Constitution of the Vamana-Purana Text

Constitution of the Vamana-Purana Text [vamanapuranapathasya nirdharanam] / By Sri Anand Swarup Gupta; Purana-Department, All India Kashiraj Trust, Fort Ramnagar / 141-194

[ sratra kasirajapuragavibhage vamanapuranasya pathasamiksatmakam samsodhita samskaranam nirmayate | asya samskaragasya nirmanakale vamana- puragasya granthaparimanavisayakah pathavisayakasca kecana prasna upasthitah | te prasna patra lekhe vicaritah | pathanirdharanaya ca katicinniyama niscita vyakhyata udahrtasca | vamanapuranasya kasmoriyakosesu vemkatesvarapresamudritavamanapurana- granthasya 23-31 adhyaya na vidyante, bamgiyakosesu daksinatyakosesu ca 23- 46 adhyaya na vidyante, uttarabharatiyadevanagarikosesu punah vemkatesvarasamskaranasya sarve eva adhyaya varttante | ato'tra prasna upasthitah - vamanapuranagranthasya vastavikam parimanam kimasiditi | naradiyapurane vamanapuranam bhagadvayatmakam dasasahasraslokatmakam ca ucyate | tatra purvabhagasya parimagam satsahasraslokatmakam manyate tacca parimagam vemkatesvaragranthe uttarabharatiyadevanagari kosesu ca vidyate | lekhe'smin imani tathyanyadharikrtya vamanapuranagranthaparimanavisayako vimarsah krtah | vamanapuranakosesu bahuvidhah pathabheda varttante, vemkatesvaramudrita- pathasya kecana sloka kosesu na vidyante, kecicca atiriktah sloka api tatra tatra kosesu vidyante | tesam pathabhedanam madhye ke pathah svikaraniya kaya ca paddhatya ityapi prasno'tropasthiyate | adhunike pathasamiksavijnane pathaniscityai kecana siddhantah svikrtah santi | vamanapuranapathanirdharanayapi tadrsah kecana niyama niscitah | srasmillekhe etadrsa ekavimsatiniyama vyakhyata vamanapuranaslokairudahrtasca | udahrtan slokamcasritya pathanirdharanavidhirapi pradarsitah, nirdharita- pathasca bahudha mahabharatadigranyanam pathah samarthitah | ] The All-India Kashiraj Trust has undertaken the onerous but essential and important task of preparing the critical editions of the Puranas through its Purana-Department. The work of the Vamana-Purana edition has nearly been completed. The text has been constituted on certain principles of textual criticism adopted

142 puranam - PURANA [Vol. IX, No. 1 for the purpose. But in the course of the constitution of the text several textual problems also arose. It is proposed to discuss here these problems and the principles of text-constitution adopted for settling the text of the Vamana-Purana. A. TEXTUAL PROBLEMS OF THE VAMANA PURANA 1. Extent of the Vamana Purana text: The first problem of the textual reconstruction of the Vamana-Purana, which confronts us, is concerning the extent of its text. Its vulgate text, as represented by the Venkatesvara Press edition, consists of 95 Adhyayas and 5,815 slokas. The position of the collated manuscripts in this respect is, however, as follows: (a) All the collated Kashmirian manuscripts (1 Sarada MS and three Kasmiri MSS, which are closely allied with the Sarada MS) omit all the nine Adhyayas (23 to 31) of the Venkt.-edn¹., which contain the first Vamana-carita narrated by Suta Lomaharsana to the Rsis, but included in the extending from Adh. 22 to Adh. 49. This first Vamana-carita is, Vamana-carita is, therefore, the secondary Vamana-carita version as compared with the main Vamana-carita version of the Vam. P., narrated to Narada by Pulastya, the main narrator of the Vam.-P., and given in the last chapters from Adh. 73 to 95, in the form of the and faff. The recently acquired Sarada MS, from the B.H. U. Library, dated Saka 1444 (A.D. 1522), also omits these nine Adhyayas (23-31). (b) All the three collated Bengali MSS, (one from the Asiatic Society of Bengal, and the other two from the Bangiya Sahitya Parisad, Calcutta), and the two South Indian MSS, (one Devanagari MS, from the Adyar Library, Madras, and one Nandi-Nagari MS, from the Srigeri Matha, Mysore) omit the whole intervening interlocution between the Suta and the Rsis, 1. The Vamana-Purana references are given here from its VenkatesvaraPress edn., Bombay, Saka 1851 (A. D. 1929). If the name of a Purana is not mentioned in the reference Vam.-P. should be taken there as referred to. For the editions of the Puranas and other works referred to in this article see Appendix.

Jan., 1967] CONSTITUTION OF THE VAMANA PURANA TEXT 143 beginning from 22.47 and going up to the end of Adh. 49. It is to be noted here that no Grantha or Malayalam MSS of the Vamana-Purana exist in the Sarasvati Mahal Library, Tanjore, Govt. Oriental Manuscripts Library, Madras, and the Keral University Library, Trivandrum. But we have received from the Sarasvati Mahal Library, Tanjore, some details of its four Devanagari MSS, D. 10419, D. 10421, D. 10422 and D. 10423, of which the last Adhyaya corresponds to Adh. 95 of the Venkt. edn., and is respectively numbered as saptasastitamo'dhyayah, pancasastitamo'dhyayah and saptasastitamo'dhyayah, which shows that these four MSS, also omit all the 27 Adhyayas (23-49) containing the interlocution between the Suta and the Rsis. Moreover, a recently acquired Telugu MS from Mysore also agrees with these Tanjore MSS both in the extent and in the name of the last Adh. as 'visnupujyayogyavastukathanam '. This interlocution of the Suta and the Rsis (from 22. 47 to Adh. 49), whole of which is found omitted in both the Bengali and the South Indian collated MSS, comprises the following topics:- (1) Adhs. 23-31. The first or the secondary Vamana-carita, which is mostly similar to the Vamana-carita of the Matsya Purana (244-246), and of the Bhavisya-Purana, Bhavisya-Parvan; and seems to be adapted from these. (2) Adhs. 32-42. The description and Mahatmya of the tirthas of Kuruksetra, which is almost similar to the Mahatmya of Kuruksetra and its tirthas given in the Aranyaka-parvan, Adhs. 81 ff., and the Salya-Parvan, Adhs. 37ff., of the Mahabharata (cr. edn.). As already mentioned above, this Mahatmya is related by Suta to the Rsis in the Vam.-P. but in the Mbh., Aranyaka Parvan it is related by Pulastya to Bhisma, where Pulastya addresses Bhisma as naravyaghra (81. 23a), rajan (81. 21c), dharmajna (81.46a ) etc., which addresses are quite appropriate for Bhisma. But in many and sometimes in the majority of the collated MSS, of the Vam,-P. too we find the same readings (352)

144 puranam - PURANA [Vol. IX., No. 1 (34. 42c), and c (35. 42a) as addresses used for the Rsis by Suta, which shows that these chapters of the Vam.-P., might have been based on the Mbb. The Padma Purana, Adi.-Kh., Adhs. 26 ff., has also this Mahatmya, which is almost similar to the Mahatmya of the Mbh. There, Narada is speaking to Yudhisthira, and hence the above addresses (c etc.) are also appropriate there. Some of the slokas of these chapters of the Vam.-P., are also similar to the slokas of the Kuruksetra Mahatmya of the Naradiya Purana (NP) II. 64, 65 or adapted from them. Hence it seems that in adding these chapters (32-42) in the Vam. P., the compiler might have utilised both the Mahabharata and the Naradiya Purana. 3) Adhs. 43-49. These contain the description and the Mahatmya of the Siva-lingas established around the Sthanu-tirtha, which is a part of the Prthudaka-tirtha; related by Sanatkumara Markandeya. These chapters have not yet been traced in the Mahabharata or in the other Puranas. But in connection with the Mahatmya of Prthudaka-tirtha the Mahabharata (III. 81. 127) says: 'gitam sanatkumarena vyasena ca mahatmanah ' Does it show that these chapters might have been contained in the Sanatkumara Samhita of the Skanda Purana ? All these 27 chapters in no way form an essential part of the main theme, the description and Mahatmya of Kuruksetra and its Prthudaka-tirtha, related by Hari to the gods, where they (the gods) are advised to worship the Pitrs for obtaining their mind-born daughter, Mena, as Himavan's wife. The main theme rather breaks by these intervening chapters (23-49) and the thread is again taken up in chap. 50. But all these chapters are given in all the Devanagari MSS of Northern India, and also in the collated Telugu MS a 1 of the South. The Kashmirian MSS, like the Devanagari MSS of Northern India, introduce after 22.46 the interlocution between the Suta and the Rsis; and in the Kashmirian MSS also one of the questions, which the Rsis put to the Suta, is about the birth of Vamana (22.48c). This query of the Rsis is given in all the Northern Devanagari MSS and in the Telugu MS,

Jan., 1967] CONSTITUTION OF THE VAMANA PURANA TEXT 145 and also in all the Kashmirian MSS as noted above. But while the Devanagari and the Telugu MSS, give the first Vamanacarita also, in reply to this query, the Kashmirian MSS, while containing the query about the birth of Vamana, omit this Vamana-carita. So this omission of the Vamana-carita by the Kashmirian MSS, seems to be a deliberate one, keeping in view of the main Vamanacarita of the Vam.-P., which is given in all the manuscripts of the Vam.-P., without exception. Now the position about the extent of the text of the Vam.-P. comes to this :- Either we have to retain the whole of the inferlocution of the Suta and the Rsis, as has been done in the Northern Devanagari MSS, and the collated Southern Telugu MS, or we have to omit the whole of it from 22.47 upto the end of Adh. 49, as has been done in the collated Bengali and the SouthIndian MSS, other than the Telugu. But unless we collate or consult some Grantha or Malayalam MSS also, which are not yet available as already noted above, we cannot be sure about the real extent of the Southern text of the Vamana-Purana. Moreover, quotations from these chapters are given in some of the old Nibandhas (Dharmasastra-Digests) e. g. Krtya-Kalpataru (Tirtha-Kanda) of Laksmidhara, which is perhaps the oldest available Nibandha, belonging to the 12th century A. D. The Tirtha-Khanda of the Caturvarga-cintamani of Hemadri is not et available. Vaidyanatha Diksita, a South Indian Nibandha kara, also quotes some slokas from Adh. 34 of the Vam.-P. in the Ahnikaprakarana of his Smrti-mukta-phala. So, nothing can yet be said with certainty about the exact nature of the extent of the Vam.-P. text in South-India. Under such circumstances it would not be safe to exclude these 27 Adhyayas from the constituted text of the Vam.-P. simply on the basis of the scanty evidence available from Bengal and the South. And as already discussed above, only the first Vamana-carita cannot be omitted on somewhat misleading evidence of the Kashmirian MSS. So we have to retain the whole portion containing the interlocution of the Suta and the Rsis. Rsis Moreover, interlocution between the Suta and the Rsis in 19

146 puranam - PURANA [Vol. IX., No. 1 the Vamana-Purana seems to be referred to by the NaradiyaPurana also as follows:- ityetad vamanam nama puranam suvicitrakam | pulastyena samakhyatam naradaya mahatmana || tato naradatah praptam vyasena sumahatmana | vyasattu labdhavamcaitat tacchisyo romaharsanah || sa cakhyasyati viprebhyo naimiseyebhya evam paramparapraptam puranam vamanam eva ca | subham || (NP. I.105.17-19). Hence it may be assumed that there also existed some text of the Vamana-Purana which was narrated by Suta Romaharsana to the Rsis of the Naimiseya forest, and the above mentioned 27 Adhs. are remnant of that text of the Vam.-P., which somehow might have crept in the existing text of the Vam -P. which is narrated by Pulastya to Narada, thus filling in the lacuna created by some lost portion of this Pulast ya-Narada Vam.-P. Many of the slokas of this lost portion are found quoted in the various Nibandhas. The Puranas have always served as the real encyclopaedia of the Hindu religion, and as such they have been revised from time to time, adding and incorporating in them whatever new or additional material could be available to the compilers and reaisers of the Puranas in a particular period and place. The compilers or the redactors of the Puranas always retained and interpreted the old, but they were never averse to the new, rather they welcomed the new currents of thoughts and thus kept the Puranas upto-date and really useful for the Hindu society. In this way their extent has increased from two lacs to four lacs of slokas. And n account of these timely additions of the new material the Puranas have been more popular and have been more widely read and recited than even the Epics. Such timely additions made by the compilers or the redactors should, therefore, be considered as the great merit of the Puranas, and should be treated as the valuable literary treasure worthy to be preserved, and should not 2. For the extent of a Purana see Appendix. Also cf. my article 'Puranas and their Referencing I' in Purana, VII. 2.

