An Unpublished Fragment of Paisachi

8/13/2019 An Unpublished Fragment of Paisachi

1/13

An Unpublished Fragment of Paic

Author(s): Alfred MasterSource: Bulletin of the School of Oriental and African Studies, University of London, Vol. 12,No. 3/4, Oriental and African Studies Presented to Lionel David Barnett by His Colleagues,Past and Present (1948), pp. 659-667Published by: Cambridge University Presson behalf of School of Oriental and African StudiesStable URL: http://www.jstor.org/stable/608723.

Accessed: 13/08/2011 08:26

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 forms

of scholarship. For more information about JSTOR, please contact [email protected].

Cambridge University Pressand School of Oriental and African Studiesare collaborating with JSTOR to

digitize, preserve and extend access toBulletin of the School of Oriental and African Studies, University ofLondon.

http://www.jstor.org
http://www.jstor.org/action/showPublisher?publisherCode=cuphttp://www.jstor.org/action/showPublisher?publisherCode=soashttp://www.jstor.org/stable/608723?origin=JSTOR-pdfhttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/stable/608723?origin=JSTOR-pdfhttp://www.jstor.org/action/showPublisher?publisherCode=soashttp://www.jstor.org/action/showPublisher?publisherCode=cup


2/13

An Unpublished Fragment of PaisaciBy ALFRED MASTER

(PLATE 3)M OSTof the available specimens of Paisaci are recorded in Lacote's Essaisur Gundcdhyat la Brhatkathd,pp. 201 f., and in my article "TheMysterious Paisaci ", JRAS., 1943, 217 f. They comprise about a hundredwords. I am now able, by the courtesy of MuniPunyavijayaji of the SvetambaraJaina Dharmasala, Baroda, who has sent me photostats of an early MS., topublish the longest extract extant in Paisaci from an unpublished portion ofthe Kuvalayamala.Parts of this work were published by L. B. Gandhi in " Apabhramsakavya-trayi ", GOS., xxxvii, in 1927. It is described in the text as a samk7rna-kathdor composite narrative, that is, one not purely religious, and was written byUddyotana, son of Vatesvara and pupil of ilafika,1and completed, presumablyat Bhillamala (Bhinmal or grimal), on the last day of gaka 699 (A.D. 778).It is written mainly in Jain Maharastri. The script of the originalMS.is namedMaratthaya desi vannaya, in all probability that to which the name Ndgariwas afterwards given. The oldest MS. was written on palm-leaves in St. 1139(A.D. 1083) six years before the birth of Hemacandra, and is the property ofthe Jesalmer Jain Bhandara (J.). Another MS. written about the fifteenthcentury is in the Bhandarkar Institute, Poona (P 1), a copy of which, datedSt. 1968 (A.D. 1912), has been lent me also by Muni Punyavijayaji (P 2).J. contains fewer mislections and omissionsthan P 2, but can often be correctedfrom it. The palaeographyof the former MS. is of interest, but considerationsof space forbid more than a limited reference to it. The aksaras of J. are usuallydistinct, but there are a number of them which are differentiated only by asmall stroke or by the position of the stroke. Thus t and n, d and v, bh and h,p and y, m and s, tth and cch,ddhand tth,th and vv,pp, yy, and gh are all liableto confusion, which has resulted in many inaccurate transcriptions in P 2. Itis, therefore, often difficult to determine the intention of the author as regardsorthography, a matter of such vital importance for the exposition of a dialect.The scribe's misreadings of the Maharastriwords are generally easy to emend,as there is a standard for comparison,and in the mixed Skt.-Pkt. dialects of J.,fol. 130-1, the readings differ very little from those of P 2. But in the PaiSacipassage the readings differ more widely. Even so, some firm conclusions canbe drawn. P 2 has viyyathara, viyyddharo,and vijjddhara,J. viyydthara, -oconsistently, establishing the latter form for Skt. vidyddhara-, and so on.Namisadhu's vidyddhara s clearly incorrect.

1 The guru's name is usually given as Haribhadra on the authority of Jacobi and MuniJinavijayaji, but the text does not bear them out.


