Chapter 1—First Varga

VARGA I. THE manuscript begins with which is evidently the H of the scribe and indicates that he must have been a follower of Jainism. It is therefore not inserted here in the text. The beginning of Udbhata's work as presented here does not contain any . Nor does the end also contain any indication that the work is finished. This, especially the former, is a very curious and unusual feature; and one may even draw the inference from it that the existing work is a fragment of some greater work which extended before the beginning and after the end of this little treatise. But Induraja at least considers this work as a whole and he has never mentioned nor even hinted at any greater work of which this fragment formed a part. The Ms. of udbhatalankaravivrti procured from the Government Oriental Manuscript Library, Madras, however, has the following verse at the beginning of the text :- sarvatra vikasvarata pramodada bhavati nayamekantam | mukuloccayanam tena kvacidapi kartavyatam dhatte || It is a beautiful arya and appears to be quite in keeping with the style and sentiment of Udbhata. It evidently serves the purpose of a in the form of vastunirdesa according to the rule of poetics asirnamaskriya vastunirdeso afy, and gives a fitting reason why a collection of vapi tanmukham, alankaras ought to be made even at this stage, though. they are like buds still in an undeveloped state. The verse is an elegant instance of every appropriately supported by most proper for the occasion. P. 1, L. 1-ga etc. The whole work is divided into six as or chapters. The first chapter begins with the enumeration of four sabdalamkara s beginning with punaruktavadabhasa and four of the beginning with . M. reads punaruktavadabhasah . 1 [K. s. S.]

Kavyalankara-sara-sangraha The author of the faaf, whoever he may be, gives a ourious reason why the alankaras are treated and presented here by Udbhata in so many groups. He says, it is to show the limited vision (knowledge) of the old Alankarikas: vargairvagairalankaropadanam cirantanalankarakrta matpadarsitam prakatayitum | Perhaps this remark of the commentator may have some reference to the undeveloped state of the science of alankaras alluded to in the first verse given above from the southern manuscript. P. I, L. 4 -- ityetaeva ... udahrtah | These are the only alankaras of speech mentioned by some. Here the author appears to refer to some of the ancient alankarikas (like Bharata) who have mentioned some four, six or eight alankaras only. Udbhata thus seems to be aware of the historical development in alankarasastra. Bharata mentions only four alankaras, viz:- upama dipakam caiva yamakam rupakam tatha | kavyasyaite hyalamkarascatvarah parikirtitah || - bharatanatyasastram - 16 . 41 . It is particularly to be noted here that Udbhata mentions • anuprasa and treats i very elaborately but altogether ignores. yamaka ; while Bharata omits anuprasa and includes yamaka among his principal alankaras. Dandin also lays great stress on anuprasa and mentions yamaka with & disqualification. Dandin's kavyadarsa 1-61 . avrttimeva samghatagocaram yamakam viduh | tattu naikantamadhuramatah pascadvidhasyate || Cf. Mammata too views with a kind of disfavour; cf. a tatkavyantargadubhutamiti nasya bhedalaksanam krtam | ka . pra . 9, p. 504. Modern alankarikas acknowledge the claims of anuprasa and yamaka in equal proportion. Rudrata in his gives more prominence to and devotes a full chapter of 60 verses to its treatment. (Vide Rudrata's Kavyalankara, Adhyaya, 3.) P. 1. L. 4 vacam | The vivrtikara explains vacam thus:- vaktiti ucyata iti ca vak sabdothasca which means 'of word, of sense and of both.' tatradyah ubhayasya sabdasyarthasyaca | tatastrayah sabdasya | sesastvarthasyalankarah |

Notes. 3 P. I, L. 14 - atranuprase rupake ... caturiticoktam | The words trigha attached to anuprasa and catuh to rupaka in the very threshold are intended to remove difference of opinion about their subdivisions, for Bhamaha has only two kinds of ME and two kinds of rupaka . Thus Induraja, the vivrtikara says that the mention of the number of subdivisions of anuprasa and is made in order to suggest that these subdivisions should be prominently understood as they are seen in many alankara works. He remarks - anuprasatridheti rupakam caturiti ca pasca- kartavyamapi vibhagapratipadanam yad uddesa eva krtam tad bahvalamkara - [ grantha- ]drstasaprabheda- lamkarastakamatra darsitamiti mukhyata evasyopadeyata dhvananaya | ata eva kaisciditi bahuvacanam | bhamaho hi dvividham rupakam canuprasam cavadit | para ityadivaksyamana bahuvacana- syayamevasayo jneyah | ... 6. P. 2, L. 1 - rupakam caturityatra bhavatityarthah | The word catuh is formed by [samkhyayah kriyabhyavrttiganane ] dvitricaturbhyah suc | pa . 5|4|17 and 5/4/18, and means "four times," not of four sorts." The termination is suc ( s ) where there is kriyabhyavrtti ito. repetition of action. The nature of 64 (Metaphor) appears permeating the four divisions of it, and thus the existence of is repeated four times in them. So rupakam catuh means rupaka ( Metaphor ) severally exists four times (in its four divisions) i. e. has four sorts. In tridha the termination is dha which means prakara (pa . 53342 ); but in catuh the termination is suc which means times or repetition. Hence arises the apparent difficulty which Induraja has thus explained very ably. vivrtikara says - rupakajatyupalaksitavyaktigatam pratiyamanam bhavati | kriyabhyavrtti- ganane krtvasujityatra dvitricaturbhyahsujiti suc | tena caturbhedam rupakamiti tatparyarthah || P. 2, L. 7 - punaruktabhasam | In the enumeration at the beginning Udbhata names the figure as punaruktavadabhasam, while in the definition he calls it punaruktabhasam for the exigencies of metre, it seems. The author puts it in neuter gender in both places as he takes it with reference to poetry ( kavya ) which it embellishes. The real name accepted by all the alankarikas (except Ruyyaka ) is punaruktavadabhasah . punaruktabhasam is the laksyanirdesa and abhinnavastvirupapadam is the laksana or definition, which is a very long and awkward compound. Induraja explains it as a of five padas including When dissolved in the two agafes of two padas each. bahuvrihi s

4 6 Kavyalankara sara sangraha. way of the commentator the whole compound means: [ & kavya ] where two padas (words) having different verbal forms, appear to be one thing i. e. having one meaning. The word vastu in this compound means 'sense' or 'artha ' and 'rupa ' means the literal form' of the word. Where the literal form of two words is different but their sense is apparently the same this alankara occurs. The vivrtikara has a rather lengthy and technical discussion on the definition of punaruktavadabhasa . abhinnam vastu vacyam yayoste abhinnavastuni te iva bhasete bhinnarupe pade yatreti kecidyacacaksire | tanutrapurapyanyo sat karikunjararudhira raktakharanakharah | tejodhama mahatprthumahasamindro harijisnuh || ityadinamasamgraho ma bhuditi samase samkhya na vivaksita ityanye | kalagaladisabdasya padatvam nasti| ' nalinatamarasapadmasampratam gaganamanirambaramalamkaroti yat || ityadi va yate iti abhinavastvivodbhasate bhinnarupam padam padantaracchabdadva, tasmadekasmadanekasma dyatretyapare | samasyati yenabhinnabhidheyam tasya abhidheyasambandhongikaryah | siddhe hi tasyabhi- dheyavatve tadabhinnabhidheyatvamasya sakyate vaktum | na catra kalagaladisabdasya abhidheya- yogah | abhyupagate va natra sitikanthadipadamabhinnabhidheyamiva abhinnabhidheyabheva | srute ca sitikanda ityadipade kalagaladisabdasyabhinnabhidheyatvabhramo na tvasrute | tasma- dvinnabhidheyamapyanabhidheyamapi va abhinnamiva bhatyabhinnavastuna iva padantaraccha dadva, tasmadekasmadena kasmadva tasmadbhinnarupam yatretyeke vivrnvate | 'tadeva suratakutam ' ityadi yadyapi nabhati punaruktih tathapi ' cakasatyanganaramah ' ityadau tasyah prati- bhasat padamarthapratitikrt natu suptinantamevetitulyakaksyadrsah | evam ca padamiti padaniti va samasaniyam | Here abhiseya = vastu = object, sense. The text of the Ms. is incorrect and sometimes unintelligible. It is hoped it has been made amenable to consistent interpretation by ramoval of errors. P. 2, L. 14- yadyapi punaruktavada ... vaicitryapradarsanartham | In the cau meration the figure is punaruktavadabhasa i. e. containing vat . But here it is without a. The aa is therefore to be understood here. The author has thus made use of both the terms in order to shew that the figure goes by [two] different names. P. 2, LL. 16-20 kacitkhalu ... tasyalamkarah | Sometimes & term is used having some suggested meaning attached to it; and at other times it is used only with the expressed meaning. In the enumeration, the name of the figure gives out the meaning of similarity by the taddhita termination vat, while here in the definition the notion of similarity is

Notes. brought out by force of suggestion. Thus the line punarukta- bhasamabhinna etc. gives the definition of poetry that contains punaruktabhasa having with it by suggestion the sense of the termination vat . On this the vivrtikara has this remark: kacinnirdisyamanarthi- nvayini kacidartham samarthyaksiptarthanvayinyabhiva bhavatiti darsanayoddese vati krtvanena prayunka ! etaddarsanam cananvayopama digamyamanasyapahnutirupakadirvacyasya bhedena bhidyata iti darsanartham | P. 2, LL. 20-26-atralamkarya ... nirdesom yukta eva | Here the figure consists of the words which appear to be repeated. It is said to be an alankara (adornment) as it is the property of poetry which is the thing to be adorned. The fruit (effect) of this round about saying is to shew that a group of words containing only an apparent] repetition is on alankara. The nature of an alankara is well defined if it is described as a dependent of poetry which it beautifies. For when it is viewed with reference to itself as separate, it no longer serves the purpose of an alankara, in the same way as necklaces, armlets, bracelets, etc. when placed in a box. So the description in the text of the igure punaruktabhasa as subordinate to poatry in order to shew its real nature, is proper . punaruktabhasa is here intended as an attribute of kavyam as it is mentioned with aeuter gender. vivrtikara has a similar remark here:- kavyasyalamkaryatvadyadyapi punaruktavadabhasanam padasthamakarastathapi sayu ...... labdhva alamkaravyapadesah | haradivilaksanataya kavyaparantryeneha alamkarateti pradarsanaya kavyamasritya punaruktabhasamiti sa napumsakah samskarah | The Ms. presents a lacuna here. P. 3, L. 2 - tadaprabhrti etc. The subject sitikanta in the stanza is to be taken with sa devah in the next stanza sa devo divasanninye etc. and is to be connected with the predicate divasanninye . M reads tatah prabhrti, and interprets as tatah kamadahadarabhya | nagakunjaro hastivarah | sitirnilah | P. 3, LL. 6-7- kalavasena nivartamana .. pida yasya sah | The Nirnaya Sagara edition makes here a needless and quite faulty emendation of the text. It has kalavasena nivartamana (ya ) sati ( tasyah ) viyogena janito yah sokavahnistadbhava pida yasya | This emendation really

6 Kavyalankara sara-sangraha makes no relevant sense. If the ad is a (returning) how can there be fc (separation) from her? And besides how can I return by the influence of time? But all this argument is not even wanted. From the context of the story and the construction of the sentence we can obviously see that thee is described as returning (a) through influence of time. The N. S. edition has needlessly separated sati from viyogena and added ya and tasyah before and after it. The is here quite explicit. He writes kalena galanti satisokanalavyatha yasya sah | P. 3, LL. 8-14 - atra nagakunjara ... punaruktabhasatvam | The words naga kuksara and sitikantha, kalagala appear to be repeated having appa rently the same meaning at first sight. But on further consideration of the sense of the whole verse that apparent repetition vanishes and the real difference in the meaning is manifest. Thus T is an elephant and is 'the best of its kind.' Similarly the possibility of repetition of the word sitikantha and kalagala as giving at first sight the meaning of 'the black-necked is set aside by properly construing the line and taking the first word of the compound to be meaning 'disappearing by time,' as TO DI satisokanalasya vyatha yasya . D punaruktavadabhasa is an ubhayalankara ie of word and sense both; of sense in as much as the alankara is not violated even when the words and are substituted by their synonyms, and of word as it vanishes when the words and are substituted by their synonyms. punarukta P. 3, LL. 15-19- nanu chekanu ... baliyasttrat | Here there is a discrepancy. The example of is placed (by the author) after the definition of . How is it then that the example is given here just after the definition of gais discussed? The answer is:-It is cited here because the meaning of the definition is not properly grasped unless an example is given to illucidate its application. Thus there is no discrepancy in mentioning the example just after the definition of gagara. In the same way hereafter no fault should be found in commenting upon the text in accordance with the sequence of the

