><f>gu, Sucpayao-Tjvos] t), and Latin e, stands for e in Bafo8Fjo, 0ijXvppos, Karevs, Ktjkcoi, KrfpcpoOpqs, Mqpos, ’Ofavrj, layrjba, sapenas, lovpatrqvot, lyaaqvos, Xafigpis’,—for af|| in Hpcoftor;—for i in ’A/fypta (with ’AjStpta), Bgaaa- pta>v (?);—for u in sagenon (?) [Indian e is also further represented by *, as above.] « stands for o in * Av8a>pans, Kaxrapfda, ActvtfUapt, 2covos, Scopa, TuxraXrt;—for au in Ocopov, Mcopuvs ; —for ava in *Hp<u8or;—for ra in 2a>irft0r)s. at, (i stand for * (ut sup.);—at, ao> for dva, in IlfvXfXatrtf, 'Ybpaurqs ;~ao for ava in *1 paos;—co for eva in AcottoXXi ;—au for au in TXauxaviKai, Glausae, Kavpapa, MautrcoXor;—ov fora, u, (ut BUp.), for n in Qv&vtos ;—tvo for eva in 8evos. [Indian o is moreover represented by a, o, &>,— au also by o, &> (as above); in Tapotar for Gaurf, there is probably a transposition from Taopiar ?] Next with reference to consonauts, the fre¬ quent use of /3 is to be noted. This letter occurs for p in Bt(3a<ris, -fiopa, Aaxiva^a8gs, Tv8a/3apa, 8t(3a, Kafiovpa (?) Kap&iadooi, Krjpe(3o0pas, paXa&aQpov, naXi/3o#pa, Ta^ao-o, Ta/Sao-troi, Tanpo^avrj, (but a great proportion of these words is probably based on Prakrit forms which already had b)for b in Bou&var, Bou88a, Kuxrapfia, 2a/3apa, Sapftot, 2i/3ot;— for bh in ftovrvpov, Bolingae, Afdgpta, A^ta-aprjs, (Brja- a-apetov ? ), ’Avbovfiaptos, Colubae, KarTvftovptvr) (?); —for v and especially as an initial, as in—Ba(o8rjo, BaXeoKovpos (P), Bapvy.i£d (?), Baacipa>va£ (?), Bqpv- Xoc, Br)<r(i8at, Btfino-ir, Bibua^rgs, Bv£avriov, as also in medials, thus—cinnabari, Avpfiaioi, Epavvo&oas, £tyyt[3epi, ’hfi.i8iov, Acovtftipe, Sai/8pa/3uTts, Tvpawo- oas, Xaftgpis;—for do' in Bapaiciy,—for y (perhaps by exchange with v) in 2apufios. [The Indian v is further represented by ph in sul¬ phur ;—by the rough aspirate in CYdaairris, 'Ytpaats by the smooth aspirate in *Epcwco-ts, ’O<ro-a8iov, Soavos, 2oaaros;—by ov in Ovtvdtov, Jlopovnpoi. Besides va appears as o>,—ava as a, e, o, ao, co,— dva as at, aco; as ova in ‘Pova&oy (Irdvati),—eva as <o, fuo (as above)]. Further, the use of <r for tho smooth palatal ch is interesting, in Glausae, Ilio-ouXai, npu<rtot, 2av- tho circnmstanco tlvit we would then have to assume a hybrid formation, and would consequently fall into the same difficulty which precludes the explanation of OP- AH0POY from ardha + athra. t The constant representation of a by a or o, t, &e., or more particularly of u by t or o, v, ov in different texts is a testimony for the homogeneity of the respective pass¬ ages ; more particularly with reference to their derivation from a common source. J With an echo purposely sought of Greek words. § If not already in the Indian word, o is to be prefixed according to Pnlkrit rules. it Rather perhaps already in tho Indian word itself: e. ■f Do. do. do. do. : o.
