arg! BT ACAI waa: ae: | TIO BOOK EIGHTH. Cuarrer First. aie tnd oir ater, FN afew edeatar Cette Saahewa taean! erat am: are Tere: eter T wre Cart era 1. From here upto VIII. 1. 15 inclusive, is to be supplied always the phrase “the whole word is repeated”. This is an adhikara sdtra. Whatever will be taught hereafter upto tea (VIIL. 1. 16) exclusive, there the phrase &iea ¥ should be supplied to com- plete the sense. Thus VIII. 1. 4. teaches “when the sense is that of ‘always’, and ‘each’.” Here the phrase ‘the whole word is repeated’ should be supplied to complete the sense, i.e. “The whole word is repeated when the sense is that of afways and each”. What is to be repeated? That which is most appro- priate in sound and sense both. Thus one qaf& becomes two, as Tata Tate ‘he always cooks’. Similarly writ mar cofta: ‘every village is beautiful’, The sdtra ‘sarvasya dve’, should not be confounded as meaning ‘the word-form sarva is doubled’. For then rules likes VI. 1. 99, and VI. 1. 100 will find no scope. The word sarva has several. meanings: (1) the tot of things (qea:) as akerym®, (2) the totality of modes (qenrt:) eatetattg: = aknar- ‘cnt meta; (3) the totality of members (orret:), as ak ree: In the present sQtra, the word sarva has this last sense: namely aif the members of a word are doubled, no portion is omitted. The force of the genitive case in sarvasya, is that of sth4na, i.e. in the room of the whole of the words like qft &c there is doubling. So one meaning of the sQtra is, that in the rvom of the one word, two are substituted. In making such substitution, we must have regard to the rule of nearness.1492 Doustine. ( BK. VIII. Cu. 1. § 1 Another meaning of the sftra however is, that it does not teach subs- titution but repetition or employment, not Ades4, but prayoga. That isto say, one word is employed twice. In this sense, of course, there is no room left for’ finding out the proper substitute. The very word-form, vate &c, is employed twice, i. e. is repeated twice or pronounced twice. The word sarvasya is employed in the stra for the sake of distinctness only. Otherwise, one may double only the last /et/er of a word by the rule of warserea, though that rule is not, strictly speaking, applicable to such cases. Obj. The word qweq should be employed in the sdtra, in order to pre- vent the application of the rule to Samasa (compound), to taddhitas, and to vakyas (sentences). Thus anqoifserrea i! Here there is the sense of vips with regard to seven leaves, and does not mean atree having seven leaves, There ought to have been doubling; but it would not take place, if we take the word padasya in the aphorism. Similarly feafeat wan, here also the sense is that of vipsa, and there ought to have been doubling, before the affixing of the taddhita affix. So also in wrt art qrafam the sentence wi qataais not doubled, if we employ padasya. So the word padasya, should be employed in the sQtra. Moreover, it would prevent our employing the word padasya again in VIII. 1. 16, Ans, We could not employ the word padasya in this stra, for then the rule would become very much restricted. Moreover in the above examples, there can be no doubling; for aq: means ‘that whose every twig bears seven leaves qf qa® eT Brea): so that the sense of vips4 is not here inherent in the word’ sapta or parna. In the case of the taddhita example, there would be no doubling, because the force of vipsd is there denoted by the taddhita affz itself, and so doubling is not mecessary. Moreover, a sentence can never bedoubled, because vips’ can take place with regard to a word, and not a sentence. Therefore the word qyeq should not be employed in the sQtra. On the contrary, if we employ the word padasya in the sitra, it would give rise to the following anomalies. We could not have qyere wrarg ; for upasarga being considered as a separate pada, only qut@ would be doubled, and q would not. So also, we have two forms gitar and gretu Here ga is added to the root gu, and @ is optionally changed tog by VIII. 2. 77, and in the other alternative there is gu As 4 and @ are both asiddha (VIII. 2. 1), the doubling would take place without making this q or g substitution. So that having first doubled the word (something as greet weet ) then optional | or ® change will take place, and we shall get wrong forms, like great iter, ret Hou in doubling. While the correct forms are grat #r€t, or #ar Hea, and not the hybrid doubling as given above. Hence the necessity of the vartika qawuferdtan- frtar (See VIII. 2. 3 last vartika),Bx. VIIL Cu. 1. §. 4] DouBLinG OF WoRDS 1493 Or the word & may be considered to be formed by ery affix of ext Tay, meaning at ara aitratte afte a4, wea e waa: That is, all operations having been first performed, then the word is doubled; so that a word in its inchoate state is not doubled. TET ATH SAT NN Ta wee, TCH, aTafeaT Ut gfe: u meg fremnen aeqt meget eaieeeat afer 2. Of that which is twice uttered, the latter word- form is called 4mredita (repeated). Thus in @rc reg, THe TTD | Tear § TAMA et, TeBeart eq, the second word being Amredita is pluta. The word wfuftq occurs in SOtras VI. 1.99, VILL. 1. 57, VILL 2. 95, VIIL. 10. 3 &c. The qea here denotes avayava-shashthi: and the word q¢ denotes the avayava. In fact, this stra indicates by implication, that a Genitive case is employed, when a member (avayava) is denoted. So that the sentences like ga sraex become valid. In the above examples, the word has become pluta by VIII. 2. 95. agardé BURN TaN Tar, aN ah: urged wv erate crater 3. That which is called &mredita is gravely ac- cented. Thus geet GSh | TPL TTT That is, all the vowels of the Amredita become anudatta or accentless. In the above examples, the root yx becomes Atmanepadi by I. 3.66: the Personal endings are anudatta by VI. 1. 186; the vikarana qa gets the acute by III. 1. 3; the sf of this 4 is elided by VI. 4. 111, and thus ‘qui becomes finally acute by udAtta-nivetti-svara (VI. 1. 161). The Amredita bhugkte becomes wholly anudatta. The word qg is formed by the affix g and is finally acute. Prerfitcadt: nen aahen Pre, freeat: ghee u Pree are carat oe aes rea KATA 4. The whole word is repeated when the sense is of ‘always’ and ‘each’. What words express ‘always? The finite verbs, and the Indeclinable words, formed by krit affixes. What ‘always’ is meant here? The word ‘nitya’ here means ‘again and again’, and this idea of ‘repetition’ is the quality of an action. That action which the agent does principally, without cessation, is called “nitya”. So that nitya refers to an action (See III. 4, 22). Thus yen® get ‘he cooks continually’, weqft weafw ‘he talks incessantly, wer gee FraRy or1494 Dovus.inc or Worps. ( Bx. VIII Ci. 1§.6 ‘rs ars arm ‘each one, when he has eaten goes away” Sec III. 4. 22. wate eatetatara Graft (See III. 4. 