RV-Prātiśākhya

Ṛgveda-Prātiśākhya Chapter 2 (Saṁhitā-paṭala) (2.1-82)

Introduction

------------------------------------------------------------------------------------

Pada-pāṭha (word-text) --> Ṛgveda-Prātiśākhya --> Ṛgveda-Saṁhitā

------------------------------------------------------------------------------------

The main topic of Ṛgveda-Prātiśākhya is to give all the Sandhi rules, general and specific, to convert the uncombined word-text (Pada-pātha) into the combined text of Ṛgveda-Saṁhitā (Saṁhitā-pāṭha).

For example, the Pada-pātha of RV 10.90.2ab is:

puruṣaḥ | eva | idam | sarvam | yat | bhūtam | yat | ca | bhavyam |

The corresponding Saṁhitā-pātha is:

Puruṣa evedaṁ sarvaṁ yadbhūtaṁ yacca bhavyam

Puruṣa indeed is all this, what was and what will be.

The relationship between the Saṁhitā and the separate words (pada) of Ṛgveda is expressed in the first sūtra of chapter 2 of Ṛgveda-Prātiśākhya:

2.1 Saṁhitā pada-prakṛtih

Samhitā has the words (pada) as its source (prakṛti).

This is the viewpoint of Ṛgveda-Prātiśākhya: transforming the separate words, which represent pieces of unity, into the wholeness of Samhitā. Why do we say pieces of unity? According to the Pāṇini-sūtras of Vyākaraṇa, each word is the united form of basic constituents: roots (dhātu), underived stems, and affixes (pratyaya). For example, the word bhavati (‘he becomes’) is derived from the root bhū (‘to be’ or ‘to become’) and the two affixes ‘a’ and ‘ti’:

bhū + a + ti => bho + a + ti => bhav + a + ti => bhavati

Ṛgveda-Prātiśākhya is not concerned with the grammatical structure of the words. Only the phonetic form (the sound) is important for applying the Sandhi rules.

We can compare Ṛgveda-Prātiśākhya with a computer program, which has the Pada-pātha (the separate words) as the input and the Saṁhitā-pātha as the output.

Why is the combined text called Saṁhitā? Ṛgveda-Prātiśākhya 2.2 gives the answer:

2.2 (It is called Samhitā) because, while combining the finals to the initials of words, it proceeds without any interval of time (pause).

Note: Saṁ-hitā is derived from the root saṁ-dhā (to put together, unite, combine).

In the recitation of the Pada-pāṭha there is a pause after each word. The pause is indicated by the vertical line ‘|’. For example ag-nim | ī- ḻe |. After the 2 syllables ag and nim there is a pause. Then follow the 2 syllables ī and ḻe and again a pause. In the Saṁhitā-pātha we have the syllables: ag-ni-mī-ḻe. There is no pause between agnim and īḻe. The recitation of the Pada-pāṭha has a stop-and-go quality, while the recitation of the Samhitā-pātha is a continuous, sequential flow of syllables. Of course, there are also pauses in the recitation of the Saṁhitā as required by the metre. For example, in the Gāyatrī metre there is pause after the first two pādas and again a pause after the third pāda, which is the end of the mantra.

Sandhi rules

The Sandhi rules explain how the final letter of a word (pada) and the initial letter of the next word become modified in the Ṛgveda-Saṁhitā. The final letter and the following initial letter can be either a vowel (svara) or a consonant (vyañjana) resulting in four classes of combinations:

1. svara-svara 2. svara-vyañjana 3. vyañjana-svara 4. vyañjana-vyañjana

Examples: 1. ātmā + eva = ātmaiva 2. brahmā + bhavati = brahmā bhavati 3. yat + aṅga = yadaṅga 4. yat + ca = yacca

Chapter 2 of Ṛgveda-Prātiśākhya gives the rules for svara-svara and vyañjana-svara combinations.

