________________
242
a, became only common in Apa. For another Apa. word already occurring in Pkt, see aha discussed in *23. In fact, the occasional occurrence of the variant tae or tue (i.e. tаě, tuě) for tai would betray attempts to remove this Apa. form from the text.
For an explanation of the origin of the variant aħdharahia found in K, see above, p. 165.
For the particle o, see Var. IX 4 (o sūcanāpaścăttāpavikalpeşu) and Hem. II 203 (o sūcanāpaścăttāpe).
of the verb udvas- Skt has only the causative, i.e. udvāsay-, meaning 'to remove, to destroy' (cf. uvvāsia 'removed' in Gatha 106). The participle uvvasia, from uvvasa-, seems to me an 'deserted, fallen in disuse' (from there possibly also 'dilapidated'; cf. uvvasia in 233, quoted in the note on *7); compare (late) Skt udvasa 'uninhabited, empty' (RajataranginT), 'solitude' (satruhjaya-Māhātmya; MW).
For the reasons to adopt hara found in k instead of ghara found in the other MSS, see above, p. 164.
For devaeva (Tp, ) beside devaa va of the other MSS, see *23.
nomalia corresponds according to Hem. I 170 to Skt navamalika 'Jasminum sambac'. Var. I 7 has nomallia, i.e. Skt navamallika 'id'. The variant omālian (Bh, K, S) undoubtedly is a corruption for nomāliam, which may be due to preceding o suhaa. As becomes clear from the quotation of part of the last pada of this Gātha in I 38 Hemacandra used a MS of the Sattas at which contained this corruption. His subsequent dealings with the word omāliañ are illustrative of his method. The context suggests that it denotes flowers offered to a deity, which, moreover, had lost their fragrance. In II 92 Hemacandra ad hoc produced a Skt word avamālya, which would be synonymous with 'nirmālya 'the remains of an offering (of flowers) to a deity' (1 38).
, P, 31 G, E, T; 653 s)
33( 70 Ma, 71 Tp; 31 Bh, 33 R; 31 K, B, Y, 32 Missing in T and W.
pahar avanamaggavisame gāmaniuttassa ure
jāā icchena lahai se niddah pallt uņa se suhah suai.
papahara Ma -- viņa (vana) R
-- maga P; magva S -- kicchena Ma;