Jan., 1967] CONSTITUTION OF THE VAMANA PURANA TEXT 147 be discarded, unless there are very strong grounds for their exclusion; or unless they are proved as interpolated by the whims of some scribes or readers and reciters. The real extent of the Vamana-Purana is said to be of ten thousand slokas (NP. I. 105.1). According to the Naradiya Purana it had also the uttarabhaga called the brhadvamanapurana which consisted of the four Samhitas, viz. Mahesvari, Bhagavati, Gauri and Ganesvari, of one thousand slokas each But this Uttarabhaga, known as the Brhad-Vamana-Purana, has not yet come to light, though quotations from the Brhad-Vamana are found in some of the Dharmasastra Nibandhas (e. g. Viramitrodaya-Pujaprakasa, and Acararatna-Devapuja-prakarana) and also in some works on the Krsna-bhakti (such as of Jivagosvamin and Rupagosvamin). So we have to be satisfied with the remaining extant Purva-bhaga only which should have traditionally an exent of about 6,000 slokas. A large number of such Vamana-Purana slokas are found quoted in the Nibandhas, as are not available in the printed texts and the available manuscripts of the VamanaPurana. But unless we get sufficient manuscript-evidence for them, they cannot be included in the constituted text; they can be given in the appendix only. 2. Lacunae in the extant Vamana-Purana Text During the course of the transmission of the text of the Vamana-Purana some portions of it seem to have been irretrievably lost; e. g. - (1) The names of the rivers rising from the two Kulaparvatas, Mahendra and Suktimat, are missing in the printed Venkt. edn., as well as in all the manuscripts of the Vamana Purana. In the Bhuvana-kosa chapters of the Puranas seven Kula-parvatas -T 3. Cf. ityesa purvabhago'sya puranasya tavoditah | srnvato'syottaram bhagam brhaddhamanasamjnakam || mahesvari bhagavati gauri ganesvari tatha | catasrah samhitascatra prthaksahasrasamkhyaya | ( Np. I. 105. 13-14).

148 puranam - PURANA [Vol. IX., No. 1 (mountain ranges) of Bharatavarsa and the rivers rising from each of them are mentioned. In the Vamana Purana, too, seven Kulaparvatas are mentioned as follows:- mahendro malayah sahyah sa ( su ) ktimanrksaparvatah || vindhyasca pariyatrasca saptatra kulaparvatah | ( 13-14-15 ) But the rivers rising only from the five Kula-parvatas- _Parivatra, Rksa, Vindhya, Sahya and Suktimat -are mentioned (13.23-33), od even here also rivers actually rising from the Malaya are wrongly mentioned as rising from the Suktimat : krtamala tamraparna vanjala " cotpalavati || suni caiva sudama ca sa (su ) ktimatprabhavas tvimah | ( 13-32-33 ) (v. 1. sini for suni ; sudhama, sadama for sudama ) Thus actually the text containing the names of the rivers rising from the Suktimat and the Mahendra is missing in the Vamana-Purana. (2) Among the rivers rising from the Himavat ( 13. 20cd- 23 ab) the names of the two important great rivers, Ganga and Sindhu, are found missing, but they are mentioned in other Puranas, eg. Matsya (114.20), Markandeya ( 57.16) and Vayu (45.94). (3) The first two padas of 40.24 of the Vamana - Purana are missing in all the collated manuscripts and also in the Venk. ed., and so here the construction and the sense are loose, as will be clear from the following :- tatah sarasvati sapta visvamitrena dhimata | sravahacchonitonmisram toyam samvatsaram tada || 22 prathasayasca devasca sarasvatim tada drstva tasmistithamvare gandharvapsarasastada | babhuvurbhrsaduhkhitah || 23 ramye sonitam samupavahat || 24 Here the $ 1.24 has its last two padas only. The first favo 4. For the rivers rising from the Malaya Bee Appendix. 5. For the rivers rising from the Suktimat see Appendix. 6. But in the Matsya (114. 29b) and Vayu (45. 103b) the river Vanjula is mentioned as rising from the Sahya.

jan., 1967] CONSTITUTION OF THE VAMATA PURANA TEXT 149 padas have been lost. But the Mahabharata (IX. 42.1 ab) supplies this lost text of the Vamana-Purana tatah sarasvati sapta visvamitrena dhimata | sravahacchonitonmisram toyam samvatsaram tada || ( 41.37 ) srathasayasca devasca gandharvapsarasastatha | sarasvatim tada drstva babhuvubhrsaduhkhitah || ( 41.38 ) evam vasisthapavaho loke khyato janadhipa | agacchacca punarmargam svayameva saritam vara || (Adh. 41 ends here) vaisampayana uvaca sa sapta tena kruddhena visvamitrena dhimata | tasmistithamvare subhre sonitam samupavahat || ( 42.1 ) Here the texts of the Vamana-Purana and the Mahabharata are almost similar, but the first two padas of 42.1 of the Mbh. are not found in the Vam.-P. which are necessary to complete the sense of the last two padas of sl. 24 ( Adh. 40 ) of the Vami. P. In the Mahabharata the name of the az, referred to in the third pada of its 42.1 above and also in the third pada of sl. 24 Abb. 40) of Vam.-P. (as quoted above), rightly and immediately precedes as vasisthapavaha (in 41.39 ab), but it is also missing in the Vam. P. here. (4) In the first Vamana-carita of the Vamana-Purana, 28.17 reads as follows :- daityanamapi sarvesam garbhasthe madhusudane | babhuva tejaso haniyathoktam paramatmana ( v. 1 paramesthina ) || But in this Vamana-carita of the Vam.-P., there is no such previous mention by God Visnu. Matsya - Purana, however, supplies this lacuna as follows :- garbhasthe mayi putranam tava yo'bhibhavisyati | tejasastasya hani ca karisye ma vyatham krthah || ( 244-48 ) And then the similar text as in the Vam.-P., (28.17): | garbhasthe madhusudane | paramesthina || ( 24452 ) babhuva tejaso haniyathoktam daityanamatha sarvesam - 70 t 60 the Matsya-Purana text is complete in this respect. 7. In some places the Vam. P. also supplics a lost text of the Mahabharata, and in some other places it gives a better reading than the Mbh. See Appendix.

150 puranam - PURANA [Vol. IX., No. 1 In the second Vamana-carita of the Vamana-Purana also we have the complete text :- - tavodare hyaham tada nistejaso daityah dakse sambhavisyami vai yada | sambhavisyantyasamsayam || ( 76-48 ) And then :- devamatuh sthite deve udare vamanakrtau | nistejaso'sura jata yathoktam visvayonina | ( 77-1 ) There are other similar lacunas in the Vamana-Purana text. All of them canot be discussed here for want of space. 3. Some unaccountable and inexplicable readings Some readings given in the majority and sometimes in all the manuscripts of the Vamana-Purana are not accountable or explainable. A few such readings are given below :- ( 1 ). tato gacchedanarakam tirtham trailokyavisrutam | yatra purvam ( vem ) sthito brahma daksine ca mahesvarah || 22 rudrapatni pascimatah padmanamabhottare sthitah | madhye hyanarakam tirtham trailokyasyapi durlabham || 23 yasmin snatastu purusah pramucyante ca patakaih | ( 41 22-24 ) Here in 22d the reading daksine is given in one MS ( de 7 ) only, all other MSS read here 'T' which does not fit here. How this reading came here in all the other MSS is not perheps accountable unless we are able to trace these two slokas in other Puranas. In the Mahabharata (Arannyaka Parvan) also we have he description of the anarakatitham, and these same four Deities are also associated with it, but the directions are not mentioned there. 8. Cf. Mbh. III. 81. 146-150 : tato gacchedanarakam tirthasevi naradhipa | tatra snatva naro rajan na durgatimavapnuyat || 146 tatra brahma svayam nityam devaih saha mahipate | anvasyate narasrestha narayanapurogamaih || 147 sannidhyam caiva rajendra rudrapatnyah kurudvaha | adhigamya ca tam devim na durgatimavapnuyat || 148 tatraiva ca maharaja visvesvaramumapatim | abhigamya mahadevam mucyate sarvakilvisaih || 149 narayanam cabhigamya padmanabhamarimdamam | sobhamano maharaja visnulokam prapadyate || 150

Jan., 1967] CONSTITUTION OF THE VAMA A PURANA TEXT 151 ( 2 ) prajapatyam netrayugme rupadhari ( 0rini ) pratisthitam | (80.8 ab) Here in connection with the Naksatras (constellations) naksatra purusavrata are mentioned as constituting the various limbs of Hari who has been conceived here as the Naksatras). The reading c u (Purusa in the from of the is contained in most of the MSS and the Groups. Prajapati is the lord of the Rohini Naksatra, as is also clear from 5.32 of the Vam.-P. given in connection with the description of the aspect of Siva, as follow sragneyamsas trayo brahman prajapatyam kavegrham | saumyarddha vrsanamedam vadanam prakirtitam || ( 532 ) Here is clearly the constellation Rohini as it occurs between (Krttika) and (Mrgasirsa), the three forming the Vrsa Rasi. So in the above text (80.8ab) Rohini (Prajapatya) is mentioned as residing in or constituting the eyes of the NaksataPurusa. But in 80.3ab Rohini is said to be the thighs of Hari or the Nakesatra-Purusa 'janghe dve rohinisthite '. Here rohini is the reading in all the MSS. Rohini (or c), therefore, cannot agai be mentioned as the eyes of Hari. But here, the two Scuth-Indian MSS and the Kashmirian MS from Jammu (T 1) help us. The two Southern MSS read : -- 'mrgasisam nayanayo rupadharini tisthati ' and Ja hmu MS also reads :- 'mrgasiro'ksini '. And this reading ( or mrgasiras ) is also confirmed by 80.24c. - mrgottamange nayane [ pujye ]; the Matsya-Purana also corroborates it as follows:- 'mrgottamange nayane'bhipujye namo'stu te rama vighurnitaksa ' (Mt. P. 54.) the The Varahi-Samhita (a work on Astronomy) in connection with the description of the Naksatra-Purusa, also confirms this reading as-'nasika magha mrgasiro netre '. Thus, in 80.8ab (quoted above) c is clearly the wrong and unsupported reading How the majority of the MSS came have this reading here is really a mystery, which cannot perhaps be solved, unless we find in some old work of Astronomy or Vedic literature mention of Prajapati as the lord of Mrgasiras constellation !

152 pu puranam - PURANA ( 3 ) dipapradhanan narakanandhatamisrasamjnakan | [Vol. IX., No. 1 tirtva sa bhayamya raja ( jyamaghah ) visnulokamagat tatah || tameva cadyapi bale margam jyamaghakaritam | vrajanti narasardula visnulokam jigisavah || ( 95, 62-63) Here in 62c all the MSS read (as vocative) in place of the reading raja . But here Prahlada is speaking to Bali, as for Bali is out of place, is clear from 63a, so this address Bali has never addressed as . How all the MSS give this reading is not clear, unless these slokas are traced elsewhere in the form of the interlocution between some different interlocutors. There are some other problems also concerning the textreconstruction of the Vamana-Purana; e.g. some readings have become so corrupt in the MSS, that their sense has become quite vague or unintelligible. Similarly some descriptions and narrations in the Vamana-Purana need to be traced in other Puranas also so that their correct text may be constituted. There are also some mutually contradictory readings of a text in the MSS; e.g. vedanabuddhimuktah, vedanabuddhiyuktah (in 5.43), vandanam nitisastranam nindanam nitisastranam (in 11.16), etc. But in this limited space it is not possible to deal with all such problems. B. PRINCIPLES ADOPTED FOR THE CONSTITUTION OF THE TEXT € For the purpose of the constitution of the text the following manuscripts of the Vamana-Purana have been collated:Kashmirian MSS TT 1 = Sarada MS, Q. 21. 224, No. 327709, from the Banaras Hindu University Library, Varanasi. First and the last folio missing. =MS No. 4012 (Steine Catalogue), from Sri Raghunatha Sanskrit Library, Jammu (Kashmir). Script Kasmiri (or Kashmirian Devanagari). With Sanskrit commentary by Pt. Ramacandra Bhatta, written in the time of Mahara Ranvir Singh of Kashmir (some 100 years ago). 9. How some readings become corrupt in course of transmission is illustrated in the Appendix.