2). The affixes ktv4 and namul formed words and the Imperative mood express the idea of ‘again and again’, only when they are repeated. While the words formed with the affix ag (Inten- sive) express this idea by the inherent force of the affix, without repetition. Thus ga: ga: Taft = grreat And when this intensive action is continually done, then this word also should be repeated; as qrqeat qrqeaz In the above, examples have been given of nitya, as illustrated by finite verbs like qafa, Indeclinable krit words like qeeq, and like ars Now for rear In what words the @rat is found? It is found in nouns (gq formed words). As finite verbs (fire) express nitya, so inflected nouns (gq) express vips4 or a distri- butive sense. What is meant by the word vipsa? It is the wish of the agent to pervade (vyAptum ichchha) an object through and through with a certain quality or action. That is, when many objects are wished to be pervaded by the speaker, with a particular attribute or action simultaneously, it is vipsA. Thus orar anit crfta : ‘every village is beautiful’. So also waqq waar crete: Tee TET PTA TT When a finite verb is repeated owing to the idea of nityata’; and we also wish to add to such a verb the affix denoting comparative or superlative degree, such affix must be added after the word has been repeated, as qai@ qafxaca But in the case ofa noun, which is repeated owing to vipsa, the whole superlative or comparative word should be repeated, as sreuatrTeNatarralt GRSSA UW TTA at: aa ater u qQaraea eieret & re anferra: u gttsrtseme Sf TATE 5. The word aft is repeated when employed in the sense of ‘with the exception or exclusion of’. As oft oft feria get ®t: (1. 4. 88 and IT. 3. 10) ‘It rained round about (but with the exclusion of) Trigarta’. Similarly qft qf ArfOa:, a ait etapa Why do we say when meaning exclusion? Observe straw aftsei Vért:—Optionally qft, meaning ‘exclusion’, is repeated when it occurs not in a compound, as git aft Fmt or aft Freq: In a compound, there is no repetition as, qfttrret Tee, because the word qf has not the meaning of exclusion only here; in fact, the whole compound word denotes here the idea of exclusion, and not the word ft alone. The word qf is here a Karmapravachaniya (I. 4. 88), and governs the fifth case by II. 3. 11. In féfrsefr the word is an upasarga. MATUT: TTT NG TATA , AH, TT, ST:, TT TTT Pe em ae aE Ca TTT E eet fete Bere gaatBK. VIII. Cu. 1. §. 8] DousLING OF WorDs. 1495 6. g, WH, 31 and gq are repeated, when by so doubling, the foot of a verse is completed. As w arantiprmen -o§ (Rig VII. 8.4), dateges qua (Rig X .191. 1), aire 8 gag (Rig |. 126.7), fx agg et@ arate (Rig 1V. 21.9). Why do we say when ‘a foot of a verse is completed there by’? Observe wt wear far This rule applies only to the Vedic verses, for there alone the Preposition may be used separate from its verb, In the secular literature, this rule has no applicability, as Prepositions are never so used. Sarat non wah sett, af, errer, erates tt at: 0 sat safe seer Carat Ete: arte Ferrier 7. sal, ara, and ata are repeated, when it ex- presses uninterrupted nearness. The word stmipya means ‘proximity’ whether in time or space. As argait ger, or sagaft ara, creat area, stand rey Why do we say ‘mean- ing near? Observe waft ‘war: Why the word is not repeated here: gait Facdt a areata The relation expressed here is not that of nearness, but that of above and below. sada geet: ef os aftaftr | sind cae wea Aiea atesray Mere the doubling is in the sense of vipsa.- arearyuaRacearqareanrearrgrearrarcerg tS vat area arti, waPaaeT, TET, THAR, BIT, Fea, WAG Ml afer no carck: ceerget rae | areaireaiaces ree: orear cafe aT Heart Tela wg aft eared afer 8. Similarly VIII. 1. 21, teaches “mand @: are the substitutes of gery starz in the plural”. The present sdtra must be read there to complete the sense, viz, when not at the beginning of a verse. Thus erira: ef, Aaa #1: Ta. But at the beginning of a hemistich, we must have the forms geara and erqara instead of qt and q@: As, ait Peat Ut gore _— ' BUT rey AerarTeTa TATE The word qrq in the sfitra refers both to the hemistiches of the sacred Rik hymns, as well as to secular $lokas. Q. Why do we employ the word gf in the stitra? Ans. The word sarvam is used in the s(tra, in order to indicate that a word, which has not acute accent on the first syllable, should also become anudatta, For the word gar (VIII. 1. 17) is in the ablative case, and shows that the operation taught in any stra governed by it, will be performed on1502 ANUDATTA [ BK. VIII. Cu. 1. §. 18 the first syllable,-according to the maxim sti: qreq i Therefore, by rule VIII, 1, 28, the word §mnf, which is acutely accented on the first syllable, (because av and fag are anudatta), will only lose its accent in Faget: Tit; but not so the verb aif whichis acutely accented in the middle (by the vikarana accent) in Waren: aia Wo -But by force of the word |, aitf® also loses its accent. Q. This cannot be the reason, for the author indicates by his prohibition in Stra VIII. 1. 29, that a verb loses its accent, where ever that accent may be. For there is no verb when conjugated in gz which has acute on the first syllable, Ans. The word ‘sarvam’ is used to prevent the operation of spitsretai@:; for where the rule sq: qreq will not apply there the other will apply: so that, if “sarvam” was not used, the rule would apply to the finals, and not to cases where there were other than initials to be operated upon. Thus the rule VIII. 1. 28 would apply to gaa: which is finally acute by VI. 1. 186, in Qrreaest Ee: but not to Fae: Hears Q. No; this cannot be the reason of employing the word ‘sarvam’ in the stra, for the author indicates that the rule is not confined to the fina/s, by the prohibition he makes in favor of @@ in VIII. 1. 51; for there is no verb, when conjugated in Lyit, which has acute on the final. All Lyit is acute in the middle by VI. 1. 186. So the erdisrereg rule does not apply. Ans. The word ‘sarvam’ is used for this reason. Had ‘sarvam’, not been used, then in those cases where there was no other rule to apply, like VIII. 1. 19, there only the word would become anudatta. But in cases where another rule also operated, there this anudatta rule would not apply, for then there would be two different sentences. For a thing which is already in existence can be made the subject of a rule ordering certain operations to be performed on it; but not so a thing which will come in existence in future time, Thus the words af and & are ordained as substitutes of yushmad and asmad by one sentence VIII. 1. 20; while another sentence ordains their anudatta-hood. .So here there is separation of sentences (vakya-bheda): which is not desireable; for a vakya bheda should be avoided, if possible. But by employing the word ‘sarvam’, this vakya-bheda is made tolerable. Another reason for employing this word is that the substitutes # and should come in the room of the whole declined forms of yushmad and asmad, with their case-affixes. Had not this word (sarvam) been used, then at and &t would have replaced only ,yushmad’ and ‘asmad’. Obj. Will not the anuvritti of the word ‘padasya’, cause the whole declined word to be replaced by at and at, not only the crude-forms yushmad and asmad? And there cannot bea full word unless it takes case-affixes; so the whole of yushmad and asmad with their case-affixes will be replaced, Where is the necessity of using the word ‘sar-Bx. VIII. Cu. 1 §. 19] ANUDATTA. 1503 vam’? Ans. This objection will apply to those cases where a Pada is a word ending in a case-affix. But the word qq is a technical term also, and applies to crude forms before certain case-affixes. Thusa nominal stem (not a full word) is also called Pada by I. 4. 