t D I. PHONOLOGY. PALATALS. 29 rare at the end of the root, appearing only in sak- ‘be able'; 2. sing. sak-as; sagh- 'be equal to': 3. sing. sagh-at; dagh- 'reach' : dagh-at (TS.); in these roots the guttural prevails throughout owing to the influence of the present stem sak-nu-, sagh-nu-2. Even in the non-thematic presents and in the perfect the palatal carried the day, though phonetic in still fewer forms; thus the guttural alone would be historically justified in the forms yunájā, yuñje; yuyója³. The palatal further regularly appears before the causative 4 suffix -áya-, where it is phonetic (= IE. éie); e. g. arc-áya-ti from arc- 'praise'5. b. As shown by the appearance, in cognate forms, of a guttural before other consonants than s, the final of the following verbs is a new palatal: anj-'anoint'; ej- 'stir'; tij- 'sharpen'; tuj- 'beat'; tyaj- 'forsake'; nij- ‘wash'; bhaj- 'divide'; bhanj- 'break'; bhuj- 'bend'; yuj- 'yoke'; ranj- 'colour'; ruj- 'break'; vij- 'shoot up'; vrj- 'turn'; sinj- 'sound'; sanj-‘attach'; svañj- 'embrace'; also in the noun sráj- 'garland'. c. Apart from being the result of the Sandhi of d+j, jj is shown by the evidence of cognate languages to be derived from a sibilant + guttural (= IE. zg) and thus to belong to the series of new palatals in majján- ‘marrow'; rájju- 'rope'; bhrjjáti 'roasts'; majjati 'dives', from which is derived madgú- (VS.) 'diver' (a bird). d. Nominal derivatives. 1. Before the suffix -a, the final of the root is mostly guttural, because the a in nearly all the cases of the noun represents IE. . The rule in the RV. is that the guttural appears before both unaccented -a and accented -á, but the palatal before accented - only6; e. g. abhi- droh-á-, druh-a- ‘injury' : drógh-a- ‘injuring'; bhoj-á- ‘liberaľ : bhóg-a- 'enjoyment'; a-yuj-á- 'companionless'; yóg-a- ‘yoking'; ruj-á- ‘breaking': róg-a- (AV.) ‘disease'; vevij-á- 'swift' : vég-a- (AV.) ‘speed'; suc-á- 'bright': sók-a- 'flame'; ruc-á- (VS.) and roc-á- (AV.) ‘shining' : rók-a- and rok-á- ‘light’7. 2. Before the suffix -as, the palatal generally appears, as it is for the most part phonetically required; e. g. ój-as- 'force': ug-rá- 'mighty'. The guttural, however, prevailed in ánk-as- 'bend'; ág-as- 'offence'; -ny-ogh-as- 'streaming'; bhárg-as- ‘brilliance'; as there were no corresponding verbs with palatal beside these words; it also prevailed in ók-as- 'ease' and ny-okas- 'comfortable', as well as -sok-as- 'flaming', though there are such verbs (uc- 'be pleased', and suc- 'shine'). 3. Before other suffixes beginning with a, the final of the root is generally palatal; thus before -ana (= IE. -eno-), vac-and- 'speaking'; tej-ana- 'act of sharpening'; mamh-ána- 'gift'8; before -ant, -ana (under the influence of Apart from roots ending in kh, see 35, note 3. 2 In ni-mégha-mana- 'drenching oneself', the gh seems to be phonetic (as -amāna -omena). In válgate (AV.) 'springs', the guttural is perhaps due to the preceding , as neither lj nor le is ever found to occur. 3 This normalization of the palatal is probably Indo-Iranian, see J. SCHMIDT, KZ. 25, 104. 4 The denominatives in -ayá- (IE. eié and oie) follow the noun from which they are derived; e. g. from aghá- 'evil', aghãyáti 'wishes to injure'. 7 A palatal before an unaccented -a first appears in dóh-a- (RV. x. 12²), other- wise dogha- 'milking'; mỗh-a-(AV.) 'delusion', beside mógh-a- 'vain'; krúñc-a- (VS.) 'cur- lew', is probably only an extension of krúnc- 5 The causative ingáyati of éjati 'stirs' is probably due to an old present base *inag-, |(VS.) ing- formed according to the 7th class; the 8 The gh of jaghána- 'buttock', is phonetic phonetic form -injayati is found in the BAU. (Gk. Koyávn). VI. 4, 23. 6 The fluctuation of words in -a- probably arose from some cases in oxytones having had IE. , others ; hence in some words the palatal prevailed throughout, in others the guttural. The agent-nouns, being mostly oxytone, show a preference for the palatal, which originally appeared in oxytoues only. Cp. WACKERNAGEL I, 128 a (p. 150, note, end). =