Jan., 1967] CONSTITUTION OF THE VAMANA PURANA TEXT 153 1=MS, No. 3571 (Steine), from Sri aghunatha Sanskrit Library, Jammu. Script Kasmiri. T 2 = MS, No. 227/425 (Mirikar Collections), from the B.I.S.M., Poona. Script Kasmiri (or Devanagari ?). Dated Samvat Sri Vikramaditya Sakah 1785 (A.D. 1728). (A recently acquired Sarada MS from the B.H.U. Library, Varanasi, No. 330188, dated Saka 1444 (A.D. 1522), has not yet been collated. It is being studied and compared with e alteauye collated sa 1. It generally agrees with TT 1). Bengali MSS a 1 = MS No. 3533-17-B-3, from the Asiatic Society, Calcutta, (H.P. Sastri's Catalogue Vol. V, No. 3989). Saka 1729 (A.D. 1807). 2=MS No. 216, from the Bangiya Sahitya Parisad, Calcutta, Some folios missing. Very 3=MS No 2086, Bangiya Sahitya Parisad, Calcutta. much worn out, hence illegible in some places. Text upo 91. 34 of the Venkt. Edn. Devanagari MSS of Northern India 1 No. 54/153, from the Sarasvati Bhandara, Fort Ramnagar. No. 14.351, from the British Museum, London. 3 No. 1447, from the Pensylvania University (U. S. A.). Saka 1679 (A.D. 1757). 4 No. 78 of 1882-83, from the B.O.R.I., Poona. Samvat 1714 (A. D. 1657). 5 No. 19 of 1873-74, B. O. R. I., Poona. Samvat 1839 (A. D. 1782). svasti sri maharajadhiraja | sri sri 5 ravala ji affariant sri mularaji va ( va ? ) canartham | sri bikaneramadhye || (From the PostColophon). o MS Wilson 127, Bodleian Library, Oxford. (Only BhuvanaKosa Chapters collated). 20

154 puranam - PURANA 7=E. 3584, India Office Library, London. (A. D. 1805). [Vol. IX., No. 1 Samvat 1862 8 E. 3585, India Office Library, London. (only Bhuvana Kosa Chapters collated). 10= Keith 6815, India Office Library, London. Samvat 1773, Saka 1639 (A. D. 1717), might be of earlier date, cf. samvat 17 || 73 || varsam sake || 16 || 306 || pravattammane pausasitadvadasyam 111111 112 113 11 bhrga di . kasi ramenatmapathanartham paropakrtaye ca mulyena grhitamidam vamanapuranam || 6 || srirastu || || ( Post-Colophon). 11=No. 54/184, Sarasvati Bhandara, Fort Ramnagar, Fairly written. Slokas numbered. South Indian MSS 1 No. D. 2263, Government Oriental Manuscripts Library, Madras, Script Telugu. Incomplete, first 12 folios missing, extent from 6.47 to Adh. 94 of Venkt, edn. 9 = No. 8B 5 (Catalogue No.), (P. M. 2463), Adyar Library, Madras. Number of Adhs. 67, the last Adh. (67) corresponds with the last Adh. (95) of the Venkt. edn. Granthas 4750. 1 = Palm-leaf MS from the Srngeri Matha, Mysore (South India). Characters Nandi-nagari. Extent and numer of the Adhyayas like a 9. Besides these, some MSS have also been consulted and their readings and other details have been compared to some extent with the allied collated MSS; viz., MS No. 4418 of the Gujarata Vidyasabha, Ahmedabad. It agrees with T 1 noted above. MS G. 844-58-B. 8 (H. P. Sastri's catalogue, Vol. V, No. 3990). Characters modern Kasmiri. It generally agree with T 2. MS No. C-433, from the Government Oriental Ly Mysore. Characters Nagari. It seems to be a transcript of some Telugu MS of the Vamana-Purana, and agrees with our 1; both these MSS Adhs. 84-88 of the Venkt. edn. are omitted,

Jan., 1967] CONSTITUTION OF THE VAMAN PURANA TEXT 155 Five Devanagari, MSS B. 1983/D. 10419-B. 1587/D. 10423, all belonging to the Sarasvati Mahal Library, Tanjore. The details of the last Adhyayas of these MSS were kindly supplied by the Library In extent of the text they (except D. 10420) generally agree with our 9 and 1, as already noted before in Sec. A., and they name their last Adh. (= Adh. 95 of Venkt. edn.) as falari. A recently acquired Telugu MS from the Government Oriental Library, Mysore, agrees, in extent and the ine or the last Adh., with the above mentioned Tanjore MS. Vamana-Purana slokas quoted in the various Nibandhas have also been collected, and slokas and topics of the Vam.-P. have been traced in other Puranas and also in the Mahabharata. This testimonia has also helped in settling some readings of the Vamana-Purana, but the main basis for the constitution of the text has been the evidence of the above noted manuscript material. According to their general agreement in readings, omissions, additions etc. the above mentioned collated MSS have been arranged in the following Groups :Group I Kashmirian Group: 1. 1.2. (5 also often agrees in readings with this Group). Coup II Bengali Group: 1. 2. 3. Group III Devanagari MSS and the Telugu MS with which they generally agree: 1. 2. 3. 10. 1. (10 sometimes de te agrees with GI.) Group IV 4. 7. 11. (These MSS sometimes agree with G II. 4. 7 often agree with each other. 11 sometimes agrees with 1 even in corruptions and mistakes.) Group V=South Indian MSS other than the Telugu 1. 9. 1. In constituting the text one of the primary considerations has pe that the constituted text should have at least some sense; Nla-Kantha in his Commentary of the Mbh. (III. 83) remarks, 'pathakramadarthamkramo baliyan ' and hence 'pathadartho baliyan '. It should be

156 puranam - PURANA [Vol. IX., No. 1 assumed that the ancient authors, compilers or the redactors of the Puranas composed, compiled compiled and redacted the texts that must have had some sense, though the sense of some of these texts might have become obscure or vague during the course of their transmission owing to the corruptions or whimsical changes in such texts due to the scribal carelessness or ignorance. Keeping this main consideration of a proper sense in view, robably genuine or the oldest text has been constituted on the following ciples 1. The evidence of the majority of the MSS, is generally accepted. 2. The evidence of the majority of the Groups or the Versions is accepted, even if the total number of MSS of these Groups or Versions may be in minority as compared to the number of the remaining MSS. The reading common to the geographically distant Groups or Versions is preferred. E.g. :- (1) In 'uttarena kurovamsah ' (13.5a) kuruvamsah is the reading in G I and G V only, but it has been adopted on account of the more geographical distance of these two Groups. (2) In 'satimenam ' (21.18c) sati menam is the reading in GI and MS 1. So it is adopted. Some MSS have gai for however, seems to be an emended reading. t,hich, ( 3 ) In ' esa vighnasahasrani devadinam hanisyati ' (54.73 cd) karisyati is the reading for ff in GI and V, and it is also a more suitable reading here in this context. It is, therefore, adopted. 4. In constituting the text of a Sloka or a group of Slokas, the readings of one Group or MS are, as far as possible, not mixed up with the reading of other Groups or MSS. rich 5. If the different Groups or MSS contain diverse readings in a text, then the evidence of that Group or MS is accepted gives a correct or more suitable Bengali Group only gives a correct [See below 7. (5)]. reading; e. g. in 57.91 a the reading, and so it is adopt.

Jan., 1967] CONSTITUTION OF THE VAMANA PURANA TEXT 157 Here it may be noted that the Khmirian Group often gives a correct or more suitable reading. It also supplies some missing texts-[See below 16, (1), (2)]. The B.H. U. Sarada MS No. 330188, referred to below, is dated Saka 1444 (A. D. 1522), and is thus the oldest MS available till now. It agrees with our collated MS 1. 6. Sometimes a correct text is preserved only in one MS or in a very few MSS irrespective of the age of the MS or MSS; In that case, naturally, this text or reading is adopt. E. g. :- (1) drstva'gnina tathaivanye devadyah pralayam gatah || (5.9cd). Here is the reading given in most of the MSS, which is clearly wrong and has no sense here. The correct reading af is given in 1 only, and it is also confirmed by the reading drso'gnina given in de 4. So drstyagnina is adopted. (2) yusmakam caprasadena duskrtena ca karmana || Here MSS read pakso'yam vardhate'smakam yatasca brahmaraksasah | (40.34-35 is the reading given only in 7; all her which has no appropriate sense here, and seems to be an emended reading. In the Mbh. (cr. edn., IX. 42. is the reading adopted, though it is marked by 16a), too, a wavy line. (3) yatra purvam (vaॅ ?) sthito brahma daksine ca mahesvarah || Here rudrapatni pascimatah padmanabhottare sthitah | madhye hyanarakam tirtham trailokyasyapi durlabham || (41.22cd-23) is the reading in 7 only, all other MSS have rsinam for daksine . But daksine is the correct reading here, for other three gods are associated with the remaining three directions. So the reading of 7 (af) is adopted. The reading rsinam is unaccountable and has no sense in this context.10 (4) tamasasya manoh putro dantadhvaja iti srutah | sa putrarthi juhavanau svamamsam rudhiram tatha || 10. See also A. 3. (1) of this article, and fn. 8 above.

158 puranam --- PURANA asthini ranakesamsca snayumajayakrdghanam | [Vol. IX., No. 1 sukram ca citrako raja (dantadhvajah ) sutarthi ceti nah srutam || (72.57-58) Here citrako, citraso, binduso etc. are the readings given in all the MSS except na 1 which reads citragau (locative of citragu which means citrabhanu or agni ; go = bhanu or rasmi ). This reading ( citragau ) of na 1 seems to be the only correct reading here. So it is adopted. the reading of a certain MSS is confirmed by some preceding or toning text of the Vamana-Purana, then that reading is adopted inspite of the majority-evidence going against it. Eg. :- (1) vasavo'stau haram drstva sasrpurvegato mune | sa tu jata saricchrestha sita nama sarasvati || ( 5. 2) sasrpuh, sasrjuh etc. are the readings given in the majority of the MSS, but 9 gives the reading : which seems to be the only ect and proper reading here, as it is confirmed by the second half of this sloka itself. (2) tadbhitam (v 1. bhanubhi ) raksasapuram tannastam ca yathecchaya || (15.38 cd) Here nastam is the reading in several MSS, (ba 1-3. de 1. 2. 10), meaning that the city of Sukesin was destroyed by the rays (bhanubhih ) of the Sun ; but drstam is the reading in several other CSS, which is confirmed by the next Sloka (39) : sa (purah ) bhanuna tada drstah krodhadhmatena caksusa | nipapatambarad bhrastah ksinapunya iva grahah || (15.39) Here the reading is given in all the MSS, and these readings drstah and nipapata here refer to the city (purah ) of Sukesin, as ic clear from the following slokas 40, 43, 59, and 61 of this Adh. ; the word g has been used in masculine by MSS in 15. 8. So in the above text (15. 38 cd) the reading is more appropriate than nastam . pura (3) tato vinakrta devah senanathena sambhuna | danavendrena vikramya nisumbhena parajitah || (21.12) 4.

Jan., 1967] CONSTITUTION OF THE VAMANA PURANA TEXT 159 kim jitah stha (stha ? ) surendrena nisumbhena duratmana | (15 ab) sa samkarat satejomsam janayisyati yam sutam | sa hanisyati daityendram subham ca sapadanugam || (20) Here in 21. 12d mahisena is the reading for nisumbhena in several MSS; in 15 b some MSS read fd, some (a 2. 4. 1) read sumbhena, but seven MSS of different versions read mahisenah in 20d is the reading for in most MSS. Bereen these two readings-nisumbha (or sumbha ) and mahisa - the reading mahisa is, however, supported by 51. 7c (g af, all MSS) and 51. 18 d (faff, all MSS) and also by Adh. 58 in which the battle between Skanda and Mahisa is narrated. It seems that there were two different demons of the same name Mahisa, the one was the son of the demon Rambha from a mahisi or she-buffalo (cf. 17. 51 ff), who was the associate the demons Sumbha, Nisumbha, Raktabija etc., and was killed by Goddess Katyayani (cf. Adhs. 19-20; and 55. 1-18); while th other Mahisa was the brother of the demon Taraka (cf. 58. 85), and an associate of the demons Andhaka, Taraka, Bana (Bali's son) etc. (58. 45ff.), who was killed by Skanda (5%. 87ff.). The two ahisas have been confused, and (or g) is the resuntant wrong reading in Adh. 21; the scribes who substituted the wrong reading nisumbhena (or sumbhena ) for the right reading mahisena might have thought that because Mahisa had already been described in the previous Adh. (20) as killed by Katyayani, he could not again possibly be mentioned in Adh. 21 as a participant a later war with Skanda. (4) vaisakhe ca yada'stamyam mangalasya dinam bhavet || tada snanam tatra krtva mukto bhavati patakaih || (41.24-25) Here yada sasthyam (for yada'stamyam ) is the reading in the Kashmirian MS o ly; besides these, 1. 4. 11. also read cc. The reading of the Kashmirian MSS (for) is supported by the further text of 41.28 cd.-zgarfq qar qgi nga

160 puranam - PURANA [Vol. IX., No. 1 where cg is the reading in all the MSS. So cat is the correct reading. (5) When Skanda was anointed by the gods as the General of their armies to fight with the demons Taraka, Mahisa, etc., he was given various Ganas or attendants by the Devas, Nagas, Yaksas, Parvatas such as Himavan and Vindhya, Sarasvatis (i. e. Nadis), Kutila, Krttikas, Rsis, and also by the Tirthas such as Prthudaka, akra-tirtha, Gayasiras, Kanakhala, Puskara, Mana sa and Ausanasa (vide 57. 60-91). Now 57.91ab reads as follows: rudramosanasah pradat tato'nyan mataro daduh | anyan Here is the reading in the Venkt. edn. and also in several MSS; 91b then will mean, according to this reading, that 'the Matrs then (a), gave [to Skanda] the others (i. e. the other Ganas); or 'the Matrs gave [to Skanda] others than these (tato'nyan ) (aats In the Kashmirian MSS and in 4. 1 the reading is 'tato mataro daduh ' where anya ( = anyah ) seems to be corrupt and has no clear sense in this context. In the Bengali MSS a 1-3, however, a', which means-'then the we have the reading 's others (i. e. the other Tirthas) gave [to Skanda] the Matrs 11 or the female attendants. The following text (57.91 ed.-102ab), in rhich the other Tirthas such as Soma-tirtha, Prabhasa, ndratirtha, Udapana, Sapta-sarasvata, Naga-tirtha, Kuruksetra etc. are mentioned as giving various Matrs (female attendants) to Skanda, supports this reading of the Bengali MSS. Hence the reading of the ba MSS only - 'tato'nye mataro daduh is correct and appropriate here. (6) At the time of the departure of Skanda to fight with Mahisa and other demons, Hari pronounced svastyayana (benediction) 11. The word 'H' here is an object of the verb '', the normative form (Prathama) has been used here in 'H' for the accusa ive form (Dvitiya). For further discussion and illustration see fn. below,