3/13

ALFRED MASTER-The question of the source of Uddyotana's Paisaci at once arises. Fromthe manner of its introduction the passage seems not to be an extract from

the Brhatkatha, although this work is mentioned, J., fol. 3.Recto 2 b. sayala-kaldgama-nilaydsikkhdviya-kaiyanassauhayamddkamal'dsanoGu.naddho arassa7 assa Vaddakahd.It is not even possible to say whether Uddyotana used the BK as a model.In any event the passage is literary in character, as is shown by the stiltedconversation, the long compounds, and conventional images, very differentfrom the lively exchanges of the monks and novices in the monastery sceneof fol. 130-1. And in the introduction to the Paisaci passage (see below) theauthor distinguishes Paisaci from the other three literary languages by theprocess of exclusion chiefly on literary grounds, with the observation that itmost resembles Sanskrit. The passage, therefore, seems to have been speciallycomposed to serve as an example of the author's conception of Paisaci. It isimpossible to say, however, to what extent the original text has been alteredby the scribe, either through misreading the characters or by substitution ofa more familiar form. The meaning of a sentence often remains obscure andit serves no good purpose to attempt emendations, unless they help to renderthe remainder of the sentence so intelligible as to make the readings certain.They cannot, in any case, be used as evidence.The language resembles Hemacandra's Paisaci more closely than that ofany other grammarian or writer on poetics, but out of twenty-two ofHemacandra'ssutras twelve cannot be applied for want of opportunity. If oftwo variant spellings we choose that which is nearer to Pali, the followingsimilarities and differences are the most striking. Hem., Su. 306, n > n:J. keeps Skt. n when not due to remote cerebralization according to the Palirule, so cunna, ramaniyya. The Maharastri narrative portions are writtencarefully with cerebral n, e.g. annena bhaniam. Su. 307, t, d > t: four ofHem.'s examples are found in J., bhagavati,pavvata (Hem. -1), satam, pateso.It is possible that Hemacandratook these examples from this very MS. Su. 310,hrdaya-> hitapakam: this is found both in Namisadhu and J. It is a mis-reading for hitayakam, cf. Prakrta-prakasa, x, 14, hitaakam. Lapitam of J.is also found in Hem., Su. 304, and the passive lappiyyate correspondswithramiyyateof Su. 315. Pali has two spellings puthavzand puthuvi. The formerappears in J. and the latter in Namisadhu's Paisaci (JRAS., 1943, p. 224).Ramaniyyais spelt in J. as in Pali both with the dental and with the cerebraln.

PhonologyJ. occasionally unvoices consonants other than d, thus approaching theCulikaPaisaci of Hemacandra and the Paisaci of the pseudo-Vararuci. It alsosometimes geminates intervocalic consonants following a short vowel, cf.Pischel, GPS., ? 194, Pkt. pdikka (pdddtika-), JM. jitta (jita-), Mg. hagge

(ahakah),and often in Apabhramsa. For the most part the phonologyresemblesthat of Pali.