2.10 In the so-called pratiloma anvakṣara sandhis (i.e. consonants precede and vowels follow), the first class consonants become third class consonants.

k + vowel = g + vowel

ṭ + vowel = ḍ + vowel

t + vowel = t + vowel

p+ vowel = b + vowel

Examples:

bhiṣak | upala-prakṣiṇī > bhiṣagupalaprakṣiṇī RV 9.112.3

tri-pāt | ūrdhvaḥ | ut | ait > tripādūrdhva udait RV 10.90.4

vi-rāṭ | ajāyata > virāḍajāyata > virāḻajāyata (for ḍ > ḻ see 1.52) RV 10.90.5

triṣṭup gāyatrī > triṣṭubgāyatrī RV 10.14.16

2.11 The Visarjanīya of eṣaḥ, syaḥ and saḥ is dropped before a following consonant.

eṣaḥ + consonant = eṣa + consonant

syaḥ + consonant = sya + consonant

saḥ + consonant = sa + consonant

Examples:

eṣaḥ | kaviḥ | abhi-stutaḥ > eṣa kavirabhistutaḥ RV 9.27.1

syaḥ | vṛṣā > sya vṛṣā RV 9.38.1

saḥ | naḥ | soma > sa naḥ soma RV 9.20.3

2.13 puraetā, titaünā, praügam and namaüktibhiḥ are the (cases of ) hiatus in the interior of words.

Note: 1.61: In the absence of a statement to the contrary (one should treat) the parts of a compound as words.

Examples:

puraetā-iva > puraeteva RV 6.47.7

titaünā > titaünā RV 10.71.2

hiraṇya-praügam > hiraṇya-praügam RV 1.35.5

kim | āsīt | praügam > kimāsītpraügam RV 10.130.3

namaükti-bhiḥ > namaüktibhiḥ RV 8.4.6

2.14 Cases of hiatus, other than these, (occur) in the (external) combinations of words.

Examples:

śruta-ṛṣim > śrutaṛṣim 10.47.3

ete | somāḥ | abhi > ete somā abhi RV 9.87.5

2.15 samānākṣare sasthāne dīrgham ekam ubhe svaram

Two simple vowels (samānākṣare) which have the same place of articulation (sasthāne) become both (ubhe) one (ekam) long (dīrgham) vowel (svaram).

a + a = ā; a + ā = ā; ā + a = ā; ā + ā = ā

i + i = ī; i + ī = ī; ī + i = ī; ī + ī = ī

u + u = ū; u + ū = ū; ū + u = ū; ū + ū = ū

Examples:

na + aparaḥ = nāparaḥ (jīvo brahmaiva nāparaḥ)

brahma + asmi = brahmāsmi (ahaṁ brahmāsmi)

yogi + indram = yogīndram

madhu + udakam = madhūdakam (RV 9.67.32)

su + uktam = sūktam

2.16 a (i.e., a and ā) when followed by i (i.e., i and ī) is, together with the following (vowel), changed to e

a + i = e; a + ī = e; ā + i = e; ā + ī = e

Examples: mahā + īśvara = maheśvara; eva + idam = evedam

2.17 In the same way a (i.e., a and ā) when followed by u (i.e., u and ū) is changed to o.

a + u = o; a + ū = o; ā + u = o; ā + ū = o

Example: puruṣa + uttama = puruṣottama

2.18 If the odd ones (i.e., e and ai) among the next (vowels, i.e., the diphthongs) follow, (a i.e., a and ā, together with the following vowel) become ai.