17, before non-sarvanamasthana case-affixes. Therefore yushmad and asmad will be pada before those affixes. Thus be- fore the affix of Dative Dual, they will be called Pada; and had gq not been used, then only “yushmad” and “asmad” will be replaced and not their case- affixes, Thus anit af daa, wavit a Tae Here ay and a would require after them the case - affix sat otherwise. araarer TNR athe areas, oH gfe: 0 orrafereren gee oarercenraranay THaTTET ateTAT A AAT enften U aarr area Prare gory STU THREAT: It 19. All the syllables of a Vocative are unaccented when a word precedes it, and it does not stand at the beginn- ing of a hemistich. Thus qafe tara, raft garg uu The Vocative is acutely accented on the first by VI. 1. 198, the present makes it all unaccented. Vért:—The rules relating to nighata ( by which all syllables of a word become unaccented, such as the present, and VIII. 1. 28 &c) and to the sub- stitutes of yushmad and asmad apply then only, when the preceding word which would cause the nighata or the substitution, is part of the same sentence with the latter word. Therefore not here:—er4 aoa, era ll “This is the staff. Carry by means of it.” Here gt does not lose its accent by VIII. I. 28, though preceded by the Noun danda, because these are parts of two different sentences, Bret Ta, Tt aASATe | STE Ts, AA AFASAfy “Cook the food, it will be for thee. Cook the food, it will be for me.” That is the rice cooked by thee, will do both for thyself and myself, Here the ¥ and ¥ substitutions have not taken place (VIIL I. 22) for yushmad and asmad, for the same reason. Another example is, wate eaten, trparres “ Vishnumitra is here. Come back Devadatta.” Here Devadatta, though in the Vocative case, does not lose its accent. Nor can you say that eaefi qafafy: will make this vartika redundant. In all the above examples, the different sentences are connected with each other in sense. Thus in the last example, Devadatta was searching for Vishnumitra, when some one says to him, ‘Here is V. come back D.” Thus the two sentences are samartha, yet there is no nighata. But the rule will apply here:—qe @ieq! mrar® auafe | warfereid (VIL. 1. 28) gett oretat & eiret aeatf@ tl In the last examples the Vocatives, the verbs and the substitutes of yushmad and asmad are not in syntactical construction with the words that immediately precede them, and yet the nighata rule & does apply: inspite of the general maxi aay: gafafy: (IL. 1. 1), for rules relating to completed words apply to such words only which are in construction,1504 PRONOUNS. (BK. VIII. Cu. 1. § 22 In the last example, the construction is qg feat arat & tree, and not wy ara. Thus ag is in syntactical construction with ferar and not yrya, yet it causes nighata of Devadatta. Similarly in 7anfeqefa a, the word aeq: is not in construction with firgfy, but with ag, ie. rar: Ha faera; yet it causes the nighata of fagfa u Similarly qratat is not in construction with 3, but with Sires, i. ©. uretat sires & qremf® ul Yet it causes ¥ substitution of yushmad. Though the preceding words are not samartha with regard to the words that follow them, they cause the changes, because the words are in the same sentence. gare: wragdifcdtarenitatard to wat w geng, a wath, TH, sat, fadtarerit:, ara, arr — wary wet Wwhagliftetrenitiqret wingarnte were aw u 20. For the Genitive, Dative and Accusative Dual of yushmad and asmad, are substituted arg and # respectively, when a word precedes, and these substitutes are anudatta. All the three stras qeea, Tang and sragre adaqratar are applicable here. Thus wrt st ert 0 wera At ena ara at tear erste Ar Mate | Ter at Tae wearer Ar seaft tt These two arqand @ come in the Dual only, because other substitutes have been taught for the Singular and Plural in the two subsequent s(tras. Why do we say “for the Genitive, Dative and Accusative’? In other cases there will be no substitution. As, anit garat gaan The word eg in the s(tra indicates that the case affixes must be express and not understood, for the purposes of this substitution. Therefore, not here: qf gerqqgat though here yushmat is preceded by a word in a sentence, and is in the Genitive case yet @ substitution (VIII. 1. 21) does not take place, because the eee is elided. agaaaed geet WL Ta I Ay aAET, Ta, 0 qf agrerrrangorent: wha iifictarerdrdarded vy 1a rare were 21. For (the Genitive, Dative, and Accusative) plural of yushmad and asmad are substituted ae and ae respect- ively (when a word precedes, and these substitutes are anudA&tta). Thus grt & 6a, WAI: ETE; ANT T Aas, rae Fr tee; ait «: Tear sored a: Teh Ra eTaTET Vi TTA, wat, TH, TaTET atau qorqenatonererararat: cetera dteraictaredant & Wr eetererteh sere: tt 22, For the Genitive and Dative Singular (of yushmad and asmad are substituted) & and ¥® respectively, (when a word precedes and these are anudatta).Br. VIIL.Ci. £. § 24 J PRONOUNS. oo 1505 Thus wre era, art Reva, ANT aw, ar etaw uw For the Accusa- tive singular, other substitutes have been taught in the next sdtra, hence & and % come only in Genitive and Dative. varelt Pratrarat: WRN war near, a feetterrer: q he: u carreaeafe a8 | fidtarn aearet eeragererickrdagen at watarn- wit waa: 0 ; 23. For the Accusative Singular of yushmad and asmad are substituted and #1 respectively, under the same circumstances. The word yarraaea is understood here. Thus arreen qeaft, ani at Teate a qeretage NW TaN A, &, AT, &, HE, Te, TH ghee 0 rag re ea Uirait Gerrerqataraed a wate 24. The above substitutions do not take place when there is in connection with the pronouns any of these:— q, ‘and’ an, ‘or’ ¢, ‘oh! se ‘wonderful’, or T ‘only’. Thus arent e8, anit TT TTI gare et ara TL ToT Tene! Te TET ATED oT Mee wat aga Gaz) gaat ag Mad erepat a dak gona a dad! eed wT Gay! wrreett a gaat! anit ai qa! gat a ceafert errat a reer) goat Teate | ratte Teaie nT TeTT aT OTT | aT a aT eT! Gaeta erat renatat err gare aC eTT orerae era omen eT fad) ant aah ar Rat! gaat tak | rand ar daa t gemy at Aaa! orang a daa arrest a aah | ery at ar qeafee | gata qeatt | erat ar ceahe | TORT AT TeaAy eT STH TEA | TARTS & ETL OTT Ae er Dag eve orrraie ern gears | ETL oreTE | eT TED | Mast ATT TL e Ray gaat daa srarat¢ dats gorda 9 Mad verde taal are | qeaftr | arte at ee teat | gate real) sree @ oealee | gerry Teas | stem Tate I Beh TTRT ATE OL | TAT AT ETL TTA OTL! STATE ETL TOTERTY ETL TST. ag era repay dae | aeTETE Rae! gIParag Mas | PATE tas) gona Qa srepane fat araeere geal! anit are Pate | gay Teale) eG pate | Tere Teahe | ETT TART | TE ATT ETT ATT eT gaat TTI STATE OTH TOTRAT FTL TRENT ETL ATTA Gas | aie aes aa ga: saner dat) srapaits fae | genaty fat repeats Mae | arenes gate | TAT ang goat | gathe Tate | TraTieT THA TORT oeaRT | eraTerT Tae | NTE ererarrafirenta( gan ei Taf ae a eT Tt TH ETL The word g=x is employed in the stra to indicate direct conjuriction. There is, therefore, where the conjunction is not direct but intermediate the employment of the shorter forms. As arqw ® e, ant-ah era 1 The particles , 1, &c denote conjunction, ‘separation’ ‘wonder?1506 PRONOUNS. (Bk. VIIL Cat. I. §. 26 &c. Where the sense of ‘conjunction.’ ‘separation’ &c are inherent in the Pronouns yushmad and asmad, and these particles are employed to manifest that sense, there the present sdtra will apply its prohibition. qeatfarerataa i MN cathe oer, ah, a, arate afer u cearat atarat at ert) oraret caffe peatiordras ve mete gerrertratararear + afar ut 25. The above substitutions do not take placo also in connection with verbs having the sense of ‘“seeing”’, when physical seeing is not denoted. The word qaarat: is equivalent to etarat:, and at means ‘knowledge’, ie. verbs denoting ‘to know’, sqraraa means perception obtained through sight ie. physi cal ‘seeing’ opposed to metaphorical “seeing”=“knowing”. The substitutions of af and #% &c. for geag and sr@q do not take place when these pronouns are employed in connection with verbs denoting ‘seeing’ (metaphori- cally) but not.‘looking’ ( physically ). Thus qrrera ef evfgarre:, anit am et others; area Raat hearer, art wer kerart aetearra:, orreett aetgarra:, ATA at eretgerera: ‘ Why do we say when not meaning ‘to look’? Observe arreeat Tema, art at areata tt Ishti:—With regard to verbs of “seeing ”, the rule should apply even where the connection is not direct: as, we have already illustrated above. eqatar: saarat Rarer Re aa ow qatar, wearer, frarst afer: u Rremarrgaterernrerrerag eas nearer RTT AT AAA anftariu gerrertt frre errata efter ro WE CT ATTRA SHATTT TTT THEA: : 26. When the pronoun follows after a Nomina- tive, which itself is preceded by another word, then the above substitutions may take place optionally. Thus af) araeet eT or WR aamerTENy UAT een U APRA ent aerate am aameng atten ape wena FR chia un aie aewat wer tad watt ores Paha Ua ores Tala uae OTA AT Tara WaT OTAT at Pata Why do we say “which itself is preceded by another word”? Observe RR AH, HeTAT A OTT Why do we say “after a Nominative’? Observe arrat wih @ ert, area Hi OTE Vart:—The option herein taught is restricted to gerg and srerg when not employed in anvadesa. But when there is anvadega, then the substitution isBx, VIII. Cu. I. §. 27] PRONOUNS. 1507 compulsory and not optional. This vartika restricts the scope of the stra, and makes it a vyavasthita vibhdshd. Therefore, there is no option here, where there is anvadesa:—arat air armeReey TZ, STAT TA aa FETT Vért:—Others say, that the substitutes af, # &c are a/f optional, when not employed in anvadega, whether the nominative is preceded by another noun or not. Thus the rules 20, 21, 22, 23 are optional. As: ayswaé e4q, or RUT OTL, ERAT A ert or aewat aq ETT Why do we say “when not in anvadeSa”? Observe tft eamar era stat Heat etal QO. If this isso, where is the necessity of the present stra at all? Ans. The present s(tra is for the sake of anvadesa: the option taught in the present sdtra will take place only then, when there is anvadeéa and not otherwise ; thus erq@t att aeqeey ETL, oF Srey NA RENT ETD, BAT AA RETA A Eee, OF ore MR aeAT AT ETE In other words, according to this vartika, rules 20, 21, 22 and 23 are all optional in anva- deéa, but compulsory where is no anvades4; but when the pronoun is preceded by a nominative which itself is preceded by another word, then the above rules are not compulsory, even in anvAdeéa: there also the option will apply. feat stare wecrarateraat: 129 Tat Ae:, ates, prem, aretererat: t ages a RararserrerenFOe eter Beers Beretta reat tear stgereth ae 27. The words *t# &c, become unaccented after a finite verb, when a contempt or a repetition is intended. Thus vet ra, sweafa irr, when contempt is meant. Here Taft srw means ‘he proclaims his Gotra &c, so that he may get food &c’. Tait is from the root af safrac® ‘to make evident’ (Bhu. 184). Similarly seq ‘rT “he repeatedly utters, his descent &c, in order to get married &c”. Where contempt is not meant, it kas the force of repetition, ie. he repeatedly utters his Gotra as one is bound to do, in marriage-rites &c. And vate vate area, seat seqft sta when repetition or intensity is denoted. Similarly vaia ara, vate Tafa qty The word tis a noun derived from the root by the affix ar, the 7 substitution for a has not taken place, as an anomaly. 1 ate, 2 as, 3 weer, 4 ree, Soe, 6 natam, 7 TTSe, 8 ATA, 9 zara, 10 g- cam, Lt Fea, 12 sree, 13 erreara, 14 afaes, 15 at art (ATT aT) 16 maT, 17 ga, The word ars optionally becomes anud&tta: in the alternative, it is first-acute. Thus vafir 31 or Tata Ar W1508 ACCENT. (Bx. VIII. Cu. 1. § 29 Why do we say ‘after a finite verb’? Observe oftad sia tl Why do we say “ Gotra and the rest”? Observe rafa sau Here art is‘an adverb, Why do we say ‘when contempt or repetition is meant’? Observe erate sit ata gan “He digs a well having assembled the Gotra”, The words ‘contempt and repetition’ in the text qualify the whole sentence or stra, and not the word stra nor the word srgara understood. For we find that whereever the word ara is used in this Chapter, it always implies the sense of ‘contempt or repetition’, Thus the word sirana is used in VIII. 1. 57 and there also the sense is of contempt and repetition, fagstere: Wen warn fag, aire Sie front reairenrererereraret Taf tt 28. A finite verb is unaccented, when a word precedes it, which is not a finite verb. Thus gryw: raf Why do we say “a finite verb”? Observe atagey Fa, OFT THA Why do we say ‘when the preceding word is nota finite verb’? Observe ana Fa ‘the act of cooking exists’ = qrx fixat wafa | Here the word waft is a finite verb, therefore the verb pachati does not lose its accent. So also aefafty safe, W@, Gen® ‘successful he conquers, rules, thrives’, wépat wit Wife “for us conquer and fight”, The word wig: is in one sense redundant, because in one simple sentence, two finite verbs cannot be employed; one sentence consists of one finite verb only. But the very fact of this stra indicates that the condition of amraarea does not apply to this sAtra; so that the two words need not be portions of the same sentence, for the application of this nighata: as we have already explained i in VIL 1. 19 Other examples are: srf@mh? guia (Rv. I. 1. 1). eae vem’ (Rv. Lr 4), fea seral ea (Rv. Ve 1. 1), Seovareren ter ai ASTN WU TNA a, SEN ata: u Géonreak afate sroaa | geet feat aged Tae 29. But the Periphrastic Future is not unaccented, when it is preceded by a word which is not a finite verb. This restricts the scope of the last sOtra which was rather too wide. Thus qt: wat, et WAY, TET we UH The Sarvadhatuka affixes 1, %, vq are anudatta after the affix wmf by VI. 1. 186, the whole affix are becomes udatta (III. 1, 3) and where the portion of agi. e. the syllable sa, is clided before the affix <1, there also the 4 of = ales udatta, because the udatta has been elided. See VI. 1. 161.BK. VIII. Cu. I. §. 31] ACCENT. 1509 AattaeaghaaarnererrgTe go N seth Fram:, ag, aft, wa, § faa, aa, a, am, afta, TH, GHA Ul See: wee eka aay re et BANA HT aly awe errr eH PEAT gars wate 30. The finite verb retains its accent in connection with the particles aa, ‘that’, ‘because’, aff, ‘if, ea, ‘also’!, ‘OP, gira, ‘well’, 4a, ‘not’, wa, ‘if’, @ if’, err (interrogative particle, implying ‘I hope’ or ‘I hope not)’, and aa ‘where’. Thus a Sate, a Sate, afe Rafe or ake Tee, eat che, gt Safe, Bry esata, gta Far, Fa Preraretr wea Sarr (Nir 1.11), Ae ge o Ae Ts The particle q4 with the indicatory % has the force of @qu Thus er aft- Gyr Ha aiteare W Other examples are eqa sw ata at Sat Saran § wait ‘if thou, Soma willest us to live, we shall not die’ atta hear PaRar cas “if he will come here, we will make friends with him.” gare want 7, Aa: gered aT «Jf Indra makes us happy &c”. This rule does not apply to the = which is a conjunction meaning ‘and’, So also: faz wee afer he, a gs, aha Why do we say “with particles”? Observe aq aft UAEL= TEU Fala waeqt Here aq is not a particle, but the Present Participle of the root qu “to go’, It means sega ‘ the car creaks while going’. Rule VIII. 1. 