BOOK VI. ASHTADHYVAYI of PANINI. PARIS XXI—XXIV. S.C. VASU Price Rs 3.OPINIONS OF DISTINGUISHED PROFESSORS OF SANSKRIT. Professor Max Miller, Oxford, gth February, -1892—* * * From what, I have seen of it, it will be a very useful work.’ What should I have given for such a work forty years ago when I puzzled my head over Panini’s Satras and the Commentaries. * * * I hope you may succeed in finishing your work. Professor Gopalji S. Desai, Rajkot in Kathiawar, 20th February, 1892—* © * The first part that has already been out before the Public shows clearly - that the author has spared no pains to make the translation as exact and pithy as possible. The work when completed will no doubt claim a high place amongst the best works by European scholars on Sanskrit Grammar. The paper, printing and the general get-up of the book are excellent, Professor T. Jolly, Ph. D., Wiirsburg, (Germany), 23rd April, 1893—* © * Nothing could have been more gratifying to me, no doubt, than to get hold of a trustworthy translation of Panini’s Ashtadhyayi, the standard work of Sanskrit literature, and I shall gladly do my best to make this valuable work known to lovers and students of the immortal literature of ancient India in this country. Professor W. D. -Whitney, New Haven, U.S. A., 17th June, 1893—* * * The work seems to me to be very well planned and executed, doing credit to the translator and publisher. It is also, in my opinion, very valuable undertaking, as it does to give the European student of the native grammar more help than he can find anywhere else. It ought to have a good sale in Europe (and correspondingly in America). Professor V. Fausbol, Copenhagen, 15th June, 1893—* © * It appears to me to be a splendid production of Indian industry and scholarship, and I value it particularly on account of the extracts from the Kasika, Professor Dr. R. Pischel, Hlale (Saale), 27th May, 1893—* * * I have gone through it and find it an extremely valuable and useful book, all the more 80, as there are very few Sanskrit scholars in Europe who understand Panini, Pandit Lalchandra Vidyabhaskar, M. R.A. S.(London), Gurus to H.H. the Maharaja of Jodhpur. ‘sofort qereprenaat ya Temes aaa MK TTC ATE MATT RATT ATTA | eqqighogingamnrent niger 1 rat try te roneRTTTRTLS Tatar ares se Fra | eqiaerTens Thani adysaretdar vee de wera <vavisty GOORleHarvard College Library ‘THE BEQUEST OFOPINIONS OF THE PRESS. The Indian Union, (Allahabad), 26th November, 1891—* * * The ori Text and Commentaries, carcfully edited with English notes and explanations, bespeak of much diligence, care and ability ; and those of the English-reading public who have a taste for Sanskrit cannot be too thankful to Babu Sris Chandra for offering them so easy an access to the intricate regions reigned over by that Master Grammarian, * * * We recommend this book to all English-knowing lovers of the Sanskrit Language. The Tribune, (Lahore), 9th December, 8gr—* * * The translator has pains in preparing a translation which may place Panini’s celebrated in the comprehension of people not decply read in Sanskrit. The paper, printing and the general get-up of the book before us are admirable, and considering the size of the book (it is estimated to extend 2000 pages) the price appears to be extremely moderate, The Amrita Bazar Patrika, (Calcutta), 18th December, 1891—* * * Judging from the first part belore us, we must say that Babu Sris Chandra has suc- ceeded well in the difficult work of translation, which seems to us to be at once lucid, full and exact. It is no exaggeration to say that Babu Sris Chandra’s translation, when completed, will claim a prominent place among similar works by European Sanskrit Scholars. The Hope, (Cateutta), 20th December, r89r—* * * The get-up of the publication is of a superior order, and the contents display considerable pains- taking on the part of the translator. The Arya Patrika, (Lahore), 22nd December, 1891—* * © The talented Babu has rendered a great service to the cause of Sanskrit literature by produ- cing the sort of translation he is engaged upon, * * * Hiseffort is most laudable and deserves every encouragement. Karnatak Patra, (Dharwar), 7th February, 1892—* © ® Mr. Vasu gives ample proof of his competency fo undertake the work which is not an easy task even to learned Sanskritists. The Punjab Times, (Rawalpindi), r7th February, 1892—* *.* Babu Sris Chandra is well-known for his scholarly attainments. He has done the translation in a lucid and clear English. We can strongly recommend it to those who wish to study Sanskrit Grammar through the medium of English, The Mandalay Herald, 31st December, 1891—* * * Students and readers of classic language of India are much indebted to the learned Pandit for his translation of this scientific work on the grammar and philology of the Sanskrit language. The Arya Darpan, (Shakjahanpur,) February, 1892—* * * Such a work has been a desideratum, It is well got up, and praiseworthy, We wish it every success.OPINIONS OF THE PRESS (continued). The Allahabad Review, January, 1892—* © © The translator has done his work conscientiously and faithfully. ‘The translation of such a work into English, therefore, cannot but be a great boon to all students of the science of language and of Sanskrit literature in special, The translator deserves the help of the Government and the Chiefs of India. The Maharatta (Poona), gth July, 1893—* * * If we may judge from the first part which lies before us, it will prove a very valuable help to the stu- dent of Panini, * * So faras it goes it is creditably performed, and there- fore deserves patronage from all ‘students of Sanskrit and the Education Department. Indian Mirror (Calcutta), 17th March, 189g—* © * The manner in which the publication is being pushed through, spcaks volumes in favour of the learning and patriotism of its editor, Babu Sris Chandra Vasu, The whole work when completed will bean invaluable guide to the study of Sanskrit litera- ture and especially to the English speaking students of the Sanskrit language, Lucifer (London), March 1896—* © © We are pleased to find that the translation is clear and easy and the extracts from the Kashika are not only very numerous but also valuable. The large number of references to the pre- ceding sdtras for the formation of ‘padas' will save much time and trouble, and obviate a difficulty which discourages many from the study of Panini, Indeed the English translation of the grammar could rightly be called “ Panini Made Easy.” © * * It can be recommended as of great impor- tance, not only to all who wish to have a general knowledge of Panini but also to the few who desire to master the great grammar thoroughly and enjoy all its benefits, —~a_ RATES OF SUBSCRIPTION Payable in advance, for the complete work, including postage, Indian. wR, 20 0 0 Foreign =. 2 200 Payable by instalments, per volume, excluding postage, Indian. = Re, B00 Foreign & o6o0 - Payable at the end of publication, for the complete work, excluding postage. Indian oe - Rs. 800 0 Foreign... ~ . 2 800 All communications to be addressed to SINDHU CHARAN BASU, PUBLISHER OF PANINI, Benares Cantonment, N-W. P. India,BOOK VI. THE ASHTADHYAYI OF PANINI. TRANSLATED INTO ENGLISH SRISA CHANDRA VASU, B. A., Provincial Civil Service, N. W. P. Benures: PUBLISHED BY SINDHU CHARAN Bose, at the Panini Office, 1897.Trac 345! 2etF HARVARD UNIVERSITY LIBRARY ‘Gurdle Sir John Gage, B. @. G, CHIEF JUSTICE OF THE NORTH-WESTERN PROVINCES, THIS WORK 1s, WITH HIS LORDSHIP’S PERMISSION, AND IN RESPECTFUL APPRECIATION OF HIS LORDSHIP’S SERVICES TO THE CAUSE OF ADMINISTRATION OF JUSTICE AND OF EDUCATION THESE PROVINCES, Dedicated BY HIS LORDSHIP’S HUMBLE SERVANT THE TRANSLATOR.vavisty GOORlearg BAT WAAC TAA: AWE: | SS 1[ BOOK SIXTH. Cuarren First. wera X waster mg cari th om, avers, B, sreteres, tt afew arfipardt smc ear TE CT rea w fread BiteorT | Tt wat ane: area RT: HER ET TR TEAM | eT FE writonaraeara | ae uniterarearparaen weMeeRTat | Tat 1. In the room of the first portion, containing @ single vowel, there are two. Upto Sétra VI. 1. 12 Inclusive it is to be understood that for the first syllable two are to be made, i.e, the first syllable is to be reduplicated, This is an adhikara stra: all the thrce words viz gare:, and % and waite are to be read in the subsequent sdtras upto Vi. 1, 12, before the rule of Samprasaraga, begins, Thus Sdtra VI. 1. 8 says “when fg follows, of an unreduplicated verbal root”. The sense of this sttra .is incomplete, unless we siipply the three words of this stra, when it will read thus: “when faq follows, there ‘até two in the room of the first portion, containing a single vowel of an unredupli+ cated verbal root”. Thus from siry:—arairere (airy + orf, LV. 3. 82st ary +r sa+ammtter VII. 4. s9-marme “he awoke”) So also garg, fara, Te from. the roots Ty, ganda The reduplication takes place through the force of the affix that follows. Thus the affixes of the Perfect Tense (ffm); the Desidera» tive (aq), the Intensive (ae), the »q vikarapa, the Aorist in ta ;all cause reduplication. The word gar means that which consists of one vowel (srt) When a stem or root consists of more than one vowel (Is a dissyllabic or polysyllabic root), then the ware aA: will be the first syllable: as in ary the portion wr is the 944 gare ‘the first portion consisting of one vowel’, When a root con- sists of a single vowel, as q ‘to go’, then strictly speaking there cannot be any portion which may be called first (mam) of ckach consisting of a single1088 REDUPLICATION. (Bx. VI. Cn 1. §. 2. « vowel). Here, however, will this rule be applied and will be reduplicated, according to the maxim “sayeftreeafer,”—" An operation which affects something on account of some special designation which for certain reasons attaches to the latter, affects likewise that which stands alone and to which therefore, just because the reasons for it do not exist, that special designation does not attach”, So also in qm there is no first syllable strictly speaking, but still the rule will apply under the above maxim. In making Reduplications the Rule I. 1. 59 should always be borne in mind, Thus qa+st (or of fe) sya t eg rete (VIL. 4. 59 and 60)=qare ‘The word fi in the sftra indicates that the very word-form is to be doubled or pronounced-twice, and not that another word of similar meaning is substituted. In fact this is not a rule ordaining substitution of two, in the room of one; but of the repetition of the one. Compare VIII. 1. 1. arargficdterer Rl want il ara, ama:, faxdrres The meafederarrarat sae srarthidtacienrst Reberaterarae | re sritier ait: errrarer Reatard t wee 2. Of that whose first syllable begins with a vowel, there are two in the room of the second portion con- taining a single vowel. ‘This debars the reduplication of the first syllable, Ina verbal root beginning with a vowel, and consisting of more than one syllable, the second syllable is to be reduplicated and not the first. Thus the garq root wiféa, (wat ate ne+ K+ a VIL. 2. 35=87q), reduplicates the second syllable and the 3rd per. sing is wie fafa, the q being elided by VII. 4.60: and & changed to by VIII. 3. 59. So also srfufirats, stftfeer@ The last form is thus evolved, w+ aq +R +a4 (VIL 2. 74)=0+E4+8 (VIL. 3. 84, and VI. 1. 51). Here now we have to make reduplication, and if I. 1. 59 be applied, then since a vowel affix x follows, the reduplication of eye should take place fe.thes the sthAnt should be reduplicated. But we have explained the sdtra firtwasfir I. 1. 59, by the phrase fedwatahrasi, ie. the vowel-affix causing reduplication, Here the vowel-affix r docs not cause reduplication, but the consonant-affix qq that does so. Therefore rule I. 1.59 does not apply, for qcitself is a aff, and the maxim applies erimguaty arat ftrera apitat “surely that which undergoes an operation can, so for as it undergoes that operation, not be made the cause of the application of a grammatical rule", Thus from the root eft is formed eftrat ; here the root wt is gunated before the augment yz. Rule I. 1. § declaring the prohibition of gunation with regard to Pisa or fem words does not apply here. For though wt has an indicatory ®, yet as it undergoes an operation itself, it cannot be the cause of the application of another rule,Bx VI. Ct 1. §. 