Page:Atharva-Veda samhita volume 2.djvu/495

951 1. Jan̄giḍá art though, jan̄giḍá; defender art thou, jan̄giḍá; what of ours is two-footed, four-footed—let jan̄giḍá defend it all.

2. The witchcrafts that are thrice fifty, and the witchcraft-makers that are a hundred—may the jan̄giḍá make them all of vanished brilliancy (-téjas) [and] sapless.

The first pāda is corrupt in the mss., and very doubtful; the translation implies yā́ḥ kṛtyā́ḥ, which is most naturally suggested by the connection, and takes tripañcāçís as an indefinite large number (like tisráḥ pañcāçátaḥ, RV. i. 133. 4), and as formed like triṣaptá, triṇavá, etc., in spite of the important objection that none of these make a fem. in ī, and that the word most naturally means 'fifty-three, composed of fifty-three,' or the like. ⌊W's conjecture, yā́ḥ kṛtyā́ḥ, nearly coincides with that of Geldner (KZ. xxvii. 218), yā́ç ca kṛtyā́ḥ. Geldner's is metrically better; and he takes trip- as an indefinitely large number (cardinal), as does W.⌋ It was this word tripañcāçá, applied to the set of dice in RV. ⌊x. 34. 8⌋ (but perhaps meaning 'thrice fifty'), that suggested the not very happy emendation in our edition to akṣa-kṛtyā́s. The mss. read mostly jāgṛtsyas tr- (with various accent, most often on -syás: p. jāgṛtsyáḥ: tripañca॰ açī́ḥ!), also jyā-, yyā-, and (two of SPP's) yāgṛtsyás; this last the comm. also has, and understands it as yā(ḥ) gṛtsyas, explaining the latter as = gardhanaçīlās ⌊SPP's pada-text accordingly, yā́ḥ: gṛ́tsyaḥ⌋, and tripañcāçīs as tryadhikapañcaçatsaṁkhyākās, both as epithets of kṛtyās (understood). ⌊With this reading, we can take gṛ́tsyas as nom. pl. fem. to gṛ́tsa and render 'what fifty-three clever or sly [witchcrafts there are]'; but gṛ́tsa, in such an application and with such sinister meaning, has rather slender support, to wit, VS. xvi. 25, as cited by BR. ii. 778.⌋ Ppp. gives yā kṛcchrā tripañcāçīç ch-, which, while it is itself (emended to yāḥ kṛcchrās) not wholly unacceptable, also favors our yā́ḥ kṛtyā́s; there is insufficient reason for the feminine words if kṛtyās be not expressed. ⌊I cannot here attach much value to the evidence of Ppp.: on the one hand, it confuses surd and sonant very often (kovidam for govidam, xix. 13. 5: cf. xi. 5. 4, note); and, on the other, the relation of its cch to ts may be somewhat like that discussed under x. 9. 23 (ṛchára, ṛtsára, etc.). The mss. are decidedly in favor of gṛtsyas as against kṛtyās; but Whitney's objection as to the omission of kṛtyās seems to me a weighty argument in favor of his conjecture.⌋ Our vínaṣṭatejasas in c was an emendation, which, now that the comm. also reads it, may be regarded as sufficiently established; the mss. mostly vinaktatéjasas (also vinaktu t-, viniṣṭat-, bhanakti t-, minaktu t- ⌊etc.⌋: pada-readings, vinaktu [as independent word or as compounded with téjasaḥ] or also vinakta॰t-). SPP. strangely contents himself with vinaktu téjasas, which certainly he would be unable to translate into anything even simulating sense.