Note: The above sūtra refers to the vowels following u and ū, which are e, o, ai and au. The first vowel after ū is e and the third vowel after ū is ai (‘the odd ones’).

a + e = ai; ā + e = ai; a + ai = ai; ā + ai = ai

Examples: ātmā + eva = ātmaiva (ātmaivedaṁ sarvam); brahma + eva = brahmaiva (jīvo brahmaiva nāparah); vasudhā + eva = vasudhaiva (vasudhaiva kuṭumbakam); parā + ait = parait (RV 10.61.8)

2.19 (a, i.e., a and ā, together with the following vowel) become au when the even (diphthongs, i.e., o and au) follow.

a + o = au; ā + o = au; a + au = au; ā + au = au

Examples: yatra + oṣadhīḥ = yatrauṣadhīḥ (RV 10.97.3); pra + aukṣan = praukṣan (RV 10.90.7)

2.20 These combinations (Sūtras 15-19) are named Praśliṣṭa (‘contracted”)

In chapter 2 (Sūtras 21-33) Ṛgveda-Prātiśākhya gives the rules for the following Sandhis: Kṣaipra, Padavṛtti, Udgrāha, Udgrāha-Padavṛtti, Bhugna, Udgrāhavat, Pancāla-Padavṛtti and Prācya-Padavṛtti.

In all these Sandhis the first letter (i.e., the final letter of the first word) is either a vowel or Visarjanīya (aḥ, āḥ). The second letter (i.e., the initial letter of the second word) is always a vowel, which remains unchanged after applying the Sandhi-rule.

2.21 A non-guttural simple vowel (samānākṣara) is changed into its own semi-vowel, when followed by a vowel.

i,ī + vowel = y + vowel; u,ū + vowel = v + vowel

2.22 (But) not (so) whenever the same simple vowel (as the preceding one) follows.

i,i + i,ī = ī; u,ū + u,ū = ū

2.23 These combinations in which the following (vowel) remains unchanged are called Kṣaipra (‘hastened’).

Examples:

vi + oman = vyoman (ṛco akṣare parame vyoman - RV 1.164.39)

carāmi + aham = carāmyaham (ahaṁ rudrebhir vasubhiś carāmyaham - RV 10.125.1)

khalu + idam = khalvidam (sarvaṁ khalvidaṁ brahma)

2.24 An unrhotacized Visarjanīya, if preceded by a long vowel and followed by a vowel, becomes ā.

āḥ + vowel = ā + vowel

Note 1: The definition of ‘rhotacized’ (rephin) is given in Sūtras 1.76-103:

1.76 The fifth Ūṣman (Visarjanīya), when preceded by a Nāmin-vowel (ṛ, ṝ, i, ī, u, ū, e, o, ai, au) is called rhotacized.

Sūtras 1.77-103 give special words ending in aḥ or āḥ, which are called rhotacized. For example: svaḥ, antaḥ, punaḥ, vāḥ.

Such a rhotacized Visarjanīya becomes r (repha) when followed by a vowel or a voiced consonant (Sūtra 4.27).

Examples: agniḥ + asmi = agnirasmi; pūrvīḥ + aham = pūrvīraham;

punaḥ +āgāḥ = punarāgāh

An unrhotacized Visarjanīya, as mentioned in Sūtra 2.24 above, can only be preceded by a or ā.

Note 2: Sūtra 2.24 above says that the unrhotacized Visarjanīya should be replaced by ā. For the correct interpretation of Sūtra 2.24 the following Paribhāṣa rule is important:

1.37 A Visarjanīya, which is not rhotacized and is followed by a vowel or a voiced consonant, (should be treated) together with its penultimate, as one letter.

According to this rule, replacing the Visarjanīya by ā means replacing āḥ by ā.

2.25 Also the last two vowels (i.e., ai and au become ā, when followed by a vowel).

ai + vowel = ā + vowel; au + vowel = ā + vowel

2.26 These (Sandhis, described in Sūtras 24-25, are called) Padavṛtti (‘hiatus between two words’).

Examples:

devāḥ + adhi = devā adhi (RV 1.164.39)

kau + ūrū + pādau + ucyate = kā ūrū pādā ucyate (RV 10.90.11)