3] REDUPLICATION. 1039 Some persons explain the word srwi as the Ablative singular (and not Genitive Singular) of spf and they cot sr the word as a Karma- dharaya compound and not a Bahuvrthi compound. According to them the word means: ‘he syllable following the initial vowel is reduplicated”. (sreraraftirerante:, erareeacenrardt tT) According to this interpretation, the word fataey is merely explanatory. wea: Garage: We TA Wa, aaT:, Gate: Nh afer fietreette atta Retedrarsy seengarat sare area aT ficken afin Wh meaner Feet Se: Sara FETT carte u wearcecrearay fea THEA re nearer at AT NER: Te aeaTETCER te FARA TAA ETH ‘are uw cedkegetta ee CA TAME Te tL aRegTataT gaa ee OR TOTAL are warrant getactenrte Ha CTRL Te Uae ATTUTATeNT TRENT 8. The letters %, {and x being the first letter of a compound consonant and being part of the second syllable, are not reduplicated. The word dvittyasya “of the second syllable”, is understood here. This stra debars the doubling of the consonants n, d, and ¢ when forming parts of the second syllable, provided these letters occur in the beginning of a conjunct (saiyoga) letter. Thus the Desiderative root of aq is sfiqy, formed by adding eand yeu The second syllable here is feyg which is to be redupli- cated by the last rule. But in doing so, will not be doubled, Thus we get the form aPaRtaT®; so also from EE and srt we have wrigheafir, and wffefarater w Why do we say “when , ¢ and are the letters"? Observe ¢faftrat from {in which a is the initial letter and has been reduplicated in to —u Why do we say “being the first letter in a conjunct consonant"? Observe mnfafirafir from the root sr ‘to breathe’, the q being changed into @ by VIII. 4.19 and 21, The phrase sraig: of the last sOtra is understood in this sdtra also, Thus aftgrar@ Some explain the word sta: understood in this stra in the same way as in the last, namely as a karmadharaya compound in the Ablative singular. They do so, in order to explain the form yfepftfawa, which is thus evolvedgaft wort = wrt a IIL 1. Seca Then wefifaghrestt = wfia+aqit Here though the second syllable consists of n,d and r, yet alone is rejected in reduplication and not ¢ and € also, as ¢ and € are not immediately after the initial vowel (erat) x in this case. Vért :—This prohibition should be stated with regard to the letter also, As aq4—zafafwafir This prohibition applies when the word is taught primarily as having a penultimate qu But when it is taught as primarily1040 REDUPLICATION. (Bk. VCH. § 5 having a penultmate g, then the rule does not apply: the ¢ should then be ordained as a substitut: of yu See Tudadi root 20. Vért:—Thera is no prohibition of the x when it is followed by au ‘Thus eqaai This is the Intensive form of the verb mg, and is thus evolved: + we (IIL 1. 22, Vart)=srg+@ (VIL. 4. 30)= an, then follows reduplication by the rules of this stra, The second syllable & has an initial ¢ which is not however rejected in reduplicatis Vart:—In the case of the verb fuife the third should be reduplicated, “Third of what"? Some -say the third of the consonants, and according to them the form is éfar@ar@ | Some say the third syllable, therefore of the root diéay ( foi + e+ 472), the third syllable q is reduplicated: Thus éfedaw, eeatee (VIL. 4, 79) the # of q is changed inta shart ¢ by VIL. 4. 79. =éRaftfer Vart:—The third syllable of the verbs aega &c. is to be reduplicated. ‘Thus mega + ict a= argtar = aghaahy; so also wT &e. Vart:—The third syllable of verbs derived from nouns sreduplicated. Thus epeftrafagfar or erfendtfawfa it Vart—Others say, any syllable of a mrmurg (verbs derived from nouns) may be reduplicated, Thus from the noun gw is derived the root gvfta, the desideratiye rpot of which is qwhf@9, which requires reduplication, Here any syllable may be reduplicated; as. 1, gywifieftt, 2, gfawtfaufir, 3. gHtfaraar, 4. geffareafir | Or there may be double and treble reduplicat it as. gyfrstttarratr qatsares 88 gr ch, seae: afte: oR ere TTT oeTiee ToT A | aH RETA heTTRT ward: gat sree: Gr Sareea ATT 4, The first of the two is called the Abhyfsa or the Reduplicate. The ward {in the nominative case is understood here, but for the purposes of this stra it is taken in the Genitive case i.e, «ards git “the first of the twa” ordained above. The ward Abhyasa occurs in sttras UII. 1, 6, VIL 1.7 &o, SH TET TET Ta, ayTETT age ee hte Tey TET gare MATTER | | PAT SH TAY Cat HET Ty s optionally 5. Tho both are collectively called Abhyasta, ‘Though the word {was understood in the stra, the use of the word ‘e% ‘both’ indicates that the word Abhyasta applics to the two taken together ‘and not to any one of them separately. The word abhyasta occurs in sdtras Hh 4 109, Vie 1. 32 &e, Thus stra VI, 1, 189 declares “the first vowel ofBx. VI.Ch.1.§8) REDUPLICATION. to4t an Abhyasta gets the udatta accent”, The Abhyasta being the collective name of the both, the accent will fall on the first and not on the second, as in Fu dédati, the accent is on the first a" So also by VII. 1. 4 sr takes the place of @ after an Abhyasta, therefore, e+ ft ="e+ mA = "aR So also em (VI. 4. 112), nirearrga: TENE aT N sre, KENTA, TEN ae 0 sere ATR Ta | ee Kee aftereaNTee Te Tea SET NAP | Se HT ot argrmereretgr indie u merereerratftere ae Wate Aft Teaehe ary 6. So also the six roots beginning with jakshi are called abhyasta. The word abhyasta is understood here. The verb jakshi and the six verbs that follow it in the Dhatupatha, in all seven verbs, get this designation. These are ae, ary, FFU, TATE, WE, HH, and tw Panini has overlooked lt and mentions only the firstsix. By getting the designation of Abhyasta these verbs get acute accent on the first syllable when followed by a sarva- dhatuka Tense-afix not having the intermediate t= and beginning with a vowel, Thus staf jécrati, tena jikshati, Hear dividrati, faren® chakasati, sirafa 44Sati, Yak didhyate, and “Geax vevyate, The present participle Yur Is irregularly formed by adding the affix (sr) and when so formed it does not take the augment gu by VIL. 1, 78, Sargiat drat ssarever WW TPT aS, atta, he:, erearerer it afer: u gartanitreteenitomraer: | are were | gartthareren w Ferea:, Teak, HIT gare garraetarwareres Gri: arg iat 7. In the room ofa short vowel of the Reduplicate of the roots a@ &e, a long is substituted. ‘There is no list of mf verbs given any where. The word snft in anf therefore should be construed as “verbs /ike tuj”. So that wherever we may find a word having a long vowel in the Reduplicate, we should consider ita valid form, Thus gga (qutarry IT. 2, 106—gqara Rig. I. 3. 6). morgen, erage, erat Ataray, eT gATA Hl This lengthening only takes place in the Vedas before some special affixes, and not every where or in secular literature, As gare wert eft fat aratcrvarees Wc arf n fff, aret:, ererearereT tt afte 0 Fe seat STEER reterarveR ETRE: etter AT marae Kae: errr Recrererrencer ere Re TeReeE Te U Rafe habre sree eC I 8. When the tense-affixes of the Perfect follow, there is reduplication of the root, which is not already roduplicated,1042 REDUPLICATION. (BK. VI. Ci. 1 §.9 The reduplication is either of the first syllable or of the second syllable according as the root begins with a consonant or a vowel. Thus ¢¥+0@= TUL oE= TTT So also gure, dere uw In the case of seh, Rule II. 1. 36 does not apply, so the Perfect of this word is not formed by adding srand the auxiliary verbs g, and wg; but regularly, In fact af is regarded asf it was 4 See stra III. 1. 36 Vart. areq aeieinr art ag atea: waren, wrere giatrarey- aratigreerg tt ‘Why do we say when rag follows? Observe art, gat! Why do we say “of a Dhatu"? Observe ageiat rgfoat, dmfiraragtedt The question arises from the fact that the root (dhatu) always directly precedes the affixes of the Perfect, no vikarana intervenes as in the case of other tenses. So there was no necessity of using the word dhatu in the sOtra. To this it is replied that by IIL. 4. 117, sometimes fire is treaved as sArvadhatuka, and then it takes vikarana. As ¥ taking the vikarana eg becomes apg; ( III. 1. 74) this whole base is not a dhatu, therefore when the affixes of Perfect are added there is no reduplication, and we have »ffevt See III. 4. 117. Why do we say ‘of a non-reduplicate’? Observe great ature quit wf: WLI atgadatary, i.e. it is the Perfect Tense of the Intensive verb atxa, and is not reduplicated again, we having already caused reduplication. So also eenrar em afiting: ending in the aq of the Perfect. Vart :—In the Chhandas there is optionally reduplication of the root in the Perfect and other tenses. As enftare anferarit or fem @qrig; Yrartr at rari or een Brenton were IE OF TAT, META Mea UTE or HITE Ht Vart:—The root srg is optionally reduplicated in the Perfect, As, Sy ree (or BERIT) ET TN STS NR TAT aE TST, I aR: artrovareaeaft a8 | area eT CT | aRTEL AEE TPATEEE ardterarren ranean at ftataes ar mararet & ara: 9. Ofanon-reduplicate root ending in we (Desi- derative) or te (Intensive) affixes, there is reduplication, The word ea wét: should be construed as genitive dual and not as Locative dual, ‘In the latter case, the meaning would be “when the affixes 4 and we follow, a non-reduplicate root is reduplicated”. The difficulty would be that while the root alone would be reduplicated. the augment would not, as in.wfeRewft, startet The phrase yratenaraea is understood here also. The reduplication is of the first or second syllable, according as the root has an initial consonant or vowel, Thus water (Ill. 1. 7) eu, reduplicate Fenaia (VIL. 4.79). SovaBe. VI. Crt. 1. § 11] REDUPLICATION. 1043, Fraftera, e—afiftata, r—aPerhtefe un So also of werr—as, srvea® (VIL 4. 83), BerUa, arasad, sreraa, Hewrqe [fa root is already a reduplicated one, there is no reduplicati: as qed, sratawe being the Desiderative forms of ara (already formed by svarthika wa III. 1. 5), and of aga (the Intensive form of g) u See II. 1. 22. st W to Woah Wear qf: 0 Sr qett smereen erdrereren merredeanre Aeefteren ar errata: 10. Ofa non-reduplicate root there is reduplica- tion when the vikarana élu (4g) follows. The agisthe characteristicsignofthe roots of the third class, and causes the elision of the vikarana agi The reduplication is of the first or second syllable, according as the root has an initial consonant or a vowel, ‘Thus BR (from ag), FAP, Fa ae WRU Tania nafs age wafer qeat sarees oretcrerer errr fitetterer er quran Erte 11. Of. non-reduplicate root there is reduplica- tion when the affix @&of the Aorist follows. The reduplication is of the first or second syllable according as the root begins with a consonant or a vowel, Thus erfirwa, siteq, (VII. 4.93 and 94). So also enfteq, snftraa, siffvqv These are aorists of the causative roots fw &c. When the Reduplicated Aorist of the Causative (emt) verbs qq &c fs to be formed, as here, we should first elide the causative sign far, then shorten the penultimate vowel, and then reduplicate. Then by VII. 4.93 the effect would be like as if era had followed, and this would not be prohibited by the short vowel being consdered sthanivat. For rule VII.4.93 says “Let the effect be as if ga had followed, on the reduplicate, if followed by a light vowel, of an in- flective base to which fir followed by sa is affixed; provided there is not the eli- sion of any letter in the praty&hara sx occasioned by the affixing of Fir u” One of the conditions for the application of this rule is that the reduplicate must be followed by a /ight vowel. We have said above that the long penultimate is to beshortened before wa Now arises the difficulty. Should this light vowel coming in the room of a heavy vowel be considered like the heavy vowel by the rule of sthanivat Adega (I. 1. 