660


4/13

AN UNPUBLISHED FRAGMENT OF PAIAACI

Vowels.-- > i Suranati, ramaniyya, nigiyyamdna, bhagavati,Bhdgirathi;but remains in nat;, puthav;, vdhinrn; r > a in katanimy,> i in hitayaka,pithula, mika, and > u in puthav ; e > i (e) lapiyyam; at > e mettakenam.Consonants.-ks > kkh nirikkhitto, > cch dupiccha, paticchita; g > knakara, mika, lavamka; g remains vigalamana, dbhogo,yugalako, tibhdga,bhagavati, Bhdgirathiand initially; j remainsjdnasi, rajo, Pavvatarajo,pajja-ramta; n > n sunesu; t > tt nirikkhitto, ramaniyyattaro(5c), satta; d > tpateso, hitayaka, -6tita (udita-), but remains initially daita, dolaka; dh > thmathukara,mathura,vatha,viyyatharo,vivitha,but remainsin a groupanuddhilta,nibaddha; ty > cc sacce; tr > t tinayana, tibhaga; dy > yy uyydna, viyyd-(but Pali uyydna, vijja); dhy > jjh bujjhamta; ny > nn ?kann ; p > pplappiyyate; b(v) remains bujjhamta,bahund, tubbhe; bh remains bho, dbhogo,bharo; y remains yati, yatha, yugalako, ramaniyd, yeva, yyeva; (ey > (i)yylapiyyarm, amaniyya, nigiyyamdna; y > iyy lappiyyate, kiriyyamdna (withshortening of z of kgrya-,cf. Pali vikiriyyati, vikiriyati).Loss of intervocalic consonants with or without ya-sruti is rare and dueto scribal lapses: eyarn,jai, pavvaya (also occurring with the occlusive);so, too, ratuppalaforrattuppala (P 2). Pdyal(a), seya (4b) seem to be corrupt.The graphy tt for t may have been induced by such near-doublets as Pa.devati divinity (pers.),devattadivinity (abs.), kakkhalatd,kakkhalatta oughness,the former being feminine and the latter neuter. The use of pp for p shows,however, that this cannot have been the only cause.The divergent spellings need not be ascribed solely to the copyists. Theymay have existed in the original MS. We do not know whether any particularauthor dictated his composition or wrote it. Some undoubted holographs,which have reached this era, contain many inconsistencies. Thus the dramaticPrakrit portions of the Turfan MSS., which may be contemporarydocuments,are not phonologically uniform. In the " Alt-Sauraseni" portions dy is foundas jj and y, rv as rvv and vv, ry as yy and riy, and intervocalic t is retained,except in one instance surada. The chief difference between this dialect andthe Paisaci (or Pesaya) of J. is the partial unvoicing of the voiced consonants,but in some respects the Alt-Sauraseni is closer to Pali than Pesaya, as in theuse of in, -ma (suffix of 1 pl. pres. ind.), I and occasionally 1for r. It is clearthat Pali Alt-Sauraseni and Pesaya were all based upon what Liiders callsAlt-Prakrit or the common elements of the Early Middle Indian stage. Pesayahas, however, been perceptibly influenced by Maharastri. There are no con-spicuous morphological, semasiological, syntactical, or lexical differences.

MorphologyThere are some very long compounds, but it is difficult to distinguishflexional endings in -a from the -a- of stems. The alternation a/a is commonin the Maharastriportions of the work and may well have found its way inhere. The scribe of J. has even made corrections in ramaniyyotaro,-otamo,evidently regardingthe suffixes of comparisonas separatewords. And although

661


5/13

ALFRED MASTER-Pischel (GPS., ? 414) states, " Das Pkt. gebraucht die Suffixe des Comparativsund Superlativs -tara, -tama ... ganz wie das Skt.", it is probable that suchforms were archaisms or had become stereotyped like mahattaraor *piyayara 1,which passed into Marathi as mahzdtr(d),piydr(d), now obsolete spellings ofmhdtdrd, ydrd. Thephraseused in this text as an alternative for the superlative,ramaniyyamramaniyyanam, s closer to the Neo-Indian use of the positive withthe locative plural. There are no conspicuousformative suffixesexcept the -kaenlargement, which appears in mettakenam,makehim,bhamamdnakehim,ndramanmyakehimith yugalako,found also in Pali.Flexions.-These tend towards the Pali types. The relative pronoun jammi(Pa. yamhi) is a middle Prakritism and, in view of the confusion elsewherebetween -a and -o, vaccdmomay stand for vaccdma. The inflexions -esu of the2nd sg. impv. sunesu, -iyya- of the passive and -ehim of the instr. pl. masc.neut. occur in Pali as late variants and are found in one or other of the Prakrits.Ujjhiiuna s a Prakritism for the ujjhituna of Pali and Hemacandra's Paisaci.It is followedby anotherPrakritism,jammi, both due, no doubt, to amomentarydistraction of the copyist. Lapiyyarn is a Pali, not a Prakrit, form (AMg.-ija, -ejja, -ejjami).Morphemesof the Noun-adjectiveCategory.

Stems.-Masc. -a, -u; fem. -a, -a, -i, -i; neut. -a, -em. Nom. sg. masc. -o;acc. sg. -am; pl. nil; dir. neut. sg. -amn;pl. -dni(rm);nom. fem. sg. pl. -a, -I;acc. sg. -am, -.m. Instr. masc. neut. sg. -enam; pl. -ehim. Dat. abl. nil. Gen.pl. masc. -anamr; fem. -dnam,-ina. Loc. nil.