Notes. subject matter and reversing the order of the work itself where necessary. For the sequence of the subject matter is always more powerful than the order of the reading of the text. According to our present knowledge of the alankara literature, Udbhata must be regarded as the originator of this alaikara. punaruktavadabhasa is both a sabdalamkara and an arthalamkara . The test that is applied in distinguishing & sabdalamkara from arthalamkara is its parivrttisahatva or parivrtya sahatva . If an alankara is solely dependent on the form of words so that the alankara is lost if the words are substitued by other words of similar meaning i. e. by synonyms, then the alaikara is a sabdalamkara ; and when such substitution of synonyms does not destroy the alankara, then it is an arthalamkara See kavyaprakasa X. pp. 767-768, also IX. pp. 534-539. Following this test Mammata calls punaruktavadabhasa a. sabdalamkara and ubhayalamkara both. Here naga kunjara is ubhayalamkara ; siti- kantha-kalagalat should also be regarded as ubhayalamkara . The words naga and sitikantha suffer substitution but the others do not. If we substitute a synonym for kunjara Or kalagalat the alankara will be lost. If we have pumgava for kunjara (pumgavarsa bhakunjarah | simhasardulanagadyah pumsi srestharthagocarah ) and samayanasyat for kalagalat then the apparent repetition will be evidently lost. Neither Udbhata nor Induraja makes any remarks about punaruktavadabhasa being a sabdalamkara or ubhayalamkara . But the vivrtikara pithily discusses the whole question: - nagasitikanthaparyayo- padanepyalamkaro na hiyate ityartho'lamkaryah | kunjarakalagalacchavdaparyayopadane tu nala- karo bhavatiti sabdo'lamkaryah ityubhayalambano'yamalamkarah | tasya rajnah sumanaso vibudhah parsvavartinah ' ityadirabhangah | ' arivadhadehasarirah ' ityadisca sabhangah | sa sabdalamkarah | This discussion obviously follows the treatment of Mammata. The inference that this must have come after Mammata would not be out of place. P. 3, L. 21 - chekanuprasastu dvayoh etc. | chekanuprasa consists in a charming expression of groups of similar vowels and consonants, two at a time. One group of letters is to be charmingly similar to the next following and only two such groups similar to each other are to occur at a time. But in all there are to be many such couples of similar

groups. Kanyalankara sara sangraha, The vivrtikara has - dvayordvayorvarnayoh susthu adure sadrsocarane kriyamane chekanupraso bhavati | dvayordvayorvayoscatra saiyokteralamkarata | P. 3-4, L. 25- dvayordvayoriti svarthe ... drastavyam | The repetition uyordvayoh is made by the rule svarthe avadharyamane anekasmin vartika on 'prakare gunavacanasya | pa . 811|12 . It means when the (parts of the) thing itself are determined out of many, the indicative word for the part is to be repeated. Here by the mention of the part, the desire of pervasion ( vipsa ) of all the parts is excluded and so all possible combinations of letters are not to be mentioned but some only. For pervas sion is meant where there is entirety, but it is excluded here by the word svartha . The word avadharyamana in the Vartika shows that only two such groups of letters at a time are to be made similar and not three at a time. The word anekasmina indicates that the repetition of such combinations made several times and not only once gives rise to chekanuprasa . Thus a series of groups of similar letters mentioned together twice constitutes the figure chekanuprasa . See Patanjali's mahabhasya with kaiyata VIII. 1, 1, for the explanation of the vartika . . The fa summarises the whole argument and puts it thus : - svarthe'vadharyamane'nekasminniti dvayordvayorityatra dvirvacanam | svarthasabdena saya nirastatvanna yavanto'tra dvikastesam sarvesameva sadrsatvam karyam | sakalye hi vipsa bhavati | avadharyamanagrahanadvayorevanekasminniti vacanaccasakrt | tenanekasyaksarasya dvirucaranam chekanuprasah | P. 4, LL. 6-16-chekah | This is explained in two ways:- (1) Birds that are addicted to their homes and do not go out at all are called :. Such birds by their lonely and undisturbed mode of living develop a charming voice and from its similarity the anuprasa is called chekanuprasa . grhasaktah paksimrgaschekaste grhyakascate ' ityamarah | (2) chekah means vidagdhah learned or clever men. This a might have been called chekanuprasa because it was especially liked by chekah, acoomplished men. 'The word anuprasa is also (anvartha ) used in its literal sense, since it consists in ( prasa ) a convenient arrangement, in & kavya, of similar letters or groups of letters congenial (3) with the sentiment, so as to embellish it rasadyanugunatvena prakrsto nyasa ityanuprasah |

Notes. 9 The whole of vivrtikara 's comment on chekanuprasa follows Induraja. P. 4, L. 13 Here the is Siva. "The God (Siva) passed his days in the valley of the Lord of mountains, served by his attendants who were foremost in the assembly of the great." -assembly. and In this verse we see the following couples of similar groups: ( 1 ) sa deva and divasa, ( 2 ) indra and kandara, (3) garistha and gosthi, ( 4 ) prathama and pramatha and lastly (5) pari ard upasa . Thus here there is dvayordvayoh susadrsokti . Only two groups of letters similar to each other are to occur. But many such pairs of similar groups (as in the present example) must be existing. Thus we have here one pair of similar groups of letters. But there ought to be many such pairs to produce a charm sufficient to call it an alankara. Here we see that there are five such pairs consisting of similar letters which produce a charm. Some read instead of ninye tasmin In that case ninye'nyasmin may also be included among the groups of similar letters. P. 4, L. 22 - anuprasah | sa ca trividho vrttisamsrayat | anuprasa or alliteration is of three kinds owing to the three styles of composition in which it occurs. 66 P. 4, LL. 24-25-37...The textual order of this Karika is after the definitions of the three. But it has been brought here to acquaint the reader with the nature of first. The meaning: anuprasa The separate grouping of similar consonants in the three styles of composition (suitable to the different sentiments) the poets always call alliteration()." P.5, LL. 1-2- atastastavadvrttayo ... gramyatvabhedat | Therefore the styles of composition capable of manifesting different sentiments and emotions are described first. z will then be easily understood from them. Those are three:, upanagarika and gramya . P.5, LL. 3-4-parusa | sasabhyam rephasamyoga ... | " The style is called rough () when it bristles with letters i and f and conjuncts with [such as, etc.] and, and ." The alliteration in this kind of style is called ogone, 2 [x. S. 6.]

10 Kavyalankara-sara sangraha. The fag has something more to say than Induraja on this Karika -- nakaravarjamiha tavargastasya parusatvabhavat | parusavarnarabdhatvaca Karika-acie parusavrttirvartanam rasavisayo vyaparah atra parusanuprasah | sarupavyanjananyasarupanuprasaparya- locanaya va senaiva senaiva vetyadyavadharanam vadanti | satrayoraikyam nasti sarupatvam tvastyaiveti sasabhyamevetyavadharanamityanye misrah | kavindrai racana alpadirghah karya ityanyatrokteriha ca keliloletyadau tatha darsanat | bahubhih sasadibhih sarvairapi ca karyeti yuktataram | evamanyata udaharanam | P. 5, L. 10-tatra toyasaya 1 "There the autumnal season shone with lakes full of full-blown lotuses and with the aspects of quarters appearing brownish with awns of rice." Here the repetition of the letter gives rise to parusanuprasa . The vivrtikara explains the compound vyakositakusesaya 88 vikasitani padmani yasyam sa which evidently ignores the causal form. P. 5, LL. 15-17 - upanagarika | sarupasamyoga | The style is 15-17-TAZIT upanagarika (ie polished ) when it consists of conjuncts of the same consonants [ such as ka, pa ca etc. ] and of conjuncts with nasals at the beginning and one of the first 25 letters from ku to m at the end [such as nka, ca, nta, nta, mpa, etc.]. 6 P. 5, LL. 21-25 - esa khalu mupanagarikanuprasah | The style is called 3car because it resembles an accomplished lady of town ( nagarika ). The compound upanagarika is based on the vartika 'avadayah kustadyarthe trtiyaya on " kugatipradayah " | pa . 22 /18 . Tha vartika is explained by Bhattoji Dikshita in his siddhantakaumudi under "jivikopanisadavaupamye " | pa . 14 .79 . P.6, LL. 1-2- sandaravindando | "Iu some places [the autumnal season appeared with bees delighted with sweet and thick drops of honey beautifully dropping from clusters of lotuses.' P. 6, LL. 3-8. - aba ikarakhyah vrtipadarthabhutam| Here the conjuncts of a and are beautifully repeated. The epithet sandra is to be taken with bindavah and not with aravindavrnda, The expression sundarasyandam syandibhih is to be understood in the sense of a particular action of beautiful dropping among the common actions of dropping, according to Panini's rule, sva pusah | pa . 324640 illustrated in expressions such as reposam pusnati . The compound becomes adverbial: sundarasyando yasminkarmani

Notes. 11 ant syattatha | syandibhih ]. Here in the explanation of the compound sundarah syando yasmin, 'the general dropping' is the anya (the third thing to be meant) and 'the particular dropping i. e. beautiful dropping' is the component part of the compound. indindirah = bhramarah bees. The vivrtikara here only states the gist of Induraja's explanation in his comment. He simply writes -- raiposam pusnatitivat samanyasyando visesitah | P. 6, LL 9-10-pramya | sesairvarnairyathayogam ... Scholars well versed in poetics describe the homely style to consist of the rest of the letters (such and others). The style is (also) called tender or agreeable. The in it is called smooth or simple alliteration. The vivrtikara reads prathitam instead of kathitam and explains the Karika thus:--vrttidvayopayuktavarnebhyo 'nyairyadibhirnibaddham komalaparabhidhanam pramyam vrttim bruvate | lakaradibhih saha ghakaradayo na karya iti yathayogamityuktam | komalavaniravyatvadatyantakantyabhavancatra komalanupraso gramyanupraso va | 66 P. 6, L. 6-faedi..... 1" Somewhere (the autumnal season) producing a false impression of the tinkling of anklets of the goddess of Beauty riding the forest, by the sweet buzzing of swarms of bees rapt in sporting." Here the letters, and are charmingly repeated. is to be connected with, st. 3. St. 3 with 4 and 5 forms one sentence with cakase as the verb. The vivrtikara reads 'khabhramah which is evidently a misreading. P. 6, L. 23- evametastitro ... vyakhyatah | Thus the three styles are explained. The d is separately composed in these with a view to help the expression of sentiments. The same is said in the Karika H.....(see p. 4, 1. 24.). The s or styles of Udbhata described here should not at all be confounded with the fs or modes of expression such as vaidarbhi, pancali, gaudi etc. given by authors like Dandin, Vamana and others. The of Udbhata cannot really be called style in the strictest sense. It is only the outward form of letters and has nothing to do with the sense. The fifa of Dandin and others is properly to be called style in the real sense of the word. The as were originally the provincial peculiarities apparent in the language of the people which afterwards were consolidated into a settled form.

12 Kavyalankara-sara-sangrapha. The anuprasa and vrtti s of Udbhata are a new feature of his work. The s do not exist in Bhamaha. His definition of anuprasa is sarupavarnavinyasa which, it will be noticed, is happily improved upon by Udbhata by putting the word vyanjana instead of varna, anuprasa always occurs with consonants; other varna s i. e. svara s are not to be reckoned in anuprasa . mata's treatment of is very similar to that of Udbhata. MamP. 7, LL. 1-3 - latanuprasah | svaruparthavisesepi ... "The repetition, owing to the difference of purpose, of the same words without grammatical terminations or of complete grammatical forms or both, yielding different meaning, though their form and their original meaning is the same, is called latanuprasa ." 00 P 7.LL. 4-7 - sapadadvitayasthitya | svatantrapadarupana ... The varieties of are described in these two Karikas. In all there are five varieties. They are:When both the words 1 st variety: are situated in two different cs. Here both the words are paratantra . Example - kacidutphullakamala kamalabhrantasatpada Here the word kamala (not pada ) belongs to two different pada . 2 nd variety :- ekasya purvavat (paratantrataya ) [sthitya ], tadanyasya svatantratvat | One word belongs to a different as before, but the other word constitutes an independent c of itself. Example- pada padminih padminigadhasprhayagatya manasat | antardanturayamasurhasa hamsakulalayat || Here the first word padminih is svatantra and the second is paratantra . So also hamsah is svatantra, while hamsa in hamsakulalayat is paratantra . 3 rd variety :- dvayorva ekapadasrayat | Both the words come together in one pada . Example - jitanyapuspakinjalka kinjalka srenisomitam | Here the two kinjalka s appear together in one pada . 4 th variety :- svatantrapadarupena dvayorvapi prayogatah | Both the words occur as different and independent cs. Example- kasa ivodbhamsi saramsiva saramsi ca| Here both the kasa s occur severally. 5 th variety :- padabhyasakramena ca | Repetition of pada s (quarters of verses) in all its orders. All repetitions-once, twice, thrice and in all possible orders-are to be included. padasamudayatmake ca pade bhavat - Induraja. Example-striyo mahati bhartrbhya agasyapi na cukrudhuh | bhartaropi sati stribhya bhagasyapi na cakuh || Here the

Notes. 13 fourth q is the repetition of the second. The first agasyapi na cukudhuh refers to the women and the second to the husbands. Hence the purpose of both is different. Thus we have here described the five varieties of latanuprasa Induraja treats them not in the order in which they are stated in the Karikas but in the order in which their examples occur. Accordingly he has treated the last two varieties first. This P. 7, L. 10- vacyasya ... taptaryabhedat | Even the original 10-gen.....mafia i meanings of the words that are repeated are the same; only their purpose in the sentence is different. distinguishes latanuprasa from yamaka . latadesanivasi etc. - This anuprasa is called as it is favourite with the people of the Latas, probably the modern Gujarat. P. 7, L. 12 - sa ca prathamam tavatriprakarah 1 Udbhata mentions the five varieties apparently without having any principle in his mind. But Induraja divides according to some logical principle. His mode of division might be stated in a tabular form as follows:- dvayoh svatantrayoh (svatantrapadasrayah ) padas. latanuprasah svatantraparatantrayoh ( svatantrapadaparatantra sabdasrayah ) Of two independent Of one independent pada and ekaikasminpade bhavat ( ekaikapadasrayah ) Occurring in two different padas. another dependent word. III Ex., padminih padminigadha0 | padasamudayatmake pade bhavat (padasamudayasrayah ) Occurring in two whole dadas or carana s • padadvitayasrayatve (paratantrasabdasrayah ) The dependent words occurring in two different pada 8. IV dvayoh paratantrayoh ( paratantrasabdasrayah ) Of two dependent words. ekapadasrayatve (paratantra sabdasrayah ) The dependent words occurring in one samasika- pada . V. Ex. jitanya I Ex. kasah kasa II Ex. agasyapi na Ex. utphullakamala- ivo0 | cukudhuh | kamalabhrantasadrapada | puspa kialka0 | Nos. I and II are the 4 th and 5 th varieties according to Uabhata's enumeration. No. III is the 2 nd variety of