Jan., 1967] CONSTITUTION OF THE VAMANA PURANA TEXT 161 on him for his well-being and victory (58. 14-25). Sloka 18 of this svastyayana runs as follows :- mariciratrih pulahah pulastyah kraturvasistho bhrgurangirasca | mrgankajaste kurutadvimangalam maharsayah sapta divi sthitasca ye | (58.18) But the name Here in 18c 'mrkandujah ' is the reading for 'mrgankajah ' in the Kashmirian MSS. Now 'mrgankajah ' means the planet Mercury ( budha ) as he is regarded the son of Mrganka or the moon. of a planet is out of place here among the names of the Rsis. Moreover, in this context, the planet Mercury or Budha has already been mentioned earlier in S1. 17 as follows :- divakarah svastikaro'stu te sada somah sabhaumah sabudho guruca | kavyah sada svastikaro'stu tubhyam sanaiscarah svastyayanam karotu || Budha (mrgankaja ), therefore, cannot again be repeated here. So 'mrkanduja !' (Sage Markandeya), the reading of the Kashmirian MSS, is more appropriate here. 8. When the reading of a Ms or of several MSS is supported by some external source or testimonia, then that reading is adopted. E.g. :Jaure (1) namo namaste'cyuta cakrapane namo'stu te vamanaminamurtte| 3.22ab) Hara praises Visnu in 3.14-23. In 3.22ab as quoted above vamanami kumam ( na 1), madhavaminakumam (de 9), and madhavaminamuttem ( ka 2 ba 1-2 de 4) are the variants for vamanaminamurte . The reading madhavaminamurte, however, is supported by the Naradiya Purana also as follows: svetagangam narah snatva yah pasyecchvetamadhavam | matsyakhyam madhavam caiva svetadvipam sa gacchati || 13 svetamadhavamalokya samipe mattya madhavam | 14ab adhaya tarunam rupam madhavam matsyamadhavam | pranamya prayato bhutva sarvan kastan vimuncati || 16 (NP., II. 56) So this reading is preferred to the other readings, and is, therefore, adopted. 21

162 puranam - PURANA (2) virabhadraya ciksepa cakra krodhat khagadhvajah || tamapatantam satasuyamkalpam sudarsanam preksya ganesvarastu | [Vol. IX., No. 1 sulam parityajya jagara cakra ' yatha madhum minavapuh surendrah || (4.48cd-49) Here in the MSS sa 1 kara . ka 1 'yatha mayo vira (ghora, kara .) vapuh surendram ' is the variant for 'yatha madhum minavapuh surendrah '. But these are the two contradictory readings. The reading 'yatha madhum minavapuh surendrah ' nowever, has some support in the Skanda Purana (V. iii. 9. 36-38) as follows : sa vivesa maharaja bhutalam sasurottamah | danavantakaro devah sarvadaivatapujitah 136 minarupadharo devo lodayamasa carnavam | vedamsca dadrse tatra patale nihitan prabhuh || tau ca daityau mahaviya drstavan madhusudanah | mahavegau mahabahu sudayamasa tejasa | (tau ca daityo = madhukaitabhau ; cf. ibid 33 cd). But no source has yet been available in support of the reading 'yatha mayo viravapuh surendram . (3) nagadantasthibhrnganam tatksanacchuddhirisyate | ( 14.64ab) Chapter 14 of the Vam, - P. deals with the ways and means of purifying various articles of daily use. This Dharmasa tra topic is also included in several other Puranas, such as the Markandey a and the Vayu. In the above half sloka the purification of ivory, bone and horn is mentioned. Here is the reading in most of the MSS of the Vam. P. and in the two MSS of the Markandeya also, which have been collated up till now. This reading, however, carries no sense, for it only means that these articles are purmed immediately. But two MSS ba 2.3 read taksanat, meaning there by that these articles are purified by scratching or abrading; (taks ). This reading is confirmed by the Yajnavalkya-Smrti, Acaradhyaya, 51 185 - 'taksagam darusrngasthanam ' and the Mitaksara explains it as 'taksanam tavanmatravayavapanayanam '. The Vayu-Purana 782 makes this reading certain by using the synonym taksana - ' tatha carmasthidarunam srnganam cavalekhanam '. for

Jan., 1967] CONSTITUTION OF THE VAMANA PURANA TEXT 163 Thus the reading is proved to be a correct one, and is, therefore, adopted inspite of the majority MSS evidence going against it. In fact, '' seems to be a wrong emendation of the lactio difficilior 'taksanat ' (4) In 13.25 of the Vamana-Purana the names and the positions of the nine divisions or Varsas of Jambu-Dvipa are given. The nine Varsas of the Jambu-Dvipa are:- 1. Ilavrta, 2. Bhadrasva, 3. Hiranvan, 4. Kimpurusa, 5. Bharata, 6 Harivarsa, 7. Ketumala, 8. Ramyaka, and 9. Kuru-varsa. Now in the MSS of the Vam.-P. a confusion exists between the positions of the two Varsas, viz. Hiranvan and Kimpurusa. The text of the Venkt. edn. is as follows:- purvadaksinato varsom hiranman raksasesvara | 11 ( 13. 3cd) purvamuttarato ramyo varsah kimpurusah smrtah || (13.5cd) [v. 1.-3c) to be given and discussed below. -3 ( most MSS ), hiranyo (G I or Kashmirian MSS) for hiranman . purvam uttarato (G III or de 1. 2. 3. 10 te 1), purvottare (G V or the two South-Indian MSS), uttarapurvamto ( ba 1. 2) for purvamuttarato . ] In 3c, only the two MSS 4.7 read ga saat af (a 4 ga- uttaratasca ) for purvadaksinato varsom, all the remaining MSS have the same text as that of the Venkt given above. In 5 cd one MS . gives somewhat defferent reading as 'bharataduttare varsam kimpaurusam manoramam ', and the ten MSS (1. 1. 234-8.11) omit 5cd altogether, so that the text containing the name and position of KimpurusaVarsa is missing in these ten MSS, except in 4 which gives after 3 cd an additional half-sloka - 'purvadaksinatacapi kimnaro varsam ucyate | ' So, according to these readings the positions of the two Varsas may be summed up as follows:Situated Hiranvan (or Hiranya) Varsa: in East-South (according to all MSS except 4.7) in East-North (according to 4.7 only).

1 164 puranam - PURANA Kimpurusa (or Kimnara) Varsa: in East-North (or North-East) (Most MSS) Situated in East-South (according to 4 only) to the North of Bharata (TT. only). [Vol. IX., No. 1 Thus, there is the confusion between the positions of these two Varsas. In order to ascertain their right positions and then to decide the correct reading we have to depend on the evidence of the other Puranas. Several Puranas contain the chapters on the Bhugola or Bhuvana-Kosa, and in those chapters the position of each of the nine arsas of Jambu-dvipa is also given". The right position of each of these two Varsas-Hiranvan and Kimpurusa-can be ascertained only when studied in relation to the different positions of the other Varsas also. Fortunately, the Puranas, containing the Bhuvana-kosa chapters, are almost unanimous as regards the position of the nine Varsas of Jambudvipa. The Visnu-Purana (II. 1. 15-32; 2. 11-15) gives the following aunt of these nine Varsas :- "The first ruler of Jambu-dvipa was Agnidhra, who was the son of Priyavrata and grandson of Manu Svayambhuva. Agnidhra had nine sons, viz. 1. Nabhi, 2. Kimpurusa, 3. Harivarsa, 4. Ilavrta, 5. Ramya, 6. Hiranvan, 7. Kuru, 8. Bhadrasva and 9. Ketumala. There were also the nine Divisions or Varsas of Jambu-dvipa which were known by the names of their Varsaparvatas, i. e. the Parvatas or Mountains which separated these nine Varsas from each other. Agnidhra gave each of these nine Varsas to each of his nine sons as follows:- 1. Hima-Varsa (i. e. the Varsa containing the Himavan mountain) to Nabhi, whose grandson was Bharata, after whose name this Varsa began to be known as Bharata-Varsa. ; 2. Hemakuta-Varsa to Kimpurusa, and so this Varsa was also known as Kimpurusa-Varsa; 3. Naisadha-Varsa (i. e. the Varsa containing the . Nisadha Varsa-Parvata) to Harivarsa, and so it was also known as Harivarsa-Varsa; 12. For the references to Bhuvana-Kosa chapters of the Puranas ad the position of the nine Varsas ef Jambu-Dvipa see Appendix.

Jan., 1967] CONSTITUTION OF THE VAMANA PURANA TEXT 165 4. The Varsa containing the Meru in the middle was given to Ilavrta, and hence it was known as Ilavrta-Varsa. 5. The Varsa containing the Nila Varsa-parvata was givenn to Ramya, and was known as Ramyaka-Varsa. 6. Sveta-Varsa (containing the Sveta Varsa-parvata) was given to Hiranvan, and hence was also known as Hiranvan-or HiranmayaVarsa; 7. The Varsa situated in the North of the Sric van VarsaParvata was given to Kuru, and hence was own as KuruVarsa or Uttara-Kurus; 8. The Varsa situated to the East of the Meru was given to Bhadrasva, and hence was known as Bhadrasva-Varsa; 9. The Gandhamadana-Varsa (situated to the West of the Meru was given to Ketumala and hence was known as KetumalaVarsa." T It is now clear that Kimpurusa-Varsa contains the makuta Varsa-parvata, and is, therefore, associated with it. Similarly, Hiranvan-Varsa contains the Sveta Varsa-parvata and is, therefo associated with it. The same account of these Varsas and Varsaparvatas are given by the other Puranas also which contain the Bhuvanakosa chapters. Now the positions of the Varsa-parvatas are mentioned in. the Visnu-Purana as follows:- himavan hemakutasca nisadharacasya daksine | nilah svetasca srngi ca uttare varsaparvatah || ( srasya = meroh : ibid. S1. 7-9). (II. 2. 10) The Garuda Purana (54.8) also mentions the same positions of these Varsa-parvatas, and in the same words as the VisnuPurana. Other Puranas also corroborate these positions. So Kimpurusa-Varsa which is associated with the Hemakuta is situated to the South of Meru, or to the South East (East-South) as given in the MS. 4 of the Vamana-Purana. And similarly, firanvan-Varsa (v.1. Hiranmaya-Varsa, Visnu P. II.2.13; HiranyaVarsa, Kashmirian MSS of the Vamana-Purana), is situated to

166 puranam - PURANA [Vol. IX., No. 1 the North of the Meru, or to the North-East as given in the two MSS 4.7. Thus, the reading of 4.7 is correct as regards the position of Hiranvan-Varsa; similarly the reading of 4 is correct regarding the position of Kimpurusa-Varsa. According to the readings of 4.7 (for the position of Hiranvat or Hiranya Varsa), and the reading of 4 (for the position of Kimnara or Kimpurusa Varsa) the positions of these nine Varsas of Jamba-Dvipa in the order of the Directions from the Meru, then, would be as follows: 1. In the Middle (of the Jambu Dvipa)-Ilavrta-Varsa (with the Meru)13; 2. In the East (of the Meru)-Bhadrasva-Varsa (with the Malyavat)14; 3. In the East-North-Hiranvat-Varsa (with the SvetaWarsa Parvata); 4. In the East-South Hemakuta); Kimpurusa-Varsa (with the 5. In the South-Bharata-Varsa (with the Himavat); 6. In the South-West-Harivarsa-Varsa (with the Nisadha) 7. In the West-Ketumala-Varsa 8. madana)15; (with the GandhaIn the West-North-Ramyaka-Varsa (with the Nila); 9. In the North-Kuru-Varsa (also called Uttarah 13. The Varsa-Parvatas as associated separately with each Varsa are given here within brackets, and according to the Visnu-Purana (II. 1. 18-23, 32), Garuda-Purana (54. 6-8) etc. The Vamana-Purana however, does not mention in this conection the Meru and the other Varsa-Parvatas. 14-15. For the association of the Malyavat and the Gandha-madana mountains respectively with Bhadrasva and Ketumala cf. Vayu. P. (33. 44-45):- varsam malyavatam capi bhadrasvaya nyavedayat | gandhamadanavarsam tu ketumale nyavedayat ||

Jan., 1967] CONSTITUTION OF THE VAMANA PURANA TEXT 167 Kuravah according to the Visnu and some other Puranas) (with Srngavat). Thus, according to the reading of 4. the description of the positions of these nine Varsas ends with the North, as in the other Puranas. Those MSS of the Vam-P. (including ≥ 7) which omit the mention of Kimnara-Varsa, also end this description with the North. But in the MSS which mention KimnaraVarsa as situated in the North-East this description ends with the North-East and not with the North. The readings of 4.7 regarding the position of Hiranvat- "Varsa and of 4 regarding the position of Kimnara or KimpurusaVarsa are, therefore, correct, while the readings of the remaining MSS for the position of these two Varsas-mentioning Hiranvat as situated in the East-South and Kimpurusa in the North-Eastare not corroborated by the other Puranas.17 9. Sometimes an originally correct reading becomes corrupt during transmission, and then later on this corrupt reading is also wrongly emended; or the correct reading itself, if it happens In that case the to be a lacio difficilior, is wrongly emended. original correct reading is settled (a) by the help of some other 16. Cf., for example, the Visnu-Purana, II. 2, 14: 'vai bharatam tatha ' 17. The Garuda-Purana, however, gives the following positions of these two Varsas:- purvadaksinato varso hiranvan vrsabhadhvaja || tatah kim puruso varso merodaksinatah smrtah bharato daksine prokto haridaksinapascime || (GP., 55.1-2) The Garuda Purana text (55. 1 cd) for the position of 'f Varsa tallies with the text of the majority of the Vamana-Purana MSS. But this text of the Garuda-Purana is wrong, for, the position of the Varsa-Parvata with which the Garuda Purana associates f(in 54. 11) is given in the north of the Meru in the GP itself (54. 8 ef) : 'nilah svetasca srngi ca uttare varsam parvatah '. The position of the 'f' Varsa is given in the South, which nearly agrees with the reading of 4 of the Vam. P. in this respect,