Examples:Masc. neut. stems. pavvaya-, tira-, ramma-, taru- (rajo in rajo-cunna-is a Sanskritism).Nom. sg. masc. pateso, 'abhogo, -rajo, yugalako, ramaniyyo; acc. Pamca-nanam; pl. nil. Dir. neut. sg. samkulam, mandalam; pl. sakkarani, katanim,patakani.Instr. sg. mettakenam; pl. bhamamanakehim, ramanlyakehim.Dat. abl. nil. Gen. sg. nil; pl. ramaniyyanam.Fem. stems. ela-, natI-, puthavl-.Nom. sg. ramaniya (58 V. 3a), samtati, Bhagirathi; pl. kamini, manini,pasutta; acc. sg. vahinim; pl. nil.Instr. dat. abl., nil. Gen. sg., nil; pl. dolamanam,taralina. Loc. nil.Personalpronouns. 2nd pers.nom. pl. tubbhe. Instr. sg. tae; pl. tubbhehim.Demonstratives. Dir. sg. neut. tarm,etam. Instr pl. masc. imakehim.

Loc.pl. fem. ?tas (conjectural). Relatives. Dir. sg. neut. yam (jammi out ofcharacter). Interrogatives. Dir. sg. neut. kim. Nom. sg. masc. kataro.Verbs.-The verbal morphemes(both infixed and suffixed)are not sufficientto justify a separate list and can be easily gathered from the examples.

1 ND. suggests *priyakdra-.

662


6/13

AN UNPUBLISHED FRAGMENT OF PAIAACiAct. pres. ind. sg. 2 janasi, pl. 1 vaccamo, 2 ullapatha, 3 patassati.Impv. 2 sg. sunesu, pl. payatthatha (cf. Pali). Opt. 1 sg. lapiyyam.Pass. pres. ind. 3 sg. lappiyyate.Pres. part. act. ullasamta, bujjhamta, ramgamta; mid. vicaramana,dolamana; pass. vikiriyyamana.Past part. pass. katanirm,lapitam, vittasita, cunna, bhinna, samaruidha.Absol. (ujjhiu.naout of character).Other words of interest are the adverbs, etc.: upanta (with the varga n),visesam, mamdamamdam, evam, yati, yatha, yyeva, cceya, ceya, bahuna.Sandhi is mostly of the Middle Indian type. Skt. type: saggavatara,viyyatharopavan', -talotita.The words ettham and kattham are not used in the Pkt. senses "here"and " where ", but "thus " and " how ", Skt. ittham, katham. Pali has ettha" now ", as well as "here ". It is possible that there was a Skt. form *katthamand the Pkt. words tena so, tahim there, etc. (Skt. tatas, etc.), show how easilyinflected demonstratives acquire adverbial force (time, place, and manner)and are liable to confusion, when there is any similarity of morphemes.The passage is introduced in JM. as follows (J., fol. 57 Verso 5a). A prince,who is on his wanderings, lies down at the base of a banian tree " and aftera moment's interval was roused by the faint sound of both harsh and soft

words 1 spoken with a trace of dialect. And he paid particular attention tothem. After listening he thought, 'In whatever kind of language is this beingspoken ? I wonder. Hm, it cannot be Sanskrit, because that is a combinationof compounds of many words, particles, prefixes, cases, genders, ingenuities,and ambiguities, difficult to understand and as hard as the heart of a wickedman. Yet it is similar. Then can it be Prakrit ? Hm, it is not, because thatis a garland of all the arts, a medley of billowy water, an ocean of folklore,a cream-panof liquid dropsof nectar proceedingfromthe mouths of the mighty,full of pleasantness like the speech of persons of position, which, when theymeet, contains words of formal compliments in compositions of various kinds.That, then, is not right. Then can it be Apabhramsa? Hm, it is not, becausethat sportsand leaps withwavesof wordsbothcorrectandcorruptof Sanskrit andPrakrit, like mountain-rivers flooded by new rain, cloud-bursts, and torrents,and charming as the bickering of lover and beloved in a lovers' quarrel. This,too, is not right. So what can it be ? ' Thinking it over again he came to theconclusion,' Why, there is a fourth language, Paisaca. Then this may be it ' ".The following text begins with the words of which the last twelve wordsabove are a translation. The recto (R) side of the folio is that which isunnumbered in the plate, the verso (V) being numbered. 3a refers to thefirst section of the third line of the first unnumberedfolio of the plate.There is no separate symbol for b in J., and v is therefore rendered as bwhere b is normally found in Prakrit. Similarly, as the aksaras for u and o1 khara-mahura. Perhaps khara-muhara,an unpolished speaker.