14 Kavyalankara sara sangraha. Chata Nos, IV and V are respectively his second and third. The fat gives the divisions in his own words thus: esa dvayoh sabdayordvayoh padayoh padasabdayorva yada bhavati tada purva tribhedah padadvaya- srayatvena [eka ] padasrayatvena ca sabdasrayo dvividhah, ekamiti pade padasamuhatmake pade ca bhavat padathryo dvividhah | evam ca pancavidho'yam | He has rather a fanciful discussion on the definition and divisions of chekanuprasa . After explaining the five divisions in the manner of Induraja he continues : taduktam | padasya labdhatmalabhasya svatantra vyavaharam krtva sa ityadineti vyakhyanamasat | yato latanuprasah punaruktih | na ca sabdapadanami drsi sa | yatra hi sabdanam ca dvih padanam ca dvirabhidha so'sya visayo nyayyah | iha tu padminisabdasyaiva punaruktih | na tu padminirityetasya padasyapi | kim ca sabdanam va padanam veti vasabdamapasya vrttih karya | adanavisirnabhidhanam ca prasajati | tasmaditya- metat | sabdanam padanam veti dvau bhedau, tatra padantare punarucyamanah sabdah paratantrah svatantro va ekapade paratantra eveti sabdanam tridha paunaruktasya yathoktau dvau bhedaviti pancavidhabhedah kramenoktah | taduktam sa ityadi | punarucyamanaparatantrasabdasrayah sa padadvitayasthitya dvayoriti laksitah | From the words taduktam which intimate that the passage following is taken from some other writer one may infer that there were some other persons earlier than the who had commented upon Udbhata's work. He reads dvayodhyekapadasrayat . $ The fa gives the examples of the five varieties of latanuprasa in the order of the definitions of Udbhata which is exhibited above in detail. He explains: - - avatamsah karnabharanam, kasah puspavisesah, aci- ksipuh - avarjayamasuh padabhyasakramena ceti padayoh kramagrahanam dvitiyacaturthaprathama- trtiyapadabhyase latanuprasasya bahulyadarsanaya, and adds yasya ca savidhe dayita davadahanastuhinadidhitistasya | yasya na savidhe dayita davadahanastuhinadidhitistasya || iti darsanat padasabdo'tra padasamuhopalaksanaparah . P. 7, L. 20 kasah kasa iva ... | "Kasa (white grass) appearing beautiful just like Kasa grass, lakes and rivers appearing only as lakes and rivers, ravished the hearts of the young." Here are again compared to. This produces & shade of ananvayalamkara (g. v. ). Here the words repeated are svatantra as they are complete grammatical forms upalabhyamanasutirupah .

Notes. 15 (absence P. 8. L. 9 - striyo mahati ... | Hore agasyapi na cakradhuh has different purposes in both places. The first tells the of anger) of the lovers, while the second shows the ana of the ladies. P. 8, L. 24 - kacidutphulla kamala ... | Here kamala, padapada and mukhara are repeated having different purports in their different places. The description of which was commenced from etc. (P. 5.) is also continued trough this verse. All the four epithets are adjectives of How the a or purport of the repeated words is different is explained in the commentry. P. 9, L. 12- Here the faces of the young women being compared with the moon by the expression, the idea of the hare or the dark spot on the moon is suggested by the word en, dark lotus. The compound jitanya ... sobhitam should be thus dissolved :- jitah anyapuspanam kimjalkah yaih te [ kimjalkah ] | tesam kinjalkanam srenayah tabhih sobhitam [asitotpalam ] | | P. 9, L. 23-... "The swans coming with intense desire for lotuses from the Manasa lake which is the abode of families of swans, internally indented the lotus plants." Here the idea is that the lotus plants on which the white swans with prominent necks sat, appeared as it were with teeth erect and prominent. The erg is quite clear here. The reading ought to be c: and not as the Nirnayasagara edition reads. The vivrtikara reads padminih . P. 10, LL. 7-9-... When a c (word, noun) is connected with another in a subsidiary relation, their direct relation being impossible, it is called . P. 10, L. 10-bandhastasya yatah ... | It is samastavastuvisaya when its construction displays all objects (constituting the 4) directly expressed in words; and when some are directly expressed and others are implied. The vivrtikara has something more to add here:-tasyetyupa- carat tatsambadhinavihocyete | upamanopameyayoh samanadhikaranyasambandhatmana rupakasya sritatvabhavattena tatsambadhinoryatah sabdo bandhah tatsamastamaropyaropavisayatmakamubhaye

16 Kavyalankara sara sangraha. vastu visayah sabdamasyeti samastavastuvisayam srutya samarthyena ca yada bandhastena karane- naikadese visesena sabdena kramena vartata ityekadesavivarti | ! P. 10, L. 11- samastavastuvisayam ... ! "A rupaka containing & string ( a series ) [ of superimpositions] is [also] called samastavastuvisaya and it is called also ekadesavrtti when the prakrta thing is described in the form of aprakrta (para ). " ..... tatra rupakata | When utpalam may P. 10, L. 14-padantarasya gunavrtteh ..... word (noun) is connected with another in a subsidiary relation it is 4. If it is so, in the expression a the word is connected within a subordinate relation as qualifying it, and here too possibility of a arise. To remove such a doubt the author has inserted the clause srutya sambandhavirahat sruti means abhidhavyapara the direct process of conveying the meaning. And when a q has no possibility of connection with another by this direct process, it assumes and thus connects itself with the other word. This d or 'the process of conveying a sense that is similar in qualities to the original sense' is a sort of laksana And when a word is used in this gunavrtti sense it becomes if it has sufficient charm in it to be called an alankara. In simho manavakah we see that it is impossible •for a boy ( manavaka ) to be a lion (simha ), therefore the word simha is used here in the sense of simhagunasadrsagunavan . vivrtikara This is the comment of Induraja on Karika 11 and it is quite simple and lucid. The has rather a pedantic dissertation bearing on this Karika. He observes - yatpadena padantaram sambadhyate tadrupakam | nanvevam vrkso bhatityatratiprasanga ityaha gunesu dharmesu vartamanam | evamapi nilotpalamityatra praptirityaha srutyetyadi | srutirnirantararthanisthasya vyaparah tatpratipadito'rtha ihopacaracchutih | kriyaya yo'bhisambandhah sa srutiprapitah | tayorasrayasrayinorvakyanniyamastvatisthata iti nyayadarunayaikahayanya pingaksya somam krinatityadau arunadikrayadyoreva trtiyaya pratipadane sambandhasya srautatvat | tadabhave- pyarunaikahayanyadinam vakyiyasamanvayasadbhavanna srautasambandhabhavamatranna bhavati gunavrttitva- miti hi mukhyarthabadhah | tena srautenarthenarthantarasyarthayo vakyiyasamanvayah tadabhave yada svagatagunasamagunavrttirbhavati | upamanopameyayoh gunairiva gunanamapi dharmantarena samyakalpanayantvanavasthapadasya sritenarthena srutya sambandho bhavatyeveti tadabhavo na vyakhyeyah| tadevam srautenarthenanupapadyamanarthantarasambandhagata eva sadharanan gunan laksayitva padarthadupasarjanamaprakaranikarthabhidhayi prakaranikarthabhidhayina padena gaunena vyaparena sambandham bhajate yattadrupakam | aropyamanarupenaropyamanavisayasya rupavatah kriyamanatvat |

Notes. 17 P. 10, LL. 19-20. - jyotsnambune ...... masicyata | Example of samasta vastu visayarupaka . "The garden in the form of the sky variegated with flowers in the form of the stars was sprinkled over by young maidens one by one in the form of successive nights with water in the form of moonlight by means of a jar in the form of the moon." rupakataprasangah | Here the rupaka P. 10, L. 21. - atrambusabdo is clear. jyotsna is the prakrta and it is called ambu . It is impossible for (water) to mean (moonlight) by the direct process. Therefore it resorts to and means ambugunasadrsagunavati . When we say that jyotsna is ambu we mean that has qualities very similar to those of arg. With ambu . this meaning, then, the term can validly be applied to. The same is the case with all the other parts of the rupaka, viz. indukumbha, tarakusuma etc. Now in nilamutpalam it is true that qualifies ; but it does not resort to qualifying 4. It can qualify object by its own direct power. possibility of c in that case. for or any other blue Therefore there is no ambu P. 11, LL. 3-13. -- nanu viruddharthabhidhayinoh ...... gunavrttitvaprasangah | If it be said that out of two words giving opposite meanings and being in the same case, which cannot be mutually connected by the direct or a process, if one of them is said to be (subordinate) then alternately each of them would become subordinate. As the word ng is said to be gunavrtti, evidently jyotsna can also be said to be gunavrtti and be connected with ag as the principal word. But the word in the Karika precludes such a possibility. It is always quite consistent that the subordinate epithet should possess gunavrttitva by laksana as it is the epithet which qualifies the principal. Now the, i. e. the thing belonging to the subject in hand, is always the principal [in poetry]; and the other thing is regarded as of a subordinate nature and so it ought to be considered the qualifying term. Therefore the word which is a i. e. not belonging to the subject of description should properly be accepted as with reference to the word which belongs to the 3 [x. .. 1 or the subject. Thus there is no

18 Kavyalankara sara-sangraha. chance of jyotsna and ambu both being taken as gunavrtti s promiscuously or alternately. The vidyutikara explains the term pradhanena and its purpose thus :-nanvevam samanadhikaranyabhave yadekasya gunavrttitvam kalpyate tatparyayena prapta- niyame karanabhavadityata aha pradhaneneti | pradhanasya ca gunavrttita na bhavati pradhanya- haneh | pradhanatvam ca prakaranikatvajjneyam | anyatha gunavrtteh pradhanenaiva sambandhe tadvacana- manarthakam syat | P. 11, LL. 14-21. - nanu ca jyotsnambune ......manyatsamasoktam - iti | Again it may be said that in the expression jyotsnambuna, ambu is pradhana and jyotsna is gauna . For in this sloka the action described is sprinkling which has totally enveloped the real action of pervading (). Pervading is the real action concerning the subject. But it is here described as sprinkling (in a hyperbolic fashion). Now the thing required for sprinkling is water which is here present in the form of sar or moonlight. Thus, taken in this light, the word jyotsna assumes a qualifying nature and water becomes the principal material. This interpretation gives the sense that the water which is required as a material for sprinkling is nothing but moonlight. This view is also supported by Rudrata. He says: 8, 40 .) When the upameya which compound with upamana that is member of it, there is is samasokta . P. II, L. 22. - tatkathamidamukta is etc. (Rudrata's is prakaranika enters into a aprakaranika and in other places there as a subordinate rupakatvamuktam | Then how is it said that the word here for the reason that jyotsna is pradhana ? We answer this objection in the following way. There are two states of things existing here. One is the superimposition of and the other the connection of jyotsna that has assumed the form of ambu with the action of sprinkling. In the former case when ambu is superimposed upon jyotsna, ambu assumes the gunavrtti by means of the qualities of erg which are to be shown a exactly similar to the qualities of . Thus in this case ambu is taken to be gauna and jyotsna to be pradhana because it is as said before]. as

Notes: 19 P. 12, L. 3-6. - yadatvasau ... gunesu vartamanatvat | But when once the word expresses (moonlight) as having the nature of water' and thus becomes connected with the action of watering the word itself in that case becomes qualified by meaning that the water which is used here as a material for sprinkling is of the nature of jyotsna . This is the second state of things referred to above. But in this case there is no . It was in the former case that ambu had resorted to gaunavrtti . Hence it assumes the shade of atisayokti in this state. For the sense jyotsna or moonlight is completely enveloped in this case by the sense 'water' in order that it may have the action of sprinkling. In the former case stood to be gauna, for jyotsna being the prakaranika was pradhana . of the was underOn this whole subject the fan sharply differs from Induraja. He denies that there are two and that there arises a shade of often the consummation of rupaka . He observes : - nanu jyotsnambunetyatrambunah pradhanyam | tatha hi | atra yeyamasicyatetikriya tayambu eva sadhanamapeksitam jyotsnaya visesyate | tena jyotsna gunavrttih| naitat | yato'sicyatetyanena vyaptirlaksyata iti yathoktameva sthitam | yattaktam- upasarjanopameyam krtva tu samasametayorubhayoh | yacca prayujyate tadrupakamanyatsamasontam || (rudratah - kavyalamkarah, 8 .40 ) -iti | etayorityupamanopameyayoh | tacchabdaprakramanusarena | sabde hi kame simho manavaka ityadau simhadirupavacchadito manavakadirapradhanabhutah pratiyate | vastava- nayanusare tu uktarityasya ( simhasya ) gunavrttitvam | ata eva sabdanayanusarenasicyate- tikriya samucita | etatkiyopadanasamarthyaccambvadinamapradhanyamiti rupakaparigrahe nyayo'stityanekalakriyollekha ityadi samkara laksanam bhava tada | evam sati dve tatravasthe | tatra jyotsnapradhanye rupakamambupradhanye'tisayokticchayeti vyakhyananasat | etasmin vakye'vasthadvayapratipatteh | atra ca sarvesamaropyamananam vastuvisayata | Unfortunately the MS. contains omissions at the end. can ascertain the general tenor of the argument. Still we P. 12, LL. 8-10. - rupakatvam catra ... danvartham drastavyam | rupaka is the superimposer of one thing over another. In the object on which superimposition is effected (aropavisaya ) is made to assume the nature (4) of the thing which is superimposed, . bharopavisayam vastu rupavat karotiti rupakam .