28 applies here. : : Q. Now by the rule of Pratipadokta &c. that aq should only be taken, ‘which is a Particle; moreover it is read in connection with other Particles like af &c, so it must be a Particle. How then can the Present Participle aq be taken at all? Ans—This s(tra indicates that the sfaqirn rule and the area rule do not apply here. Thus the word arg formed with the affix "ax (V. 2. 39) is also included in the word arma of this stra. As arrgdt 1 stand, cet arty Spar AT UAT saree sarr TEEN Why do we use the word gvf “ in connection with”. Observe aw 5 we Te ae CTE TAT Other examples are: 4 ag afte (Rv. I. 1. 4.) ‘what offering thou protectest ’. am a: Gf Fat vg: ‘ whither our fathers of old departed *. srar afta aft argure’t difer ‘ let me die on the spot, if I am a sorcerer’, arama gfe, fH sara “ come on, let us share up this earth ”. matt aera deta ‘if a Brahman has grasped her hand ', 4 ex aft ga sion ‘ that the sun may not burn thee with his beam’ saUfr: GAT ara ‘will he come hither for our praises'(Whitney’s Grammar, Para 595). AE TATE WA WaT Ne, TETRA Ul ate: Ure gait AR MATER reat TTT TAA1510 ACCENT. ( BK. VIII. Cu. 1.°§ 34 31. The finite verb retains its accent in connec- tion with ag when employed in the sense of forbidding. When something urged by one, is rejected insultingly by another, then the reply made by the first tauntingly, with a negation, is pratyarambha. Thus A says to B: “Eat this please”. B rejects the offer repeatedly, in anger or jest. Then A in anger or jest says ‘No, you will eat"—as ag taal Here Arae retains its accent, which is acute on the middle, for & becomes accentless as it follows srgyty (VI. 1. 186), and eq becomes udatta by the qeraeae (III. 1. 3). Another example is mgréasta ‘No, you will study’, Why do we say, ‘when asseverative’? Observe ag Sftim are’ Ufeotiestg “Verily in that world they do not wish for fee”. Here it is pure negation. aftqa is first acute by fz accent, a& is final acute because it is formed by era of gare (IIL. 1. 134), &ferar is first acute, because it is a Pronoun ending in et (eargterrenaTa, Phit H. 6) and peer®r is anudatta by VIII. 1. 28, Set rR WR TATA N wera, TH att W werttatrt gen Preah agers waft 1% tt 82. The finite verb retains its accent in connection with eq when used in asking a question. Thus qed Ae 4@ ‘Truly will you eat? geppaeay ll! Why do we say ‘in questioning’? Observe gr gear arge ‘I shall tell the truth, not falsehood’, Seater a & saftey eres it sreraritearee aR TATE U aT, sire Ht gir ng cert oh Rreernniaes ren argent wars 83. The finite verb retains its accent in connection with ey when used in a friendly assertion, Anything done to injure another is prati-loma, opposite of this is apratiloma, or friendliness, In fact, it is equal to anuloma. Thus erg¥=t ‘yes, you may cook’, Here anga has the force of friendly permission. So also srg &= But when it has the force of pratiloma, we have:—erg wat § awe 01 vert aren are “Well, chuckle O sinner! soon wilt thou learn, O coward”, Here wg is used in the sense of censure, for chuckling is a thing not liked by the person: and Is pratiloma action: for pluta-vowe] see VIII, 2. 96, fewn gen cart n ft, wn qi: u fret om Reeernferars aged rere 34. The finite verb retains its accent in connection with f& when used in a friendly assertion.BK. VIII. Cu. I. §. 37] ACCENT. Sit Thus a fg, aft Fr, afe fs Why do we say ‘in friendly assertion? Observe a fe EH TI! tart grETs eT STAG SHAT NAG TATA GeThe, ata, af, aret- v Ne RARere ora Fae ak freee crarERTeNAAs gars TTT 85. In the Veda, the finite verb retains its accent (but not always), in connection with yg, when it stands in correlation to another verb, even more than one. That is sometimes one verb, sometimes more than one verb retain’ their accent. Thus of more than one verb, we have the following example:— are TT Seater rent ot 1% $m’ ‘Because the drunkard tells falsehood, therefore sin will make him impure: i. e. he does incur sin”. Here both verbs yar and frgan@ retain their accent: and fg has the force of aq ‘because’. According to Kaiyyata the meaning of this sentence is qearma awtsqd wat, weTTaTaT Tor @qsaa i.e. a drunkard does not incur the sin of telling a falsehood, because he is not in his senses, See Maitr. Sanhita I. 11.6. As regards one verb ina correlated sentence retaining its accent and the other losing it,we have:—sfraf@ qialsaq aPritsyemsq Here the first retains its accent and the second not. The force of ig is to denote here cause and its effect. So also:—srw wrRGnae ara (or eria), er ar Faery wPrerea® (Taittarlya Samhita IV. 2. 10, 4). The word erwiae is first acute, the augment we having the accent; while IAAT, is all anudatta. Urererarsarg 3g vant ll area, TaTEATT Hl aft: w arraur qrarat gh Raat az ANT 36. A finite verb retains its accent in connection with qraq and qu it Thus arnt wate, aang ee; eens’ cantie, tree: Gefe ave, tere: See atu The meaning is that the verb retains its accent, even when arnq and zat follow after it. Another example is qur faq werTTAg The word wrraq is the Imperative (#1) Second Person Dual of the root aru arr BP aASTT- orn: u The verb 9q: is Imperfect (m#) Second Person singular of the root a ‘to scratch’. The accent, is on sf (the augment). Garat arraay 39 gana Gaara, 4, warATH II aha U aTGaT raat genre freer Gerad ea ager TeTe Ce ale aTgETRTT 37. But not so when these particles qraq and aur immediately precede the verb and denote ‘praise’.1512 AccenT. [ Bk. VIII. Cu. 1. §. 40 That is, the verb loses its accent, and becomes anudatta, Thus arrq wafs Gna, aa cafe Terr, ara HANA AWE It Why do we say when denoting ‘praise’? Observe aay geek, Tat Tew Wt Why do we say ‘immediately’? Observe araq tara: Safe anv, wat Wars: athe ae Here the verbs retain their accent by the last stra. srentsata SBS gah overt, saa, = Il afer: u arrerqnat gs erento tet are PM arRTe a ATA Fan aN TT TTT AA 388. A finite verb loses its accent when it denotes ‘praise’ and is joined immediately with arrq and quar through the intervention of an upasarga or verbal-preposition. The last sOtra taught that the verb loses its accent when swmmediately preceded by aprq and aay ut This qualifies the word ‘immediately’ and teaches that the intervention of a Preposition does not debar immediateness. Thus eT TRO are, aU TNS TT, are TRA ATE, are ATE ATTA, The word ‘immediately’ is understood here also, Thus araq tava: shaft, Gert, aar few fre: wating are i The upasarga ¥ has udatta accent. Prof. Bohtlingk’s Edition reads the sAtra as eqyq 2 qarare 8k Naat lg, ceW, TENE, we: qaraa tl ahr: ug Tea Tea ore eae reat age wah Tara Pree 39. A finite verb retains its xccent in connection with g, qe, qaqa, and ag, when meaning ‘praise’. Thus arated See GT; Tea AOE TE TAA, PT ATT TER Gerry, ore ores wee irr Why do we say ‘when meaning praise’? Observe, Ta wit wate tt The repetition of the word garara here, though its anuvritti could have been supplied from VIII. 1. 37, shows that the negation of that s(tra does not extend here: for the gsTava of that sAtra is connected with two negatives, the q of that stra, and the # of VIII. 1. 