56), or not? It should not be con- sidered so, for the reasons given in I. 1. 57. For it was established there that the sthanivadbhava will arise then only when some operation is to be done to a thing anterior to the non-substituted i.e, original vowel. Here no operation is to be applied to something anterior to such original (srufte) vowel, but to1044 Repvpiscation. (BK. VI Cu. § 12 ing of such substituted (enfye) vowel. Therefore, the light vowel red non-existent for the purposes of aaxara it If this order of operation is not followed, there would arise incongruities, ‘The order must be this: rst the elision of t@ (VI. 4. 51) 2nd the shortening of the penultimate (VII 4.4) 3rd the reduplication. If the reduplication take place first and the vowel be shortened, then the reduplication having taken place before the subs- titution of the vowel had taken place, such substitute will now be considered as sthanivad: the /ight vowel being considered as still Heavy, will prevent the application of gerare The difficulty will be in the form of eirgtway, where the short is ordained after the fiw If this be the order of operations, there will not be the reduplication of the second syllable (@ in the word wrfeeq? This is, howwer, done by I. 1. 59. which see. area arerg igiar nM TATA MATRA, UTETA, HigTE, TM aftr: werner rere signee oeareorater arerat afetor fargrereey enferernc ut agerdtat a ree CA TRIAL Te tl afcatsraferantat frerrerarepareeatt ere FR HRSA he MRR are Mere ATR oT Meee I 12, The participles dasvan, sthvan and mtdhvan are irregularly formed without reduplication. ‘The word grata is from the root apg ‘to give’ with the affix @g (III. 2. 107), here the reduplication and the augment ygare prohibited irregularly. As qatar wae: gy (Rig I. 3,7) The word ergrg is derived from the root ae ‘to endure’, by adding the affix gq (III. 2, 107), the irregularity being in leng- thening the penultimate, not allowing the augment ¢¥ and the reduplication. ‘Thus angry waren: So also stearq comes from fag ‘to sprinkle’ with the affix gy (III. 2. 107) the irregularity consisting in non-reduplication, non- application of ¥g, the lenthening of the penultimate vowel, and the change of winto wn As steratrara wrara wea It is not necessary that these words should be in the singular always, in their plural forms also they do not redu- plicate. Vart:—Reduplication takes place when wm &c, are followed by the affix qu Thus fan waa-B+a—woe, fettafefere uw The affix & comes after wry and fig with the force of the affix Eu Vé1t:—The roots "1, 4%, WL, and 8x take reduplication when follow- ed by the affix era (III. 1. 134) and the reduplicate (abhyasa) takes the aug- ment se i! The final consonants of the Abhy’sa (reduplicate) are not elided in these verbs, in order to give scope to the augment, for if the final conso- nants be elided by VII. 4.60, then the addition of the augment becomes unnecessary; as there is no difference between the augment and the Adesa (shortening of the vowel) taught in VII. 4. 59. ‘Thus rat, Tare, FATT, WeTeE!UK. VI. Cu. 1 § 13.) VOCALISATION. tgs Vért:—The aliove vastika is optional, so we have the forms ‘tt Ter) ar ca, ea TET KE. Vart :—The root wa is reduplicated before the affix srg, and the aug+ ‘iment wre comes after the Abhy4sa, and @ Is the substituté of @ of the Abhyasa, Thus (q+ sint~ ot+ wg + OE+ STH TATE: (The sécond @ is change ed into 4 by VII. 3. 55), as in the phrase garaa: sinrrwdottara, Vétrt :—The causative root qt is reduplicatéd before the affix wm, there is clision of fit (sign of the causative), and #€ is the aiigment of the Abhyasa, and it is lengthened. Thus qf? 4-ereq= age: Wt STS: Grace Gareiteacgey UM VTE cas, Araneta, gs, welts, aegEt APH: n gee He cerearemereareteret THT Te: SrETTA ATT Ut atten uo: gemerein rarer: FE: 13. There is vocalisation of the semivowel a of the affix ete (IV. 1. 78) when followed in a Tatpurusha com- pound, by the words g¥ and 4ft tl When the words yx and qa are tlie second members, fofming a Tats purusha compound, there is samprasarana (vocalising the semi-vowels) of the affix emg of the preceding. That is @ is charigéd intog « Thus tfqeey ered. ‘sea=ttening (a Bahuvrthi compound taking the samAsanta affix or rather substitute ¢ by V. 4. 137) anteqrqeqras,=anthrniea + erat (1V. 1.92) = aettremed: The feminine of this will be formed by adding tag (IV. 1. 78). Thus we have ‘writermat (see IV. 1. 78). Now in forming the Tatpurusha compound of this word with gm or ff, the final @ will be changed into ¢ and we have erérqaft- g%, arderafefe! 1 The er of at becomes merged into y (VI. 1. 108), and the short is lengthened (VI. 3. 139). So also @@mmeitga: o Ragredeta Why do we say “of the affix tg”? Observe parg!, @rTarg™: i Why do we say “when followed by ¥¥ or 98”? Observe ariqraat- Boy, Fegweargay Why do we say “when forming a Tatpurusha compound”? Observe. emrdtrearafacea area = artermarafircd Ara It isa Bahuvethi compound, The affix ea@ is here the feminine affix q followed by TE (st) (see IV. 1.77 and 74). A general maxim relating to all affixes is “an affix denotes whenever it is employed in Grammar, a word-form which begins with that to which that affix has been added, and ends with the affix itself: weramq wera PiftrenPrerrart ae” N This maxim, however, does not apply in case of feminine affixes, where we have this rule “a feminine affix denotes whenever 21046 VocaLisation (Bk. VI CHL § 14. it is employed ina rule, a word-form which ends with that affix, but which need not necessarily begin with that to which the affix has been added, but where the word form is subordinate: eftaaiy argrest 4” u Thus we have orarftenaatan gw: = qorereredige: and so also qorartemaftaft: i! Not so when the word is an upasarjana or subordinate ina compound, As siftrareat serdtqrearg= wfrartteraar, ceIgT: = eftrntreraargys, srfratreeraMraTe tt This vocal takes place when gw and gf@ alone, not compounded with any other word, stand at the end: not so when a word beginning or ending with these words follows. Thus ardanmargay, artrerearrerge: &c. Though a word ending in ex may have may semi-vowels, yet the vocalisation takes place of the affix 4 (srg) only, according to the maxim fatteanrteaivar itt “substitutes take the place of that which is actually enunciated”, The word dwar has regulating Influence upto VI. 1. 44 inclusive. The rules of vocalisation (change of semi-vowels into vowels) is contained in these sftras, Vart:—Prohibition must be stated when gf and gq follow the word Prat n As franarge:, ttrararara: instead of sirarefge: &c. wegt wget nee aan argh, egret Pu cree: Soreronfceergetg | aeqUETeRTe TEA ore Ore tomer UT Te WATT ATERAT tt 14. Thore is vocalisation of the affix St# when the word eq follows in a Bahuvrihi compound. Thus artteragrerqrer=artrmditarg: So also érgyredtarg: tt ‘Why do we say “when the compound is a Bahuvrihi”? Observe aré- qraaran aeg: = artreraareeg!, which is a Tatpurusha compound. Like the last sOtra, we have here also qorartenmafarg:, but wfrardqaaarary:, wrerareareey- et, erQerrear eA: tt Though the word avgft is exhibited in the sQtra in the neuter gender, it is in fact a masculine word, Vart:—There is vocalisation of ere in a Bahuvrihi compound with area, AIgR and Arq optionally: as antreradtans: or sntrenmarand:, ardhereutarys: or ardteraararga:, arcereatarar, arermararar tt The indicatory of rae makes the word take the udatta on the last syllable (VI. 1. 163), thus debarr- ing the especial accent of the Bahuvrihi (VI. 2. 1). All Bahuvrthi compounds ending in W@ take the samAsanta affix @y, so ATg would have become arg by force of V. 4. 153, so the separate enumeration of arg and "Tg™ here shows that am is also optional,Bx. VIL Cr 1. § 15] ‘VocALISATION. 1047 wef ereerhreramretiatt RVR WRG TAR Ufa, ate, otra, Fae Ren Serena Tae ore ea Pep Re ee foe wa AP Teneo Brey 08 | carga, ea teperdreR TART TTT AT MAT | Ae ATEN Gert Fare remy Te: STEEL ATT 15. The semivowels of the roots 8%, erg and qant verbs are vocalised when followed by an affix having an indicatory ® The anuvritti of eae does not run into this sOtra. The root #w in- cludes the ww qftargt of the Adadi class (II. 54) and the wa substitute of g (IL. 4. 53) 4g is the root ftyeag g® of the Adadi class (II 59). The wanty verbs are the last nine roots of the Bhvadi class viz. ay, 4g, 4%, TH, WH, Gy, HH, WE and Pg The semivowels of these eleven verbs are changed into the cor- responding vowel, when an affix having an indicatory mis added to them. Thus with the Past Participle affixes *& and srg we have the following forms:— 1. aw, TTA, (VITL 2. 30). 2 erg—gqn:, gHATL 3 THe: LeTE, (VILL 2, 36) 48—an;, INT, 5 —ae!, SEAT (VIIT. 2 31, 40, VIL. 8.13 and VI. 3. 111) 6 wa —aten: aftaary, (VI. 2.56 VIII. 3.60) 7 1 —an:, zaary, 8 ci —eite:, etry 9 LT, rT, 10 ee aie, TTA, EE, ET avar ever’ eer aT FrgTait!—When in a Grammatical rule, an operation (dependant on an affix)is tanght with regard to a root (dh&tu) by mentioning particular verbs (dhatu) specifically,and not by using the word “dhatu",generally, then the operation takes place only when the root is followed by such an affix as can be added to roots by emunciating the word wrat: 1 The existence of this maxim is inferred in the formation of the word sitorerg in VI. 4. 174, which see, Therefore, there is no vocalisation of ¢ in the following wreafay, uifeart Here ‘areaft is formed by adding wag to are (areiieafit) Wary is formed by Tt added tow, the vowel being lengthened and samprastrapa being expressly prevented by Unadi II. 57. Now, it is a general rule that a root taking the affix feng, Pre, oF FHeRdocs not lose its character of a dhAtu. Therefore are is a root (gg). and it ought to take vocalisation before the affix exe But artis not enunciated to come after a dhatu, but aftera gaa (gqeTET: er, LIL. 1. 8) in the rule ordaining its affixing. So also mfwa: formed by adding BE to aE (See Vart. V. 3, 83). aitvrrafrsatrafetratrratagesngrrdiat fete a neem aarti wR, car, aft, sai, af, frat, gat, qoute, aroacitare, ef, il: aft: ue cen, ar eater, Hei e:, eT area, TT RT, OL east, fore BR. reo stearal, wea oh, eat TET AeA wet TeTMaTATeRRT erat ‘otorer u fart: eerererenafenf yng ferareerea: me1048 VOCALISATION. (BK. VI. Ci. 1. § 16 16. There is vocalisation of tho semi-vowels of the following verbs, when an affix having an indieatory § or € follows:—ug ‘to take’, sar ‘to become old’, asx ‘to weave’, oat ‘to strike’, aa ‘to shine’, saq ‘to deceive’, Rat ‘to cut’, mew ‘to ask’, and yes ‘to coak, to fry’. By force of the word ‘¢ ‘and’, the anuvritti of rar is read into this sftra, Thus 1. ay—egtes:, état (by se and swag), ceonfr (I. 2. 4), were (by a¥ of the Intensive), 2. ser—sftqt, tray (VIII. 2. 44 @ changed to a), fama (I. 2. 4), the short ¥ is lengthened by VI. 4. 2 and is shortened again by VII. 3. 80. Svitay (ag, 3. sfe—The root Wy is replaced by gy when fare follows (II, 4. 41), This #f@ can have no fay affix after it, it takes only fi terminations, which as we know are farm. (I. 2. 5). The examples, therefor given will be of fq. affixes only, Thus xaq:, xg! Now arises this ques: tion twhy do you enumerate sf, for is not xx (for which af is substituted) already included in qaf@ class of verbs given in the last stra, and by force of that sOtra, sf& will get samprastrana before fing affixes”, The reply is “fa is numerated here in order to shaw that the prohibition of sampra- strana with regard to was taught in VI. 1. 40, does not apply to its substi- tute qfq in the Perfect Tense”. Thus while the Perfect of 4H is wit, Ta, TE, the Perfect of its substitute ay is sara, gg: and Sew More over VI. 1. 