663


7/13

664~~~ALFREDASTER-are the same, the final aksara u is rendered o to correspond with the finalconsonant-vowel aksaras -to, etc., which are distinguishable from -tu, etc.

"Are, atthi cautthai bhdsd Pesiiyd. Td sii imii hohi tti. Ettha vada-piyavavare 1 Pisdydna ullivo akkhai"1 tti. Tii pun~aPisei~iean~iyaya-bhSj3K7a>e "Bho, eyaM~ tae lappiyyate yathd tubbhehimipavvaya-niati-tira-ramma-vanakdnan' uyyana-pura-nakara-pattana-satta-samnkulamn uthavi-mandalamn bhamamiinakehimnkataro patesoK7c>ramaniyyo nirikkhitto tti. Etthamnkimgapiyyam~? Tam, abhinav'ubbhinna~c-i~ta-mamitari 3b-cont.>kusumo,aralinia mam~damamdam~oIamdndmup,inta' piitapamn, arala-sdkhii-samnghattavittiisit'acch'acchara0 tnara-nara ndyamdna tanutara pakkhat samptati,vighatft.anuddhfita vicaramiinaraocnabin-iaaalvgl td-nibaddha-dolaka-samar-udha sura-siddha-viyy,ithara-kamnta-kiiminIanamp-dolam-ana-gita-ravd-kanni8 sukkha-nyibbharapasutta; K5c> anaka-mika-yugalako Namdana-van'dbhogorama-niyyattaro "19 tti.Avaren~abhaniiamn"Yati na jiinasi ramaniyyo ramaniyyd K6a>nam~~,visesam~d sunesu.20 Uddiima-samcaranta-tinayana-vasabhaK6b>hem~kkern-taravuppittha2-bujjhampta-Gori-Pampcdnanamj,dsava-savitinuia-vikkama-ni-pdta-pdtita tumga-tuhina-sita-sisira-sikharo llemaipto22 ramanlxya-tamo " tti.

Annueniahan~iamn Na hi, na hi, velhi-taram~ga-ramngamnta-salila-tlevutth-iUiat-sis1ra-mdru3 vikiriyyamdn'elhi-lavaipka-kakkolaka-kusuma-bahala makaramndamutita4-mathukara-kala-kalii-Iliva25-rdv'uggiyyamiin'ekekka6-pita-kusuma-bharo,mo yyeva Veldvan'-Tbhogorama niyyatamo "127 tti.

1 J. p& aoae P2pta 2 J. ?kkhai; P 2 akckhai samufllvo are wanting.3This portion at first omitted in J., is in place in P 2. 4P 2 etamn.5Up to this point Maharastri cerebral for dental n has been used.

6 P 2 lappijjate. 7 J. ubbhinava. 81P 2 dolamulnannava.9 J. upiinta pcltarusasMgqhafa; 2 peltapafl tarala-s&khed-sarnthavva.10 P 2 citt&tsittabaccara;. a second nara is inserted by another hand, agreeing with P 2.11 J. hitapaka; P 2 hatapamklc. 12 J. viyyTitharo;P 2 viyyathara.13 P 2 niccam&nauditagotta-gotta-keittan'. 14 P 2 -c& eya.15 P 2 pat&lanii. 16 J. ratu-. 17 J. -yyotaro; P 2 -jjataro.18 J. kann4n& P 2 kannana. 19 J. -iyyotaro; P 2 -ijjataro.20 P 2 tClautesuvi8esaiM. 21 vuppittha P 2 om. 22 P 2 Himavamnto.23 P 2 melruta. 24 P 2 -ramndamatita. 251 J. talakalelveva P 2 kalel2.261 J. ekekke; P 2 ekkekkcama. 27 J. -niyyotamo P 2 -nijjatamo.

664


8/13

(a)

57 V

BSOAS. XII]


9/13

KUVALAYAMAL,A MS. J,(A.D. 1083)

(b)


10/13

PLATE 3.