29, but the gwrargof this stra is con- nected with the general # of VIII. 1. 29 only. Another example is ergy erarag griee Feu wet TN oN aa Mat, TN ater u erat eect ark feat Taare ATS Tara Te 40. A finite verb retains its accent when in con- nection with erg meaning ‘ praise ’. Thus er@r ae: Fate Gort, He Pregine: al ae wu The separationBx, VIII. Cu. I. §. 43 ] ACCENT. 1513 of this etgr from g &c, of the last stra, indicates that the next rule applies to ergt only. Had arér been joined with g &, then Rule VIII. 1. 41, would have applied to q &c, which is not desired. Hence the making of it a sepa- rate sftra, fret Reaver ee Warr art, Reaver afar: u erat cata ae Reread UR Per aaa aTe 0 41. A finite verb retains its accent optionally when in connection with aét in the remaining cases (i.e. where it does not mean praise). What is the @q alluded to here? The @¥ means here senses other than gar or ‘praise’, Thus qenar ated or went sitet u This isa speech uttered in anger or envy and not in praise (sq Tai) u So also aa TpreaPe The word eq is employed in the sQtra simply for the sake of distinct- ness; for stra 40 applies to cases of puja, while this sQtra will give option in cases other than puja. gos wetearara eR TATA Nl GU, |, wtearaTT Nl ata: 0 go peter oe reat Tierra Pret araad rahe 42. wareard aa Wafer ‘Himself eats the rice, and makes the preceptor eat the gruel’. Here in both examples, the nighata of the first verbs is prohibited, The verbs become svarita-pluta by VIII. 2. 104. wrt afer ge TATA Hare, A, FA AA, agPer: 0 re eer Tene reo ere entrgereT aeatt Pere TET TUTE Ararat .61. In connection with erg, the first verb retains its accent, when it refers to various commissions, (as well as when a breach of good manners is condemned ), The word fafaaret means sending a person to perform several com- missions. The word @ in the sAtra draws in the anuvyitti of férar also. Thus ef We ari Tes, ot oe eG TES So also when fir is meant,BK. VIII. Cu. I. §. 64] ACCENT. 1523 as erang at ate, senda oale maf uw eraalet Gey, TIead HTT qraaft ¥ The prohibition of nighdéa and prolation are as in the last apho- tism. . ereetrarernarerg aR TTT Ss, aE, AA, ce, ef, ae aroorg ahr iu hy ere era FragrtFenatgeren wate PearaeeTTo eRe 62. When q@ and 8 are elided, the first verb still retains its accent, when aq with the force of limitation, takes their place. When does this arq take place ? Where the sense of or srg is connoted by the sentence, but these words are not directly employed, there is then the elision of wand wy There the force of w is that of aggregation ( agwa ), and of wr is that of ‘only’ ( #4). The q is elided when the agent is the same, and 87g is elided when the agents are several. Thus where * is elided:—Qaqq 9a ort freog, tara Greed esq = ATT aTed FT TST Ut So where wg is elided: as:— rea yt ari hesg, mare Fa Tera OT = wT BT, CN ATs Why do we say srqyrcmt ‘when limitation is meant’? See ¥aya: Prt aaae, tare: Sra weeay | The word yq here has the sense of ‘never’ ‘an im- possibility’. The first sentence means 4 eatery Weak e+ eter by TET (VI. 1. 94 Vart ). caren fear eg cana ia, aly, ata, ara afer u are rerrertrgenctt wetter cere; Fat BR var Rrehraetaren wat Fearer 63. When &, ( a, & ae and aw) are elided, the first verb optionally retains its accent. Thus with + @re:—age ateat Safer or Sea, Rear or ersare geht H ‘Here waft optionally may either lose or retain its accent. So also when at is elided, as :—sftfeft Sta or 284, a Fea So also with the remaining. erate score Wee Tea Sars, £8, «, STE A ohare eceat gear mre Freres are argereT ee orate FH 64. Also in connection with & and ara, may optionally, in the Chhandas, the first verb retain. its accent. Thus ory arama orate, (or srreftt), afregeors sirety weer Sarat Gite eireftey (or eee), woerAaragUT (Taittariya Sanhita V1.4. 10, 1. but with out the particle ® ), srg aa een Sreftey (or errefter ), Ree wrrefteg tt1524 ACCENT. : (Bk. VIII. Cr. I. § 67 ‘w has the force of epe and erat, and arq that of afaia and epe ul aera TAA NR TANT eR, seATSATH, eATat ATT NN Pn gen eter aarat erabat gen rer Prehrtnttaren agen waft Sees GR 65. Also in connection with aw and aq, optionally in the Chhandas, the first verb retains its accent, when there words have the same meaning (‘the one —the other’). Thus seater Rieafe (or Rast ), aSPrarcafte urate Revs ender ( or Gf ), wrmMed erect (Rig Veda I. 164. 20, Mundaka Upanishad III. 1). Why do we say qmafat ‘having the same meaning’? See gat ¥arg- qifeer ' Here gx is a Numeral and has not the sense of eq ‘the one another *. The word amy is used, in fact, to restrict the meaning of q™m, for it has various meanings: while there is no ambiguity about the word eran gatsard mat 7 ae RA TT TT AST Ceara yay aqreniirers een carte aa, ware, Peng afer u rer woreda Pepe, Prarefatrs gaa tt anf tree, u merenree Bf Tene 66. In connection with ag in all its forms, the verb retains its accent always. The anuvyitti of seat and geqfe ceases. The prohibition of nighata, which commenced with 4 #¢( VIII. 1. 29) is present here also. In what ever sentence the word wg occurs, that is called agqe ' The word qf denotes here the form of af in all its declensions with case affixes. According to Kasika, ‘wat and waa are not included, according to Patanjali they should be included. See also the explanation of fad in VILL. 1. 48. Thus a get, & Ardara, ar TEE, aot ne, eer wT TET: ( Rig Veda X, 121, 10 ) aging radia (T.S. V. 5. 1.1.) ec arg: daa For the form ayaa, see VI. 3.92. Though the sOtra is in the Ablative ( ayer) and therefore requires that the verb should i#-mediately follow it, yet in aqam arg sift, the intervention’ of arg: does not prevent the operation of this rule, according to the opinion of Patanjali. Vart:—Optionally when the sense is that of ‘wheresoever’ or ‘when- soever’, The word arqratsa means qwyes ‘as one wishes’, without regard of time or space. The nighdta is prohibited here also. As ay ea 47 9¥e OT Raat TT Tae Ut qreregfrerrgare HreTRT: WOO TATA N wre, ger, ey- area, wrenficea: it afar u qetem: arenes secre Ferra aR eqftrena w aetTTBx. VIII. Cu. 1. §. 67) PRONOUNS. 1525 67. After'a word denoting praise belonging to wren? class, the word whose praise is denoted, becomes aundatta. This refers to compounds, the first members of which are praise-de- noting words, The word arerftat, is added to the stra from a Vartika. Thus qrerearra:, Reh RTR:, THUTEAT TE, TEM TT tt BATAAN | STAAL! | ATTTMTRTR! | GATEATTR! | STATA TRE: | WAN TAMAR | aH | ACRE: To aoTeTTR: | are Tremere: | gy Berean! TOT TTTeATER! FE ere: fee | areremT: | a | ater: | ait Pree: | ETSATTR! | TTL TATSATTH: | HEATON | HEATTEATT A! Ut Véet:—The final q should be elided in forming these words. The word arest is an adverb, and therefore in the accusative case, likeaia wa? In such a case, there can be no compounding: hence the elision of q is taught. This is the opinion of Vartika—kara Katyayana. According to Kagika, there is compounding under mat sqaaify rule, and so @ is elided by the general rule of samasa. This becoming of stgqra takes place in the compound, and after composition. In fact, it is an exception to the general rule by which a com- pound is finally acute ( VI. 1. 