vai + aham = vā aham

2.27 But that (unrhotacized Visarjanīya), if preceded by a short vowel (and followed by a vowel) becomes a.

aḥ + vowel = a + vowel

2.28 So the last two vowels which precede the next to the last (vowel become a, when followed by a vowel).

e + vowel = a + vowel; o + vowel = a + vowel

2.29 These (Sandhis, described in Sūtras 27-28, are called) Udgrāha.

2.30 If followed by a long vowel (the same Sandhis are called) Udgrāha-Padavṛtti.

Examples:

puruṣaḥ + eva = puruṣa eva (puruṣa evedam sarvam - RV 10.90.2)

agne + ā + yāhi = agna ā yāhi (RV 6.16.10)

ye + it = ya it; te + ime = ta ime (RV 1.164.39)

vāyo + ukthebhiḥ = vāya ukthebhiḥ (RV 1.2.1)

2.31 When the two vowels (i.e., a and ā), which arise from labials (i.e., o and au) are followed by a non-labial (vowel), there is an insertion of v (which is called) Bhugna.

Note: Labial vowels are u, ū, o, au

o + non-labial vowel = av + non-labial vowel

au + non-labial vowel = āv + non-labial vowel

Examples:

vāyo + ā + yāhi = vāyavā yāhi (RV 1.2.1)

tau + indrāgnī = tāvindrāgnī

mitrāvarunau + ṛtāvṛdhau + ṛtaspṛśā = mitrāvarunāvṛtāvṛdhāvṛtaspṛśā (RV 1.2.8)

2.32 The two guttural (vowels) become a when followed by (the vowel) ṛ. This (Sandhi is called) Udgrāhavat.

a,ā + ṛ = a + ṛ

Examples:

mahā + ṛṣiḥ = mahaṛṣih

viśva-śambhuvā + ṛtavarī = viśvaśambhuva ṛtavarī (RV 1.160.1)

pra + ṛbhubhyaḥ = pra ṛbhubhyaḥ (RV 4.33.1)

2.33 When forming the first element (i.e., the finals of the preceding words) of the Udgrāha Sandhis and when followed by a, the two (letters, i.e., e and o) remain unchanged, (while) the former (i.e. ah) becomes o. These Sandhis are (called) Prācya-Padavṛtti and Pancāla-Padavṛtti; those (Sandhis) in which the preceding (vowel) is a labial belong to the Pancālas.

Pancāla-Padavṛtti: aḥ + a = o + a; o + a = o + a

Prācya-Padavṛtti: e + a = e + a

Examples:

ṛcaḥ + akṣare= ṛco akṣare (RV 1.164.39)

yo + asmai = yo asmai (RV 8.31.2)

te + agre-pāḥ = te agrepāḥ (RV 4.34.10)

In chapter 2 (Sūtras 34-50) Ṛgveda-Prātiśākhya gives the rules for the Abhinihita Sandhi:

2.34 Now begins the Sandhi (called) Abhinihita (‘absorbed’).

A in the beginning of a Pāda, becomes one with these primary and secondary (vowels, i.e., e and o; they (i.e., e and o, remaining) here as the result of the Sandhi.

aḥ + a = o; o + a = o; e + a = e

Note: In the previous Sūtra 2.33 the rules for the Sandhis called Panchāla-Padavṛtti and Prāchya-Padavṛtti have been given:

aḥ + a = o + a; o + a = o + a; e + a = e + a

In the Abhinihita Sandhi the second vowel, a, is dropped. According to Sūtra 2.34 the Abhinihita Sandhi will apply only when the a is in the beginning of a Pāda. What is a Pāda? A Pāda is a section of a ṛchā. For example, in the Gāyatrī metre a ṛchā consists of 3 Pādas, each containing 8 syllables (Akṣara), and in the Triṣṭubh metre there are 4 Pādas with 11 syllables each. After the 2nd and 3rd Pādas of Gāyatrī there is a pause and therefore Sandhi rules do not apply. In Triṣṭubh the pause is after the 2nd and 4th Pādas. This means that in Gāyatrī only the 1st and 2nd Pādas are connected through Sandhi-rules. In Triṣṭubh the Sandhi-rules have to be applied between the 1st and the 2nd Pādas and between the 3rd and the 4th Pādas. Since the Abhinihita Sandhi only applies to the beginning of a Pāda, it would apply to the beginning of the 2^nd Pāda in the case of Gāyatrī and the beginning of the 2^nd or 4^th Pādas in the case of Triṣṭubh.