P. 12, LL. 11-15. -- atra copamanavartino ... yukta iva drsyate | Here as the properties similar to those which exist in are seen in the cc, the superimposition of the nature and word of the upamana is made on the upameya . Here there are three views: Some say that there is sabdaropa first and artharopa afterwards. First the things are verbally said to be the same and afterwards the sameness of their qualities etc. is understood. Others say that auf is made first and then sabdaropa . But still others say that sabdaropa and artharopa are made simultaneously. The last view seems to be proper. For the same view is accepted in the Karikas that follow. The vivrtikara has here similar remarks to offer. atra keci • cchandaropapurvamartharopam bruvate | apare viparyayam | anye yaugapadyenobhayam | esa eva ca yuktah paksa ityahuh | tasya ca vaksyamanarupabhedabhyam saha namadvitve'pi prak pratijnatam catu- rupatvam sthitameva | P. 12, LL, 16-19. -- sabdopacarattadrupam | upamanagunaistulya ... | 1 st view : First sabdaropa and then ruparopa . 2 nd : First ruparopa and then sabdaropa . 3 rd : sabdaropa and ruparopa simultaneously. " P. 12, L. 22. - tasya ca rupakasya ... The two kinds are: (1) samastavastuvisaya and (2) ekadesavivarti . They are dealt with in order. When all the for all the component parts of the thing described are directly expressed the rupaka is samasta- vastuvisaya sruti means the direct process of abhidha . P. 12, L. 26. - jyotsnambunendu ... Here all the objects jyotsna, indu, tara, ratri and vyoma are directly mentioned as being the same as ambu, kumbha, kusuma, kanya and udyana respectively. No metaphor is left to be understood. saritam = variegated. P. 13, LL 9-10 - yatasca srutyarthabhyam iti | When the rupaka is made up of some actually expressed and some understood it is called ekadesavivarti . Induraja dissolves and explains the terms samastavastuvisaya and ekadesavivarti thus : - atah samastavastuvisayatvam| samagrani hyatra rupyatvenabhimatani vastuni svakanthenopattasya rupakasya visayah | ... ekadesavisesena svakanthoktatha vartanat ekadesavivartitvam | p. 13. The vivrttikara explains much in the same manner- tasmatsamastamaropyaropavisayatmakamubhayam vastu visayah sabdamasyeti samasta T

Notes. 21 vastuvisayam | srutya samarthyena ca yada bandhastena karanenaikadese visesena sabdena kramena vartate ityekadesavivarti | $ P. 13, L. 12. utpatadbhih ... | avijyanta | Though the root viju is not enumerated in the dhatupatha of Panini still it is to be accepted as a valid root, like mith khac klav, ksap which are accepted and used by the learned. In the verse only two rupana s are expressly mentioned viz. piccha is mentioned as bala and sarah &s nrpa . But there is no metaphor supplied for rajahamsa s and sarada . We have to understand camara as the upamana for rajahamsa and nayika for sarat . The meaning of the verse is: The kings in the form of lakes are fanned by (the maiden in the form of) autumn by means of (chowries in the form of) swans having clusters of hair in the form of rows of feathers. The unexpressed are also made clear in this. as The vivrtikara gives another instance of ekadesavivarti rupaka, cakravakakucabhoga nilanirajalocanah | ambhojininitambinyo hasanti sarasiruhaih || and adds the remark ityadavekadesadvivartate calatityekadesavivartitvam | The stanza is composed in the style of Udbhata and appears as it were to belong to his kumarasambhava . The fat gives here other views of interpretation:- anye tvahuh | camaranamarthatvam rajahamsanam tu sabdatvamityekadesa upameyatmani visesena vartanadekadesavivarti | evam ca picchalibaletyadi samastavastuvisayavinabhavi samasta - vastuvisayam tu nirapeksam | jyotsnambunetyadau tu savayave rupake avayavanam yadyapyavayava- peksa tathapi rupanam paratantram | P. 13, LL. 23-24. yadi va ... bahuni rupakani | Or malarupaka should be called samastavastuvisaya by the derivation : samuccayena asyante bahuni rupakani atra . For a series of different rupaka s are superimposed upon one thing in this case. cin Karika 13 introduces other kinds of rupaka The vivrtikara has - yadva samuccaye . nastani ksiptani vastuni visayo'syeti malarupakam samasta vastuvisayamucyate | ekasyaiva rupakamalanyasacca malarupakatvam | P. 14, LL. 1-2 - vanantadevata ... | "The rows of bees shone as braids of hair of sylvan deities, as (iron-) chains of death to the ladies whose husbands were away and again as swords of the warriors of cupid." The three different s

Kavyalankara sara-sangraha. + upon the same subject of description viz. bhrngamalah are very clear. P. 14, L. 8. -- ekadesavivarti tu ...| ekadesavrttirupaka is where the prakrta object is made to assume the nature of the para or aprakrta one. Here the reading ought to be ekadesavrtti and not ekadevivarti as will be clear from the sequel. slesa . P. 14, LL. 11-12. - asaradhara ... Here the word prasadhyate stands both for 'decorating' and for 'acquiring' by & tinge of . Out of these the sense of decorating is the prakrta here. But it is quite ignored in the cumulated meaning of the sentence. In asarajyam prasadhyate the verb prasadhyate naturally means "is acquired" following the object which appears more prominently than asa . Thus the upameya or aropavisaya (which is asa ) in this case merges into and assumes the form of the upamana or aprakrta by the power of slesa . The vivrtikara 's comment on this stanza is very interest ing :- atra prasadhyate iti bhusane vartate saradvarnanasya prastutatvat | yattvasyanekarthatva- darjanamarthantaram tena bhusane rupite sati visikhadibhirupapadyate rupanam | spastamidamudaharanam- " merurukesaramudaradigantapatrama mulalam pracalasesasariranalam | | yenoddhrtam kuvalayam lalata salilamuttasakarthamiva patu sa vo varahah || " iti | evam ca ' sanandajaladhisasilekhe ' tivat (Rudrata's kavyalamkara, 8 .51 . The printed edition reads janananda instead of sananda ) kuvalayam kuvalayamevam- bhutamiti rudratokta paramparitarupakabhametat | atra sabdamukhenarthasya vicitratve'pi mukhyatah kavipratibha samrambhagocarasya sabdasya vaicitryamiti sabdalamkaratvam yuktam | pranthagaurava- bhayattu na pratanumah | kimcittu slesasthane vaksyamah | In the verse meruru0 the word kuvalaya is slista . kuvalaya = (1) the terrestrial globe, (ku = prthvi ); (2) a lotus. The meaning lotus' is aprakaranika which when superimposed gives rise to the metaphor. As ekadesavrtti rupaka is dependent upon slesa and as slesa in such cases is a sabdalamkara the vivrtikara is inclined to regard this as a sabdalamkara and not an arthalamkara . It will be also clear from his comment that he regards the ekadesavrttirupaka of Udbhata as similar to the paramparitarupaka of modern writers. Thus he says that when bhusana is rupita by the sense uparjana then alone asaradhara etc. can be rupita by visikha, nrpa eto. P. 14, LL. 15-1. - yaduparjanam tada ... tenatraikadesavrttitvam | The com mentary here is rather obscure. The obscurity will be

Notes. 6 23 from their removed if we displace the words present position and connect nrpavisikharajyabhumyatmakam with karaka- kadambakam which it qualifies yena tadpavat kriyate or better, tena ( parena ) yadrupavatkriyate (tadrupam ) is only a paranthetical clause explaining the word 4. The meaning of the passage is :-The action of ' acquiring' is here what is aprakrta and hence para or anyat The complete form (64) of this sense of acquiring', which is the and sense, is the group of Rs (i. e. substantives having some relation with the verb viz. zal, a, a, sampradana, apadana, and adhikarana ) which consists of nrpa, visikha, rajya, and sangramabhumi nrpa or king is the subject (karta ) of the action of acquiring, visakhah ( arrows) are the means (karana ), rajya is the object (karma ) and samgramabhumi is the adhikarana of the action of acquiring. Thus all this paraphernalia of ES forms the of the sensei.e. becomes T. And by this i.e. the the verb (in the sense of acquiring) the superimposition upon balahaka, asaradhara, dik and nabhobhaga - all these being the objects intended to be superimposed upon-is effected in order. Thus nrpa, visikha, rajya and samgramabhumi are respectively superimposed upon balahaka, asaradhara, dik and nabhobhaga . In this way there is here. The whole play is on the word pararupa . of P. 14 LL 19-21. - ekadesavrttityatra ... syabhimatam | The interpretation of the term ekadesavrtti is rather curious: ekada isah (vakyarthah ) tadvrttitvam | " To be in accordance with a sense which is at one time the principal sense of the sentence". From this interpretation the nature of becomes somewhat clear. When two distinct senses are expressed in one sentence in the manner of a metaphor one of them is prakaranika and the other is aprakaranika . In the instances such as etc. cited before, both the senses are expressed by two distinct sets of words and the verb applies to one of the senses only. Thus the whole purport of the sentence is single and concordant. But in a sentence where a metaphor is expressed by means of a fag verb the situation is different. There only one sense can be prominent at one time, for there is a rule of the as that one word can mean only one thing at a time, sakrducaritah sabdah sakkadartham gamayati . In the example jyotsnambuna etc. the verb asicyata

has only one meaning and the meaning of the whole sentence is made convergent to the meaning off by resorting to gunavrtti . But in the example asaradharavisikhaih (st. 14) the verb prasadhyate has two distinct alternate meanings. When the prakaranika sense of prasadhyate viz. ' decorating ' is brought out then the sense 'that the quarters are embellished by the white clouds' is prominent and the metaphor is thrown into the back-ground. But when the sense 'acquiring' of the verb is brought into prominence. then the metaphorical sense viz. visikha, nrpa, rajya and samgramabhumi is brought to the forefront, anyatra anyada ca prabhavisnuparjanam aprakrtam hi tat chesavasenatra nitam (P. 15, 1.1.). Thus it will be seen that here rupaka exists along with the sense (viz. acquiring) that is at one time (not always) prominent in the sentence. ekada isah yo vakyarthah tadvrttitvam rupakasyabhimatam . This peculiar sense of ekadesa is given by Induraja on the authority of Udbhata who, he says, explains the word ekadesa in this manner as ekada isah in his bhamahavivarana . The fan has another explanation to put forth of the ekadesavrtti rupaka . He also oriticises the explanation of Induraja though the point of his criticism is not perfectly clear. He says- tatha parenarthantarena yo rupo rupanam arthatprakrtasyarthasya tena nimittena yadrapunam tasmadekadanyadrso yo'rthah tatra vrtiryasya tat | parasyarthantarasya yadrupakam karakavrndam tena rupanaditi tu vyakhyanam ayuktam | evam hi visikhairbhusyante smetyasambaddho'rthah syat | On the whole the variety seems to have proved a stumbling block to the commentators. What is exactly meant by the term is not clearly discernible. Nor is Udbhata's difinition of ekadesavrtti viz. pararupena rupanat more clear. Udbhata's divisions are always logical and precise; his definitions are always terse and epigrammatic. But all these qualities seem to have reached an excess when he created this variety of rupaka . The nature of ekadesavrtti rupaka is extremely difficult to understand and no other alankarl. ka has accepted this variety. In this way Udbhata's rupaka is of four kinds: 1. samastavastu- visaya, 2. ekadesavivarti, 3. samastavastuvisaya ( malarupaka ), and 4. ekadesavrtti .

Notes. 25 On the whole Udbhata's treatment of chas not much to recommend in its favour. Of course he has made an advance in respect of the definition beyond Bhamaha. But his treatment of this important alankara and the commentary of Induraja upon it have not been very happy. He has brought in an obscure variety such as ekadesavrtti which is not accepted by any author either of the old or of the new school of alankarikas. His nomenclature of the malarupaka as samastavastuvisayarupaka has not also proved acceptable for no one has followed him in calling the malarupaka by the title samastavastuvisaya after him. His definitions of the alankara and the varieties must have been, however, of his own invention for he is not indebted to Bhamaha in that respect. Bhamaha's q is- upamanena yattatvamupameyasya rupyate | yunai akai gar zub ara afag: 11 samastavastuvisayamekadesavivartti ca | dvidha rupakamuddistametattaccocyate yatha || bhamahalamkara - 2, 21-22 . And we must really congratulate Udbhata upon having constructed such definitions solely by his own genius. Even Mammata has thought fit to take some thing that was made ready by Udbhata: compare the definitions of of the varieties of rupaka in kavyaprakasa, X. p. 594- samastavastuvisayam srauta aropita yada | srauta arthasca te yasminnekadesavivarti tat || The indebtedness of this Karika to that of Udbhata viz, TEJENEY YA: YUI yuqizzi a aa az samastavastuvisayamekadesavivarttim ca || is quite clear. It can also be seen here how the neatness and system of Mammata have caused improvements. Mammata gives one variety completely in one half and the other in the other half. But Udbhata gives the sof both in the first half and the names of them in the latter half which looks rather awkward and becomes more difficult to understand. The divisions of adopted by Udbhata do not seem to be exhaustive. He appears to have left out what is 4 [x. s. s.]

called paramparitarupaka which is recognised by rudrata, ruyyaka, mammata and later writers. gives a peculiar charm and deserves a special mention. Bhamaha gives only the first two divisions and Udbhata seems to have somewhat improved upon them. He gives four varieties of which is one which bears a sort of resemblance to fg. paramparitarupaka . Perhaps this ekadesavrttirupaka of his might be the precursor of the of the Moderns, as it consists of a principal word having a tinge of . The has vivrtikara hinted at this in his comment. He says rudratoktaparamparatarupaka- bhametat . Mammata also in his vrtti on the first kind of paramparitarupaka remarks that the slistaparamparitarupaka is called ekadesavivarti . by other Alankarikas- ekadesavivarti hidamanyairabhidhiyate . Here the edition of the Kavyaprakasa gives two other readings in the foot-note on P. 471, viz. and • savartidamanyairapyabhi0 . The second reading is in point and seems to be the correct reading of the author, as it exactly tallies with Udbhata's nomenclature. By he seems evidently to refer to among others. His definition of viz. [niyataropanopayah ] syadaropah parasya yah is as it were a paraphrase of the definition pararupena rupanat of the ekadesavrtti rupaka given by udbhata . Thus udbhata may be said to recognise the paramparitarupaka in this way, and in fact he should be called the originator of this important variety. P. 15. LL. 4-12. dipakam | adimadhyanta ... | That is dipaka where properties belonging to the c as well as the having an implied sense of similarity in them, are mentioned in the beginning, middle or end of a Kavya. yatrantargatorthasamarthya ...samanadhikaranyam | Here a kavya is called dipaka by means of e, apposition or identification, as both the terms are put in the same case. atra ca dharmanam ... hetutvaddipakata | The properties should be mentioned only once; for when they are repeated they give rise to as the author as himself says further on. Thus two or more standing in relation to one another as and and having common properties mentioned once, and both of them being thus illumined by the dharma s stated only once (taduddipanahetutvat ) constitute what is called dipaka . P. 15, LL. 12-15. - yavacca tesam ... praptarthatvat | Here as the properties are mentioned with reference to two as stand-