38 teaches that the a of qq is not vocalised in fag affixes, which therefore, implies Gfapaka) that the other semivowel i. e. ¥ of a will be voc: 4. Sayfa, feaeara, and with fea affixes Foamy, wna u 5. Ta—afere:, efeaara, and with faq affixes wes, auiea 6. era — fife: treme fyi, Aieeaig By a Var tika under I. 2, 1, the word gris considered to belong to gent class, and therefore all affixes after it, other than those having an indicatory @, = or the affix TH, are considered as fea, and therefore, there will be samprastrana before these affixes : as, afétean, aiatage, sfatracaan 7. TERM, THT How is the final -g of aq changed into % for by VIII, 2, 36 ought fo have been changed into q before the genft allix am? To this we reply the affix = is replaced by 4 (VIL. 2. 42) and as this is not a gevfy affix, Rule VITL 2, 36 does nat apply. This is done an the following maxim : “The substitute of the Nishth@ should be cansidered ta be siddha or effective when applying the rules relating tq the change of a letter into w, to accent, to affix, and ta the addition of the augment Fx”. But when wis to be changed into w, the substitute is cansidered asiddha (nt to have taken place). Thus we have the forms qam: &o, Before faq affixes we have qpafa, wtaweat 8. mes—ge: (VIL 2. 36), qeara, Fea—qeore, vagesde n The forms qaeg and ease would have been evolved by the simple rules of Reduplication even, without theBx, VI. Ci. 1.§. 17 J VocatisaTion. 1049 application of this rule, But gwg+ae (IIL. 3. 90)=meqt Here there is no vocalisation, because Panini himself uses the word get: in sOtra III. 2. 117 showing that this is the proper form though irregular. 9, wq"—we: (VIII. 2. 36), qeara, feaq—asnia, wth ssa The @ of wem is changed first into q by mat wa en (VIIL 4. 53), and then ¢ is changed into = by eit egar eg (VIIL. 4, 40)- feerreetratare nol want uBR, aye, THT, (@srarcora) aha: want ereanttat eendtat + Fae TOA Sateen eoraTCe TTR 17. There takes place vocalisation of the semi- yowel of the reduplicate (abhyasu) of both vachyadi (VI. 1. 15) and grahdi (VI. 1.16) words, when the affixes of faq follow. Thus qw—are, gata, gery, seers, T%—cars, cafes, gee wary, wafer As regards weit verbs; ag—serg, wafer (there is no speciality in case of this verb, as these forms would have been evolved without even this rule). war—firsar, fafseea, sft—aara, and afer; sea—ficary, freafry, su— wre, eats, ory—frcare, Preafre, at—aww and yafwril Some say that with regard to mw, it is equal whether there existed this present rule or not. For they argue thus, ae +UR@au+AEtt reste e (ett we VIL. 4. 60)=4am 1 To this we reply, this form would be evolved no doubt had this rule not existed. But when this rule is applied, and you make reduplication without first vocalising the z, then you will have to vocalise q by force of this rule, Hence the necessity of this rule with regard to mw also, for having changed ¢ into ™, we have gag and then change @ into wt by VIL. 4. 66. Then this yy substitute becomes sthanivat to w (I. 1. 59), and therefore there is no vocalisasion of this 4, for rule VI. 1. 37, prevents the vocalisation of a letter standing before one which has already been vocalised, There is no speciality with regards to seg and west verbs, This vocalisation of the reduplicate is taught with regard to those affixes which are not fai With regard to Para affixes, the Rule VI. 1. 15 will apply. And as the rule of vocalisation is subsequent to that of redupli- cation in order, therefore by the maxim of qteq I. 4. 2, vocalisation will take place first and then reduplication, according to the maxim ga RU Pearman “occasionally the formation of a particular form is accounted for by the fact that a preceding rule is allowed to apply again, after it had previously been superseded by a subsequent rule”. ‘Thus qrq+ erg: (which is Paya I. 2. 5)= =, FOG MTL TTt MG HUT, HT tt1050 VocaLtsaTion. (Bk. VI. Cu. 1. § 20 Though the phrase gwaera could have been supplied into this sdtra by the context and the governing scope of the preceding sOtras, its express mention in this sdtra is for the sake of indicating, thet the rule of vocalisation supersedes even the rule of waft try: VII. 4. 60, the vocalisation must take place at all events. Thus erg+ogaeq+erg+er Here by VII. 4.60, the second consonant @ of sq ought to have been elided, and the equation would have stood q+eqg+m, and there would have been vocalisation of ¥ by this rule, But that is not intended; there is vocalisation of q and we have Frearg t In fact, the universal maxim of vocalisation is:—The samprasarana and the operations dependent on it possess greater force than other operations which are simultaneously applicable”. mercer apr Bri AT eaters nen want WN eart:, ale, (@arercora) Wt aA eeaHEER eaedee er em UF Ge Stree Hea 18. Of the causative verb eaft “to cause one to sleep”, there is vocalisation of the semivowel, when the affix ‘¥ of the Reduplicated Aorist follows. Thus the Aorist of emit is regu, EEETATE, HAETE The vocalisa tion takes place before reduplication, then there is guna of the penultimate short vowel, then this is again shortened by VII. 4. 1, then there is reduplica- tion, and then lengthening of the vowel of the reduplicate by VII. 4. 94 Thus ent +4m=gt + TRH ET (VI. 4 51) 4am (VII. 3. 86) arg + ee gr+ SE (VIL 4. eget gute (VL 1 1)=gara (VIL. 4.94) which with the augment # becomes arava Why do we say ‘when wa follows? Observe eae’, emfta: i The anuvritti of fapfat has ceased, that of fet@ however is here. caftreatretert aFe nen caf H eala, e, store, aS, (Qrearcora)Mt af: Feu, a eg ea OT ey, Sn eae, CeATT TT af TET: EOTETCOL THR 19. There is vocalisation of the semivowel of the verbs ea ‘to sleep’, eq ‘to shout’, and 2% ‘to cover’, when followed by the affix ¢& of the Intensive. Thus argrat, Sferat, tae (VIL. 4. 25 the short x is lengthened). Why do we say ‘when a@ follows? Observe erway formed by afwe IIL. 2. 172, 7 AT Roll afer: uw aati we statardfe sea: ert Ta 20. There is not vocalisation of the semi-vowel of 7, when the affix wg follows.Bx. VI. Cit. I. §. 23] VocatisaTion, 1051. The word af is understood here. As qrrach, erreu®, street Why do we say “when a@ follows”? Observe @:, vuft | See VI. 1. 16 to which this rule is an exception. sare: SU TTA are, ET UR weet RT| Pa EPEAT arene em ate wa 21. 'The verb ¥ is substituted for the verb tz ‘to worship, to observe’, when the Intensive affix 7 follows. The phrase af& is understood here. Thus Saiten, Sata, Satter ‘The exhibition of aft in the stra with a long ¢ indicates that there is long vowel even in tenses where a is elided, as fra: "For if the sOtra had enunciated fx with a short gas the substitute of ra, the forms watay &c would have been still valid, Thus fe +age Prt Peta=%+Petaqeetatta (VIL 4.25 causing the lengthening of the short ¢). But then the Nishtha would have been feat which is wrong. eara: Sh Praag WR TT eare:, hr, Prerere WAU ewe Strearat arhrenes ureitterat sem: extrait wa 22. The eat is the substitute of err ‘to swell’, when the Nishth4 affixes follow. ‘Thus este, efrrrgu Why do we say “when the Nishtha affixes follow”?Observe extfit: formed by faxq ut In the phrase egret af, the word eqnit is the feminine of the word emf formed by fant The phrase Preram, “when the Nishtha follows” governs all the subsequent s(tras up to VI. 1. 29. eer oqder 93 ogre eer, 3, Gea, (@erETCOrA) It aft uw Prarartatty Tae eomeefen 1 ear Cem Tea | ea OR Gergaresr area @ ater: eee TAR MET, | Rar eR TTC eTaUrad Tem! eTETT ware 23. The verb eat (zt and a) when preceded by 5 changes its semivowel toa vowel, when a Nishthé affix follows. ‘The phrases “when the nishtha follows”, and “there is vocalisation” are understood here. The anuvritti of er does not run here. The roots #& and @¥ both assume the form fat and are included here, Thus 9+ ear+sn=m+ fer +8 (VI. 1. 108)=9edta: (VI. 4.2), and gefteargi The @ of nishtha affix would have been changed into by VIII. 2. 43 in.as much as the root ear has a semivowel and ends in long 1" But by the vocalisation of a, the condition of awaq for the application of VIII. 2. 43,no longer existing, the affix % is not changed to But wis optionally changed to # by VIII, 2. 54 as mefim: and aetna t ‘Why do we say “when preceded by 8?” Observe freeartt (VIII. 2 43), aearrara [fit was intended that vocalisation should take place when g1052 VocaLisation. [Bx. VI. Cu. 1 §. 24. singly stood before, then the stra could well have run thus we: The use of the word g¥ea implies that the rule applies when ¥ stands first, though other upasargas may intervene between it and the root. Thus weetta?, netefterara ‘The compound wage should be explained as a Bahuvrihi=w: gat aea urgent ‘eaqaea & wrk (that which consisting of root and upasarga is preceded by wis called myg). Therefore the rule is made applicable to weirdta: &c. weaafererhreat: eer WE Taree ener, eases, wT: Fhe aergett arreandaee en erate Are enftemee Tat arat Te! eereTe TT 24, There is vocalisation of the scmivowel of the root wat (xt) ‘to go’ when the nishthd affixes follow, when the sense is “coagulation ” or “cold to touch”. The word gagft means hardening of a fluid, by coagulation &c. Thus ett wa, star vat, ett Ha: “a coagulated butter, grease &c”, The w of nishtha is changed to 4 by VIII. 2. 47. But when the sense is that of ‘cold’, the q is not changed ; as eft ea, eat era, efteaeaa.w The word afte is here used both as a noun meaning ‘cold weather &c’,and an adjective denoting ‘cold’, ‘There is no vocalisation when the sense is not that of ‘coagulation’ or ‘cold’, as tearit qfwra: ‘the rolled up scorpion’, The short ¢ is lengthened in wtat &c, by VI. 4. 2. wea WRG Tere N ret:, @, (GraTTorA) It age 0 ee fe THR | sreeTeee eareTerat Tem ETC TT Ht 25. There is vocalisation of ga preceded by the upasarga mf, when the nishthé affixes follow. Thus afirdta: afadtrarg.u This stra applies to cases where the words need not have the meaning of “coagulation” or “cold”, Rrarar ssreaqer wg ara Reaver, a, aa, Ger, (rare afer: na fae eat | ait ore Carrere saraafieral forrer earch aT 26. There is optionally the vocalisation of wat followed by the Nishtha affixes, when the upasargas afiy and aq precede it. Thus wfritet or wprRarr gd, eerie or srasarta, qfirat uw This option applies even when the word means “coagulation” and “cold”. As raid or srrgat ga, HE Werte or sramaT arg’, Ke. aN or STTRATATBTE The a4 in this stra serves the same purpose as in VI. 1. 23. Thus anfacett or sifirdgart, sreieti, sratreart | According to the author of Sid- dhanta Kaumudi, the word g# shows that it is a vyavasthita vibhasha, hence there is no option allowed when wf# and #1 are in the middle, as, eafoart, amrqqart Here we cannot have the alternative forms efit &c,BK. VI. Ci. 1 §, 29.) ‘VOCALISATION. 1033, et TH WOM TAMA N oTTA, TTS, Ht aR u Roreagray oar Te TET UTAPsrreMTOTAT T TH SHI TERY Tet pT Proves Pore 27. Optionally »t is formed in the sense of ‘cooked’, by the vocalisation of the semivowel of wt before the nishth@ affix = ‘The word frat is understood here. The roots & ‘to cook’ of the Bhvadi class, and "at ‘to cook’ of the Adadi class, and "at ‘to cook’ of the Bhvadi sub-class Ghat4di, are meant here. In all these, »y replaces wt, whether causative or not. Thus >t eit, opt gett This is a vyavasthita vivhashd 5 so that wt is invariably changed to » when referring to wt and ef, but not anywhere else, Thus stam aang, sitar wary: (VIL. 3. 36 and VI. 4.92) This word does not take double causative though the sense may require it. As waft Gre Fare aA “Devadatta through Yajiadatta has caused the mille to be cooked”. The mr is intransitive. When the sense is that of Reflexive, or Causative, there the form >qq is to be used. As ya drt erane, ort aC trea a cara: nen want nw care, dt AR: Pear | Tread TareRveR area erae Roary tate eee 28. dt is optionally the substitute of the root cart, ‘to increase’ before the Nishth& affixes. The root sireardt Gar belongs to Bhvadi class, The indicatory wt shows that the affixes # and aa are changed into @ and awq (VIII 2.45) Thus ofa gen, BY Ty, Gage U The option here is also a regulated option (vya- vasthita-vibhsha). ‘The substitution takes place sxvariably when the root is without upasarga, and never when it is preceded by an upasarga. Thus gear. ‘wearameamt: tt The substitution, however, must take place when the preposition ware precedes the root, and the words wy! and syq are in composition : as ending: erdragt facrstar WRN oar faz, TST, w, (A) afte u Pearere Frege cara: ftetreeerarg meat | PARE ah gem: Care: dtenarTt wai w 29. There is substitution of for cara, when the affixes of the Perfect (f@) and the Intensive (#@) follow. The anuvritti of the word frraret ceases. The phrase care: @ of the last sOtra is drawn into this by force of the word = ‘and’; thus wf, sit Frama, rf N The substitution of # a subsequent rule, would have debar-. red reduplication which precedes it in order. The reduplication, however, takes place after the substitution, by force of the maxim of ga! WER &c men- 31054 VocaLisaTion. Bx, VI. Cu. 1 § 31.) tioned in VI. 1. 