(c)

[to acep. 664


11/13

AN UNPUBLISHED FRAGMENT OF PAIAACIAvarena bhaniam, "Are, kim imakehim yyeva ramaniyakehim; yam

paramaramaniyakam, tam na ullapatha tubbhe. Saggavatara-samanamtara paticchita naval-tibhaga nayana bhatta-kataghata2-nivasasasi-kala niddhutamata3-nivaha-mathura-dhavala-taramga-ramgavali-vahinlm pi bhagavati Bhagirathi 4ujjhiuna, jammi papa ka-sata dutthap-pamo pi.5 Kim bahuna mitta-vatha katanim pi patakani sijjana 6 mettakenamyeva sata sakkaranipatassamti.7 Tasu (3a) cceya 8ramaniya Suranati " tti.9Tao savvehi vi bhaniyam, "Yadi 10evam, ta payattatha.11 Tahim ceyavaccamo " tti, bhanamana uppaiya seya-khagga-nimmalam gayanayalaPisaya tti.It may be considered premature to offer a translation of a text in a raredialect which has not been fully elucidated. But even an imperfect translationgives the reader some idea of the subject-matter and aids the checking of wordsby their contexts. The style is Gaudi and may be compared with the lesselaborate portions of works such as the Vasavadatta of Subandhu. I shouldlike here to express my thanks to my student Mr.Prabodh B. Pandit for somevery useful suggestions.

[Fol. 58 R 3a] " Here at the great banian tree the Pisacas are conversing.aThen again he stood thinking, "What is their conversation ? "Another Pisaca said in his own language, " Sir, you are saying whichregion you regarded as (the most) attractive during your wanderings throughthe world full of hundreds of mountains, rivers, shores, pleasant groves, woods,gardens, cities, towns, and ports. So, what can I say ? This ... (where thereis) the blossom of the clusters of newly-sprouted mango-shoots, a tree nearfair ladies slowly swinging, a succession of ....b Apsaras with eyes troubledby the collision of moving branches,proud beauties disdained sobbing out theirhearts bruised and torn by passion, wandering unharmed by their distresses;it is the extensive garden of the Vidyadharas,c the delight of Vidyadharaschosen by their own choice is the most attractive."Another said, " No, no (It is where are) divine damsels wanderingat theirsweet will, beloved ones singing, those who thrill with joy in praises of theirlineage, white waters . . . the King of Mountainsd with his noble heightschequered with red lotuses and broad slabs of gold is more attractive."[5a] Another said, "How, sir, is this which you have said, well said ?(Where) lovely maidens of the Gods, the Siddhas, and the Vidyadharas aremounted on swings tied by swaying creepersto various Trees of Plenty, virginschanting songs which make men disquieted are lying replete with happiness,the broad Nandana grove e with its pair of golden deer is more attractive."Yet anothersaid, " If you know not what is the most attractive of attractive

1 P 2 adds nabha before nava. 2 P 2 jatdka.tdpotara. 3 J. -tdmana.a p 2 vahinirmpi bhagavatiBhdgirathz. 5 P 2 -sataruddhapasopi.6 The ak*aras in J. for ?si and ?jjd are irregular and uncertain; P 2 sijdna.7P 2 panassamti; J. omits m. 8 J., P 2 tasavveya. 9 P 2 nammayd Suratattzti.10 J. jai. 11 P 2 payaddatha.

665


12/13

ALFRED MASTER-things, then hearken closely. (Where) Gauri and Siva awaken roused by thebellowing of the bull of the Three-eyed One roving f unrestrained, Hemanto gwith its tall, chill, white, icy peak humbled by the descent of Indra's mightvouchsafed is the most attractive."

Another said, " No, no (Where there is) a weight of bloom on every treewhich singswith the hummingsand buzzingsof bees fragrantwith the abundantpollen of cardamom, clove, and kakkola flower scattered by the cool breezesblowing h from the waters which toss with tidal billows, this, the tract ofCoastal Forest is the most delightful."Another said, " Oh, enough of such attractions You speak not of whatis the supremeattraction. After her descent from heaven ... Lady Bhagirathileaving her stream with its stages of white waves sweet with their store ofnectar cast off by the moon-beams, in which we destroy a hundred sins. Whysay more ? Sins of the slaughter of friends have been committed. ... J Amongthem Suranadiis (the most) attractive."Then all said, " If it is thus, then proceed. Let us go there." So saying thePisacas flew up to the skies clear as a white sword.