Notes. 27 ing in relation to each other as upamana and upameya, and as upameya being prakaranika is pradhana and upamana is gauna with reference to it they are a fortiori the properties of pradhana and itara that is gauna . Hence the expression pradhanyetarayoginah in the text is a mere repetition as its sense is already brought out by the aupamya . The far here passes strictures upon the explanation of pradhanyetarayoginah given by Induraja. He does not endorse the opinion that the sense of pradhanyetarayoginah is already expressed by antargatopamah and that pradhanye0 is only an / anuvada, The whole of his comment will bear quotation. samarthyava seyopamanopameyabhavo yesu satsu tadrsah visayasabdasyanyatrabhavarthatvadadyabhave ca madhya evanta eva va natvasakrdavasthita dharmah kriyadirupah yatra dipane taddipakam | prakasavadekastha maneko pinaddipakam | antargatopama ityanenaiva gatarthatvat pradhanyetara - yogina iti vispastarthamupameyam hi pradhanam upamanam tvittaraditi vyakhyanamasat | tulya- yogitabhidhanarthametasyopattatvat | tatra hi sarvani prakaranikanyevaprakaranikanyeva va | prakaranikarthavispastarthamupameyam | prakaranikanantupameyaddipakam | The goes even to the length of calling in question the correctness of Udbhata's illustrations. He asserts that Udbhata's examples of are really examples of tulyayogita, and therefore gives his own examples. Compare his remarks : - sarvesamevatra prakaranikatvat tulyayogitaya etanyudaharanani yogyaniti rehalamihirena naham marasenakasam sarena jovvanaam | amaena sudhi dhavalo tumaenaranaha bhuvanaminam || ityadinyudaharanani deyani | The threefold division of adopted by Udbhata as adimadhyantavakyavisaya is of no consequence and is not accepted by later writers. They recognise a twofold division as kriyadipaka and karakadipaka which gives some charm. Compare Mammata's definition of dipakah- sakrdvrttistu dharmasya prakrtaprakrtatmanam | saiva kriyasu bahvisu karakasyeti dipakam || Udbhata seems to follow c in his threefold division of dipaka Bhamaha's definition of dipaka runs thus :- adimadhyantavisayam tridha dipakamisyate | ekasyaiva vyavasthatvaditi tadbhidyate tridha || (- bhamahalamkara, 2 . 25 )

28 Kavyalan kara-sara-sangraha. P. 15, L. 20.-... Here the property of kadamba destru tion is mentioned with reference to the two (1) floral beauty of and (2) the happiness of ladies whose husbands were absent, bringing out similarity between them, the former being prakaranika and the latter aprakaranika as the time of autumn is the subject of description. The word attached to, by means of a tinge of slesa means time as well as antaka destroyer, or yama . Here the property of Er is mentioned in the first sentence and is to be understood consequently in the second: therefore it 18 adidipaka . | P. 16, LL. 5-7. - madhyadipakam | videsavasatih | Here sarad is prakaranika and videsavasatih and yatapatikajanadarsana are both aprakaranika and there is implied similarity between them. The property which occurs here in the middle of the sentence in the form of yatapatikajanadarsanam duhkhaya kevalamabhut applies subsequently to the other two and thus gives rise to madhyadipaka . videsavasati, yatapatikajanadarsana and sarat are all causes of pain to travellers. P. 16, LL. 13-15 Here the property sarvasukhahetutva mentioned in the last vakyartha as kasya nanandako bhavat gives rise to antadipaka . P. 16, L. 20. Here there is a question. In the Karika 2. of the text, was mentioned before and as such it ought to have been defined first and then dipaka, following the rule yathoddesam laksanam the definition should follow the order adopted in the enumeration. The answer kumarasambhava as examples and there the instances of a occur before (those of 3). To keep up the sequence of narration, the order of enumeration is abandoned. The order of enumeration itself is not made to follow the sequence of examples for fear of violation of the metre. If we change into dipakamupama caiva, the anustubh metre is spoiled. Similarly the same explanation is to be given in the following treatment where the order of enumeration is violated in definitions. The has similar remarks to offer, is: The author has given part of his on " A

Notes. 29 P. 17, LL. 1-15-...... That charming similarity, existing between and which are expressed by words and whose pravrttinimitta s viz. desa, kala, jnati, guna, &c. are mutually different from each other, is 344 A). Bhamaha's definition of 4- viruddhenopamanena desakalakriyadibhih | upameyasya yatsamyam gunalesena sopama || - bhamahalamkara, 2 .30 . Udbhata seems to have borrowed the idea of mitho vibhi- nakaladisabdayoh from Bhamaha's expression desakalakriyadibhirviruddhena . The words and occur in both. Bhamaha says samya is upama while Udbhata speaks of sadharmya as upama Both mean the same thing but express the idea in different words and which are accepted as identical in common parlance. Mammata appears to have improved upon both and lays down his pithy, accurate, and elegant yet simplest definition of upama as sadharmyamupama bhede . Here he brings out the notion of desakalakriyadiviruddhatva of Bhamaha and mithovibhinnakaladivacyatva of udbhata existing in upamana and upameya by one short word bhede, and leaves out the words upamana and upameya altogether, intending to supply the idea by mere implication. He has thus improved upon both enormously and made his definition very short and sweet. Thus the most essential characteristics of upama are only two: bhinno- pamanopameyatva and sadharmya similarity. These two are further developed by natural process into the four necessary requisites of upama viz. 1. upamana 2 upameya, 3. sadharanadharma and 4. upamavanvaka . The idea of similarity involved in the definition of c is expressed differently, it will be seen, by various writers on Alankara, if we compare their several definitions. It will swell the bulk of the volume if we quote. all the definitions here. More important of them are given by R. B. K. P. Trivedi in his edition of Ekavali in the Bombay Sanskrit series at P. 530; and by Dr. Belvalkar in his edition of Dandin's Half PP. 75-76. Here we find bharata-sadrsya, dandi -sadrsya, udbhada-sadharmya, Part second, Second using the word-Y, vamana samya, rudrata samana- rudrata-samana-

30 Kavyalankara sara-sangraha. gunadi, khyyaka-sadharmya, mammata - sadharmya, bhoja - samanyayoga, visvanatha - samya, vidyanatha - samya, vidyadhara- sadharmya, appayadiksita--sadrsya, jagannatha - sadrsya, hemacamdra -- sadharmya, vagbhata - samya, vagbhatalamkara - sadrsya, alamkarasekhara - sadharmya . citramimamsa - sadharmya, sahityakaumudi - sadharmya and alankarakaustubha -sadrsya, The word sadrsya is used by 6 authors, samya by 7, and sadharmya . by 8. The words samya, sadrsya and sadharmya are grammatically formed in the same way by adding the c termination syan to the bases sama, sadrsa and sadharma all in dual number, in the sense of prakara which means dharma (property) or sambandha ( rela tion or connection ) : prakrtijanyabodhe prakaro bhavah . On the authority of this interpretation or derivation, some understand the words in the sense of dharma i. e. sadharmya meaning samanadharma, and others in the sense of sambandha i e a relation between the anuyogi and pratiyogi of that sambandha, these latter being in the present case upameya and upamana, the sambandha between which being brought about by a samanadharma found in both : samanah dharmah yayoh sadharmanau | tayoh sadharmanoh bhavah (dharmah ) sadharmyam or tayoh sadharmanoh bhavah (sambandhah ) sadharmyam | But the rule of the grammarians samasakrttaddhitesu sambandhabhidhanamanyatrarudhyabhinnarupavyabhicaritasambandhebhyah (See kaiyata 's gloss on the Mahabhashya on "tasyabhavastatalau " pa . 5.1.119 ) governs the present case. Most of the Alankarikas follow grammarians and take upama to be a kind of sambandha, whose anuyogi is upameya and pratiyogi-upamana, the sambandha being brought out by the existence of the common property (samanadharma ) found in both. The words samanena dharmena sambandhah or dharmatah samyam occurring in the definitions and the s clearly show that it is a or relation brought out by means of or through the common property and not the (connection of the common property i. e. sadharanadharma as pratiyogi on the one hand with the upameya and upamana both as anuyogi on the other. The latter interpretation was first started by the late Mr. T. D. Chandorkar in his edition of the Kavyaprakasa, Ullasa X. and was followed by the late Pandit Vamanacarya in his Balabodhini on the Kavyaprakasa, third edition. They considered upamana and upameya both as the anuyogi and sadharanadharma as the pratiyogi of this sambandha which, in this way, is merely reduced to the position of a primary a between and varmim which is technically known as samavaya in the phrase logy of the Naiyayikas. Now all this is against the

Notes. 31 acceptance of almost all the Alankarikas and their commentators. Neither Induraja nor the vivrtikara / has anything to say upon the nature of the sambandha which sadharmya indicates. The dism cussion was developed later and the early authors appear to have no notion about it. The far has some remarks to offer about the propriety of mentioning the words upamanopameyayoh in Karika 15. He also makes a passing allusion to the well-known of the grammarians in this counection. He begins his comment thus- samasakrttaddhitesu sambandhabhidhanamiti samanena dharmena sambandhah | upamitirupama | sa ca na karyakaranayorbhavatyapi tupamanopameyayoreveti gatarthamapyupamanopameyayoriti vacanamarthe samye prasiddhiranumantavyeti pratipadanaya | tena yatha mukham kamalamiveti kriyate na tatha sadharanadharmaprayogam vina mukham kumudabhiveti | 'upamitam vyaghradibhih samanyaprayoge ' ( pa . 2.1.56 ) - itivadekataropadane'pi (i. e. mentioning either upamana or upameya ) ayamarthah sidhyatityubhayopadanam dvayorapi siddhasvabhavatve upama karyeti darsanaya | tena nirjagmuh karmukattasya patrinah prajyatejasah | varidharah sahasramso jvalantyo mandaladyatha || - ityadernirasah | Here evidently the nature (svabhava ) viz. prajya tejastva of patrinah or jvalattva of varidharah is not siddha, already established. As usual the whole of vivrtikara 's comment on the upama is elaborate though learned and characterised by the independent views of the author. His two chief characteristics viz. his antagonism to Induraja and his indebtedness to Mammata are also conspicuously apparent. On a perusal of his commentary on this portion we find that he refutes and criticises the following points from the Laghuvrtti of Induraja. 1. cetoharityanuvadah | praptarthatvat| ( P. 18. L. 7 of the text ) 2. The interpretation of mitho vibhinnakaladisabdayoh 28 mithah parasparam vibhinnah kaladayah pravrttinimittabhuta yayoh sabdayostathavidhau etc. (see P. 18, LL. 7-18). The vivrtikara separates the words as mithah avibhinna instead of mithah vibhinna and has altogether a different explanation to offer. 3. tadevamesa saptadasavidha granthakrta upama pratipadita | ( P. 29. L. 4. ) Induraja divides into 17 varieties. But far on

32 Kavyalankara sara sangraha. the authority of the same Karikas of Udbhata extends the number of varieties to 21. 4. Induraja thinks that 'ayah sulikah ' can be regarded as an example of taddhitavaseya upama . vivrtikara avers that this example cannot be considered as constituting upama ; it might be regarded as an example of atisayokti, (following the opinion of mammata ). P. 17, L. 23. - samipe ksipyate tasmadupameyam | The Mahabhasyakara, Kaiyata and others interpret the word this way : upa samipe natyantaya miyate paricchidyate yattadupameyam - that which is nearly and not completely measured is c (See Mahabhasya and Kaiyata on 'upamanani samanyavacanaih ' pa . 2.1.55 ). Induraja explains it as tat upamanena sadrsyapratipadanadvarena samipe ksipyate . The explanation seems to be somewhat beside the mark. He takes upameyam in the sense of upadiyamanam for which there seems to be no authority. P. 17, L. 25. - tena dharmena sambandhah | sadharmya is the sambandha brought out bythe property which is common to both upamana and c. Here the instrumental case occurring at the end of the word shows instrumentality proper (and therefore some authors use synonimously for it), and not sahitya accompaniment in the sense of dharmena saha (sambandhah ), as some modern interpretors suppose. dharmena saha sambandhah would make the relation exist between the on the one hand and upamana or upameya on the other which, in fact, the Logicians would call samavaya, a nityasambandha . 3 P. 18, L. 1. - samipyaparicchedahetutvadupama | It is called upama because it serves as the means of nearly measuring the by the standard of measurement (upamana ) on account of similarity between them. P. 18, L. 1. -tasyascalamkaradhikarat | As it falls under the general heading or subject of alamkara s. P. 18, L. 2. - gunavyatiriktatve sati alamkaratvat | The line gives a logical definition of an alakara-as gunabhinnatve sati kavyasobhavaha- dharmatvam alamkaratvam . P. 18, L. 3. - kavyasobhahetavo dharmah gunah | These are only three madhurya, ojas, and prasada according to Bhamaha, Udbhata and