168 puranam - PURANA [Vol. IX., No. 1 preceding or following slokas, or (b) by the help of some other external evidence, or (c) by the context and sense. E.g. :- (1) aruhya valabhim tastu samudaiksanta sarvamsah || (65.115cd) Here 11 is the reading in several MSS; in some MSS, however, the Anusvara has disappeared and the reading, therefore, has become 'a' which has been emended in some other MSS, however, as 'agar'. But cf. Sl. 119- ityevamuktva vacanam valabhya sravatiyam ca | samasanta'gratah sambhorgayanti gitakanchubhan || (65.119) The reading 'r' in 115c is, therefore, confirmed, and the other readings 'balabhitastu ', bahubhitastu are corrupt or wrongly emended. (2) satadrava lalitthasca paravatasamusakah (13.39 ab) These are the peoples or Janapadas of the Uttarapatha. In the Kashmirian MSS we have the reading 'sitadravah ' for 'satadravah '. But in the Mark.-P. (57.37a) the reading is ':'. So, in 39a above 'satadravah ' ( satadru +srana ) is the correct reading ; 'sitadravah ' is a wrong emendation and is, therefore, corrupt. (3) tatah sitavanam gacchenniyato niyatasanah || (85.44 cd) The Venkt. reading 'a' is contained in 1.2.1 also, but in some MSS the reading is '' and in some other MSS ( \ 47 ) it is 'sitavanam '. It is clear here that the reading 'sitavana ' got corrupled as 'a' and then again this latter reading was emended as 'sitavanam '. Or, reverse may be the case. The original reading might have been the 'sitavanam ', which was corrupted as 'sitavanam ', and then '' was emended as ''. In the cri. edn. of the Mbh. (III. 81. .) we have also the reading 'aa'. The Venkt. edn. in 34.5 cd also reads 'g This reading '' here (in 34.5 cd of the Venkt. edn.) is contained in T 1. T 1.2.4.10 also, but 2.3 read and 1.7.11 read aari.

Jan., 1967] CONSTITUTION OF THE VAMANA PURANA TEXT 169 Several MSS of the Vam. P. viz. sa 1 ka 1.2 de 1.5.11, read in 35.44c 'sivavanam ' for 'sitavanam '. But the reading sivavanam " is not found in the MSS of the Mbh. (4) purvam gayena ksitipena yatra isto'svamedhah satasah sudaksinah | manusyamedho'pi sahasrakrtvastatha pura dujayanah suraribhih || (76.15) Here ':' (=cc:) is the reading in 11 and a 3 for 'manusyamedhah ', which gets corrupted or wrongly emended as 'narendrasuryah ' (in Kashmirian MSS and 5', and ':'. etc. (5) hanau punavasusrokto nasa maitramudahrtam || (80.7 cd) Here the Naksatra is mentioned as the nose of the Naksatra-Purusa Visnu. Mitra is the lord of the constellation Anuradha; so here Anuradha has been supposed as the nose of the Naksatra-Purusa. It is the reading in several MSS also. But 80.5a reads as (?) and again in 80.17 c we have 'anuradhasu vakso'ya ' and in the Matsya-P. also 'anuradhasu unalameva (Adh. 54). So the reading above is wrong. But in some MSS we have the reading '' for ''. Pitys are the lords of the constellation Magha, and in 80.24 a we have g. So the constellation Magha is the nose of fhe NaksatraPurusa, and '' is the correct reading in 83.7 d above, which was easily corrupted paleographically as ''. 10. Older or less familiar readings are preferred. E. g. :- (1) trailokyarajyamacchidya balerindraya yo dadau (1.1) Here is the reading in the MSS of G III and GV, 'akrsya ' in the MSS of the Kashmirian group and aksipya ' is the Treading in ka 2. dem 4.5.7. The reading 'aksipya ' in the sense of 'taking off' or 'taking away' seems to be older or less familiar than sracchidya or prakrsya ; Monier Williams refers the use offer in this sense to Mahabharata. So this reading fc is preferred here. (2) brahmahatya'smi samprapta mam praticcha trilocana || (4.4 cd) Here in 4c 'r' is a variant in Several MSS. This 22

170 puranam - PURANA [Vol. IX, No. 1 reading is preferred to '' which is a more familiar reading, compare also: dustara paravadhya'pi svavadhya'pyatidustara || ( 17.49 cd) Where all the MSS read paravadhya and svavaghya . (3) 'nabhisthanam subhakaram ' (6.99 :) Here the reading 'r' is contained in 1-4. 9. 10 also. But the MSS sa 1. ka 1. 2. ba 1. 2. de 1 read 'nahasthanam ' (meaning the place where the lower garment is tied from (v). The reading '' being less familiar, besides being contained in different versions, is, therefore, preferrable. 11. Grammatical aberrations abound in the Epics and the Puranas. They are sometimes due to their archaism, sometimes to the influence of Prakrta, but mostly they are due to some metrical exigencies, for, grammatical correctness has often been sacrificed for the correctness of metre. But such grammatical aberrations are often found changed to their regular Paninian forms in several MSS; and sometimes they are even wrongly emended in some other MSS; hiatus between two vowels is also generally removed by the insertion of such particles as a, a, f etc.; similarly, cases of double-sandhi-3 are found cmended by change of construction or in some other ways.18 Hiatus, double sandhis and other grammatical aberrations are, therefore, older than such emendations. In the constitution of the text, therefore, hiatus should be preferred to vowel-sandhis or to the insertion of particles (,, f etc.), and the irregular double-sandhis to the regular Paninian Sandhis; similarly, other grammatical aberrations, if not proved to be the scribal mistakes, should be preferred to their regular Paninian forms. E. g. :- (1) Hiatus: M tairvimisra janapada mlecchascaryasca bhagasah || ( 13.19 cd) 18. For the detailed discussion of hiatus, double Sandhis and other grammatical aberrations my article puranesvapaniniyaprayoga ' in Purana, IV. 2 (July, 1962), pp. 277-296,

Jan., 1967] CONSTITUTION OF THE VAMANA PURANA TEXT 171 The Venkt-reading '' is also contained in some the MSS, but some MSS have the reading : while the MS 4 reads 'Now it is clear that this reading of 4 containing the hiatus must be the original one, for an attempt seems to be made later on to remove the hiatus by inserting the particle 'a', and thus the reading '' became ' mleccha caryasci '. But this emended reading 'mleccha caryasci ' is wrong according to the rules of the Sandhis, so it was further emended to become '. been the process. Thus mleccha prayasci > mleccha caryasca > mlecchascaryasca must have Hence the original reading was ' mleccha prayasci ', which contained the hiatus, although the sandhi was regular according to the grammatical rules of Panini. (2) Double-Sandhi: jalopari mahiyam hi naurivaste sarijjale | tasyopari ca deveso brahma sailendramuttamam || karimgaka karamatyucca sthapayamasa sattamah | ( 11.31-32 ab) The reading in de 4 is 'tasya (mahyah ) upari ', while in de 7 it is 'tasyamupari ' for 'tasyopari ' in 31 c. Both these readings are the results of a later attempt to emend the double-sandhi existing in 'f' which is the reading in most of the MSS also. This reading can be analysed as follows: tasyopari < tasya upari < tasyah upari Thus, the reading 'a' containing a double-sandhi in upari is older and is therefore preferrable. The reading 'tasyopari ' is probably due to the metrical exigency. (3) Grammatical aberrations (other than those of sandhis): (a) Linga-vyatyaya (Reversion of gender) : (i) tato gatayam saradi prabuddhe caiva kesave | (2.7 ab) All the MSS here rcad 'vyatite saradi ' for 'gatayam saradi '. The word (Autumn) in Sanskrit is feminine, so the reading

172 puranam - PURANA [Vol. IX., No. 1 'vyatite saradi ' in masculine is a case of linga-vyatyaya, and may be original or older reading. (ii) 'vrajatsu yositsu catuspathesu ' (3.32a) Here '' is also a case of linga-vyatyaya, which is perhaps resorted to owing to the anuprasa in the combination 'vrajatsu yositsu ' de 4.7 emend it as 'yantisu yositsu '. (iii) sa esa nunam tapatam varistha rtadhvajo natra vicara gam'sti || 65.117 cd) The reading in most of the MSS and in the majority of the versions is 'vicaramadina ' for 'vicarana'sti '. The word 'vicara ' (masculine) is used here as neuter, and so is emended as 'vicarana . The reading 'vicaramasti ' is therefore, preferable. (iv) 'pancamam mukhamabravita ' (2.32d) Here most of the MSS read 'pancamo'tha mukho'bravit ', which is preferrble. (b) Vacana-vyatyaya (Reversion of number) mayapurvam mayapurvam vivadantau parasparam | sragamyocurmahesaya brahmane madhavaya ca || (58.106) The reading ':' in plural is used here for the two persons (Indra and Skanda) as the reading 'faral' shows It is given in most of the MSS. The reading ':' here is due to the metrical exigency only and is preferrable. (c) Vibhaleti-vyatyaya (Reversion of case) (i) hrsadvatyam narah snatva tarpayitva ca devatah | agnistomatiratrasya phalam vindati manavah || (36.48) The reading in many MSS is 'agnistomatiratrabhyam ' in 48 c, in which trtiya or caturthi (instrumental or dative) has been used for the sasihi (or the genitive case). This reading is also similar to and thus corroborated by Mbh. III. 81.73. Compare also :- tatah pariplavam gacchet tirtham trailokyavisrutam | sragnistomatiratrabhyam phalam prapnoti bharata || (Mbh. II. 81.10) Compare also the following sloka of the Mbh :- gamanadeva rajendra dirgham satramarimdama | rajasuyasvamedhabhyam phalam prapnoti bharata || ( III. 80.117)

Jan., 1967] CONSTITUTIONS OF THE VAMANA PURANA TEXT 173 Here also trtiya or caturthi has been used for sasthi, as is indicated by the following text of the Mbh :- gatva hi sraddhaya yuktah kuruksetram kurudvaha | phalam prapnoti ca tada rajasuyasvamedhayoh || ( III. 8. 6 v. 1.) Here, we have the usual sasthi in 6 d. (ii) sakrasyahrtya ca gajam yamyam ca mahisam balat || (55. 13 cd) Here most of the MSS read as :- sakrasyahrtya (v. 1. krsya ) ca gajo yamyasva mahiso balat | where prathama (nominative) has been used for dvitiya (accusative), which is perhaps due to the influence of Prakrta in which the • forms of nominative and accusative are generally the same; e. g. compare the reading 'nidhayah samdidesa sah ' (78.36 a) which is given in most of the MSS.1 (iii) 'jagmuhumsta rathebhyaste ' (65.106 a). Here, all the MSS except 7.9 give the above reading in which caturthi (dative) has been used for trtiya (instrumental) as is clear by the reading 'f' given in 7.9. This reading of de 7.9 is an emendation of the reading rathebhyas . (iv) tatah sakunina panigrhito yaksakanyaya | (65.162 cd) is also the reading in most of the MSS. Here trosya is used for sasthi. This reading is also due to the metrical exigency. (d) Other Grammatical Aberrations: (i) 'sa'tha mam praha kim kurmo ' (55.38 a) The reading in many MSS and versions (1. 3, 10. 1-3, na 1) is 'kurmi ' for 'kurmo ' ; in sa 1. ka 1. 2 and de 4 the reading is karomi, while in 7 it is 'f'. The reading '' is older or archaic, and is, therefore, preferable to the other readings which are perhaps its emendations. • (ii) 'hasanti suratham praha ' (65. 116c ) The readings in many MSS is 'suratham hasati praha ' which contains a grammatical aberration in '' (=). This reading is due 19. For further illustrations see Appendix.