17, Thus @+Pe-M+a+a=Rteate (VI. 4. 82 the @ being substituted) = srrfae% with the affix wrt. So also in the Intensive, as wrfatai, sinha, ender 0 * flare ga Wo GaN Arar, 93: Ae free TOR THR ree TE ats oe eater eoreTTOE TR | 30. There is optionally the vocalisation of tho semivowel of fia before the affixes of the Perfect and the In- tensive. ‘The phrases far adr: and gover are to be read Into this stra, Thus Ut or ferara, TIVE: oF FUT! u So also in the Intensive as enya or weit- Ga The root fig would not have taken vocalisation before ag by any pre- vious rule, this sQtra teaches optional vocalisation. The root would have taken vocalisation before ff, which is a fara, affix, invariably by VI. 1. 15, this stitra modifies that by making the substitution optional. In the alternative, when the root does become vocalised, the reduplicate is also not vocalised in spite of VI. 1.17, This explains the form fapara, which by VI. 1. 17 would have been sparen fer+og=(Wartmg VI. 1. 30)=g+g (VI. 1. 108) =u +048 (V1.1 8) gee oft Graret 1 8e TAHT NR, &, AL, TST, ( GTEC, ) I A Rome RR HN eRe re a oe ee: Taha ere 31. There is optionally the vocalisation of tho semivowel, in the causative of f#, when followed by the Desi- derative a and the Aorist we affix. The phrase fewer 8 is understood here. In forming the Desiderative and the Reduplicated Aorist forms of the Causative of fig, there is optional vocalisation, ‘Thus qurefasfit or ferarafagia w So also in the Reduplicated Aorist svg, as wrgaeg or sifirrag By the maxim émercet eared wetar araftr (see stra VI. 1. 17), the antaranga substitution of Viiddhi &e is supers eded by the samprasdrana and the subsidiary operations relating to it here. The vriddhi and the substitntion of wir take place after the samprasdrana has, taken place, The stra VII. 4. 80 teaching the substitution of ¢ for the ¢ of the reduplicate, when qq follows, shows by implication that the substitutior caused by fir is sthAnivat though it itself does not cause reduplication (I. 1.59). Thus fat fra +e =sTe+t=nfa nt In reduplicating wafa we cannot form sapara, but must consider the substitute equal to the original ¢ of ft and must reduplicate it. We thus get firarafwafi i! In the case of vocalisation, the rules of Vriddhi &c are postponed, Thus fig +1@-+04( VI. I. 108 Jog+ f+ waa OT FT THR Hee Oat TN Herewe requ:Bx. VI. Cu. 1 § 34.) VOCALISATION. 1055, ire reduplication. Here the substitute erg in urg though not caused by aa, is treated sthAnivat under I. 1. 59 tog, which iscarricdin reduplication. Similarly the Aorist: fT+M+SE= T+ C+ Hgts THe Regt OE +e+megtottete (VIL 4 egtwarete (VL 4. st )=aqurt (VIL 4.94) Grae WRN TATA He, SoreacoTT OR 0 oe TET TH | GTC TE Ge oir GUT: GTEC ATT 32. There is the vocalisation of the semivowel of the causative of Ave (®) before the Desiderative and the Redu- plicated Aorist affixes. ‘The whole of the last stra fs to be read Into this sOtra. Thus gera- faeit and qerafae:, werafiata; argon, srEeTAL and eragey i The root gr does not take the augment ® required by VII. 3. 37 before the affix fir, because the Samprasarana rule is stronger. The repetition of the word gwarcor in this attra, though its anuvritti was present, indicates that the force of the word fearet has ceased. Though this and the next stra could well have been made one, their separation shows that the samprastrana does not take place when ano- ther affix, not causing reduplication, intervenes. As grenfteart= graxtar®; the Desiderative of this verb is Fargrattizaf® 1 ATE TNA TATA TTR, &, ( rac ) tt BRE Tae, ep AeA cafe, PAETET A KAR, MEPTETET Kale:, arco, RapmerareTe RAR! NPA RATATAT ae tt 33. There is vocalisation of the semivowel of gr in the reduplicated form, in both the syllables. The abhyasta means the reduplicate and the reduplicated,both the syllables. The vocalisation takes place before reduplication, ‘Thus aqry, ait, and gaia u- This and the last sdtra are one, in the original of Panini, they have been divided into two by the authority of a Vartika, age ote ae Tea aged, Sate, ( Grercre ) tt We eee TH | eft Pete Raa aes GorATOH eT UL 34, In the Chhandas, the semivowel of this root is diversely vocaliscd. ‘Thus gi or gamit, as qeareft ge ‘Invoke Indra and Agni’. Yar ecerdt "rN The form gris Atmanepada, Present tense, 1st Pers, sing. the vikarana ‘Uz is elided, then there is vocalisation and substitution of 7g So also, gamit nee firs, wart Preaa yr So also gy: as pftera( Rig I. 2. 1. ) ‘hear the invocation’,“1056 VOcALISATION, (BK. VI. Cu. 1 § 36. Sree TW AL TaTa HN aTe:, SH Re ays orretfa Tae | raeatar ore Prod aes eran wate 35. For rz is diversely substituted = in the Chhandas. ‘Thus Pagar mifeeg:, ated Pregt Piregcaa lt These are forms ending In the affix aq of the Perfect. Sometimes there is no substitution. As erftr- waffaftrarea: 0 arpasgragirsaitreanmna: firsarsicratra: ak watt i regia, weg, ag: freq’, fears, wear, fran, raft, araftat:;, TAe wep tare TTS: oTZE: Prog’ fears, rat, Pree, orrett, errata, CaP Brn Fea 36. In the Chhandas, the following irregular forms are met with :—aqegtura, aay; atgg:, (say, fears, wan, frag, araftt and srreftd: 0 . ‘The word graf is understood here, From the root eq ‘to chal- lenge’, is formed erqeqivura, being the Imperfect (a), 2nd Pers, Dual, Atmane- pada : there is reduplication of the root, vocalisation of {, and the elision of wicregularly. As, ware feet aereqrareu In secular language the form is wraifargn Some say, it is derived from erg with the preposition spy, the vocalisation of ¢, the elision of wt, and the non-prefixing of the augment we in the Imperfect (VI. 4.75). The counter-example of this will be erqrerat- ‘argu From srt and wg ‘to respect, to worship’, are derived emrqg: and 7 Gin the Perfect before the 3rd per. pl. zw, there being vocalisation of ¢ and the elision of er irregularly. Then there is reduplication, then gg changed to w, then the lengthening of this w, then the addition of the augment, as : wet eg SETHE HRT HTT tT (VIL. 4. 66 eo teretag (WIL 4. 7o)=et tater toa (VIL 4. 71)=eTeg: Ww The irregularity consists {in the samprastrga with the elision of wu Thus agar srainrqy:, MEATY! Wt The secular forms will be wrrag:, ster: The form faregy is the Perfect 2nd per. sing. of the root eg to go’: there is vocalisation of the reduplicate, and the non-addition of the augment f= before the affix Wu This is the ir- regularity. The regular form is qegftt The form farars is the Perfect of ay, the vocalisation of the reduplicate is the irregularity. The regular form is qears From the root sft ‘to cook’, is derived sgrar before the Nishtha affix, ft changed to wt irregularly. As saver qerarar: uw The form fxd is ‘also derived from the same root by shortening the vowel with the same affix. As att the afta, Fret at qe: 1 Some say the ay substitution of wt takes place when the word refers to aa, in the plural, and fi when it refers to otharBx. VI. Cu. 1§ 37] VOCALISATION. ios? than Gr n Sometimes the word *arat is seen in the singular, referring to objects other than tw Thus af sant gata u In fact, the exhibition of the word -srar: in the plural in the stra is not absolutely necessary. The words erat and wrretet: are from the same root sft, with the prefix sg and taking the affixes fyLand w respectively. Before these wtis replaced by eX, and the non- addition of in the Nishtha is irregular, As, rand yy emdht waa, Wrehere wares TATA ATTA ON TATA A, ATA, THTATCT It gfe neta ea: ghee a: eet TT | attra afte remeernieaeeorefs Wate WSR cL 37. When a semivowel hase been once vocalised, there is no vocalisation of the other semivowel that may pre- cede it in the same word. ‘Thus eqy; has two semi-vowels ¥ and @; when qis once vocalised into & the preceding % will not be vocalised into zu Thus we have re: From exa— Parra, fax—adta: Though the rule of vocalisation does not specify what particular scmivowel is to be vocalised, yet according to the maxim that the operation is to be performed on the letter nearest to the operator, the second semivowel in the above cases is vocalised and the first is not, by force of this rule, Had the first semi-vowel been vocalised, then there would be no scope for this stra, because then no scmi-vowel will be found preceding a vocalised letter. This stra is a jftapaka that the vocalisation commences the second of the conjunct semi-vowels. Though the anuveitti of gare was understood here, the repetition of this word shows that the prohibition of double vocalisation in the same word applies even when the semivowels are not contiguous. Thus by VI. 1.133 there is vocalisation of the word gva, the being changed to z, the @is not changed as gat It might be objected that when the @ of g, and @ the sam- prasarana of q coalesce into one by sandhi rules, then there being the substitu- tion of one long & for the two @’s and this is sthanivat to the original, the and (of gra should be considered in fact as contiguous and not separated by an intervening letter, and therefore, the explanation given above does not hold we reply. The substitution of one long vowel for two vowels, is not considered sthanivat (Sce I. 1. 58). Even though it be considered as sthanivat, it is still a separating letter. Vart:—There is vocalisation of the semivowel of ft when followed by ‘4, and there is elision of the of wa, when it refers to Metres. As fim weet afert= ge ges UY aT The word ge takes the samasanta affix st by V. 4.74. Why do we say when referring toa metre? Observe eget af t1058 VocaLisaTIon. (Bk. VI. Cu. 1§ 40 Vart:—In the Chhandas there is diversely vocalisation of the semi- vowel of efi followed by the afi age as tf + m= CCF ASCH CH ALS CF EL (VILL 2. 15)= 8a: as, er eviag a Fee: Sometimes it does not take place, as efaara gfewien: i The a here is not changed to @ as required by VIII. 2. 15. Vart-—There is vocalisation of yar before ma when a Name is meant: as watered a shfir: This Vartika is unnecessary. See VIII. 2.12 where the form arefina_ is given, fare ert aT ae oT nl Pete, ara:, a: (G@rETCOnT a) safe wt eircom 1 FRE sea ar RTC TET ETH 38. In the substituted root wa (IL. 4. 41), the @ in the Perfect is not vocalised. The phrase a aware is understood here. ‘Thus zara, 3aq:, 5g: The word fag is employed in this aphorism for the sake of subsequent sftras, this fone could have done well without it even. mrarerraacent Fer WAN TATA Ne, , we, AeTATEATA, FTA af Ree ere at rare PR FE sea RET EATTEATE 39. Before the tense-altixes of the Perfect that have an indicatory ® (I. 2.5); for the qofaq may optionally bo substituted a & Thus sag? or 5rg?, RE: or ay: According to PAtanjali, the phrase ware of this stra couldhave beendispensed with;this much would have been cnough:— eqemacem rarer: W Thus + erra= at + a+ ge = wag: and ag: (the vocalisation being prohibited), In the alternative: q+ gq=a+uqu-a+-are+ MAE (VI. 1.77) =3"g', 8G: uw Here there is vocalisation. In the case of ga substitute, the ais never vocalised, so we have sag, &&: | Thusall the three forms have been evolved without using aret Ber WoW Tari asr:,(Gsrarcora, 4) tt Ou Fabregas oy eegetarreaten eran? cea: Soret tafe 40. The semivowel of % ‘to weave’ is not vocalised in the Perfect. Thus ei, exg:, WEN This root belongs to aft class and would have been vocalised before fiat affixes by VI. 1. 