Notes. 33 Induraja. This view is adopted and maintained with great force and argument subsequently by Mammata. Dandin and Vamana accept 10 gunas. (See Introduction P. 7. ) P. 18, L. 4- sobhatisayahetutvam | Induraja seems to distinguish gunas from alankaras in this way - kavyasobhahetutvam gunatvam | kavyasobhatisayahetutvam alamkaratvam, and brings in Vamana's authority in support of his view in his words kavyasobhayah kartaro dharma gunah, tadatisayahetavastvalamkarah | (See kavyalamkarasutra - 3, 1, 1-2-3 .) Mammata distinguishes gunas from alankaras in this way: guna rasam vina navatisthante avatisthamanascavasyam rasamupakurvanti ; alamkara rasam vinavastante avasyam rasam nopakurvanti rase angadvarena tisthanti | rasasyotkarsahetavah acala sthitayo gunah ; santam rasam ye'ngadvarena jatucit upakurvanti te'lamkarah | (See ka . pra . 8, pp. 464-470.) P. I 8, L. 7 -cetoharityanuvadah | The expression cetohari (charming) in the definition of 34 is redundant as the notion of charmingness is already implied by considering upama as falling under the head of alamkara, and the term alamkara carries with it the notion of charmingness or strikingness. The vivrtikara has a very elaborate note on the word cetohari in Karika 11. He rebuts the contention of Induraja that ceto- hari is simply anuvada and assigns an important cance to that word. He says that is put in the definition in order to keep off the upamadosa s viz. jatipramanagatanyunatva, adhikata va tadrsi, anucitarthatva etc. His statements are based on Mammata's treatment of upamadosa s (Vide K. P. X. pp. 772-775). He observes -' candalairiva yusmabhih sahasam paramam krtam | iti jatinyunatve, 'vahnisphulinga iva bhanurayam cakasti | ' iti pramananyunatve, sa munirlanchito maujaya krsnajinapatam vahan | vyarajannilajimutabhagaslista ivamsuman || ' - iti tadidrpadharmanyunatve, 'visantu vistayah sighram rudra iva mahaujasah | - ityadhikye, 'patalamiva nabhiste stanau ksitidharopamau | venidandah punarayam kalindigatrasamnibhah || iti pramanadhikye, 'salanchanendubimbabham mukham bhati mrgidrsah | iti dharmadhikye, 'sarasmicancalam cakram dadhadevo vyarajata | sabandhavagnih savartah srotasamiva nayakah || - iti dharmanyunadhikatve, 'bhayeneva smareneya- maho svidyati vepate | ' iti prasiddhivirodhe ca upama na vidheyeti pratipadanaya cetohari - grahanam | na tvalamkaraprastavat sattvaprameyatvadisadharmyanirakaranarthatvenopayujyate iti vispastartham | The otherwise of the has again a long note on the faultiness or when it undergoes a change as it 5 [. . .]

is connected with the upameya and upamana under different circumstances of kala, purusa, vidhi, linga, vacana, pramana etc. He closely follows the argument of Mammata. He says- 'nanu rama ivayam raja bhati 'ti, 'bhasi tvam haririvavaninathe 'ti, 'dahatu bhavandahana iva dvisah samagrani 'ti, 'satrivayam naiva ganyata ' iti, 'sarasijavanamiva nayane tasyah svara- muccakaistanute ' iti, 'kim dustam mano dustam evamvidhe visaye pratiyamanadharmanimittopamana sambandha- danantaramupadiya manava menabhisambandhadanyatha duhsamanvayat | nanu yudhisthira ivayam satyam vadatityupatenaiva samanena dharmenopamayah pratiteh kathametaduktam samgacchate | yudhisthira iva satyavadi satyam vadatiti pratipatteh samgacchata evaitat | satyavadi satyam vadatiti ca na paunaruktayamasankaniyam | repo pusnatitivadyudhisthirasatyavadanena satyavadyayamityarthantara | etacca sthitesu prayogesu samarthanam na tu sarvatha niravadyam | 'kantimavadanam kante rakakamuruvattava ' ityadau krtvapyupamane vibhaktiviparinamam sadharanadharnasya lingaviparinamah karya iti dustatvameva | 'saroruhavadenaksi mukham te kamala- kulam ' ityadau tu sadharanadharmo na viruddha ityupamanaparinamo'pi na dosa iti kecit | yadanekapayodhibhujastavaiva sadrso'sya hinasuratarasah | nanu valajitah katham te sadrsastadasau surudhikrtah || tatha- pararudhiramoditaruci tvayi samamasipatrametadabharanam | vedanavasudha bhavataiva bhavanmatih prabho tulya || iti nipunaprayogadarsanatsadosa eva parinama ityanye | The whole passage is a paraphrasis of Mammata's remarks on the as of Upama given at the end of the Kavyaprakasa, Ullasa X Compare lingavacana vidhyadi ) bhedopi upa- manopameyayoh sadharanam cet dharmamanyarupam kuryat tada ekatarasyaiva taddharma samanvayavagateh savisesanasyaiva tasyopamanatvamupameyatvam va pratiyamanena dharmena pratiyate | (K.P. X. P. 774) So also - yatrapyupattenaiva samanyavarmena upama avagamyate, yatha- 'yudhisthira ivayam satyam vadati iti ; tatra yudhisthira iva satyavadyayam satyam vadatiti pratipatsyamahe | satyavadi satyam badatiti ca na paunaruktyam asankaniyam, raiposam pusnatitivat yudhisthira iva satyavadanena satyavadyayamityarthavagamat | satyametat, kimtu sthitesu prayogesu samarthanamidam na tu sarvatha niravadyam, prastuta vastupratitivighataditi sacetasa evatra pramanam (K. P. XP. 781-82). P.18, LL. 7-9 - natyantam sadharmyena | upamanopameyabhava does not exist where there is complete similarity resulting in identification as in the instance gaurivarya gauh . Therefore the expression mithovibhinna sabdayoh is inserted in the definition. P.18, LL.9-19 - pravrttinimittabhutah | Here the words kala, desa &&c, are intended to convey the idea of the motives for

Notes. 35 the currency or use of words ( pravrttinimitta ) showing time, place, action &c. Some words such as and others show time, praci and the like show direction, go &c. genus, sukla &c. quality, gacchati &c. action, rajapurusa &c. the relation of a servant to his master. Our pravrttinimitta s ie motives for use of words may thus be multiplied mitho sabdayoh | The whole expression is a Bahuvrihi upon Bahuvrihi, thus:- mithah vibhinnah kaladayah yayosti | etadrsau sabdau vacako yayostayoh (upamanopameyayoh ) | na pravrttinimittabhedah | - In the instance gaurivarya gauh there is no difference in the pravrttinimitta ( gotva ) of both the gosabda s as gotva cowness is the same pravrttinimitta of both of them. There can, therefore, be no upamanopameyabhava in this instance. The fa strongly criticises Induraja's interpretation of the second half of Karika 11. He separates the words as mithah avibhinna and rebuts Induraja's theory of pravrtti- nimitta . He says - 'rama iva bhati rajeti kalabhede, 'haririva bhasi rajan ' iti purusabhede, 'pratapatu bhavanprabhakara iveti vidhyadibhede, 'candrakaleva sugara ' iti linga- bhede, 'pibatyananametasyah puspam madhuliho yatha ' iti vacanabhede yada sadharano bhedasabdah tadapi na karyeyamiti darsanaya mitha ityadyuktam | mitho'nyonyagatatvena vartamano'rthat sadharanah avibhinnakaladih sabdo yayorityakara praslesah | na tu bhinnapravrttinimittayoriti vyakhyeyam | jatyadinam pravrttinimittatve kalatvatathatvaprasiddheh, mukhyasabdatyage nopala- ksakasabdagrahanasya nirarthakatvapatteh, mitha ityasyanupadeyatvaprapteh so'svah suva [sa ]svena samah sasattva ityekatvabhave'nanvayatvanupapattau upamarupatvaprasangaca | P. 18, L. 18 - upama tu tat | The particle tu separates upama from another alamkara . P. 18, L. 20- esacopama | Divisions of upama- upama sampurna chuva (trividha, sakalyena pancavidha ) (sakalyena dvadasa kara ) vakyavaseya (dvividha ) sriti samasavaseya (ekaprakara ) taddhitavaseya (dvividha ) arthi (avyayavaseya ) arthi (sadrsadipadavaseya ) P. 18, LL. 24-27. - avyayam hi without a case termination, is not decidedly connected srauti bharthi avadyotayati | avyaya, being

36 Kavyalankara sara-sangraha either with or with and therefore shows the upamana upameya relation (sambandha ) in the form of karma, karana or bhava (object, instrument or state of being) as the case may be, the latter being the subject of the action of comparing the upameya with upamana ; thus upama is srauti there, as is said by the author yathevasabdayogena sa srutthanvayamarhati | vasabdadeh .. P. 19, L. 1.-C. The particles and others also bring out upamanopameyabhava just as yatha and iva do L. 6. - padminim vanyarupam | Here va is used in the sense of iva as anyarupam padminimiva . L 8 - ksanam - anavasthitam - unsteady. L. 13. sritena rupena merely by word or hearing. ubhayadharam - the relation upamano- pameyabhava existing in both. LL. 16-17 - netrairivotpalaih ...... | sarahsriyah with utpalaih, padmah and cakravakaih are the upameya s, tarunyah with netrah, mukhaih and stanaih are the upamana s, bhanti is the sadharanadharma, and iva is vacaka ; therefore it is srauti sampurna vakyaga upama . P. 19, LL 19-22. - ya tupama tenasau arthi | But when an occurs with words and the like, there the relation between and is not known merely by hearing the words but after the sense is apprehended, as • words like sadrsa, tulya are grally connected with only one of them, its connection with the other being understood after its similarity is recognized with the other; therefore it is an arthi upama . anyatha means otherwise than sritena rupena 1. 6. arthena rupena . P. 20, L. 3- pratimasabdah | Here the word pratima brings out the relation between upamana and upameya . It is grammatically connected with 3c when the compound is taken as a ; and in this case it brings out its similarity with ccnot merely by hearing but by understanding the sense; for we know is f , residing in two objects. But when the compound with pratima is taken to be trtiyatatpurusa, then is directly expressed in and indirectly in upamana ; thus in both cases the upamanopameyabhava is artha . P. 20, 11-12 LL 11 - 12 - aparamapi sadharmyam | Here by the force of the property kinjalkalinasatpadatva found only in the upameya -kanana, another common property viz. is suggested though not expressed. As quis spotted with black owing to the apparent image of a hare in the middle of the disk, so

Notes. 37 kumudakanana is also marked with black spots owing to the bees lying in the filaments. Therefore owing to the property not being expressed in respect of the disk of the full Moon, the instance may also be taken as of dharmaluptopama . P. 20. LL, 16-17 - tesamanupravesat | In the example netrairiva &c., though the common properties dirghatva, nilatva &c. of the members netra, utpala &c, of the subordinate upama s are not expressed, still they enter into the common property which is expressed, therefore the upama is purna, as all the four requisites are present. sadharano dharmah |-bhanti is here the sadharana dharma which even extends to (includes ) the qualities of dirghatva, nilatva &c.; therefore this upama is only sampurna . But the upama in prabodhaddhavalam &c. is purna and also lupta owing to the absence of the common property asitodaratva . Also indubimbapratimam is a com pound; therefore this is samasaga upama . If it is to be made vakyopama, then simply change the expression as akhandenenduna tulyam &c. P. 21, L. I - taddhitavaseya sampurna srauti is when the taddhita termination vat is made under the rule tatra tasyeva (Panini 5. 1. as in mathurayamiva mathuravat sune prasadah and caitrasyeva caitravat maitrasya gavah . It is eff when the same termination is made under the sutra tena tulyam kriya cedvatih (Panini. 1. 115 ) as in brahmanena tulyam brahmana- vadadhite ksatriyah . In the first case, vat shows the upamanopameyabhava of both in the same breath merely by hearing as in the case of a; but in the latter case in the sense of is grammatically connected in the first place with the action of the upameya, and then by cognisance of the meaning, it is connected with that of the upamana, and thus brings out the upamanopameyabhava of both; therefore it is artha . P. 21, L. 8- taduktam | The author brings out the sense by the words vatina catadabhidheya in the Karika. karmasamanya- vacano vatih = the termination vat is enjoined when the kriya is tena tulyam by Panini 5. 1. 115. vacyeti | - (sa upama ) vacya means srauti . L. 14 - purvasya udaharanam | that is of sampurna arthi taddhitavaseya . L. 19-afa...... Here the termination is affixed by the rule tena tulyam kriya cedvatih | patitanabhradrstya tulya patitanabhravrstivata (tisthet ) |

38 Kavyalankara sara-sangraha. P. 21, L. 16 - utkalikah = utkanthah = longings. bhadvat = mama iva = liko mine. akandoddamarah = suddenly provoked or furious. pracanda is another reading which means violent, enraged. = P. 21, L. 27 to P. 22, L. 4- asmadarthah | aham in madvat is the upamana ; tacchabdarthah--sa in tasyah is the upameya . nirargalah- unbrialed, unchecked. samarpitah = bestowd. syuh - bhaveyuh = would be Here utkalikakartrkam bhavanam in bhaveyuh is the sadharanadharma . vatisca | - Here the termination cat is applied under tatra tasyeva ( Panini 5. 1. 116); thus it is sampurna srauti taddhitavaseya . S P. 22, L. 4-5 tasya apitipathe | - Here when the reading is tasya api, the vat termination is to be applied to the base of aham in the genitive case; when it is tasyamapi it is to the base in the locative case. L. 9. - pancaprakara &c. - Thus sampurna is of five kinds. See the table on page 35 P. 22. L 10. tasyasca pancavidhatvam | luptopama is also of five kinds. See the subjoined table. luptopana krdavaseya taddhitavaseya (catusprakara ) (dviprakara ) (ekaprakara ) vakyavaseya samasavaseya subdhatupratyayavaseya (ekaprakara ) (catusprakara ) P. 22, L. 14. - luptaikadesatvamapi | Here (in the example akhandenenduna etc.) when we take asitodaratva as the sadharanadharma, it is not present in both; so it is also an instance of a luptopama .. L. 17 - samyavacakavinyuteh | When the sadharanadharma is dropped. tadva- civirahena | when ivadi is dropped. tadvaci = upamavaci . L. 21- trayam anusajyate | Here in the whole section the three expressions - samksepabhihitapyesa, kvacit samase and nivadhyate are to be understood wherever they are required by the context. L. 22 - samya- vaciviyogena | By dropping the sadharanadharma . L. 25 -sabdena nopattam | Here the common property is not expressed by words, as it is to be understood by implication. P. 23. L. 5 kadambah = kalahamsah = a kind of goose with dark gray wings. sarojakarnika gauram | beautiful like the pericarp of a lotus. L. 8- yah samasastatsamarthyat | Here the compound karnikagauri is to be taken as karnika iva gauri like sastrisyama by the rule upamanani samanyavacanaih (Papini 2. 1. 55.) ; therefore it is vacakalupta . L. 10 - evamekalope sati dvividha | 1 dharmalupta and 2 vacakalupta