174 XT puranam - PURANA [Vol. IX, No. 1 to the metrical exigency; i. e. for avoiding a long 5th syllable it. this pada of the Anustubh metre, which should always be short according to the metrical rules). It has been emended by changing the construction as 'hasanti suratham praha '. (iii) purvam gayena ksitipena yatra isto'svamedhah satasah sudaksinah | (76.15ab) The reading 'g is contained in several MSS also, but 'g' is the reading in the majority of the Groups-in 5. 7.11. sa 1. ka 1. 2. ba 1. 3. na 1. How the reading 'g' occurs in so many MSS and yersions is not certain. Is the reading ':' an emended form of '', or is 'g' a corrupt form of '' or is the reading 'yasto ' adopted for avoiding the haitus in 'yatra isto ' ? (iv) krtanjaliputo bhutva harah stotramudairayat || (B. 13). Here 'harah stotramudirayatu ' is the reading in the majority of the MSS. The reading 'a' is grammatically irregular. According to the rule of the Panini-grammar it should have the agama at and the form should be then udairayat as given in the reading of the Venkt, and some of the MSS. But the reading being irregular and also being contained in the majority of the MSS is older, and '' is an emended reading. Readings having grammatical aberrations occur in large number in the Vamana-Purana as in the Epics and the other Purinas. The above are only a few illustrations to show how such older irregular uses were being regularised in the MSS by the scribes or the readers. 12. Sometimes some MSS, and even the majority of MSS, contain grammatically wrong readings which are generally due to the ignorance of the scribes, and, therefore, are not really old grammatical aberrations or archaic forms. Such grammatically incorrect readings are adopted in their grammatically correct forms. E. g. :- (1) dhanisthadham satabhisa prosthapadamsa katrayam | (5.41ab) The reading prosthapada meaning here the prosthapada (or bhadrapada ) Naksata. is wrong. According to the rule of Panini gca' (VII. 3. 18)

Jan., 1967] CONSTITUTIONS OF THE VAMANA PURANA TEXT 175 the vrddhi in the uttara-pada (second member of the compound ''prosthapada ') is correct only when the word 'prosthapada ' means born under the Naksitra prosthapada . So in the sense of the Naksatra prosthapada (or prosthapada and also prosthapadya, if we add gran or syan here in svartha) prosthapada is the only correct form. MSS give here various readings such as prosthapadya, prausthapadya, prosthapada (de 10 only). prosthapada, prausthapada, Out of these readings the correct realing 'gqa' is acceptable. (2) brahmacayam sada satyam japam jnanam ca raksasa | niyamo dharmaveditvamarso dharmah pracaksate || 11. 22) Here is the reading in almost all the MSS also. But it is grammatically wrong; the correct form should be in passive as 'pracaksyate '. This correct reading 'pracaksyate ' is given only in two MSS 1. 10. The correct form '' is adopted. W (3) According to the rule an when occuring in the (AMS) :' middle of a pada should invariably be changed into the c, but in most of the MSS, except the Sarada MSS, a is not changed into the parasavarnam . Similarly according to the Sutras 'rasabhyam no nah samanapade ' (VIII. 4 1) etc. should be changed to T, and according to the Sutras (VIII. 3.57 etc) should be changed to . But in the MSS owing to the scribal whims these changes are not regular and uniformly made, conscquently such forms as 'a occur. The Rules of Sundhis according to the Sutras (VIII. 4.40), 'toli ' (VIII. 4.60), 'naschavyaprasan ' (VIII. 3.7), 'che ca ' (VI. 1.73) etc., also are often subjected to the scribal ignorance, carelessness or whims. And such readings, therefore, as 'brahman sailasrrmge ' (Vam. P. 8.3ca), 'bhagavan lolah ' (3.40c), 'bhagavan jatah ' (2.19c) etc. occur in the MSS. In such cases grammatically correct reedings only are adopted either from the MSS or by emendation. prose-formulas, 13. In the case of the texts of the containing uvaca, ucatuh, ucuh, also MSS are not uniform. Some give as 'pulastya uvaca, ' some 'pulastya u0 ' and some as 'pulastyah ' (generaly Sarada MSS). Even a particular MS also does not always observe the uniformity in this respect. Sometimes wrong readings as We

176 puranam- PURANA [Vol. IX, No. 11 ar sara' are given. In the constituted text, however, the uniformity is followed in this respect, and ':' etc. are given. 14. In course of time MSS often simplify an original difficult reading by rightly or wrongly changing its form or by substituting an easier synonym for it. A difficult reading (lactio difficilior), therefore, is held to be older, and hence it is preferrable according to the recognised principles of textual criticism. E.g. :- (1) krtvopavasamastamyam navamyam snanamacaret | masi margasire snanam rudra'ca dadhija smrta || (16.39) The mode of Rudra-worship to be performed in each month is described in this chapter. The general procedure is (a) on the Astami (b) snana with gomutra, gomaya, godugdha, gohavis etc., on the Navami, (c) Rudra-worship with some herbs or flowers, (d) c or incense, (e) naivedya, (f) daksirana and lastly (g) namaskara to a particular form or aspect of God Rudra, such as Virupaksa, Hiranyaksa,, Sthanu etc. Separate names or forms of Rudra are prescribed for each of the twelve months; similarly separate articles or dravyas to be used for performing these various angas of af in each month ruenre prescribed (16.30-65). ean. Here, in the sloka 16.39 (quoted above) this reading of 39cd -'masi margasire snanam rudrarca dadhija smrta ' - is given only in the Venkt. But here the article with which the T is to be performed is not mentioned; then the af or worship of Rudra is to be performed with some herb or flower, and not with af or curd which should be used for the TT as other similar articles- (32b), (35), (i. e. or cow-milk) (37a)-have been prescribed for the snana in the months of bhadrapada, prasvayuj and kartika respectively. So the Venkt. text of 39cd is clearly wrong. Now in the MSS 3 and 1 we have the following reading in its place:- 'masi margasire snanam dadhna'rca bhadraya smrta 1" The reading ' ' is given in the majority of the MSS, and so it is correct and also quite appropriate, but '' is the reading in 3 and 1 (1) only, which means that the

Jan., 1967] CONSTITUTIONS OF THE VAMANA PURANA TEXT 177 is to be performed with 'T'. The word 'T' in this context is synonym for the names of several herbs as Krana, ananta, aparajita, rasna, etc. (bhadra vistau nabhah sariti katphale | krsna'nantarasnasu, Hemacandra; Anekartha-Samgraha; also cf. M. W. Dictionary under the word 'T'). Any of these herbs is quite appropriate for the worship (f) of Rudra, and so the reading also seems to be a correct one: The reading c' has several variants in the MSS, such as rudraya, rudriya, rudraja, ksudraya, tatraya etc., which are the results of the efforts on the part of the scribes to simplyfy the lactio dificilior 'bhadraya '. In the variant 'ksudraya ' (given in ka 1.2 only) '' is also the name of several herbs, but cT is more preferrable, and it is mentioned in two distinct versions also. (2) srutva tadvacanam skando mayuram projbhaya tatksanat | bhadra pradaksinam padacari kattu turnam taro'bhyagat || (58.103) and The reading '' means '' or leaving, discarding ete. Many MSS read in its place. It is a difficult reading, but it seems to be the correct reading. It is formed from means 'pushing away' etc. (See M. W. Dictionary). It was not easily intelligible, and so it was changed to c etc. in the MSS. (3) mudgare vitathe jate pasamadaya vegavan | praciksepa naragryaya tam ca ciccheda dhamamjah || (0.22) Here prasamavidhya is the reading in ba 2 de 4.7.9 na 1. The word f here means 'swinging' (+√= 'to throw', 'to fling', 'to swing'--M. W. Dictionary. The use of +√ in the sense of "to swing" is referred to the Mahabh.). Compare also:- pasamavidhya gadam pragrhya ciksepa daityah sa jalesvaraya | ( 10.42cd) So the reading 'fa' being a lactio difficilior is preferrable here in 8.22 also. 15. In the domain of knowledge ideas often develop gradually, and knowledge generally becomes more and more improved and detailed in course of time. So the text containing less improved or less detailed ideas may be an earlier one, and 23

178 52 puranam - PURANA [Vol. IX, No. 1 hence it is preferrable to that which contains more improved or detailed ideas: E.g. :- brahma muraristripurantakari bhanuh sasi bhumisuto budhasca | gurusca sukrah saha bhanujena kurvantu sarve mama suprabhatam || ( 14.23) Here, in the Kashmirian MSS and also in de 1.3.10 'sanirahu- ketavah ' is the reading for 'saha bhanujena '. This reading 'sanirahuketavah ' seems to be a later improvement. The reading 'saha bhanujena ' is therefore, preferrable. 16. As far as possible break of symmetry (pralerama-bhanga) is avoided in settling a text. E.g.- prthvi sagandha sarasastatha''pah sasparsavayujvalanah sutejah | nabhah sasabdam mahata sahaiva yacchantu sarve mama suprabhatam || (14.26) Here in the reading 'sutejah ' (qualifying jvalanah ) there is a break of symmetry, as all the other qualities (Gunas) of the remaining Mahabhutas are compounded with (which becomes 'sa- ' in a compound). But in many MSS we have the reading 'satejah ' also, which is in symmetry with the other readings 'sagandha ' 'sarasah ' 'sasparsah and 'sasabdam ' and also with 'mahata sahaiva '. So the reading 'satejah ' is more suitable here than 'sutejah '. 17. Sometimes a text is omitted only in a few MSS, but the omission is justified by the preceding or the following text. In such case that omission is adhered to. E.g. :-- nasayamasa sa ( = katyayani ) yatra danavam mahisasuram | namaram raktabijam kacasau mahiso ka'sau katyayani namaram raktabijam ca ca tatha'nyan surakantakan || 38 narada uvaca nama raktabijadayasca ke | nama ya jaghne mahisasuram || 39 tatha'nyan kasvasau mahiso nama kaste jatasva kasvasau raktabijakhyo namarah kasya etadvistaratastata yathavad surakantakan | kasya sah || 10 catmajah | vaktumarhasi || 41 (Adh. 17)

Jan., 1967] CONSTITUTIONS OF THE VAMANA PURANA TEXT 179 Here the half sloka 39 ab is omitted in some MSS ( 3 9 1). The omission is justified by the following text of 39 cd-41, which is more in conformity with sl. 38. The text of 39 ab, here, is obviously superfluous. 18. Sometimes the MSS of a certain Group only supplies a missing text. In that case the additional text given in these MSS, supplying the lacuna, is adopted. E. g. :- ( 1 ) srabhojyah sutikah sando marjarakhu ca kukkutah | patitapaviddhanagnascandaladyadhamaca || ( 14.82 ) Here a list of those persons is given whose food is prohibited to be taken. Among these 'Marjara' (literally meaning 'a cat') and an 'Akhu' (literally meaning 'a mouse') are also included. The reading ' ' is given in the Venkt. edn. only. Most of the MSS read 'g' which is grammatically wrong. These abhojya persons are defined in the same order in Slokas 84-92. After the definition of Marjara in S 1. 86 Akhu is defined in Sl 87, and then immediately after it, Kukkuta is defined in $1 88 in the Venkt. edn. as well as in all the MSS except those of the Kashmirian Group (G 1). The Kashmirian MSS give the following additional sloka between the definitions of Akhu and Kukkuta - yah paresam hi marmani nikrntanniva nityam paraguranadvesi sa svana iti bhasate | kathyate || C Now on the basis of this additional text of G 1 MSS we can reconstruct the text 'marjarakhasca kukkutah ' ( 14.82b), as 'marjarakhu ‍T:'. Here the original reading--was wrongly written as 20 and thus the text has become corrupt in almost all the MSS. But the definition of '' or '' is missing in the printed text as well as in the MSS. Only the Kashmirian MSS have supplied this missing or lost text, which also leads to the construction of the correct text in Sl. 14.82b as shown above by suggesting the substitution of '' for ''. 20. In MSS, and specially in Devanagari MSS, and are often found wrongly written for a and respectively.

180 puranam - PURANA [Vol. IX, No. 1 (2) In Adh. 56 of the Vam.-P. the battle between Goddess Katyayani and two Assuras, Sumbha and Nisumbha, and their great Army-Commander, Raktabija, is narrated. When Raktabija with his large army attacked Goddess Candika (Katyayani) she roared the lion's roar and then from the different parts of her body the various Matrs (or Saktis) were produced. The following Matrs or various forms of the Goddess are described or mentioned in the Venkt-edn. S13-12 and in the MSS of the Vamana-Purana : 1. brahmani from the mouth of the Goddess, 2. mahesvari, 3. kaumari (the Saleti of Kumara or Skanda), 4. vaisnavi from her two arms, 5. varahi from her back, 6. narasimhi from her heart and 7. siva sivaduti from her body. The text of 56. 20 cd-22 ab of the Venkt edn. reads as follows:- - mahajala ksepahataprabhavan brahmi tatha'nyanasuramcakara || 20 c d mahesvari sulavidaritorasacakara dagdhamsca paramca vaisnavi | saktya kumari kulisena candi (?) tundena cakra ेna varaharupini || 1 nakhairvibhinnanapi narasimhi srattattahasairapi rudraduti | 22 a b In 21c all the MSS read 'caindri ' for the Venkt-reading 'candi '. So according to the MSS the ' is also included among the Matrs mentioned here. Now compare the similar text of the Devi-Mahatmya (9.37cd-40) of the Markandeya Purana :- srasuramstamstatha kali sivaduti tatha'paran || 37 cd kaumarisaktinibhinnah kecinnesumahasurah | brahmanimantraputena toyenanye nirakrtah || 38 mahesvari trisulena bhinnah petustatha'pare | varahitundaghatena keciccurnikrta bhuvi || 39 khanda khandam ca cakra ेna vaisnavya danavah krtah | vajrena caindri hastagravimuktena tatha'pare | 40 (Devi Mahatmya, Adh. 9) Here in 40c 'vajrena caindri ' is equal to the reading 'kulisena caindri ' (MSS) in 21c of the Vam. - P. above. Thus, the MSS-reading 'caindri ' is fully confirmed. But in the description of the birth of