15; and before non-farm affixes the Reduplicate syllable of the Perfect would have been vocalised by VI. 1. 17. Both vocalisations are prohibited here. wafe ane naar n cafe, a,(arercora, a) 0 ghey cara cafe we get As SrETCOt a NT 41. The semivowel of & is not vocalised when the Participial affix era follows.Bk. VL Cu. 1 § 45] VOCALISATION. 1059 ‘Thus gare, tvara ll The separation of this sdtra from the last, is for the sake of the subsequent sdttras, into which the anuvritti of emg only runs. SaaT NBN TATA Nee, @, (SaTCONT a) Ut WV: weedeat ean aera Tete eT: rT ATE 42. 'The semivowel of sat ‘to grow old’ is not vocali- sed when the affix ev follows. Thus wvara, s7sara tt SeT NBR TAT Hse, , (AorETTOTAT, #7) ogee: etter | Saray error eTATEAT TOR: CETETC TTR 43. The semivowel of sat (a) ‘to cover’ is not vocalised when the affix 1a follows. As vara, syeura tt The separation of this sdtra from the last, is for the sake of the subsequent stra, in which the anuvritti of & runs. frarr at: nee oT Bran, a, (@erercora, a) I af eae oy spire Tint Teper AARC TTT Te rare eT A we 44. The vocalisation may. optionally take place when s@ preceded by af takes the affix eran ‘Thus 9ft#ta grt or qieary tt The augment q™ presented by VIL. 1. 71 Is debarred by VI. 4. 2. which causes the lengthening of the vowel: since VI. 4-2 is subsequent to VI. 1. 71. arta sagT SA VG TTA Narra, Te: TATA, TANT gfe: rar FRE CR wT ae TAUNTS erearenrrte arate Fale ye TT 45. Ina root, which in the system of grammatical instruction (i. ¢. in the Dhatupatha), ends with a diphthong (a, 2, af and a), there is the substitution of at for the diph- thong, provided that no affix with an indicatory g follows it. The word writ: is to be read into this sOtra from VI. 1.8 Thus a= sare, erga, Taree, St—Prare, Praraes, Marre! Why do we say ending with an Tx (diphthong)? Observe wat from w, and ga from gn Why do we say in upadesa or Dhatupatha? Observe "at, wirar where & and ei are not. the forms taught in the first enunciation of the root in the Dhatupatha, but are secondary forms. Why do we say, before non-firt affixes? Observe #-1-0%, +fiq= rarer, ere | Exception is, however, made in the case of the affix mol the Perfect: as aH, 4 U This is explained by interpreting the word fer as fret, that is the affixes having an indicatory W in the beginning: has qat the end. This is done on the maxim afery Pitvermrarrg agit “when a term which denotes a letter is exhibited in a rule in the form of the Locative1060 VocatisaTion. [ Bx. V1. CH.T§ 48. case, and qualifies some thing else which likewise stands in the Locative case, that which is qualified by it must be regarded as beginning with the letter which is denoted by the term in question and not as ending whith it ‘The word wfara is an example of gasaufaru:. a simple prohibition of acontingent case. Therefore, the substitution of wt for gq vowels, is not caused by the affixes that follow, but must take place prior to the occasion for the application of the affixes arises. Thus III. 1. 136 ordains ® after roots ending in #17; so that * and #% are presupposed to end in em when applying this affix: thus gra, ges: Similarly byIIL. 3. 128 we apply 34 to #% and 3 by presupposing it as siya roots: as TaTH:, Gra: It The word wr is understood in all the subsequent sAtras upto VI. 1. 57 ar aeit Ff weg Tia Hh a, Ba, Fesf, (seen) gee ul Say eater era Ter rrr AT Tafa 46. There is not the substitution of at for the Diphthong of the root #t when the affixes of the Perfect follow. Thus dfeeara, afeafirr u The reduplicate of the Perfect is vocalised here by VI. 1.17, The Vriddhi in dfsata takes place by VII. 2. 115 before the for affix org RHUISHSAHAT WON Teh eRTS, FeM:;, afk, (aa) BAR: ure Nf TAT Lege eGR TaTTTrated er ey a oe STEER HATA 47. In the roots €§T and Pw ‘to move’, there is the substitution of the #t for the diphthong when the affix ‘wet follows. Thus Freere: instead of fete, and Freer: instead of feetia: uw By VIII. 3. 76, the optionally changed to q after f%, as faemre:, and FeeRtT: U strestat ot ween wqTAa Hat, wa, StretTar, oft, (erat) ft ugar earth, cm TE, Fae ate, RAGE TETAS: Ce AT TET rTeRTCTT wat 48, The substution of ar for the diphthong takes place in the causatives of the roots mi ‘to by’, ¢ ‘to study’ and ‘to conquer. Thus servaft, eereafe and argafy The augment gis added by VIL. 3. 36, since these roots end in long 8 Rereadcarcalt aa wee TTA reas, staTeaheE afte ona | Fog Hear Rene urreaTeah PRN aT: EAT TCT eRe TTT TTTBK. VI. Cit. 1 § 50. ] wr SUNSTITUTION. ro6t 49. The substitution of at takes place for the diphthong in the causative of fra, when it does not refer to the next life. The word grat is derived form quia ‘the next world’ by adding the affix #5y with the force of ‘for the sake of '(V. 1. 109). The double Vyid- dhi takes place by VII. 3. 20. The sense of the word fet must refer to srr or non-intelligent things for the purposes of this substitution. Thus wrt eraatt, ‘art araafa nw Why do we say when not referring to the next world? See eqearred traf, ereaet aeiiien Brain The force of fe is here that of know- ledge, arva: feifa = aritrrararaia | The ascetic acquires certain know- ledge through austerities, the knowledge so acquired produces its result in the next world (qU&@#) i.e. the next life, therefore, the Wrat of the ascetic Is for the sake of the next life, and so there is no substitution of sir for gu Why the substitution docs not take place here: srt qraate, eretparrenix ‘he causes food to be prepared for the purpose of giving to the Brahmanas’, Here though the gift of the food produces effect in the next world, yet as that if the effect of qa and not of the verb firg, the substitution does take place. ‘When the verb Firg directly and not through the mediation of another action, produces piraloukika effect, then the substitution does not take place. The verb ft here belongs to the DivAdi class and not to the Bhvadi, as the form Rraife in the stra shows it. _ atarfaRrifardret aft use Ween afte, Rreitfe, reve, waft, awa aA: U ore aera eae ay Rrarata, Por ae, Ae are, aR TTT ale fata sareires Fredy aires HR TETAS CT STAT ATR 50. And there is substitution of at for the finals of at ‘to hurt’, f ‘to scatter’, and € ‘to decay’ when the afix gag follows, as well as before those affixes which demand this substitution for the diphthong. By force of the word ‘and’ in the stra, the substitution takes place of the diphthongs also of these verbs before all affixes other than rea, and this substitution takes place, before the occasion for the applying of the affixes arises (avruraturara stret 9f@) "So that these verbs should be under stood as if they were enunciated with an wf, so that all rules of affixes relat- ing to sqt will apply to them. Thus qatar, Tataeak, TATE, ware, Pavan, Pome, Rerasan, fara, TTA, THAT, THTTL, THA The substitution of wr be- ing understood to have taken place in-the very ¥¢a (in the dhatupatha)-of 41062 wr SupstrruTiON (BK. VI. Cit. 1. § 52. these roots, the affixes relating to roots ending in x or ¢ do not apply to these atall, Thus aqqrar vai, faqrart, formed by wy and gw by taking fra and adding these affixes (III. 3. 128 and LL. 3. 18) and not the affixes ra (ILI, 3. 56) and wa (IIT. 3. 126). ferarar area: We cae nt Rrra, aera: A eatiee we, arte Ser Tt RorNTTA Fee: At ere maTreeM ATT sears ar PE: ea | ataraterodrs a TaN Fare BTC ATET TE RATT FTA ATT oer rea eran Fro Fteat exert: wae aR: 51. There is optionally the substitution of = for the final of #1 ‘to adhere’, in the very dhatupatha, when the affix eqq follows or such an affix as would demand this substi- tution of the diphthong. The final of at will take gupa substitution before far affixes, and will become %, this incipient diphthong q is changed to amt by this rule, The same is the case with all the roots subsequently taught, thus fag will be &, and then @ changed to sit, t= sit and sit changed to wT &c, Therefore we have employed the anuvritti of we ‘diphthong’ in all sdtras. The words rary and sittat sata are understood here. The roots # belonging to Dividi and Kriyédi are both included here. Thus frorat, fearas, Fraratat, fers, fatar Rear, fears and fata u Vart:—The wt substitution does not take place when the affix wry (IIL 3. $6, IIL. 1. 134) and ea (IL. 3. 126) come after fa, 4, ft and a: as, atone: Prova A | deere, emai ea | hee ose, Pree The #f invariably takes sq substitution when the sense is that of ‘showing respect,"deceiving’or ‘insulting’, As atergarra®, saatt afrargarcat 0 The option allowed by this aphorism isa restricted option (vyavasthita-vibhasha) ‘The substitution of wrt for the final of € is optional when the sense of the root is not that of ‘showing respect’, one of these three senses, the substitution is compulsory. See 1 + Petar WR TET Nat: GTA afer wR ere tearrnen Teer: eH SRT Fre Rare saree BET HAAR - 52. There is optionally the substitution of at in the room of the diphthong of the verb férg ‘ to suffer pain’, in the Chhandas. The word faurqr is understood here. Thus frst Frearg or Freq In secular literature we have Ferst aft w ‘subduing’ or ‘deceiving’. But when it has any 70.Bx. VI. Cit. 1. § 56] war SUBSTITUTION. . 1063 TTD TAS WZ TNA WaT, TT:, oats, (aT) ght 0 get carted orarergtes ahs ser Tat ea rea rare eT Ta | 53. There is optionally the substitution of atin the room of the diphthong of the root 4t ‘to exert’ when pre- ceded by the preposition #¥ and taking the affix HR 4 The affix w94 forming the Absolutive Participle is added by ITL 4. 22, and the word formed by it is repeated. Thus orqarenqaren, oF sryererasite So also this affix is added by III. 4. 53: as, Breage Gave or wreqire gure “ they are fighting with raised swords’ Pregacit n qe naar n Re, eRe, or tt gf 0 Fey Ege Ceaairatearett sear CE ea PraraT STATES AT 54, There is optionally the substitution of at in the room of the diphthong of the roots f and "Re when in the Causative. Thus sraait, waa (VIL. 3. 36 for the addition of ¢) so also eatait or enreafai saat fara: wy hoger wae, area afte u omit aa A mR TRTTETAAT Cen TAT THA TATE Tea A TST HRTTTT ae | There is optionally the substitution of a in the room of the diphthong of the root #t in the causative, when meaning ‘to conceive an embryo’. The root # Adadi 39 means ‘to go, to conceive, to shine, to eat, and to desire’. The substitution takes place when it means ‘to impregnate or conceive’, Thus git arate: wargaf or aaraaie= ai ureaft The word #7 means the receiving of the embryo which in course of time will lead to the birth of a child. Praxezas ign agri aaa, Ty AWW AP u onraia Ta? Ponrea ae batts ToT: ATTEN MAORI TEAL, TAT aaey area, aaa REAL! ae Tearreg He areae Tato Tea Pera TATE wa 5G. There .is optionally the substitution of at for the diphthong of the root at ‘to fear’, in the Causative, when the fear is produced directly through the agent of the causa tive. The words er and fer are understood here. The word ¥q in the sQtra is the technical ¥ meaning eqaeaea TatTR ([. 4. 54 and-55.), When the1064, wr Susstirution. ° — [ BK, VI. CULT. § 58 Hetu Agent himself is directly the cause of the fear, that fear is called the Wepra “the fear caused by the Hetu”. Thus gra arma (VIL. 3. 36) or afear (VIL. 3. 40°. So also sfeat arqait or viqaz This root takes the affixes of the Atmanepada, by I. 3. 68; and the augment g is added by VII. 3. 40, which does. not come when there is ai substitution, for the ¥tin VIL. 3. 40 is equal to r+ tiie. At ending in § and means sf ending in € takes the augment § U Why do we say ‘when the agent of the causative is the producer of the fear? Observe giswatt wraafa term: tt Here the fear is produced from the giéqar and not from Devadatta the Agent. Fret erat: Go rer fete, TAT: Hh af Wome Ta, RTE oe Rremreenforans Pree fore trmerrarer arate wad ott qeat Prermencteat ATF 57. There is invariably the substitution of a for the diphthong of the root fet ‘to smile’ in the Causative, when the astonishment is produced directly through the Agent of the Causative. The words & and Yara are both understood here. The word Freq shows that the anuvritti of frarar ceases. Thus gedraftat ar fremraa u Other- wise we have @feenatt fretaag uw The word wa here is taken to mean erat i. €. ‘wondering, feeling astonished’, See I. 3. 