Notes. 39 samasopama . L. 12 - samyatadvacivicyavat | When only the upamana and upameya are mentioned, and sadharanadharma and vacaka are dropped simultaneously, it is samasavaseya dvilopa . L. 13-laksyate | Here by the word samya, samyavacakasabda is to be understood by laksana, as the idea of similarity () is impossible to be dropped in upama . P. 23, L. 15 --The two instances of dilopa samasavaseya are tridasadhisasardula and nilotpaladaleksana . In the first instance tejasvitva or some such dharma is to be understood as sadharanadharma by implication and in the other, dirghatva, nilatva as sadharanadharma . tridasavisasardula is 3 tatpurusa compound to be dissolved as tridasavisah sardula iva (surah ) by the rule upamitam vyaghradibhih samanyaprayoge (Panini 2. 1. 56 ); while nilotpaladale iva iksane yasyah sa nilotpaladaleksana is a Bahuvrlhi compound. • P. 24, L. 1-tritayalope tu tadvaciviyogat | trilopa samasaga occurs where the words expressing sadharanadharma, upameya and vacaka are simultaneously dropped. L. 3-4 - atrapi ... laksyate | Here, too, by the words samya and upameya, words expressing them are to be understood by laksana as before. The example is sasi- cchayavadana . Here the compound is a Bahuvrihi upon Bahuvrihi and to be dissolved thus : - sasicchayaya tulya chaya yasya tat | etadrsam vadanam yasyah sa | In this instance sasikanti - the upamana, vadanakanti the upabheya, ahladakatvadi the sadharanadharma common to both, and sadrsya all the four are cognised; but alone is expressed in words as sasicchaya, the other three being understood by implication by the power of the compound. P. 24, L. 15-subdhatupratyayavaseya | This upama is of three kinds and is brought out by terminations such as kyac kyan and kip which are applied to pratipadika s or nouns. That which is formed by is also of two kinds:-one formed by applying the termination to a noun-root which is the object () of the action and which serves as upamana in a certain subject when behaving ourselves towards the object' is the intended sense, by the Sutra upamanadacare (Panini 3. 1. 10.), and the other formed similarly under the same circumstances, when the noun shows location (adhikarana ) by the Vartika adhikaranaceti vaktavyam on the same Sutra. (See Panini 3. 1 10, and the Vartika on it).

* P. 24. LL. 18-22 - yatha samase ...... nibadhyate | Just as a luptopama in a samasa is formed and understood by the force of the compound, in the same way, the c involved in an expression which is formed on the authority of kyaca termination enjoined in sutra ( sautrah ) upamanadacare (Panini's 3. 1. 10.) and in the vartika on it (aupasamkhyanikah ), adhikaranaceti vakta applied to in a certain subject, which is the object or location of an action respectively when 'behaviour towards' is the intended sense, this is formed similarly and is understood by the power of its formation. P. 24. L. 26 to P. 25, L. 4-: Here (meaning-one overpowered by extreme poverty) is the upamana, bhagavatah sivasya atma · is the upameya, acara ( behaviour towards ) is the sadharanadharma given by the kyacpratyaya . duhsthamivatmanam acaran behaving towards oneself (showing onself) as a poverty-stricken beggar. atra ca ...| Here upamana and sadharanadharma are expressed in words, and and are to be understood by implication; therefore it is dvilopa subdhatupama . This is the instance of a subdhatupama when the object is the upamana . narakiyati kanane is the instance of subdhatupama when location is the upamana . L. 6-atra narakah &c. Here naraka is upamana, kanana is upameya, acara (behaviour towards it, brought in by termination) is the sadharanadharma, upamavacaka ivadi is not used as the relation of upamana and upameya is known by the power of kyac termination; therefore it is ekalupta, that is vacakalupta . subdhatupama . kyan P. 25, L. 11-The c brought out by the force of the termination is when the doer of the action serves as upamana, as the author says karturacare kyana sa . L. 11- tatheti &c. / The expressions tatha, upamanadacare and kacita mentioned before are to be understood here. So the whole sense of the sentence is that a c is sometimes formed by the termination kyan (by the rule kartuh kyan salopasca Panini 3. 1. 2.) in the sense of 'behaviour towards'applied after that which is the doer of the action and serves as GcH at the same time. P. 25, L. 17 - khadyotah = jyotirmalika = a line of light. It also means a fire-fly or the sun, a distinct knowledge

Notes. 41 of the real nature of things. L. 1: The 19 vayapratyayopattah common property is an (behaviour) set forth by the termination kyan Here is not used as its sense is implied by the force of termination, therefore it is ekalupa, that is, vacakalupta suravatunama LL. 21-25- kamanikakapa This is formed by the termination which is similar to ksad in having the karta as upanana . 'sarvapratipadikebhya ityeke ' - This is the wording of the vartika found in kasika vrtti . The as of Katyayana discussed in the Mahabhasaya is 'sarvapratipadikamyah kivva vaktanyah ' on the Sutra 'kartuh kyan salopatha (Panini, 3. 1. 11) which means upamanavacakebhyah kartrvacibhyah sarvebhyah pratipadikebhyah acare vipratyayo vikalpena vaktabhyah | P. 26, L. 1-is the present participle of the no in-root formed by applying termination to krsanu meaning fire. L. 4-na khalvasruyamanasya vipah- Here acara, the, is to be understood by implication. It is not expressed as the termination kinu which is applied for acara, itself vanishes. g P. 26, L. 6-... Thus the formed from (noun-roots) is of three kinds by the application of the terminations 1 kyac 2 kyad and 3 kipu . That which is formed by is of two (2) when adhikarana is upamana ; those formed by kyan and also by faq have no subdivisions. So all these taken together are four kinds of luptopama . kinds-(1) when i and P. 26, L. 11- krtpratyayasamarthyavaseya | The luptopama occurring in the gerundial formation by which is a termination (i. e. a participial termination) is of two kinds-when the termination is applied to the object or subject of an action serving as upamana by the rule 'upamane karmanica ' (cat kartari ) - (Panini, 3. 4.45 ) as is said by the author himself upamane kumana va kartari va yo namul kasadigatah | tadvacya sa | kasadigatah- The verbal form of the same root of which the gerundial formation (a) is enjoined, is to be used immediately after it, in the case of the roots and others, by the Sutra yang ygaan:' (Panini, 3. 4. 46.). 6 [K. B. &]

42 Kavyalankara sara.sangraha. | P. 26, L. 15. - itaramanodaham | This is an instance of luptopama formed by the gerundial termination namul applied to the root dahu the object of which is the upamana . Here itaramanah = prakrtamanah 'ordinary man's mind' is the upamana, bhagavanmanah is the upameya, dahyamanatva the property or condition of being burnt is the sadharanadharma . All these three are expressed by words, sadrsya being implied by the force of the termination, in the absence of the particle or the like; therefore this is the instance of a luptopana (where only one requisite viz. vacaka is dropped) formed by namul in which karma is upamana . L. 23. -- samanya pranicintam | samanyaprani = an ordinary (being) man void of excellent qualities. sah = bhagavan cintayitrtvam = 'contem plating' is the common property. pranicintam is the namulanta gerund. P. 27, L. 3 -kalpabadestaddhitasya prayogat -- By the application of the aftermination and others of like nature in the sense of isadasamaptau i. s. sadrsya (See Panini, 5. 367). L. 5- prabhrtisabdena - The word prabhrti includes the terminations kan and others of similar nature which are enjoined by the rule a pratikrti ' (Panini, 53.96 ) in the sense of 'when the image is like the original', as asva iva asvakah, asvasya pratikrtirityarthah LO kalpa pratyayena - The idea of similarity is set forth by the termination kalpu . L. 11 - prakrtyartha sadrse ... vidhanat - The terminations and others of like nature are prescribed in the opinion of the revered Katyayana in the sense of 'an object similar to the original' (Of Katyayana's Vartikas on Panini, 3. 3.67). L 14 - asvakah ityatra | Here the termination kan ise joined by the rule 'ise pratikr ' and 'samjnayam ca ' ( Panini, 5.3.96-97) in the sense of the image or being like the original' to which the termination is applied as asva iva pratikrtih asvakah a being or image like a horse. P. 27, L. 15- ayahsulikah is a taddhitavaseya tritayalope luptopama . ayah sulenanvicchati is the explanation of the taddhita termination thak which is applied by the Sutra 'ayah suladandajinabhyam thakthani ' (Panini, 576). anvicchati = vyavaharati behaves, aims at tiksnah upayah ayahsulam (ayah sulenavyavasitah ), tena anvicchati vyavaharati iti ayahsulikah, sahasikah krurah ityarthah | Compare mahabhasya and kaiyatah -ayah sulamivayah sulam | yo mrdunopayenanvestavyanarthat rabhasenanvicchati sa ucyate ayah sulikah | mahabhasyam

Notes. 43 (5.2.9.) || yathayahsulam tiksnamevanyopi tiksna upayo'yah sulamupacaraducyate | kaiyatah || L. 18 - atra ...... | Here ayahsulam is upamana, kascidarthanvesanopayah is upameya, tiksnatvadih sadharanadharma and upamanopameyabhava - all these four requisites are found. Out of these only the upamana ayahsula is expressed in words; the other three are understood by implication. P. 27, LL. 21-27 - natu catra &c. Now here is a question. Here as the upameya is swallowed up by the upamana ayah sula, and as an anvesanopaya the latter appears completely identified with it, it must be an instance of atisayokti and not of upama, then how is it quoted as an instance of upama ? The answer is :-As in the instance sasicchayavadanam, although vadanacchaya beauty of the face is completely shaded or hidden by beauty of the moon, by adverting somehow to the knowledge of ¿ difference between the two, it is given as an instance of a kind of I, so here too a similar argument may hold; thus there is no harm in taking it as an instance of upama . So, in the opinion of Induraja, ayahsulikah is a taddhitaga trilopa luptopana . But Mammata severely criticises this view, and by turning the tables by the same argument rejects this as an instance of trilopa luptopama and establishes it as an atisayokti . (Compare Mammata's at the end of the explanation of the Karika 'trilope ca samasaga .' Kavyaprakasa, P. 578.) The fa also does not forego this chance of criticising Induraja. Following Mammata e avers that :- sulikah cannot be regarded as an example of upabha He Says:- ayah sulenanvicchatityayah sulikah ityatrayah sulasyopama nasyaivopamanadartharjanopayasya kasyacidupameyasya, taiksnyadeh sadharanadharmasya, dyotakasya ca lopatrayalopa upameti na yuktamabhidhanam | artharjanopayasyayasularupatayadhyavasanadatisayoktitvam hyacitam | sasicchayavadana mitivadbhavisyatiti ca na yuktah pariharah | atrapi sasi cchaya yasyeti vivaksayamupamavisrantavapi rupakatvameva | rupakadavapyasti hyupamavisrantih | Those who compare the remarks of Induraja on this topic with those of the quoted here will not fail to notice how the views of the two commentators are directly antagonistic to each other. P. 28, L. 1 evam sva mumursati, kulam pipatisati ... And in the same way, if, in the instances sva mumursati and kulam pipatisati, there be something which would bring out the idea of difference

between the wish, expressed by the termination, which is the and the tendency of dying or falling which is the 3, the latter being swallowed up by, and thus completely identified with, the wish, then these may be given as instances of lutopana . Otherwise they may be regarded as instances of atisayokti . For a complete discussion of all see Patanjali's Mahabhasya on 'dhatoh karmagah sama kartrkadicchayam va ' (Panini, 4. 1. 7 ) together with the gloss of Kaiyata on it:-these cases mahabhasyam - asankayamacetanesupasamkhyanam | asankayamacetanesupasamkhyanam kartavyam | asma luluthite, kulam pipatisati | kim punah karanam na siddhyati | evam manyate cetanavata etadbhavaticcheti | kulam cacetanam | acetanagrahanena narthah | asankayamityeva | idamapi siddham bhavati sva mumursatiti | na va tulyakaranatvadicchaya hi pravrttita upa- labdhih na va vaktavyam | kim karanam | tulyakaranatvat | tulyam hi karanam cetanavati devadatte kule cacetane | kim karanam | icchaya hi pravrttita upalabdhirbhavati | yopyasau katam cikirsurbhavati nasavaghosayati katam karisyamiti | kim tarhi samnaddham rajjuki lapulapani drsta tata iccha gamyate | kulasthapi pipatisato lostah siryante bhida u jayate desantara- mupasamkanati | svanah khalvapi mumursavah ekantasilah sunaksasca bhavanti | upamanadva siddhama | upamanadva siddhametat | luthisata itra luluthisata | pipatisati iva pipatisati | na va tinantenopamana nasti | evam tarhi icchatreccha sarvasya va cetanavatvat | athava sarve cetanavat | kaiyatah | asankayamiti | asanka sambhavana | | luluthisata iti | | vartamanasamipya iti lat pratyayah | sva mumursatiti | sunacaitanyepi jivi- tasya priyatvadyadhyadyabhibhavepi tiryaktvanmartumiccha nasti | na ceti | karyenecchanumiyate ..... karyanumeyecchasrayega san prayujyate ityarthah | upamanadveti | yadicchanimittam karyam tadacetanesu nasti | yaccasti na tadicchayah karyamiti matvapariharantaropanyasah | na vai tinanteti | tinantarthenetyarthah | kriyayah sadhyaikasvabhavatvadanispannarupatvadidam taditi- paramarsavisayavastugocaratvadupamanopameyabhavasyedam taditiparamarsabhavaditibhavah | iva- sabdaprayoge tu abhyarapatu vidyate | roditiva gayati nryativa gacchati devadatta iti | paripurnena ca nyunasthopamanam bhavati | kriya ca sarva svasraye samadeti nyunatvasam bhavasta- syam | | | sarvasya veti | atmadvaitadarsaneneti bhavah | ..... | vedah sarvabhavanam caitanyam pratipadayatityarthah | P. 23, L. 6- asankayam | These three are the Vartikas of the Venerable Katyayana on the Sutra 'dhatoh karmanah samanakartrkadicchayam va ' (Panini, 3. 1. 7) which enjoins the termination to be applied to a root, in the sense of "desire."