Example:

Word-text of Pāda 1 and Pāda 2 of RV 1.1.9:

The last word of Pāda 1 is sūnave and the first word of Pāda 2 is agne. Because a, the first letter of agne, is at the beginning of Pāda 2, the Abhinihita-Sandhi has to be applied:

sūnave + agne = sūnave ‘gne

In Devanāgarī the dropped a is marked by the Avagraha sign represented here as ‘.

If a is not in the beginning of a Pāda, the general rule is that the a remains: aḥ + a = o + a; o + a = o + a; e + a = e + a. But there are many exceptions which are given in the following Sūtras 2.35-50.

2.35 (Also) in the interior of a Pāda (a becomes one with the preceding e or o), if, in the Saṁhitā text, a light (laghu) syllable beginning with y or even v follows the light vowel a.

Note: A syllable is light, if a short vowel is not followed by two or more consonants.

Examples: āpaḥ + ayanam = āpo ‘yanam (RV 4.33.7); juhūre + avase = juhūre ‘vase (RV 1.48.14)

Sūtras 2.42-48 give certain pairs of words, in which the a at the beginning of the second word is absorbed.

2.47 So also in the following combinations: svādhyo ‘janaya… pādo ‘sya, yo ‘ti, brāhmano ‘sya, yo ‘nayan.

Examples: pādaḥ + asya = pādo ‘sya (RV 10.90.3); brāhmanaḥ + asya = brāhmano ‘sya (RV 10.90.12)

Pragṛhya vowels remain unchanged (i.e. not subject to the usual Sandhi rules) before a following vowel (Sūtras 2.51-67)

The definition of Pragṛhya (to be taken or pronounced separately) is given in the 1^st chapter (Sūtras 68-75). O at the end of a word is Pragṛhya (1.68-70).

Examples: indo, o, eṣo, uto, atho. But this does not apply to the o that occurs at the end of the first part of a compound (go-iṣṭaye).

The three long vowels ī, ū and e, when standing at the end of a dual, are Pragṛhya (1.71). Examples: indrāgnī (Indra and Agni), indravāyū (Indra and Vāyu), dve (they two), dadhāthe (they two put). The two long vowels ī and ū at the end of a locative (saptamī) (1.72) and the words asme, yuṣme, tve and amī (1.73) are Pragṛhya.

In the word-text all words ending in a Pragṛhya vowel are followed by the particle iti (atho iti, indrāgnī iti, etc.) The particle u, which is also Pragṛhya according to the teaching of Śākalya, is lengthened and nasalized before the following iti (ūm̐ iti) (1.75). In Devanāgarī the symbol for nasalization is the half-moon with the Bindu (dot) inside placed above the ū.

2.51 The so-called Pragṛhya vowels remain unchanged, when occurring before the word iti.

2.52 Also when vowels (svara) follow in the Saṁhitā text.

Examples:

eṣo iti | uṣāḥ | apūrvyā --> eṣo uṣā apūrvyā (RV 1.46.1)

indravāyū iti | ime --> indravāyū ime (RV 1.2.4)

punarvasū iti punaḥ-vasū | asme iti | dhārayatam --> punarvasū asme dhārayatam (RV 10.19.1)

nṛpatī iti nṛ-patī | asti --> nṛpatī asti (RV 7.71.4)

dadhāte iti | oṣadhīṣu --> dadhāte oṣadhīṣu (RV 7.61.3)

2.53 The first (Pragṛhya, i.e., o in the vocative, is, however, treated in the Saṁhitā text as is said (above: 2.28, 31, 33, 34, etc.)