Jan., 1967] CONSTITUTIONS OF THE VAMANA PURANA TEXT 181 the Matrs in 56.3-12a b of the Vam. P. no mention of the birth of 'aindri ' is made either in the Venkt.edn or in the MSS. So the text containing the mention of the birth of seems to be lost. But fortunately we have here the following additional text after S 1. 56.8ab in all the Kashmirian MSS ( GI) : vajrankusodyatakara jata - nana'lamkarabhusitah | gajendra prsthastha mahendri stanamandalat || So, according to this additional text in the Kashmirian MSS the Mat! mahendri (ie, aindri ) was born from the stana-mandala ('breast-orb') or the rounded breast of Goddess Candika. Thus, the lost text mentioning the birth of the Matr has been supplied by the Kashmirian MSS only, and it is adopted. ( 3 ) ghrtodad dvigunah kraunco dadhyo ( ? ) do dvigunastatah || 37 c d samudrad dvigunah sakah sakad dugdhabdhiruttamah | 38 a b (Vam.-P., Adh. 11) all the MSS, which also have been the Here, in 38a 'samudrad ' is the reading in leads to the supposition that there might reading 'samudrah ' (?) immediately preceding the reading samudrad ', so that this reading 'samudrad ' in 38a might have some propriety here. Now, in all the Kashmirian MSS ( GI) and also in de 2 we have the following text in place of the above text of Sl. 37cd ghrtodad dvigunah proktah krauncadvipo nisacara | tato'pi dvigunah proktah samudro dadhisamjnitah || This additional or the amplified text 'tato'pi dvigunah proktah samudro dadhisamjnitah ' of the Kashmirian and de 2 MSS for the shorter text 'dadhyodo dvigunastatah ' of the remaining MSS, immediately preceding the text 'samudrad dvigunah sakah ' (in 38a), supplies a lacuna and corroboartes the reading 'samudrad ' (in 38a) as quite appropriate. 19. Sometimes a text is found displaced in all the MSS, and so its sense does not fit there. In such case that displaced text is restored to its proper place required by the sense and •context, according to the principle 'pathakramadathamkramo baliyan '. ' 21 21. Cf. Nilakantha- " tatha ca pathakramadathamkramasya balavattvat 'prthivyam naimisam punyamantarikse ca puskaram ' ityatah param 'ye vasanti kuruksetre te vasanti trivistape ' ityayam grantho drastavyah | " (Comm. on Mbh. Vana-Parva, 83. 203ff.) 1

182 puranam - PURANA Compare the following text (11.31-42) :- yojananam pramanena pancasatkotirayata | jalopari mahiyam hi naurivaste sarijale ||| 31 sthanani dvipasamjnani krtavamsca prajapatih | tatra madhye ca krtavanjambudvipamiti srutam || 33 tallaksam yojananam ca pramanena nigadyate | tato jalanidhih ksaro bahyato dvigunah sthitah || 34 tasyapi dvigunah plakso bahyatah sampratisthitah | tatas tviksurasodaca bahyato valayakrtih || 35 dvigunah salmali dvipo dviguno'sya mahodadhih ( dheh ? ) | surodo dvigunastasya tasmacca dvigunah kusah || 36 ghrtodo dvigunascaiva kusadvipat prakirtitah | ghrtodad dvigunah kraunsvo dadhyo (?) do dvigunaratatah || 37* samudrad dvigunah sakah sakad dugdhabdhiruttamah | dviguna samsthito yatra sesaparyankago harih || 38 taramacca puskaradvipah svaduda tadanantaram | ete ca dvigunah sarve parasparamavasthitah || 39 catvarimsadimah kotyo laksasca navatih smrtah | yojananam raksasendra panca cati suvistrtah || 40 jambudvipata samarabhya yavatksirabdhirantatah | kotyascatasro laksanam dvau pancasaca raksasa || 41 puskaradvipamano'yam tavanante mahodadhih | laksamandakatahena samantadabhipuritam || 42 [Vol. IX., No. 1 Here, the names, position and the extent of the seven Dvipas and the seven Samudras which surround them are given. The extent of the Earth, as mentioned in S 1. 31 above, is fifty crores of yojanas. The Earth is divided into seven great Divisions called the Dvapas which are named as jambudvipa, plaksadvipa, salmalidvipa, kusadvipa, krauncadvipa, sakadvipa, and puskaradvipa . Each of these seven "Dvapas [* For 37ed the MSS of GI read :-- ghrtodad dvigunah proktah kauscadvipo nisacara | tato'pi dvigunah proktah samudro dadhisamjnitah || Sec 18.(3) of this article above]

Jan., 1967] CONSTITUTIONS OF THE VAMANA PURANA TEXT 183 is surrounded (as given above in Sl. 33 ff.) by a circular Ocean as follows:- 1. jambudvipa by 2. plaksadvipa by ksarajalanidhi (Ocean of salt) iksurasoda (Ocean of syrup) 3. salmalidvipa by suroda (Ocean of sura or wine) 4. kusadvipa by ghrtoda (Ocean of clarified butter) 5. krauncadvipa by dadhisamudra (Ocean of curd) 6. sakadvipa by dugdhabdhi (Ocean of milk) 7. puskaradvipa by svaduda (Ocean of fresh water). Thus, these seven Dvipas are separated from each other by seven distinct concentric circumambient Oceans. The extent of each Ocean (except of the arg) is mentioned here, in the Vamana-Purana, as double of the extent of the Dvipa surrounded by that Ocean; and similarly, the extent of each Dvipa (except that of the gc) is double of the puskaradvipa extent of the Ocean surrounded by that Dvipa. The extent of the gun, however, is not mentioned as double of the extent of the dugdhabdhi (or ksirabdhi ) which is surround by it (the puskaradvipa ), but it is, as given in $1 41cd above, four crores and fifty two lacs of yojanas. Similarly, the extent of the arg is not also double the extent of the g which is surrounded by this Ocean, but it (the svaduda ) is of the same extent as the puskaradvipa . The size 23 22. According to the other Puranas, however, the extent of each Occan is of the same size as that of the Dvipa which it surrounds, but the extent of a Dvipa is double the size of the Ocean which that Dvipa surrounds, and thus the latter Dvipa is double the extent of the former Dvipa, and similarly the latter Samudra is double the extent of the former Samudra; cf. Visnu. P. II. 4. 87 :- evam dvipah samudraisca sapta saptabhiravrtah | dvipascaiva samudrasca samanau dvigunau parau || 23. Here also the Vamana-Purana differs from the other Puranas, since according to the other Puranas, the extent of the g is also double the extent of the TT which it surrounds, and hence double the extent of the former Dvipa, (Visnu. P. II. 4. 71-72). But the extent of the which surrounds the go is mentioned in the Visnu-Purana also as equal to the extent of the gun (ibid., §1. 86).

184 puranam - PURANA [Vol. IX., No. 1 of the gac, the first Dvipa in the enumerated order, is one lac of yojanas and the extent of the last-mentioned us (which is also regarded as a Dvipa by the Katha-sarit-sagara-MW.) is also one lac of yojajanas, the same extent as that of the first Dvipa, the Jambu-Dvipa. And the total extent of all the seven Dvipas and the seven amudras plus the extent of Anda-Kataha (Shell of the Mundane Egg) should be equal to the extent of the Earth, which is fifty crores of yojanas. The above statement may be more clearly presented as follows:Dvipas and Samudras Extent in Yojanas 1. jambudvipa 1 lac 1,00,000 - - ksarasamudra 2 lacs 2,00,000 2. plaksadvipa 4 lacs 4,00,000 - iksarasoda 8 lacs 8,00,000 3. salmalidvipa 16 lacs 16,00,000 - -suroda 32 lacs 32,00,000 4. kusadvipa 64 lacs 64,00,000 -ghrtoda 128 lacs 5. krauncadvipa 256 lacs 1,28,00,000 2,56,00,000 - dadhisamudra 512 lacs 5,12,00,000 6. sakadvipa 1024 lacs 10,24,00,000 -grafer 2048 lacs 20,49,00,000 Total 4095 lacs 40,95,00,000 Thus, the total extent of the first six Dvipas and Samudras, from jambudvipa to dugdhandhi or ksirabdhi (Ocean of Milk), comes to forty-crores and ninety-five lacs of Yojanas, which exactly tallies whith their total extent given in Sl. 40-41ab above, as follows:- catvarimsadimah kotyo laksasca navatih smrtah | yojananam raksasendra jambudvipat panca catisuvistrtah || 40 samarabhya yavatksirabdhirantatah | 41 ab Now, add to this the extent of the remaining seventh Dvipa ( puskaradvipa ), and of the remaining seventh Samudra ( svaduda ) and also the extent of the andakataha surrounding the last Ocean svaduda, as given in $1. 41cd-42 above, which is as follows:-

Jan., 1967] CONSTITUTIONS OF THE VAMANA PURANA TEXT 185 kotyassrcatasro laksanam dvau pancasasca raksasa || 11 cd puskaradvipamano'yam tavanante laksamandakatahena mahodadhih | samantadabhipuritam || 42 (That is, the extent of ge is four crores and fifty-two lacs of Yojanas, and of the same extent is the last Ocean, and the space of one lac of Yojanas is filled all round by the andakataha ). Then the total extent of all the seven Dvipas and Samudras together with the extent of the Anda-Kataha comes to fifty crores of yojanas, as shown below:Total extent of the first six Dvipas and Samudra s ( from jambudvipa to ksirabdhi ) 40,95,00,000 Yojanas Extent of puskaradvipa Extent of svaduda Extent covered by andakataha 4,52,00,000 4,52,00,000 1,00,000 99 99 Grand Total 50,00,00,000 Yojanas Thus, this total extent, fifty-crores of yojana s, comes to be the same as the total expanse of the Earth, which is also mentioned as fifty crore yojanas 'qafe' (Sl. 31). According to this calculation, therefore, the text in sl. 41cd- 42 ab which gives the extent of go and g is quite correct But the text in 39cd - 'ete ca dvigunah sarve parasparamavasthitah ' must also refes to the extent of the puskaradvipa and of the svaduda, for both are mentioned here in the immediately preceding text of 39 ab as ' tasmacca puskaradvipah svadudastadanantaram '. But in fact, the text of the Line 39 cd - 'ete ca dvigunah sarve parasparamavasthitah ' - refers to the extent of the first six Dvipas and the first six Samudras only, as is clear from the line 41 ab - 'jambudvipat samarabhya yavatksirabdhirantatah . The text of the line 39 ab, therefore, occurs in a wrong place. Its proper place should actually be after the line 41 ab, as the next two lines 41 cd and 42 ab mention the extent of the gr and its surrounding Ocean T. Line 39 ab therefore, should be restored to its place, i. e. after 41 ab. 24

186 puranam - FURANA [Vol. IX., No. 1 The constituted text of the slokas from 37 cd to 42 cd of the Venkt. text, then, would be as follows :- ghrtodad dvigunah proktah krauncadvipo nisacara | tato'pi dvigunah proktah samudro dadhisamjnitah || 37 samudrad dvigunah sakah sakad dugdhabdhiruttamah | dviguno samsthito yatra sesaparyankago harih || 38 catvarimsadimah kotyo laksasca navatih smrtah | yojananam raksasendra panca cati suvistrtah || 39 yavatksirabdhirantatah | jambudvipat samarabhya tasmacca puskaradvipah svaddastadanantaram #1 10 kotyasrcatasro laksanam dvipancasacca raksasa 1 puskaradvopamano'yam tavadeva tathodadhih | * laksamandakatahena samantadabhipuritam # 41 [* * The reading 'tavadeva tathodadhih ' is given in MSS sa 1 kara . ka 1.2 de 2 ; de 7 reads 'tavadantam tathodadhih ; de 4 reads tavanantastathodadhih ] 20. Sometimes a reading, even if it occurs in all the MSS, has to be emended, so that the text containing that reading may give a correct and appropriate sense. (The emendation is indicated in the constituted text by an underline below the emended reading). Compare the following text of the Vam.-P., Adh. 11 (already quoted above) :- tasyapi dvigunah plakso bahyatah sampratisthitah | tatasutviksurasodaca bahyato valayakrtih || 35 || dvigunah salmalidvipo dviguno'sya mahodadhih | surodo dvigunastasya tasmacca dvigunah kusah || 3 || [Here, the reading 'mahodadhih ' is given in all the MSS also. ] The sense of this text is quite confusing. It may mean :- "The Dvipa plaksa is Either double of that (i. e. of the ksarajalanidhi, and is situated on the outside. Then 99 referred to in 34 cd), the iksurasoda is also [situated] outside in & circular form -Sl. 35. 'The salmalidvipa is double [of the extent of iksurasoda ]. The mahodadhi (which ?) is double of this (asya) (i. e. of the salmalidvipa ). "