Notes. 45 asankayama cetanesupasamkhyanam - As iccha is a cetanadharma, the termina tion showing cannot be applied in the case of a or inanimate objects. A is such an inanimate object. To include the cases of inanimate objects, g- yam &c. should be inserted in the sutra, asanka - sambhavana possibility. na va tulyakaranatvat - Or asankayam may not be inserted. For, the same reasoning can be applied to observe which is inferred from or tendency in the case of inanimate as well as in the case of animate objects. When get is thus inferred from tendency, it may be found both in inanimate as well as animate objects equally. For instance, when a potter wishes to make a az jar, he does not proclaim that he is doing so, but we infer that he wishes to make a when he is found concerned with clay, wheel, rod &c. In the same way when we observe that stones are falling and changing their places and that cracks are produced, we infer that the bank is about to fall, i. e. wishes to fall ( fa). A dog, also, when he is about to die (ar g), prefers loneliness and his eyes are morbidi swollen. upamanadvasiddham - Or kulam pipatisati may be taken to mean Afandi fagfa wishes to fall as it were i. e. is likely to fall, as an animate being wishes to fall, by taking the case of an animate being as upamana . P. 28, L. 8-atra hina va Here by the expression etc. in the Vartika, determination of identification between the two objects is suggested; but by the vartika- 'upamanadva siddham ', the upamanopameyabhava of them is expressed. L. 10- iyam ca ...| In the example kulam pipatisati and the like, as the termination is applied to a c, and as the dhatu, sannantadhatu becomes a dhatu again, the upama involved in it may be said to be a dhatucatupama like a subdhatupama (ie a namadhatuna ma .). P. 28. L. 12- va manasamipyadi | The instances like in the present tense for the immediate past or future which are erjoined by the rule 'vartamanasamipye vartamanavadva ' (Panini, 3. 3. 131) may be taken to be examples of a kind of or an according to one's own clear apprehension.

P. 28, L. 14. - curnikarasya tu astiti | curnikarah vyakaranamahabhasyakarah | curnikrdbhasyakarah syaditi trikandasesah | But in instances like these, the Mahabhasyakara prefers to have a kind of ff by adverting to the determination of identification, as, he says, 'na tinantenopamanamasti ' -upamana can never exist coordinately with a fc, i. e. in 3c or comparison & kriya can never be the upamana . In an upama, a sense ( an object) which is siddha is alone cited as an usmana . As kriya is always (in the process of being accomplished) and never f it cannot be introduced as 34. Compare the Karika- upamana . siddhameva samanarthamupamanam vidhiyate | tinantarthastu sadhyatvadupamanam na jayate || Though tinanta can never be aa upamana, it can be an upameya with another, in a few cases, such a: (Compare the whole passage in the Mahabhashya on Panini's Sutra :: &c. 3. 1. 7 quoted above). Mammata seems to follow the Mahabhashyakara in all these instances. P. 28, L. 15-For this reason Dandin has said with a great display of reasoning that there is a kind of utpreksa involving atisayokti in the instance limpativa tamongani &c. limpativa tamongani is & murddhabhisikta example of utpreksa and is discussed by almost all Alankarikas. This reference to Dandin strikes one as most opportune and proper. Dandin has an elaborate discussion on the example fo in which he refutes the arguments in favour of the opinion that it constitutes an c, and establishes with great force of argument that it is a clear instance of . (Vide this utpreksa . kavyadarsa-2 226-233 .) P. 28, L. 18 -tena kada devadatta ... Thus in instances like kada devadatta gramam gamisyasi - esa gacchami, the future tense is identified with the present tense being very close to it; hence it should be regarded as an instance of atisayoki . L. 21 The same reasoning may be applied in the case of other instances also. P. 28, L. 22 to P. 29, L. 3- dvividhatvam | The whole passage summarises the various divisions of rupama .

Notes. They are given in the subjoined table :- vakyavaseya 1 sriti upama 100 sampurna (pancavidha ) lupta ( dvadasavidha ) samasavaseya dadvitavaseya 3 adhi 2 arthi 4 srauti 5 arthi (avyayaya seya ) (sadasadipadavaseya ) 1 bakyavaseya samasavaseya 47 sundhatrpratyayavaya krdavaseya 12 taddhitavaseya ekalopa 4 dvitayalopa 5 tritayalopa 10 karmopamanika 11 karmupamanika 1 2 dharmalupta 3 vacakaluta kyajavaseाkyatyayavaseya 19 kimvaseya 8 kapama nika 6 karmopamanika 7 adhikaranopama nika P. 29. L. 4 - tadevamesa ... | Thus 17 kinds of upama are explained by the Commentator 5 kinds of sampurna + 12 kinds of lupta = 17. Mammata recognises 25 kinds of upama, 6 purna and 19 ha . Uabhata's divisions of luptopama as well as of sampurna are based on the difference of the substratum (0. 4. vakya, samasa etc.) in which the 34 is found; while Mammata's divisions are mainly based on the principle of the of one, two or three of the requisites of upama . It is sometimes interesting to see how two commentators differ in their interpretations of the same text. We have pointed out many instances where the fa criticises the views of Induraja; but the occasions where the vivrtikara vitally differs from the actual interpretation of the text given by Induraja have been comparatively few. We have already alluded to one such instance, viz. the different interpretations of mithovibhinnakaladi etc. (Karika 15) given by Induraja and the vivrtikara . The present case is another instance in point. Induraja sees in the Karikas of Udbhata only 17 varieties of Upama described, while the vivrtikara can bring out as many as 21 varieties from the same Karikas. Such divergence is perhaps rendered possible by the

laconic and incomplete expression of Udbhata. It would seem that all minutiae of division were not settled in Udbhata's time and that he did not even intend to state or imply any definite number of the divisions of Upama. But the commentators, coming later, according to the developed ideas of their times, try to fix upon a definite number of divisions and ascribe those to the original author. Thus the discrepancy between the interpretations of Induraja and the far would only indicate that the divisions of had increased in number and in scientific precision in the age that passed between Induraja and the modern commentator. The vidyutikara thus states his 21 divisions- eva (upama ) ca dvibheda samksepabhihitapyevetyukteh | purna luna ca | aya triya vakya samasa-taddhita- gamyatvat | pratyekam srautatvarthatvabhedadvaividhyam | evam bamda purnabhedah | 6| luptayastu vakya- samasa sutratu krttaddhitavaseyatva pancavidhyam | vakyagamyayah srautatvarthatvanedaidai- vidhyam | 2 | samasa gamyayacaka-dvi-tri-lo pabhedatsadharanadharmalope sratatvarthatvamedadadvau bhedau dyotakalope tveka ityekalopasya bhedatraya sadbhavat pancavidhyam | 5 | sudadhatugamya- yastu kyac kyan kip pratyayagamyatvat kyacpratyayagamyayasca karmadhikaranopamana- svat karmopamani kayancaikavilopameyatvatvavidhyam | 5 | karmakarmupamanatvena krdavaseya dvidha | 2 | taddhitagamyatvekaiva | 1 | evam pancadasa ] luptabhedah (2+5+ 5+2 +1 = 15 ) purnabhih saha ekavimsatih | (15 +6= 21 ). It will be seen that the recognises two divisions. (viz. srauti and arthi ) of pugo vakyagamya and lupta vakyasya while Induraja recognises only one variety viz. ff of these two vakyagamya upana . Thus the vivrtikara 's number of varieties is increased by two. Again the far puts down two divisions sramti and arthi for ekalope sadharanadharmalupta samasavaseya, while Induraja here also recognises only the arthi division. The vidyutikara is thus enabled in all to increase his varieties by three on the strength of the example natrairiva palaih padmanukhi on Following the dictum of the grammarians viz. ivena samaso | vibhaktayalopasca the vivr tekara optionally regards netrairiva as a com pound ; accordingly this becomes, in his opinion, samasaga upama which, as it is accompanied by iva, is to be regarded as a srauti variety. When we regard the samanadharma as upatta in the form of the verb bhanti this becomes an example of purna samasaga arthi . When we take asitodaratva etc. as the samanadharma which is

Notes. 49 here anupatta, this verse becomes an example of ekalope dharmalupta samasaga or vakyaga according as we consider netrairiva as forming or not forming a samasa . The increase of three varieties is thus accounted for. The addition of the fourth variety is made by regarding the karmopamanika kyajavaseya luptopama as two-fold while Induraja regards it as one. karmopamanika dviva ekalope dvilope ca bhavat | dvilope yatha- 'saduhsthiyan krtartho'pi etc. (St. 26.) | ekalope 'putriyati sisyam '- ityadyudaharyam | vivrtikarah | -------- | It is to be remarked that Induraja does not seem to be aware of the dictum ivena samaso vibhaktayalopah purvapadaprakrtisvaratvam ca | It is a vartika on 'saha supa ' (pa . su . 2. 1.4), and thus possesses supreme authority, only next to that of Panini, for the grammarians. It is really a puzzle how Induraja fails to notice it, as the cannot be a spurious one being noticed by the mahabhasyakara in his exegesis on 'saha supa ' (pa . su . 2. 1. 4). Compare mahabhasya, 2. 1. 2 ivena vibhaktayalopah purvapadaprakrtisvaratvam ca | iveta saha samaso vibhaktayalopah purvapadaprakrtisvaratvam ca vaktavyam | vasasi iva kanye iva | P. 29, L. 4 - tadahuh- | The commentator Induraja gives a quotation in support of his 17 divisions of upamah-krttaddhitasama- sebhyah &c. The first line gives the five principal divisions of luptopama ; the second line, the two broad divisions of upama-purna and ; the third gives the 3 divisions of based on the lopa of one, two or three of the requisites of upama at a time. The first two of these are of two kinds each, the third is of one kind only having no subdivisions, L. 8-purvau bhedau dvidha | ekalopa is of two kinds :- dharmalupta and vacakalupta ; dvilopa also is of two kinds:-dharmavacakalupta and upameyavacakalupta . L. II-esa copama ......| Although this upama is of varied kinds it assumes the character of a only where it pleases the mind and not otherwise. P. 29, LL. 14-17 - prativastupama | upamanasannidhane ... | (Karika 22) yatropamanopameyayoh That is called prativastapama where the word expressing similarity is used severally with reference fo the upamana and the upameya both. The vivrtikara explains the karika in a few words thus :-- matropamanopameyayordvayorapi samipe sadharanadharmopadanam sa prativastupama | ata eva catu 7 [3.5:80]

vakyarthah upamanam vastuno vakyarthantarasyatra canekavakyatve ivaderaprayoge kimupameya kivopamanamiti prakaranaddhi vyavasthetyaha prakaraniketareti | I P. 29, LL. 17-20-a...... Now there is a doubt. If the word expressing similarity is used more than once, then there must be more sentences than one. The words and the like are not used in different sentences, as their power of expressing the meaning is known to exist when they are used in one sentence. Then how can the idea of similarity be understood there in the absence of and other words of similar nature? To remove this doubt the author says:- prakaraniketaratva ... sunyatvam (Karika. 23 ). L. 23. nanavakyatvat | Here though the words like iva are not used as there are different sentences, the relation of and is determined by the force of the meaning when the nature of the prakaranika as well as of the aprakaranika is clear to the mind. The vivrtikara reads upasavacisunyeyam instead of upamavacisunya- tvam and adds in the comment yatprakaranikam tadupameyam itarattapamanamiti | sabdasunyeyamityato'nantaram drastavyah | P. 29. L. 25-ivaderapratitapi nibandhana | iha prakaratrayena ...| upama (comparison) is known in three ways in poetry: In some places 3 is apparent, being distinctly expressed by the sorce of the words and the like, as in the instances : &c. In other places, it is cognised as indicated through the door of the expressed meaning of the words assisted by the peculiar power which they assume when they are in a particular position, i. e. in a compound, as in the instance sastrisyama . See the Mahabhasya on 'upamanani (Panini 2. 1. 55). P.30, L. 5 ...... Here the words are and form one expression by compound, which possesses a peculiar power of expressing a specific meaning; and through the medium of this direct meaning, an indirectly indicated upama is apprehended. L. 6 - kacittapamanopameya ... 1 While in other places, its cognisance is apprehended merely by the force of the meaning, in the absence of a word or a particular operation (a) capable of bringing out the relation of upamana and upameya, as in rupaka, dipaka, prativastupama etc. L. 8- bhatavasyam ...| Therefore this prativastupama, the relation of upamana

and Notes. 51 is apprehended by the force of the meaning alone; so there is no fault or discrepancy (in the definition). LIIP. 30, L. 11- tadrsah | Nominative plural of tadrs feminine gender. fare: ...... Rare are the treasures of good disposition and beauty of that kind in this world; how many nights are there in a year, in which the moon is full? sampadah =sambharah = treasures. tadrsah sakalalokotkrstah =surpassing all men, excellent. viralah = katipayajanajusah = found in some men, rare. L. 17 - viralatvam is the common property which is expressed by the word kiyatyah in connection with the upamana- nisah and by viralah in connection with the upameya - 0 sampadah . L.. 18- ivadyanupadanepi | Here the relation of upameya and upamana is known by comprehending the nature of the and of the aprakaranika though no words like iva are used. 1. 19 - vastuni vastuni | prativastupama is so called because the common property is attached severally to each vakyartha (vastuni vastuni ) Mammata derives the word prativastupama in a different way-vastuno vakyarthasya upamanatvat prativastupama . (See Kavyaprakasa, 10, P. 634.)