223) But there is no elision in grewwearen: &c. and there is no loss of accent also of the second word. By the Vartika ‘qarqa’, this further fact is also denoted, where the case - affix is not employed and so the mis not heard, there the second member becomes anudatta. When there is no compounding, there is no clision of q as arene, TREAT! I Though the word gwa’would have implied its correlative term gfia, the specific mention of gfwa in the aphorism indicates, that the word denoting: Gfwa should follow immediately after the word denoting gat In fact, this peculiar construction of the sitra, is a jfidpaka of the existence of the follow- ing rule:-—ae sence Teertt PRISA ATTaTdaTatay “In this subdivision or context, though a word may be exhibited in the Ablative case, it does not follow that there should be consecutiveness between the-Ablative and the word indicated by it”. This has been illustrated in the previous rule of agqara Pra, in. explaning forms like awag arity &c. Though the anuvritti of ‘anudatta’ was current, the express employ- ment of this term in the sdtra indicates that the prohibition (of anudatta) which also was current, now ceases, On this subject, the following extract from the Commentary on SiddhAnta-Kaumudi, will give the view of later Grammarians :—The words wry &c, are all synonyms of erga, meaning wonderful, prodigious: and are words denoting praise, This is an aphorism appertaining to sam4sa subject. In the examples the compounding takes place under the rule of Mayura- vyansakadi.4526 ACCENT. ( BK. VIIL Cu. 1. § 69 Vart:—The elision of q should be mentioned. qramm+ RearyR, in making the compound of these two words, the elision of the case affix, in, this case g,is natural. The vaitika, therefore, teaches nothing new, but only repeats this general rule in a particular form. This is the opinion of the authors of Kasika. But according to Kayyata, the commentator on the Great Bhashya, this aphorism is not a samasa rule: and the words arema &c, are adverbs not admitting of sam4sa; and so the rule applies to these words when they are not compounded. There is no authority for holding these to be compounds under the Mayuravyansakadi class, Haradatta also says, had this been intended to be a samAsa rule, the word qar@ would have been used in the sQtra and this is valid, There is no adhikara of samasa here, that could have caused samasa and in this view, the vartika atte also becomes effective : had it been a samAsa rule, the vartika would have eed redundant. 1 mare, 2 gram, 3 stararge, 4a, 5 staraa, Oergara, 7 HEN, 8 waa, 9 BTA, 10 srame, 11 au, 12 are, 13 gee, 14 7H, 15 g, 16 ere, 17 ear anfercht Aen Es cant Na, af: or, Pere aha u anfrenaat grrr: arate: 1 ght rest TA 68. (After such words denoting praise) the finite verb (which is praised) becomes anuditta, even along with the Gati, if any, that may precede it. ‘Whether a finite verb is compounded with a gati or stands single, both the compound and the simple verb lose their accent, when it is qualified by the adverbs ara &. Thus aq ana sult, aq ard avare un By VIII. 1. 28, the finite verb would have lost its accent after the word arg, but this loss was prohibited by VIII. 1. 30 in connection with aq; the present stra re-ordains the loss, by setting aside the prohibition of VIII. 1. 30. The word anf ‘along with its Gati’, indicates that the Gati even loses its accent. The aig Gati here is restricted to Upasargas. Therefore not here ay ae ara root avtrfer OR te word fh aes is used in the stra to indicate that the words qualified by ang &c in the preceding stra, were non fag words—i. e. were substantives, The rule of warqw of that sAtra, therefore, does not apply here. grat a geninrdt Wee cans Gea, @, TA, ited ght n cerfefes Pron aati fahehe eta uo gear a gee aitasa ger aT Peat fre srefcrareh af Siftrara, u Perargeert of Tene TAT gfe MATAR TaMie THREAT I are u Raritan merger seeifee TRC UL ee es ' egret ae 69. A finite verb, along with its preceding Gati, if any, becomes anudatta, when a Noun, denoting the fault of the action, follows, with the exception of ataj&e.BK. VIII. Cut. I. §. 70.) ACCENT. 1527 The anuvyitti of qyrg (VIII. 1. 17) ceases. But the anuvyitti of the last stra is current. Thus gufir off, rate gf, cafe fran, see® Pra Why do we say gra ‘denoting the fault of the action’? See Fait Wt LH Why do we say gf@ ‘a noun’? Observe Safe fern? ut Why do we say with the exception of iw &c. Sec Saft ana, Sate BTL, THE TTT Vért: It should be mentioned that the ‘fault’ mentioned fn the sdtra, must be the fault relating to the mode of doing the action, denoted by the verb, The rule will not apply, if the gee refers to the agent and not to the action, Thus Saft qiatrem:, rea afar: u Vért: It should be stated that q@ has an indicatory wu The effect of this is that the word gf is finally acute, because of the indicatory #u The word gf@ is not a [emg formed word, because it is not feminine, as we find it in sentences like yfa@cq; nor is it a word formed by rare affix, because this is not a wat word; therefore, it is a word without a derivation. Therefore by sifafaa ere (Phit I. 1) it will be end-acute. The vdrtika, therefore, indicates that when fr causes the loss of accent of the verb, then it is end-acute, but in other cases it is acute on the beginning. According to Padamanjari, gf is derived from q by adding the UnAdi affix firg, diversely (Ug IV. 180), and is first acute. Vért: A finite verb in the plural number, loses its accent optionally : when it loses its accent, then gfa is end-acute, Thus rafay of, or water he, sqft fat or erate Sf: 0 Kérikd. The following noun denoting fault must refer to the action. The elision of ¥ is intended only in the case of non-verbs, because it is so said by those of old. The word gf has an indicatory ~, but it is optionally so when the verb is plural. The elision of mentioned above refers to the elision of qin reo sre TH: = aeoTeaTTH? | The words wards, mean srersqcrarenreaa- a anfatratt 90 anit aes, ret air wakrtar cat sqareit waft 70. A Gati becomes unaccented, when followed by another Gati. Thus spat, satmraft, sifueaiif eft un Why do we say mf: “a Gati becomes &”? Observe ¥erw: waft ll Here aya is a PrAtipadika and does not lose its accent, Why do we say ‘when followed by a Gati’? Observe sit aeyftn eftfr alfe age Taf: uw Here #1 is a Gati to the verb mfy, the complete verb is wra@ But as st is not followed by a Gati, but by a Pratipadika wry, it retains its accent. Had the word wér not been used in the s(tra, this st would have lost its accent, because the rule would have been too wide, without any restriction of what followed it.1528 ACCENT. ( BK. VIIL Cu. 1. § 7r «fer Rrqrerear 9g rath Rahs, a, SaTerafer tt SRY: wearer eter ree TET TOR fteTETsTT are 71. A Gati becomes anudatta, when followed by an accented finite verb. The word mfg: is understood here. Thus amy 9 fe, ay wafer u Why have we used the word fale in the sQtra? In order to restrict the scope of the word zqrwaft; so that a Gati would not become accentless before every udatta word, but only before udatta verbs. Thus eq does not become anudatta before mvy: in wit af Rew eftfr alfy [fit be said that the word aff is a particular name which the Particles get before verd only, and therefore sf would always refer to its correlative term ver, and not to noun, like as the