Jan., 1967] CONSTITUTIONS OF THE VAMANA PURANA TEXT 187 The g is double of that (tasya) (i.e. of the lafa ?), and the kusadvipa is double of that (suroda ) - 51. 36. "The Or situated outside in a circulr form is double ( dvigunah, 36a) [of the plaksadvipa ]. 1. 35cd. The salmalidvipa is double (fd, 36b) of this (asya) (i. e. of the ). The Mahodadhi suroda is double of that (tasya) (i. e. of the salmalidvipa ), and kusadvipa is also double of that (tasmat, i. e. from suroda )". Here as we see the first sense is confusing. The second sense some-what more reasonable. But the defect in this latter sense also lies in the fact that the word Mahodadhi (af:) will have to be construed with the following word g as its 'case-in-apposition'; but this breaks the symmetry, as no such qualifying word has yet been used for the names of the other Oceans, mentioned before g, , as their 'case-in-apposition'. Moreover, the word 'afc:" if construed with 'g' will become quite redundant or superfluous. Hence the text is defective in every respect. But if we emend the word 'f' as ':' (sasthi or genetive), the sense of the whole text will be quite clear and appropriate. 'mahodadhih ' There are also some other readings which have also to be emended. 21. There are certain readings which may be considered as less than certain. Such readings may be of any of the following three categories :- (A) Readings which have almost equal manuscript evidence; any of them may be adopted, therefore, in the constituted text In the absence of any deciding factor. (B) A reading which, even if it is contained in the majority or even in all the MSS, does not give an appropriate and correct sense, but against which no internal or external evidence is available, and which, in the absence of some such sound grounds, cannot also be emended. (C) A reading which is contained in only a single MS or only in a few MSS, but which gives a suitable and correct sense,

188 puranam - PURANA [Vol. IX., No. 1 while the other readings, though contained in a larger number of MSS or Groups, do not have any proper sense; such reading is considered as less than certain, if it cannot be ascertained whether this reading is genuine or an emended one; if it is proved to be an mended reading then it is discarded. All such less-than-certain readings, when adopted in the constituted text, are marked with a wavy line (). A few illustrations of such readings are given below:- (A)- (1) sadhuvadam tada cakru daityadanavayuthapah || (10.22 cd) Here sadhuvadam daduhrsta [:] is also the reading for 'sadhuvadam tada cakra But both of these two readings have nearly equal manuscriptevidence. So, any of them may be adopted in the constituted text, if there is not any deciding factor in favour of any one of them. (2) jalesam tadayamasa mudgaraivajrasannibhaih || ( 10. 26cd) Here the MSS contain several different readings, in the transposed order, for 'vajrasannibhaih ', such as kutharai ( kutharai, sa 1 ka 2 ) rapi mudgaraih ( ka 1), kunapairapi ( ba 1. 2), kanapairapi (ba 3 de 9), and 'parighairapi ' (na 1). The reading ':' is given only in the Devanagari MSS of GI. kuthara, kunapa, karanapa, parigha are all weapons of war, and any of them may be meant here. There is no deciding factor of sense, context etc. in favour of any one of them. So any reading which is adopted will be considered as less than certain. (3) raibhyo mariciscyavano ribhusca kurvantu sarve mama suprabhatam || ( 16. 17) The reading 'fy: is also contained in Devanagari MSS. But it is wrongly spelt here. 1 and in several Its correct form is : as is given in 4, for sometimes is found wrongly written as fe in manuscripts. Other readings are ':' (a 11), fig: (de 7), nisadduh (sa 1. ka 1. 2), 'rusadguh ' (ba 1.3) ; etc. Any of these reading may fit here. (B)- (1) esa kramasta gadito nabhadau svapatam mune | svapatsu tatra devesu pravrtkalah samayayau || ( 16.17 ) EX

1 10:1 Jan., 1967] CONSTITUTIONS OF THE VAMANA PURANA TEXT 189 The reading ccc' in past-perfect is not perhaps correct, for the context does not require the past tense here. But it is given in all the MSS except 4 which reads '' (perhaps an emended reading). So if the reading adopted on the basis of the manuscript-evidence, considered as less than certain. (2) gorocanayuktagudena caiva devam samalabhya ca pujayettu | is to be will be But the reading 'bhavantam ' And if there is no other priyasva dono'smi bhavamstvamosa macchokanasam prakurusva yogyam || ( 16. 46) The reading 'priyasva dono'smi bhavantamisa ' is given for 'priyasva dino'smi bhavamstvamisa ' in almost all the MSS. here is not justified by sense or context. evidence or ground against it, it is, then, to be adopted in the constituted text on the ground of the manuscript-evidence. It is not yet certain how this reading ' occurs here. It needs support from some external source also, which is not yet available. So the reading remains less than certain. (C)- (1 ) dhanyesu salidvipadesu viprascatuspade gausca yatha mrgendrah ( 12.50 ab) sense. The reading 'catuspade gausca yatha mrgendrah ' does not give any proper The other readings are 'catuspadesveva yatha mrgendrah ' and 'catuspadem gauh svapadam mrgendrah ". The latter reading ('catuspade gauh svapadam mrgendrah ) is contained in the MS 2 only, and it has an appropriate sense also, and therefore is decidedly a better reading. But it is not certain whether this reading is. genuine or emended. So it is less than certain. ( 2 ) tatra snatva ca sampujya putram 3 hyaditi devamataram | janayate suram sarvadosavivarjitam || ( 34-13 ) Here almost all the MSS except de 7 read 'pitva ca ' for 'sampujya ' in 13a above, 7 reads 'gra'. Now, the reading at a has no sense here. The reading gras given in a 7 is quite appropriate, but it is just possible that it may be an emended reading here. So if the reading 'gr a' is adopted on the ground of its suitability, it will be a less-than-certain reading.

190 puranam - PURANA [Vol. IX., No. 1 We have discussed and illustrated here the several principles adopted for the constitution of the Vamana-Purana text. But the application of these and the like principles to the settlement of a text requires patience, right judgement and hard labour on part of the editor; and even then it cannot be said with certainty that the text settled and adopted is the only correct or the proper one, for it may not suit the sense of what is right and what is wrong, of the critical reader. But the Critical Apparatus appended to a Critical Text provides an ample scope to the critical reader for the exercise of his own free judgement and choice. APPENDIX [The footnotes which were not so essential as to be necessarily given in the respective pages, or which were too long to be accomodated there, are given in this Appendix.]. 1. EDITIONS OF THE WORKS REFERRED TO Puranas- Agni-Purana Bhagavata-Purana Bhavisya-Purana Garuda-Purana Markandeya-Purana Matsya-Purana Naradiya-Purana Padma-Purana Skanda-Purana Vamana Purana Vayu-Purana Visnu-Purana Other works Mahabharata Vana Parva of the Mahabharata (with Nilakantha's commentary) Harivamsa Anandasram, Poona Gita Press, Gorakhpur Venktosvara Press, Bombay Pandita-Pusta kalaya, Kasi Mor edn. Calcutta Anandasram Sanskrit Series, Poona Venkt. Press, Bombay Anandasram S. S., Poona Venkt. Press, Bombay Venkt. Press, Bombay Mor, Calcutta Gita Press, Gorakhpur Critical edition, Poona Chitrasala-Press, Poon Chitrasala Press, Poona 916

Jan., 1967] CONSTITUTIONS OF THE VAMANA PURANA TEXT 191 Varahi-Samhita Yajnavalkya-Smrti (with Mitaksara and Vira-mitrodaya com mentaries Chaukhamba, Banaras 2. EXTENT OF A PURANA-TEXT According to the Vayu-Purana the extent of a Purana-text is 12,000 Slokas ('evam dvadasasahasram puragam kavayo viduh ' 32. 66 cl). The Bhavisya also says - sarvanyeva puranani samjneyani nararsabha | dvadasaiva sahasrani proktaniha manisibhih || ' (I. 1. 103 ). If this is so, then the total extent of all the 18 Puranas was formerly about two lacs of slokas, which later on went on increasing, as the Bhavisya further remarkes- punavrddhi gataniha akhyanaivividhairnrpa ' (Phid 104) and thus came to be of the four lues of Slokes- caturlaksamidam proktam vyasenadbhutakarmana ' (Matsya 53.57). 4. RIVERS RISING FROM THE MALAYA krtamala tamraparni puspaja hyatpalavati | | malayaprasuta nadyastah sarvah sitajalah subhah | (Matsya, 114. 30 ) Cf. also Vayu (45.105), Markandeya ( 57.27), etc., where these same four rivers, with some variants, are mentioned as rising from the Malaya; only puspaja (1.1. puskala, prthula ) is mentioned in place of the as of the Vam.-P., but see fn 6. The Vam.-P. has added two more rivers to this list-at and sudama ; but in the Matsya the suni (v 1. sini ) is mentioned as rising from the Rsyavanta (or Rksa-), and the TT is not found mentioned elsewhere. 5. RIVERS RISING FROM THE SUKTIMAT kasika sukumari ca mandaga mandavahini | krpa ca pasini caiva suktimantatmajastu tah || (Matsya, 114.32) (v. 1. rsika for kasika ; palasini for pasini ). These same six rivers are mentioned as rising from the Suktimat in the Vayu (45.106) and Mark. (57.29f) also; but for kasika Vayu reads rsika and the Mark reads rsikulya, while for pasini both read palasini .

192 puranam - PURANA } [Vol. IX., No. 1 7. MAHABHARATA LACUNA SUPPLIED AND TEXT IMPROVED BY VAMANA It is not only the Mahabharata which supplies some missing text of the Vamana, but the Vamana also, as if in turn, supplies sometimes a missing text of the Mahabharata. E.g. :Mahabharata- tato gaccheta rajendra brahmanah sthanamuttamam | brahmodumbaramityeva prakasa bhuvi bharata || 58 tatra saptarsikundesu snatasya kurupungava | kedare caiva rajendra kapisthala mahatmanah || 59 brahmanamabhigamyatha sucih sarvapapavisuddhatma brahmalokam prayatamanasah | prapadyate || 60 (III. 81 ) Here the text of 59 ab is perhaps incomplete in sense and construction ; the Padma Puran ( adi. Kh., Adh. 26?) also has the same text as the Mbh. But compare the Vam. - P. :Vamana-Purana- 1 tato gacchecca viprendra brahmanah sthanamuttamam | brahmodumbaramityevam sarvalokesu visrutam || 7 tatra brahmarsi (v. ]. saptarsi ) kundesu snatasya dvijasattamah | saptarsinam prasadena saptasomaphalam labhet (bhavet ? ) | 8 (Adh. 36) ) Here we see that the line 36.8cd completes the text of the preceding line 8ab, which is the same as III. 81.59ab of the Mbh, as quoted above. Compare also the Naradiya-P. II. 65.62ab.- tatra brahmanikundesu snatah somaphalam labhet | In a few places the Vam.-P. also improves the text of the Mbh; e. g. :- abhigamya sthalim tasya ( = vyasasya ) gosahasraphalam labhet || (Mbh. III. 81.82cd ) * abhigamya sthalim tasya putrasokam na vindati || (Vam. P., 36.61cd)

Jan., 1967] CONSTITUTIONS OF THE VAMANA PURANA TEXT 193 If we read with these lines of the Mbh. and the Vam.-P. the preceding lines of both the works, narrating how at this place (vyasasthali ) Vyasa was overwhelmed with sorrow for the loss of his son, and then consoled by the gods, we will at once know that the reading of the Vam.-P. 'ga farafa' is better than that of the Mbb.'' (no variants,). 9. CORRUPTION OF TEXT DURING TRANSMISSION ILLUSTRATED tato vinasanam gacchenniyato niyatasanah | gacchatyantarhita yatra meruprsthe sarasvati || ( Vana -Parva, 82.11 ) The The Sarasvati did not disappear at the Meru, nor is the place Vinasana on the Meru. So the reading is wrong. The Agni-Purana (109.13) also reads Egg in this context. original reading was 'g' (in the desert), as is given in the Critical Edition of the Mbh. (III. 80. 118 d), although some of its MSS, even of different versions, read :'. 12. BHUVANA-KOSA CHAPTERS IN PURANAS AND THE POSITIONS OF THE NINE VARSAs OF JAMBU-DVIPA. The Bhuvana-Kosa Chapters are contained also in several other Puranas; such as Agni, Bhagawata, Brahma, Brahmanda, Kurma, Linga, Padma, Siva, and Devi Bhag. These chapters state the positions of the nine Vargas of Jambu-Dvipa as situated in different Directions, and also their association with the different Varga (or Maryada-) Parvatas. The Bhagavata-Purana statement is very clear, which is as follows:- 300 yasmin navavarsarana navayojana sahasrayamani || uttarottarenelakrtam nilah svetah srngavaniti trayo ramyakahiranmayakurunam varsanam maryadagirayah pragayata ubhayatah ksarodavadhayo| evam daksinenelakrtam nisadho hemakuto himalaya iti pragayata yatha niladayo'yutayojanotsedha harivarsa- kimpurusa- bharatanam yathasamkhyam || tathaivelavrtamaparena purvena ca malyavad- gandhamadanau (gandhamadanamalyavantau ?) sranilanisadhayatau " ketumalabhadrasvayoh simanam vidadhate || 25 (Bhag.-P., V. 16. 8-10)

194 puranam - PURANA [Vol. IX., No. 1 19. USE OF prathama FOR dvitiya FURTHER ILLUSTRATED. (1) tato'gnayastribhitrai duhsamam samavaiksata | ( 5. 25 ab) (2) sapta sarasvatah pradanmatarascaturo'dbhutah | (57.92 cd) (3) etani bhutani garanamsca mataro drstva mahatma vinatatanujah | ibid. 102cd) (4) devanam mataro drstva devapatnyastathaiva ca | skannam sukra maharaja brahmanah paramesthinah || (Matsya, Jivananda edn., 191. 7) As alredy stated, this use of prathama for dvitiya is perhaps due to the influence of Prakrta. See for detailed discussion my article referred to in fn. 18 of this article.