68 where also this meaning has been extended to wa by its connection with stfen together; the proper word ought to have been Were in connection with fq, and Fra in connection with ¥f w afrentderafaret 4c cate ah, wath, ae, ane, afer DU er Prat, eft Hater, eaettateretorerenat rae Feat serever eaTA t 58. The augment epg (a) comes after the 9 of ay ‘to create’, and ea ‘to see’, when an affix beginning with o wa letter (the Mute and the Sibilant) follows: provided that it has no indicatry =u Thus $E+ GI T+A+T+Y=ag nom. sing. wey, so also wey, wea, Wer, KE EUR! This sry augment prevents the guna substitute ordained by VII. 3. 86, But in srarefta, and ergreft, the Vriddhi takes place in the Aorist, after the augment wa had taken effect, Why do we say when beginning with a Mute or a Sibilant?- Observe qarra, viva with gu Why do we say not having an indicatory a? Observe get, we: before the affix wn The forms of roots being exhibited in the sdtra, the rule applies when affixes relating to verbs con fter these, and not when other affixes relating to noun follow. Thus Farrar Mere the words are used as nouns rather than verbs, For the maxim wat: eqHet worm ari applies here,Bx. VI. Ci. 1.§. 61 J why SunsTITUTION. ° 106$ FATT AHTTETHTTTEATAT NGA TATA UM TATA, @, HA, T TIE, TTATeATT I UC: Tee aeRe THR, reaP aA 1 AT SAT eTET MTT TTT ATTRA TTT, eT sraercearTareT wafey 59. The augment am comes optionally after the % of those roots which are exhibited in the Dhatupatha as anudatta, when such roots end in a consonant having a 9 as penultimate, and are followed by an affix beginning with a. mute or sibilant and not having an indicatory & The word g¥qw is understood here, so also geaafafit "Thus ear or afar or ant, wat, eivar or gRT HW ‘The roots Ga ‘to satisfy’, and gy,‘ to be happy to release’, belong to Divadi class, sub-class RadhAdi, and they take the aug- ment ¢g optionally (VII, 2. 45), so we have the three forms given above: for these roots are also anudatta in their first enunciation. Why do we say “which are anudatta in the Upadesd or the system’ ‘of grammatical instruction”? Observe aat, ga, Teen, from we gama which is exhibited as war, and because it has an indicatory & it takes Ye optionally. Why do we say ‘having a penultimate m letter’? Observe en, Bar! Why do we say ‘before an affix beginning with a mute or a sibilant’? Observe wien, toma Why do we say ‘not having an indicatory &’? Observe qm, ‘em i) Before affixes not beginning with a gr consonant this augment will not be inserted, as arform, aor sitteserfer gol cart Ul ita, are tu atti werrat Froude eared weft Pees Prarae tt 60. The word ire is found in the Chhandas, This word is another form of fart: and means ‘head’. This is not a substitute of fart in the Vedas, for both forms are found therein, Thus reat fe aa airs ata eeiea, aH eftsoif Iitrawu In the secular literature there is only one form fae: Serafea need waa, w, wea afte uw eihina waa | ertrahteas emir afeant fe ot Petraes erat spt Premepabarata | wearer areas ser: Racrerer Utara rata afar ut Rag 61. There is the substitution of this stem sit for farcq when a Taddhita-affix beginning with q follows. The word wt is understood here. This rule teaches substitution, The original for which this substitution comes is not given in the stra, we must infer it. The appropriate original is fang u Thus wtiwat & gait wae,1066 Some SunstiTUTION. (BK. VI. Cu. 1. § 62 The affix aq ishere added by IV. 3.55. The word wim retains its original form before this affix aq, the final sq not being replaced by anything else (VI. 4. 168). Why do we say when ‘a Taddhita-affix follows’? Observe far geafa=fareaia, here @ is not a Taddhita affix, and so there is no substitution, Vart:—The substitution is optional when meaning ‘hair’, As id= ear: ut: or fircea: Sar: Wt RT WR oar n arf, athe DR: uw orardt arate Arce: etree rita wares 62. There is the substitution of sitt for farcq when a Taddhita affix beginning with a vowel follows. Thus eftatredrstet = writer: formed by adding the Patronymic affix T= (IV. 1, 96). So also euefirer (as etrenftia Had the word been eft (instead of wit as taught herein) then it would have retained its.final before these affixes and would not have given the proper forms (VI. 4. 168). Again in forming the feminine of erfeadftie: by adding exe (IV. 1. 78), arises this d culty :—cag (a) is a Taddhita affix beginning with a, when this is applied to wrfearsfift:, we must apply the last rule and change the etd into ftir (for etd being the substitute of farce, is prima facie a substitute of dt also for the purposes of that rule VI. 1.61). The form which we get will be this, erfta- Wie: + ea =eattin+ a (VI. 1. 61) = erfeavindoar (VI. 4. 168). But this isa wrong form, the desired form is @rfeaditeat How do we explain this? Thus afeatine + ore = erferate + a (the x is elided by VI. 4. 148 and in the room of ¢ we substitute a sero or artim) Now this adeSa becomes sthanivat to © thus the affix q not being directly applied to etd, because this zero intervenes, wht is not replaced by et¥1 as required by VI. 1. 61, for it is not followed by an affix beginning with q but by a zero sthanivat to qszalfix. This sOtra is not of Panini, but is really a Vartika raised to the rank of a sOtra by later authors, TAAL TT TAC TESS HATA ATEMTTT WER TTA MN UZ TH AA, AG, Ta, AY, aaa, qa, Ta, aHI, Tea, sea, area, we, afag UGC u ore ea Ter are Cra Pre TS BN ae ABT UET Tre TTT Kae TET Bart Cer dfamaTag TH: IE TL LATO Ae ra BT aT RT OT TEL TTL ARR wrt UTE afFA wren oafty aierergTaaarag ware UTE Nera AAT EET Te ate woleeeerte TemeBk. VI. Cu. L. §. 64.) & SUBSTITUTION. 1067 63. In the weak cases (beginning with the ac- cusative plural) the following stems are substituted :—qg for wrt, &a for qea, TA for arfieet, arg for are, eq for wa, Fra for Fron, aera for at, gre for qa, Taq for gre, THT for TEA, wed for wea, sax for seH and sag for areT The Kasika gives seg as the substitute for saa, the Sidhanta Kau- mudi gives the original as sia which has becn adopted in the above. Some say that these substitutions take place in the Vedic Literature only, others say, they are general. Others read the word ‘optionally’ into this stra from VIL 1. 59 and hold that these substitutes are optional and not compulsory. Examples :—1. aq —faewait wife, varema gee 2 eal —ar Tet MTT Te are 3 ae eRTeTETET U4 AT-M CRT TAT HS Eee TT ATET arePere | 6 Frat—srarrentat Fafer are 7 wre meeRT era» 8 BT aT grantor goon ater w 9 erat -aae Hem hiram 0 10 aRa—amT Taf UL wae waa Tara 0 12 gya--s Pr Feaeg ATT A 0 13 Terra Fy ae EEA Why do we say when the weak terminations ger &c follow? Observe ara & sfatteat are Bit As examples of this substitutions in the secular literature also, the following may be given :— sararengoor are TxATgT ACH S| eaneat arvana Weare ‘These substitutions take place before other affixes also: as, ear feo, RBI args aeTT Vart:—The following substitutions also take place: aie for wie, 7% for waar, and eg for arg: as atari niewstear: for niwaerat: (Yaj. XXV. 36): Te mera for geargaaa, 9 a feat a gfesar wTveay for wAraTEy Vért :—The @ is substituted for aftrat only when the affixes wm and rand the word gy follow : as, mewn, mee, nq GE: The wa, is taught in IV. 3. 55 and V. 1.6; atzis taught in V. 4. 45. Vart :—The substitution of 4 for miftrat before the affix wy does not take place when it refers to ‘letters’, or ‘ cities’ as: rftrsat att ‘a nasal letter” sfered are AaTT TT WRB TATA areas, T, GT gr w ararert: ences cute area Te nite u gorgeaerendiad wait aA G4. ‘There is the substitution of @ in the room of the ¥ being the initial of a verbal root as enunciated in the Dhatupatha. The roots exhibited in the DhAtup’tha with an initial , change it for Su Thus ve—aga, fra—firefa - Why do we use the word ‘root'? Observe1068 @ SUBSTITUTION, (BK. VI. Cu. 1. §. 65 Niza, sfem, gexs un Why do we say ‘initial’? Observe—afa, graft Why: have then roots been exhibited in the Dhatupatha with an initial g, when for all practical purposes this q is to be replaced by &, would it not have been easier to spell at once these words with a@? This appears cumbersome no doubt, but the spelling of the roots with q is for the sake of brevity. Certain roots change their qinto q when preceded by certain letters: those roots which thus change their letter e for g have been at once taught with an initial , and thus by VIII. 3. 59 their q is changed toqu Thus from fer¢—instead of faq we have raqait Thus root is exhibited in the Dhatupatha with a g, and thus we know that the must be changed town Otherwise a list of such roots would have to be separately given. One must refer to the Dhatue patha for a list of such roots. As a general rule, however, all roots beginning with aq and followed by a vowel or a dental letter have been taught in the Dhatupatha as beginning with an initial g, as well as the following roots fe, ‘emg, fer, tere and era, though followed by q org The following roots though followed by a vowel ora dental are not taught with an initial ¢ viz a FS, &, &, Ge, and git . Vart:—Prohibition must be stated of the roots derived from nouns, and of firgand eyex u Thus detait, geareit are roots derived from nouns, and though these roots begin with © are not changed: so fea—reafa, qeR—eqeR% It The word fet contains in it two roots reg and fer, one with g, another with vw, and thus we have two forms in the reduplication -—aefeay, zufeat The substitution of & for takes place in the case of the root that has It re We carte oT:, a I ott u ardrerefterg rae | areitcrtenrcen repre writ wave 65. There is the substitution of % for the initial q of the root in the Dhatupatha. The phrase uraiy: is understood here. Thus ofts—aaft, om—naf, ‘wera u But not s—sraf ut This does not apply to roots derived from nouns, wardiieaia=orattara The roots are exibited in the Dhatupa- {ha with 9 for a similar reason as they are exibited with € "By VIII. 4. 14, these roots change their # into ot when preceded by certain prepositions. All roots beginning with a should be understood to have been so taught, with the exception of the following :—a, afey aff, me, af, ary and arg BrIeT SS 0 ER WoT Ste, sr, af af u uratitia set aq ararentita grataaeniagag | Ba adterdrs earceareat- fe cea ra ATT 66. There is lopa-substitution (elision) of the q and q when followed by any consonant except %01BK. VL Cit 1. §.67) Tit ELISIoN OF @ AND qu 1069 ‘The final ¥ oF @ of any stem, be it a root oF not, is elided by an affix &e, beginning with a 7% consonant i.e. any consonant except au Thus E+ et (ILL 2. 107) =fry+ feq+4a=fafera nom. singular fefera, fiesta, fefeata: ‘Thus s+ a= aa, Spun’, FAta=Ea ‘made a noise’, So also the affix — comes after rat in forming patronymic (IV. 1. 129). Of the affix aa, ¥ is replaced by ea (VII. 1. 2). Thus arar+qa+e=tie (the # of the affix being even clided before €) So also in qireq, aaey the € of fae (IIL. 4. 102 ) is clided before e(III. 4. 105)So also wis elided in the follo affix afirq with the negative st There being diversity in the applica- tion of Unadi affixes (IIT, 3. 1), there is not 37% substitution for as required by VI. 4. 19. Why do we say “before any consonant except 4?” Observe saat, eat when @ is not clided. Why the word @t is placed first? The elision of qand % should take place prior to the elision of the apryikta % taught in the next sdtra, Thus arga+Pag=arg a (Vi. 4. 