Page:Vedic Grammar.djvu/248

238 I. ALLGEMEINES UND SPRACHE. 4. VEDIC GRAMMAR. dýśas, yakṣa-dýśas ‘having the appearance of a Yakṣa', su-dŕśas, svar-dŕšas; tveşá-samdrsas, su-samdŕšas; viśva-píšas, su-píšas well adorned'; sú-sadṛśas 'handsome'. - f. díśas, visas; a-disas, ud-disas (VS. VI. 19) 'upper quarters', upa-sprsas, pra-diśas, vi-díśas (VS. vI. 19) “intermediate quarters', sam-djśas. A. m. spáśas; ahar-dŕšas ‘beholding the day', bhīmá-samdrsas ‘of terrible appearance', svar-disas, híranya-samdrsas. f. disas, visas, vríšas; ā-díśas, pra-disas, sam-dýśas. I. m. su-samdygbhis. f. padbhís (IV. 2¹2) 'with looks', vidbhís. D. f. digbhyás (VS. VI. 19). - Ab. f. digbhyás, vidbhyás. G. f. diśám, višám; ä-diśām. - L. f. dikṣú (AV. VS.), viksú. - 1 6. Radical stems in -h. 351. This declension comprises some 80 stems formed from about a dozen roots. All three genders appear in its inflexion; but the neuter is rare, being found in only two stems and never in the plural. Of monosyllabic stems six or seven are f., one m., and one n. All the remaining stems are compounds, about three-fourths of which are formed from the three roots druh-, vah- and sah- (over 30 from the last). The origin of the two stems usníh (AV.) a metre, and saráh-2 'bee' is obscure. a. The distinction of strong and weak appears in compounds of vah- and sah-³, the vowel being lengthened in the N. A. sing. and N. pl. m.; also in the N. A. du. m. forms indra-váha, indra-váhau, anad-váhau; and in the f. sing. N. dakṣina-vat and A. havya-váham. The strong stem -vah- twice appears in weak cases, while it is metrically shortened 18 times in strong cases 4. The word anad-váh- 'ox' (lit. 'cart-drawer') distinguishes three stems, the strong one being anaḍ-váh-, and the weak anaḍ-úh- before vowels and anad-út- 5 before consonants. b. As h represents both the old guttural aspirate gh and the old palatal jh, it should phonetically become g and d respectively before bh. But the cerebral appears for both in the only two case-forms that occur with a -bh ending: sarádbhyas from saráh-, and anad-udbhyas (AV.) from añad-váh-, where the dental d takes the place of the cerebral by dissimilation. Before the su of the L. pl., k would be phonetic; but here again, in the only form occurring, the cerebral appears: anadútsu6. On the other hand the phonetic k appears in the N. sing. in the six forms-dhak, -dhúk, -dhrúk, rúk, -spŕk, uşník (AV.)8; while the unphonetic appears in the two forms -vát and -sát. The word anad-váh- forms, instead of *anad-vát, the anomalous anadván as if from a stem in -vant. C. The stem mahá- is perhaps a transfer to the a-declension from the far more frequent but defective máh- 'great'. Several cases are formed from it: sing. N. mahá-s, G. mahásya, L. mahé; pl. N. mahá and maháni, n., G. mahánām. The D. sing. maháye, used as an infinitive, is a transfer to the i-declension from máh-. Inflexion. 352. The forms actually occurring, if made from sáh- 'victorious', would be as follows: ¹ BLOOMFIELD is of opinion that here, as 7 Occurring respectively in usá-dhak 'burn- well as in the 5 other passages in which ing with eagerness', in three compounds of this form occurs in the RV., it means 'with duh- ‘milk', and in five compounds of druh- feet': Johns Hopkins University Circular, 1906, 'injure'. These three forms, together with | P. 15-19. usar-bhút, are the only examples of the resto- ration of initial aspiration in the declension of the RV. 2 That the h here represents an original guttural is shown by the N. pl. sarághas (SB.) and the derivatives sarágha- and sáragha- (TB.). 3 The Pada text has always vāh- on the one hand, but săh- on the other. 4 Cp. LANMAN 498 (middle). 5 For anad-ud- by dissimilation. 6 The dental again by dissimilation for the cerebral t. 8 The derivation of this word (AV. VS.), is uncertain; it occurs in the RV. only in the extended form of uṣṇihā-. 9 When the final h becomes t, the initial sis cerebralized.