Examples:

vāyo iti | ukthebhiḥ --> vāya ukthebhiḥ (RV 1.2.1)

vāyo iti | ā | yāhi --> vāyavā yāhi (RV 1.2.1)

Note: When in the word-text a word ending in a vowel is followed by iti, then we know that the final vowel will be unchanged before a following vowel in the Samhitā-text.

This is the general rule. In this Sūtra an exception is given. What is the reason to have vāyo followed by iti? This question has to be directed to Śākalya who is said to be the seer of the word-text (Pada-pātha) of Ṛgveda.

The words u, su, pūṣā and twelve certain other words beginning with śraddhā remain unchanged before a following vowel, in specified cases only (Ṛgveda-Prātiśākhya 2.56-59).

2.56 U, when preceded by ‘y’ which is the result of a Sandhi, and also when following a hiatus (vivṛtti) (remains unchanged), only in the Saṁhitā text.

y + u + vowel = yu + vowel

vowel + u + vowel = vowel + u + vowel

Examples:

prati | ūm̐ iti | adarśi --> pratyu adarśi (RV 7.81.1)

vi | ūm̐ iti | āyuḥ --> vyu āyuḥ (RV 10.27.7)

tasmai | ūm̐ iti | adya --> tasmā u adya (8.66.7)

dhana-sāḥ | ūm̐ iti | īmahe --> dhanasā u īmahe (RV 10.65.10)

Note: The combination i,ī + u + vowel --> y + u + vowel appears only twice in the whole Ṛgveda Saṁhitā. The Śākala Ṛgveda Saṁhitā consists of 1,017 Sūktas, 10,472 ṛcās or Mantras and 381,199 Akṣaras (syllables or vowels). This rule in Sūtra 2.56 is the demonstration of the point value of Natural Law.

2.59 The words śraddhā, samrājñī, suśamī, svadhā, ūtī, pṛthujrayī, pṛthivī, īṣā, manīṣā, ayā, nidrā, jyā and prapā (remain unchanged) if followed by the vowels a, i or ī.

Examples:

svadhā | avastāt --> svadhā avastāt (RV 10.129.5)

manīṣā | iyam --> manīṣā iyam (RV 7.71.6)

ni-drā | īṣata --> nidrā īṣata (RV 8.48.14)

The words specified in the following Sūtras 2.60-66 will remain unchanged before (certain) vowels in the Saṁhitā-text. According to Sūtra 2.67 the final unchanged vowel of these words, which all end in a or ā, will become nasalized. The symbol for nasalization (Anunāsika) is the half-moon with the Bindu (dot) placed inside. It is pronounced like ng in sung.

2.60 The word sacā (remains unchanged) before a following vowel, which is at the beginning of a Pāda.

Note: The metre of RV 1.51.11 is Jagatī, which has 4 Pādas with 12 syllables each. The first Pāda begins with mandiṣṭa and ends with sacā. The second Pāda begins with indraḥ.

2.61 The word ā (remains unchanged before a following vowel, which is at the beginning of a Pāda) if preceded by a word ending in ṣu, the words joṣam, carṣaṇī and carṣaṇibhyaḥ, a word ending in e , the words mitrayoḥ, asmat, īvat and namasyuḥ.

Note: The metre of RV 9.12.5 is Gāyatrī, which has 3 Pādas with 8 syllables each. The first Pāda begins with yaḥ and ends with ā. The second Pāda begins with antaḥ. The word antaḥ is written in the word-text as antariti, in order to show that the Visarga at the end of antaḥ is rhotacized.

2.62 The two vowels a and ā followed by the vowels e and o (at beginning of a Pāda) (remain unchanged) before the hymns of (Ṛṣi) Luśa (RV 10.35-36).

Examples:

Note: The metre of RV 1.35.6 is Triṣṭubh, which has 4 Pādas with 11 syllables each. The second Pāda begins with ekā. The metre of RV 10.34.5 is Triṣṭubh. The second Pāda begins with emi.

