________________
CHAPTER VII]
BOWER MANUSCRIPT
Ixvii
well-known prakritic diaeresis of a conjunct with an antecedent r are darisaya (for daríaya), V 1 203. pradarisitas (for pradariitati). V 54 207. varishe (for varshe). V 60 207. and kirishôrtti (for firshártti), VI 4 223; but the normal form varsha also occurs in V 4 204. 17 205, 40 206. Examples of the change of p to o we have in avi (for api), IV 11 193, upar adyaté (for upapadyatê), V 11 204,57 207. Once we have pi (for api) after an anusvâra, in sura i pi, VI 3 224; but the normal api also occurs in IV 3 192, V 9 204. Other miscellancous prakriticisms are chichcha (for chitya), VI 1 223; singhasya (for sisihasya), IV 1 192; dukkha (for dulkha), V 12 204, 21 205; satta.khutto (for sapta-kritvah), VII 6 236 ; also dvêtiyaka (for dvaitiyaka). VI 2 223, and selâya (for sailâya). VI 4 223. Mare to the category of prakriticisms in sandhi belong the following examples: apetu (for a paitu) in VI 2 203, and upaishyati (for upêshyati), IV 20 193. A final consonant is almost always elided; thus, karaná (for karanat), IV 3 192 6, 20 193 40, 43 195; tasma (for tasmat), IV 16 193; acharê (for acharêt), VI 16 225; kârayê (for kârayệt), V 48 207; avocha (for avôchat), VI. 1b 222; so also chira, V 38 206; and kichi, IV 35 194, 52 195 V 27 205 36 206; but the normal chirât occurs twice in IV 29 194 44 195, and the normal kisichit once in IV 20 193. In the noininative and accusative singular neuter of pronouns, the elision of t ord, alternates with the anusvara. Thus we have ta in V 28 205; êta, V 25 205 37 206 47 207; ya, V 1 2033 204; but also the equally prâkritic forms tai, V 24, 25 205; étai, V 4, 7, 14 204 28, 31 205 43 206 52 207, and yari, V 47, 60 207. On the other hand, the normal forms tad, étad, yad occur before vowels, e. g., tad-avapoyasi, IV 13 193; étad-uvácha, VI 1 222, yad-ipsasê, IV i 192; but exceptionally also before consonants, tad-yatha and yatsatyani, V 3 203. Occasionally the anusvâra is added to the end of a word, as in karayain (for karaya), V 6 204, dêsêm (for deil), V 58 207; or it may take the place of the final visarga, as in tatan (for tatal), IV 3 192.108 But more often such a visarga is dropped altogether; as in labha (for lábhal), IV 48 195, vritta, V 36 206, ananda, VI 5 222; priti (for priti!). V 21 205; prahu (for prahul), V 2 203, bhikshu, VI 2 222; sumitrai (for sumitrail), IV 30 194, daratai chórai nainsitikai, V 17 205; hêtô (for hêtôl), V 27 205. Or it combines to 6, as in itô (for ital) shashthé, V 13 204. Or, its omission may give rise to double sandhi, as in tatottamah (for tata-uttamah from tatah-uttamah), IV 10 193. Occasionally some consonant is inserted to avoid a hiatus, or a vocalic sandhi. Thus (a) n in nai-n-ritikail (for nair-ritikail), V 17 205, maitrî-1-airavanêshu (for maitry-airavanêshu), VI 1 224; (b) m in pari-m-apanaya (for paryapanaya), VI 4 223; Vasukind-m-api (for Vasukinäpi), VI 2 224; for other examples, see page lxvi C in jani-r-u padravail (for jany-u padravail), IV 3 192; for another example, see above, p. Ixvi 109; (d)s in gamanê-statha, V 21 205, janê-s- tatha, V 50 207; but in these two cases the reading is uncertain,
Examples of prakriticism in inflexion are the following: (1) With nouns: in the nomnative singular masculine, a final # may be dropped at the end of a verse, as in maha (for mahan), V 36 206, or before a consonant, as in bhagava (for bhagavân) chchhrávastyal, VI 1 222. But the normal form also occurs, as in bhôgavân śramanêraka), VI 6 224; and before vowels it is used always, as in mahân artha), IV 10, 11 193, bhagavan ayushmantan, VI 1 222. Similarly curtailed forms, however, occur also in the Vedas, see M. Ved. Gr., 315, p. 193. In the case of dhanava-s-cha, as indicated by the interpolated é, the omission of the anasvâra is probably a scribal error, and the reading should be dhanavam. In the accusative plural masculine, ami and ûm replace an and ûn, respectively, before consonants, as in kamani
108 In padans-tu (for padas-tu), V 5 204, the anusvåra is a mere clerical error. 109 In svati-r-bhikshwili, VI 5 222, the reading is doubtful.
