Page:Atharva-Veda samhita.djvu/41

Rh of the Kashmirian text antedates that of this work, the reasons why the facsimile was not used by me should be consulted at p. lxxxv.

Accentuation of Sanskrit words.—In the reports of the readings of accented texts, the words are invariably accented. The Kashmirian text is reckoned as an unaccented one, although it has occasional accented passages. The author frequently introduces Sanskrit words, in parentheses or otherwise, into the translation, and usually indicates their accent. The editor has gone somewhat farther: he has indicated in the translation the accent of the stems of words which happen to occur in the vocative {so sadā́nvās, ii. 14. 5), except in the cases of rare words whose proper stem-accent is not known (examples in ii. 24); and, in cases where only one member of a compound is given, he has indicated what the accent of that member would be if used independently (so -nīthá at xviii. 2. 18, as part of sahásraṇītha; -kṣétra at iii. 3. 4, as part of anyakṣetrá; cf. ii. 8. 2).

Cross-references.—Apart from the main purpose of this work, to serve as the foundation of more nearly definitive ones yet to come, it is likely to be used rather as one of consultation and reference than for consecutive reading. I have therefore not infrequently added cross-references from one verse or note to another, doing this even in the case of verses which were not far apart: cf., for example, my reference from vii. 80. 3 to 79. 4 or from vi. 66. 2 to 65. 1.

Orthography of Anglicized proper names.—The translation is the principal or only part of this work which may be supposed to interest readers who are without technical knowledge of Sanskrit. In order to make the proper names therein occurring more easily pronounceable, the author has disregarded somewhat the strict rules of transliteration which are followed in the printing of Sanskrit words as Sanskrit, and has written, for example, Pūshan and Purandhi instead of Puṣan and Puraṁdhi, sometimes retaining, however, the strange diacritical marks (as in An̄giras or Varuṇa) where they do not embarrass the layman. To follow the rules strictly would have been much easier; but perhaps it was better to do as has been done, even at the expense of some inconsistencies (cf. Vritra, Vṛitra, Vṛtra; Savitar).

Editorial short-comings and the chances of error.—Labor and pains have been ungrudgingly spent upon Whitney's work, to ensure its appearance in a form worthy of its great scientific importance; but the work is extensive and is crowded with details of such a nature that unremitting care is needed to avoid error concerning them. Some striking illustrations of this statement may be found in the foot-note below. Despite trifling