Sūtras 2.68-77 gives all the irregular combinations (Sandhis) which should be taken as exceptions to any previous Sandhi rules. Shown below is a sample of these combinations:

2.68 sedu, sāsmin, semabhi sābhi vegaḥ, sedṛbhavaḥ, sopamā, sauṣadhīr anu,

sāsmā aram, sota naḥ, sendra viśvā, seti, sāsmakamanavadya, sāsi

Note: In all these combinations listed above the first word is saḥ.

Examples:

saḥ + it + u --> sedu (RV 1.32.15,..., 10.2.3)

Why is sedu irregular? According to the regular Sandhis we expect --> sa idu. Instead of the hiatus a further contraction takes place. Sedu appears ten times in Ṛgveda Saṁhitā.

saḥ + asmin --> sāsmin (RV 10.44.5) (instead of so asmin)

saḥ + īm + abhi --> semabhi (RV 9.74.7) (instead of sa īmabhi)

saḥ + abhi + vegaḥ --> sābhi vegaḥ (RV 10.27.1) (instead of so abhi vegah)

saḥ + it + ṛbhavaḥ --> sedṛbhavaḥ (RV 4.37.6) (instead of sa idṛbhavah)

saḥ + upamā --> sopamā (RV 1.31.15) (instead of sa upamā)

saḥ + auṣadhīḥ + anu --> sauṣadhīr anu (RV 1.103.5) (instead of sa auṣadhīr anu)

saḥ + asmai + aram --> sāsmā aram (RV 1.27.6) (instead of so asmā aram)

saḥ + uta + naḥ --> sota naḥ (RV 2.24.1) (instead of sa uta nah)

saḥ + indra + viśvā --> sendra viśvā (RV 2.23.11) (instead of sa indra viśvā)

saḥ + iti --> seti (RV 2.12.5) (instead of sa iti)

saḥ + asmākam + anavadya --> sāsmākamanavadya (RV 1.129.1) (instead of so asmākamanavadya)

saḥ + asi --> sāsi (RV 2.13.2) (instead of so asi)

2.70 nū itthā te, ... , ūtyarvāk

Examples:

nu + itthā+ te --> nū itthā te (RV 1.132.4) (instead of nvitthā te)

ūtī + arvāk --> ūtyarvāk (RV 10.15.4) (instead of ūtī arvāk; according to 2.59 ūtī remains unchanged before the vowels a, i, or ī.

2.71 udū ayān, ... , rathoḻha

Examples:

ut + u + ayān --> udū ayān (RV 6.71.5) (instead of udvayān)

ratha + oḻha --> rathoḻha (RV 10.148.3) (instead of rathauḻha)

2.73 asvineva, ... , svadhāmitā

Examples:

aśvinā + eva --> aśvineva (RV 8.9.9) (instead of aśvinām̐ eva; according to 2.62 a word ending in ā remains unchanged before e at the beginning of a Pāda appearing before 10.35.)

svadhā + amitā --> svadhāmitā (RV 5.34.1) (instead of svadhā amitā; according to 2.59 svadhā remains unchanged before the vowels a, i, or ī.)

2.78 The euphonic combinations in the three cases:

śunaścicchepam niditam (RV 5.2.7), narā vā śamsam pūṣanam (RV 10.6.43), and narā ca śamsam daivyam (RV 9.86.42) are not according to the succession (of words in the Pada text).

Pada text:

(5.2.7): śunah-śepam | cit | ni-ditam

(10.6.43): narāśamsam | vā | pūṣanam

(9.86.42): narāśamsam | ca | daivyam

Notes

a ā i ï ī ī3 u ü ū ṛ ṝ ḷ e ai o au ṁ ḥ m̐ ' |

k kh g gh ṅ c ch j jh ñ ṭ ṭh ḍ ḍh ṇ t th d dh n p ph b bh m

y r l v ś ṣ s h ḻ ḻh