48)-arg+zearg: So tet +feaq= arg uw Why 1 of awis not elided before { which is a qq consonant? It is not elided, because it is so taught. Had the clision of # been intended, the root would have been enunciated as tw instead of aw t If you say the w is taught for the sake of forms like gat by samprasarana. and ww by reduplication, here also the ‘lopa’ would have applied, as being an Antaranga rule, while samprasarana, and the clision of ¢ by "enfeaq: are Bahiranga, ArgweT hl GO TENA NA, HET Ml aR: uate ele Tata Ath RerereaT ProeagerargeysaarTetet yer | Teer AT wea 67. There is elision of the affix fr when reduced to the single letter qn The affix includes Ferg, flora, Fer &c. In all these, the real affix is 4, which being an aprikta (I. 2. 41), iselided. Thus sgrer, qorqr (ILI. 2. 87). Here the affix Ferg is elided. So also yaere, Rear (IIL 2. 58). Here the affix ferg is clided. So also wnfare apie, afrae (III. 2.62). Here the affix fq is elided. Why do we say “of an aprikta—an affix consisting of a single letter” ? Observe aft formed by the affix Pr (Ft being the real affix) ; so also arafet formed fer, see Unddi Satras IV. 53. 44. No root can become a noun unless some kyit affix is added to it (see I. 2. 45 and 46); hence the necessity of these imaginary affixes, in order to raise certain roots bodily, without any change, to the rank of nouns—from Dhatu to a Pratipadika the way lies only through an affix. And though these imaginary affixes are after all totally elided, yet by 51070 ELIsion OF f&, AND gn [ BK. VI. Cu. 1. § 68 1. 1. 62, they leave their characteristic mark behind, namely the derivative word becomes a nominal stem &c. Thus fet words are adverbs (Gati) and Indeclinables. eqerert gabe om ach want Wo, TuTET, arate, g, fe, fe, wea, wR gen are ce whe ae ata Rarreeee aera Mee Bereefe aT | Eee eeraercare Rafett g far Fe CaTTTA UR TAA Keb s—eeiterere i Fe RTT Ra | rey re: eae 68. After a consonant there is the elision of the nominative-afix @ and the tense-affix @ and fe (when reduced to the form of an:l @) being consonants; and so also after the long vowels ¢ and at of the feminine (affix # and arg), there is the elision of the nominative aftix &0 The s@tra translated literally means :—After a word ending in a con- sonant, or #¥, or sira when a long vowel, the affixes 3, fa and fey when reduced to a single consonant affix (srg @R) are elided. But (% and fit are elided only after a consonant and hence we have translated it as above, As examples of the elision of g after a consonant :—asa+q-aara (VI. 4.8)+0=T8T, wer Toad, THEN After a gt:—as, Fad, a, uNFeH M After aTE:—aEE wgaan, art Tear The elision of f and F& tikes place only after con- sonants :—as, Bfeontare (2+ mE + E+ a4 Et T+ E+ T+a (VIL. 1. 10) mortattata (VIL 4: 66 and Lt. 51 )=ertatata (VIL 4. 60)=sr +e tata (VIL 4 54)eetietata (VIL 4. 76)=ettietarta (VIL 3. 84 ewer +q=wfeAt So also ware waren In both these cases q of the Imperfect has been elided. The mis elided in the following :—stfratsm, wafesatse un (fretee+faqu The ¢ is changed intoy by VIII. 2.75 and mis elided by this rule). Why do we say “after a consonant, or a feminire affix § and er”? Observe armft, ari w Why do we say “after a long vowelled feminine affix”? Observe facdrarfea:, winwg:, where the feminine affixes have been shortened. Why do we say “when followed by g, ft and fir”? Observe #i- eet The far being read along with fit, does not include fare, but refers to farqonly. Why do we say ‘when reduced toa single letter’? Observe fiat: fisatat u Why do we say ‘the consonant is clided’? Observe fairy, faery i: Here the aprikta affix of the Perfect is not elided: though it being the subs- titute of fir is like fit, and ought to have been elided, had the word g% not been used in the stra,Bx. VI. Ch. 1. § 69) ELIston OF fir, fi, gt ro7t Why has the clision of @ (3), a (1a%) and w (fer) been ordained after consonants, as they would of themselves have been dropped by VIII. 2. 23, being the finals of a compound consonant? Then the forms wr, Rat could not be evolved. As umq+q-amra+s, (VI. 4. 8)=aarg and by cliding & by Rule VIII. 2. 23, the form would have become arama, and the final % could not have been elided by VIII. 2. 7; for it is a maxim enunciated in the very opening of the Second chapter of the 8th Book, that in the last three chapters of that Book (2nd, 3rd and 4th Books), a subsequent rule is as if it had not taken effect, so far as any preceding rule is concerned ; thercfore, the rule VIII. 2.7 ordaining the clision of ¥, does not find scope, since VIII. 2. 23 (ordaining elision of =) is considered as if it had not taken effect. {o also. in the case of eraq and qensal aerterrt feng IL 2. 76not applying. ‘Thus freq gartrest Tara, “let the girl stay. Take the umbrella of Devadatta”. Here garg and 1% are not in construction, but qm is stll added optionally. Vart :—The-augment @ is optional in the Chhandas, after the words Prevaa &. As, PreraTes eH or frqraaoy, TeaTat FTL or TOTAL ReaTTTA eer aoa Wee Tata ee:, ay, ae safer wafer acer ar aur oT Ht erent: Gaaiea eesaRareard aoe eareER: : 77. The semivowels g, 3, ¢,@ are the substitutes of the corresponding vowels g, g, ag and % (long and short), when followed by a vowel. This sdtra is rather too wide. It must be restricted by VI. I. tor, namely the following vowel must not be of the same class as the preceding forBK. VL. Cut. 1 §. 80] TUE SANDHT RULES. 1075 the application of this rule. Thus efqtore= qm, my+eT=me, Bet wiqaatin. eft onitsenid, east sensi Vart:—This semivowel substitution of vowels takes place, when pre- ceded by a prolated vowel, even to the supersession of VI. 1. 101 requiring lengthening. Thus wat § 4 Gq =3Tat 8 PTL, TET YT TEHL= TET | TENT, STAT QU asa YATE, ery TTT Ter YT a a era ya The phrase ‘when a vowel follows’ exerts its influence upto VI. 1. 108. ey STATA: WS TTT MN Ost, ate - TS - TA BATaT H TE ee ear SA seat Sa Te BT BITE Tae UT AUTet AAT 78. Wor the vowels &, & sir and aft are respectively substituted #, ary, #4 and #1 when a vowel follows. Thus fq tege=%+ srr =saam, + aan; Yr aree:, Ae seree So also wad, sid, aretenta, sae (ILL. 2. 148)=t + aye Sra cet HORN TNA Naa: &, TAT ae: draive: ear eearta strarces sre vitarceary & waa WeTa ToT TTT aaftrana, u sing orf nate seafarers 79. ‘The substitution of aq and arg for sit and af also takes place before an aflix beginning with 40 The area are those which end in ¥ viz. stg and sgn Of the four substitutes tausht in the preceding stra, those which end in q (viz. eytand sx) also come when an affix with an initial @ follows. It follows that the substituted letters must be sit and ai Thus wy+an=aniv+a=aret: (VI. 4. 146 and TV. 1. 105). So alsonrezea:, greene, frrereq: arate, area V.4.01) en Why do we say “erg and srg"? Observe trafieata=tara, no change of % be- fore a Why do we say “before q"? Observe tiparg, ‘para Why do we say “an affix”? Observe rraraa, tara i Vért:—The word et is changed before gf in the Vedas. As st+af eeregrt, as srrir frat Ter Vitegitrarra.n Why do we say ‘in the Vedas’? Observ> erate Vért:—This substitution takes place when referring to the meceur ofa road: —as, -gficnrrmad a: This is in the secularliterature, aegft meat. ig ates ardteaPairreta icon cari Nard, aferfirees, 7 il gtr wee aft raat et fa cert erate we AP aTIOReTaahteTere amare Fae Tea aT Ae 80. For the final diphthongs a and & of a root, are substituted #4 and ara respectively, before an affix begin- ning with a, then only when such diphthong has been itself first evolved by that affix.1076 Tie SANDIIT RULES. BK. VI. Cut. I § 82. ] The words ew, Sa wey are understood in this stra. The word ertartret means ‘caused by that’ i, e. caused or occasioned by that affix begin- ning with qt Thus g forms its Future Passive Participle by a@(III. 1.97), this affix causes the guna of x by VII. 3. 84: Thus g+a=ar+a, which according to the present stra becomes meqmut So alsoq—dr+ a= qeamtt + oa(lI1.1.125) ot + a= sree area and rama grea Why do we say’of a root’? This rule should not apply toa nominal stein. For then, though it may be all right in the case of ay +4 =anat + a= enaea:; it will not apply to cases like tit4-q= meq, At += ATER, where wit and wt arc not caused by the affix,but are integral parts of the stem before the affixes were added. Why do we say ‘caused by that affix itself"? The substitution will not take place, when the change is not caused by that affix. Thus the Passive of ¥ with the upasarga sm is st+¥+ag+¥=Here by Samprastrana(VI.t.15)% becomes as tt+7+a+@; now by sandhi s+a—=sir VIL 1. 87), we have vit +a =shrai tt Since sit is not caused by @, there is no wa substitution. So also vitae, Srantfa:, dram: (IV. 1.95). The word gq in the aphorism has the force of limitation, with regard to roots. In the case of roots, it and it before # are changed then only to sq and wTq when q has caused the production of sit and sit; in case of nouns there is no such limita- tion. Here the substitution takes place whether the has caused the production of wit and sit or not. emeereett eaT kN TT Meee - Tea , FT - arf afte: Fat Fae eaetrcfesrr raT ce: eaTRE TeT eTCaTa TR Prorat 81. In weq and aeq there is substitution of a for @ only then when the sense is that of “ to be possible to do”. The roots fir and fy before the affix wa (III. 1.97) assume these forms when meaning to be able to do the action denoted by the verb, As wet aq =ere (fitt+a-%+a); so also em Why do we say when mean- ing “to be possible to do"? Observe aa ara, Sat que W Here the meaning is that of ‘necessity, wrererg MR TATE Hore, TTS gir u atonateteereat mrarit erate aR matt et Sart Prevent 82. In meq there is substitution of at for & when the sense is that of ‘exposed or put out for sale, saleable’. The word mea is derived from aft ‘to buy’, with the affix aq; the guna g being changed tom The word wmf means ‘for the purpose of that”Bx. VI. Ci. 1 § 84) ERADESA. 1077 Le, for the purpose of being bought. As meat iit, mem west Why do we say ‘when the sense is that of saleable’? Observe deg it area 1 fet mI, “we want to purchase corn, but it is not put out for sale”. arora TRA QT Ml stem - Tae, wT, UTA, ge Aaatatat: eater ot cataer af wey To eeTe free suite Pravere at re, w eCaT By TTS TATA, 83. ‘The forms weq and sqzar are found in the Chhandas. . The word wea is derived from A+aq, and srear from +at+aq The guna gis changed to sta Thus wea Pratetay u reeratt wear The ay is added to #t with the force of Ablative by virtue of the diversity allowed by ware wes (111. 3.113) «Thus fete ereng=weam “frightening or fear- able”. The word wqcar is always used in the feminine: in other places sta is the proper form. Why do we say ‘in the Vedas’? Observe am, Prag in secular literature. Virt -The word gyeat should also be enumerated when referring to water. As gt aa= areata: The affix a is added by 1V. 4. 110 (e+ a= grata) . com: qdaeat: Wee haan Now, qe - Tear, gfe: erranrdtent | wererrerceaie MTOR THE ATTEN TACT TTT CATT eat CHIE aT TTETT 84. From here upto VI. 1. 111 inclusive is al- ways to be supplied the phrase “for the preceding and the following one is substituted”. T an adhikara stra. In every sOtra upto VI. 1. 112 (excluding the last), whatever we shall teach, there in the room of the two, namely, the preceding and the succeeding, it should be understood, that the substitution is one, These form the well known ru'es of ekddesa, one letter or form replac- ing two consecutive letters &c. Thus VI. 1. 87 teaches ‘There is guna substitution, when 3 or 31 is followed by a vowel’, We must supply into that sdtra the phrase qart g4veat: i. €. one guna is the substitute for the final st or er and the initial vowel. Thus ett =a u Here q is the single subs- titute of the both preceding letter sit and the succeeding letter ¢u The words gi 9 show that the substitute operates simultancously on both. Otherwise the substitute would have come in the place of one only or of each one separately. ‘Thos in sre (aif) =“after 8t or mr, there is guna, in a vowel”. Here wiry is in the ablative, and by I. 1. 67 the guna operation would have taken place on the letter following it: so also 1% is in the Locative and by I. 1. 66, the 6