NOTESa Akkhai (text, footnote 2) is a pure guess, but clearly represents " is " or

" continues ".b In this difficult passage pakkha might be for yakkha,but P 2 also readspakkha.c Six regions are discussed: The Vidyadharas' Paradise, the Parvata-raja,the Nandana-vana, Hemanto, the Coastal Forest, and the River Ganges.The Vidyadharas are described by H. v. Glasenapp (Der Hinduismus, 109)as a kind of spirit, who resemble men in shape and appearance and oftenassociate with them, living in separate cities in the mountains of the north.They possess great magic powers, which enable them to change their formsand to fly through the air at will. He might have added that they did notexercise these powers until they were taught the Vidyas or Sciences, whichthen became personalities like Ariel or the familiars of witches. The name istranslated by E. Washburn Hopkins (Epic Mythology,175) as Wizard, a nearbut not exact equivalent, as Vidyadharas had no power to bewitch others.They could lose their powers in certain circumstances,and were usually warnedbeforehand by their Vidyas. They then became in all respects like humanbeings. Sometimes they merely suffereda minor penalty, as in the Vasudeva-hindi, where a Vidyadhari uses her powers of flying to chase a pea-chick andso is parted from her lover. The belief in Vidyadharas was no doubt derivedfrom dreams, in which flying and change of form play a large part, especiallyas the dream-life was one of the three planes of existence in Hindu ontology.Vidyadharas are mentioned in the Epics, but their powers are not clearlydefined. They are, however, said to be able to diminish their size and to bevihaga or spirits of air.

666


13/13

AN UNPUBLISHED FRAGMENT OF PAIgACiSanskrit literature, as a rule, is content to leave the characteristics of such

superhuman beings confused and in the text itself Pisacas are said to fly likeVidyadharas. It was left to the Paumacariya of Vimalasuri, the Brhatkathaof Gunadhya, the Vasudevahindi of Safghadasa, and the Brhatkatha recen-sionists to develop the idea of the Vidyadhara. Haribhadrain the Samaraicca-kaha uses it sparingly, but gives the Vidyadhara the additional powers ofinvisibility ahd of uprooting trees. Somadeva develops the idea very fully inthe Kathasaritsagara, no doubt because the Vidyadhara was the main themeof the Brhatkatha.

d The King of Mountains is ordinarilythe Himalayas, but in Jain literatureis sometimes used for Meru. Hemanto later seems to refer to the Himalayasas the resort of giva, so Meru is more probable here.e The Nandana is the garden of Indra and the Kalpadrumawas one of thetrees in his paradise. The kanaka-mrgadecoyed Rama away from Sita in theRamayana, but the pair of golden deer is unknown.f " Roving": J. samcaramttaor samcaramnta,P 2 samcararmta. J. usesboth mt and nt, so the second m in the second word has been taken to be anerror. Dhemkkemtdrav': f. Mar. dhemkbellowing of bull, Pkt. dhemkicrane,Guj. Hin. dhemk-crane, water-machine (creaking noise).g Hemanto might be a misreading for Hemafigo. Apte gives Hemafiga,Hemadri, Hemagiri as synonyms for Sumeru. But Hemanto seems better(n. 4).h " Blowing": some such word as ubbhiutaor utthita is involved. P 2has viruta, which apart from suggesting virudhagives no help.i For this theme, cf. fol. 51 (L. B. Gandhi,op. cit., p. 109), " Gleaming whiteis that holy Ganges of ascetics. Do thou enter it. Then thy name is clearedfrom the charge of betrayal of friends ", and J. 57 V 4b, piya-mitta-vaha-kalusiyam attdnayam vdavemi," I will destroy myself defiled by the murderof my dear friend ".j The consonantal portions of the first two aksaras are illegible in sijjdna,which appears to be a gen. pl. Patassanti, Pkt. padassanti, is to be preferredas the difficilior lectio to panassamti. Yadi should probably be yati, but thetext is never consistent.

667

An Unpublished Fragment of Paisachi

Documents