of various maladies; and, when one of these remains unhealed, boils and sores etc. (? piṭakavraṇādīni) show themselves. Also, that kalās are anupādeyāvayavopalakṣaṇa, and worthless parts of cattle etc. are collected in old pits. And in like manner collected ill-dreaming is made over to an enemy. That is his idea, and a wholly unacceptable one, of the general meaning of the verse. ⌊The verse is prose, no triṣṭubh; but may be stretched so as to count as 43 syllables.⌋
3. Embryo of the wives of the gods, instrument of Yama, excellent dream; the evil [dream] that is mine, that do we send forth to him that hates us.
The mss. all read devā́nām pátnīnāṁ gárbha (one pada-mss. -bham) yamásya kárayo bhadrásvapnaḥ; the translation implies no further emendation than to gárbho and káraṇo; ⌊the minor Pet. Lex., iv. 249, accepts bhadrásvapna as a descriptive compound, although the accent (Gram. §1280 c) is very exceptional;⌋ SPP., following the comm., changes to dévānām patnīnāṁ garbha yámasya kara yó bhadráḥ svapna. Our devápatnīnām and káraṇas were suggested especially by the devajāmīnā́m and káraṇas of vi. 46. 2 and xvi. 5. 6, of which neither the comm. nor SPP. take any notice. In the second division of the verse the two editions agree, save that ours emends tát of the mss. (which SPP. follows) to tám; and the latter is supported by our P.M., and by the comm. But the mss. have at the beginning samámayaḥ, and the pada-mss. resolve it into samám: ayaḥ. The Anukr. and comm. and SPP. add to this verse what in our text is the first division of vs. 4; our division is that of our first mss., and is preferable on the ground of the sense. ⌊The prose verse, according to the division of the Anukr., may be made to count (8 + 10: 13: 13?) as 44 syllables.⌋
4. Thee that art "harsh" by name, mouth of the black bird (-çakúni)—thee, O sleep, we thus know completely; do thou, O sleep, as a horse a halter, as a horse a girth, scatter him who is not of us, the god-reviler, the mocker.
⌊Prose.⌋ The translation here is of no authority, including various venturesome emendations of the text; it follows our text except at the end, where, instead of badhāna, it implies the (unsatisfactory) vapa of the comm. and SPP.; all the mss. read vápus ⌊or vápu⌋. At the beginning, the pada-mss. give mā́tṛṣṭā: nāma: asi: kṛṣṇa॰çakune: múkham; and the saṁhitā-mss. agree with them, with worthless variations of accent ⌊and some slight differences besides⌋, and with -kuner in one or two. SPP. reads, however, mā́ tṛṣṭā́nām asi kṛṣṇaçakunér múkham, won, as he claims, by adding accents to the comm's text; but this differs from the mss. only by ⌊the word-division and⌋ by -nāmasi and -ner; how the comm. divides and understands mātṛṣṭānāmasi is unknown, as his explanation of the words is wanting (though SPP. notes no lacuna). So much (to múkham) is, as was noted above, added to vs. 3 by Anukr., comm., and SPP. In the next division of the verse, for kakṣyā̀m, the mss., the comm., and SPP., give kāyám, which might mean 'body'; the comm. is apparently imperfect here, reading açvo yathā svakīyaṁ rajodhūsaraṁ [kāyaṁ] dhunoti yathā cā ’çvo nīnāham palyāṇakavacādikam avakirati: with kāyam is perhaps omitted also çarīram, its gloss. Our mss. end vs. 4 with nīnāhám, and it was our emendation to add the next clause; but this the comm. does also, ending with vapa, while SPP. goes on to gṛhe without making a verse-division; the sense (so far as we can be said to understand it) favors our division and the comm's. The latter reads avā ’smākam, finding thus an ava...vapa, which he