word father refers to its correlative term son and not nephew: and that, therefore, wqraaftr must refer to the verd like erfy and not to a noun like we: ; then also we say that the employment of the term fa is necessary, in order to indicate that the vers must be a finite verb, and not a verbal root. So that though a verbal root be udatta, yet if in its conjugated form (firear) it is not udatta, the uff will not lose its accent. Thus in ay 7 @afe, the root ® is anudatta, but the fear form dfs is udatta, hence the rule will apply here: which would not have been the case had gqrwafe not been qualified by afew For the maxim is aqfrargen sreaa fet ot aft segue wee: Therefore in a faery, the designation of «fy is with regard to arg or verbal root. Obj: If fafe is used for this purpose, then the rule will not apply to an wry ending forms, like y vaftetra and gyafeeare, for these are not farar; but as a matter of fact, we find that q loses its accent, in these forms also, How is this explained? Ans. Here there are two views: some compound the Gati # with the completed emar form Tefal! According to them, this would get the accent, on the rule that an Indeclinable first member retains its accent (VI. 2.2); so that even if the word fag was not used in the sdtra, the form qafirraa, being the second member of a compound, became anudatta; and so % being followed by an anudatta never loses its accent. According to them, therefore, the Gati never loses its accent in fqeferera &c. Others com- pound the word ending in qq (eafyac) with the Gati, and having formed avafirat, then add the affix wrau According to this view, the sq accent debars all other accents, on the maxim, ‘the accent of the /ast prevails’ (afeige), and so q is anudatta, not by this rule, but by ergareet Teereat ‘According to them the word is ayafirmd. According to both of these views, this sQtra is not necessary for the purposes of myqfaurg &c. But there is a third view which makes this s(tra necessary even for this purpose. There is this maxim: aftercare we: ae GHETTT we gE “It should beBx. VIIL Cu. I. §. 72] ACCENT. 1529 stated that Gatis, Karakas and Upapadas are compounded with bases that end with Krit-affixes, before a case-termination has been added to the latter”. This maxim itself has been explained in two different ways, one saying that the compounding takes place with krit-formed words only before the addition of case-affixes ; but with words formed by Taddhita affixes, the compounding does not take place before a case affix has been added. The other view makes no such difference between kyit-formed and non-kyit formed words, This latter view is not necessary for our purposes, According to the first view, the Gati 7 can never be compounded with qaftat as it is not a gar ll So both q and qatweraa having different accents, the present sQtra became necessary to cause the loss of accent of ¥ It Why have we used the word sqrmaft? See Haft, ade Here the verb loses, its accent by VIII. 1. 28, hence the Gati retains its accent. arated qatar 198 Tate N arate, ee era waa At: 0 eaters glafronrrrcne, aferratt vend or TTR tt 72. A preceding Vocative is considered as non-ex- istent, (for the purposes of the accent of the following; word; and the enclitic forms of gurg and aeng). Such a Vocative is treated as if not at all existing, it is simply ignored. The operation which its presence otherwise would have caused does not take place, and that operation takes place which would have taken place had it not existed. What are the particular purposes served by considering it as non- existent? They are (1) the absence of the accent-less-ness of the subsequent vocative, which the first, taken as a qq, would have caused under VIII. 1. 19. As kya! kere! Here the first Vocative trea does not cause the second Vo- cative to lose its accent, but it remains first acute by VI. 1. 198. (2) The accent less-ness of the verb required by VIII. 1. 28 is prevented: as, Yara r@ 0 (3) The substitution of the shorter forms of geaq and weqq, required by VIII. 1. 20-23 is prevented, as ware a4 (not &) ara: erm, Breet AA (not %) ara: ere (4) The application of VIII. 1. 37 takes place, in spite of the intervention of the Vocative between the Particle and the verb; such intervention is not considered as taking away anything from the immediateness ( srqrreg) of the Particle from the verb: as, aray fava THe (5) For the purposes of VIII. 1. 47, though a Vocative may precede arg, the latter is still considered as wita- armgd and VIII. 1.47 applies, as 3aw arg Fafetu (6) Soalso in the case of VIII. 1, 49, as sirer Fave Gera, rarer Fare TH, 1 no option is allowed here by VIII. 1. 50.(1530. ACCENT. (Bk. VIII. Cu. 1. §. 73 Why do we use the word ‘as if or "in the stra, instead of saying ‘altogether’? In other words, why do we say “it Is considered as if non-ex- istent”, instead of saying “it is considered altogether non-existent"? The vocative does produce its own particular effect. Thus in wa ar qeea! the vocative Wt is considered as one word (qateat) for the purposes of separating atta {rom the vocative twee under VIII, 1. 55. This is the opinion of Patanj- but the opinion of the author of Kasika is that #¢ would have been con- sidered as‘erferarrar, but for VIII. 1. 73. Why do we say ‘a Vocative’? Observe @aya: taf¥ 1 Why do we say qa ‘with regard to the subsequent word’? The vocative itself will not be considered as non-existent, for the application of rules that would apply to vocative as such. Thus in ¥a¢e Sarre, the vocative Yue gets its accent by VIL. 1. 193 also, In fact, the word g4 connotes its correlative qt ‘subsequent’; and the vocative is considered as non-existent, for the purposes of the opera- tions to be performed on such subsequent term, whether such operation be caused Ly the vocative itself, or by any other cause; but it is not to be considered non- existent for the purposes of operations to be performed upon itself. Therefore in Fara Fene, Devadatta does get the accent of the vocative. In gir H ny at ‘erent Ugix the first vocative sig is considered as non-existent with regard to mg, and, therefore, ag is considered as following immediately after the pada % and thus aw becomes anudatta, not because of mf, but because of %; simi- larly qven@ and ug are anudatta, not because of the preceding Vocative, but because of ¥ In other words, the intervention of the vocatives does not stop the action of ® 0 avahad carartract erarccrara 193 Tether Na, arafera, warafirnct, area Tay Il Pt 0 eftarcrrerer afta u srfrart earmtactrer garatcard arrest antrerarrraate : u 73. A perceding vocative, when it conveys a gene- ral idea, is not to be considered as if non-existent, for tho purposes of the subsequent Vocative,which stands in apposition with the former. This stra prevents the operation of the last s(tra in the particular case when the two Vocatives are in apposition, and the second qualifies the ‘first. Thus Si Yeuk, froragizeanarse u The first vocative being considered as existing, second vocative loses its accent. Why do we say ‘the vocative subsequent’? Observe tare Se here ‘the verd does not lose its accent. Why do we say ‘standing in apposition or earmfanch? Observe yare Sfeee amr, here the word gee qualifies wae, and is not in apposition with fave, and hence it retains its accent.