Nor let it be supposed, that the writers of our MSS. were so careful and so conscientious that they would, when copying MSS., regulate every consonant or vowel according to the rules of the Prâtisâkhya. This is by no means the case. The writers of Vedic MSS. are on the whole more accurate than the writers of other MSS., but their learning does not seem to extend to a knowledge of the minute rules of the Prâtisâkhya, and they will commit occasionally the very mistakes against which they are warned by the Prâtisâkhya. Thus the Prâtisâkhya (Sûtra 799) warns the students against a common mistake of changing vaiyasva into vayyasva, i. e. by changing ai to a, and doubling the semivowel y. But this very mistake occurs in S. 2, and another MS. gives vaiyyasva. See p. xlvii.

i.e.

If these arguments are sound, if nothing can be said against the critical principles by which I have been guided in editing the text of the Rig-veda, if the fourfold check, described above, fulfils every requirement that could be made for restoring that text which was known to Sâyana, and which was known, probably 2000 years earlier, to the authors of the Prâtisâkhyas, what can be the motives, it may fairly be asked, of those who clamour for a new and more critical edition, and who imagine that the editio princeps of the Rigveda will share the fate of most of the editiones principes of the Greek and Roman classics, and be supplanted by new editions founded on the col

lation of other MSS.? No one could have rejoiced more sincerely than I did at the publication of the Romanised transliteration of the Rig-veda, carried out with so much patience and accuracy by Professor Aufrecht. It showed that there was a growing interest in this, the only true Veda; it showed that even those who could not read Sanskrit in the original Devanagarî, wished to have access to the original text of these ancient hymns; it showed that the study of the Veda had a future before it like no other book of Sanskrit literature. My learned friend Professor Aufrecht has been most unfairly charged with having printed this Romanised text me insciente vel invito. My edition of the Rig-veda is publici juris, like any edition of Homer or Plato, and anybody might have reprinted it either in Roman or Devanâgarî letters. But far from keeping me in ignorance of his useful enterprise, Professor Aufrecht applied to me for the loan of the MSS. of the two Mandalas which I had not yet published, and I lent them to him most gladly because, by seeing them printed at once, I felt far less guilty in delaying the publication of the last volumes of my edition of the text and commentary. Nor could anything have been more honourable than the way in which Professor Aufrecht speaks of the true relation of his Romanised text to my edition. That there are misprints, and I, speaking for myself, ought to say mistakes also, in my edition of the Rig-veda, I

know but too well; and if Professor Aufrecht, after carefully transcribing every word, could honestly say that their number is small, I doubt whether other scholars will be able to prove that their number is large. I believe I may with the same honesty return Professor Aufrecht's compliment, and considering the great difficulty of avoiding misprints in Romanised transcripts, I have always thought and I have always said that his reprint of the hymns of the Veda is remarkably correct and accurate. What, however, I must protest against, and what, I feel sure, Professor Aufrecht himself would equally protest against, is the supposition, and more than supposition of certain scholars, that wherever his Latin transcript varies from my own Devanagarî text, Professor Aufrecht is right, and I am wrong, that his various readings rest on the authority of new MSS., and constitute in fact a new recension of the Vedic hymns. Against this supposition I must protest most strongly, not for my own sake, but for the sake of the old book, and, still more, for the sake of the truth. No doubt it is natural to suppose that where a later edition differs from a former edition, it does so intentionally; and I do not complain of those who, without being able to have recourse to MSS. in order to test the authority of various readings, concluded that wherever the new text differed from the old, it was because the old text was at fault. In order to satisfy my own conscience on this point, I have collated a

number of passages where Professor Aufrecht's text differs from my own, and I feel satisfied that in the vast majority of cases, I am right and he is wrong, and that his variations do not rest on the authority of MSS. I must not shrink from the duty of making good this assertion, and I therefore proceed to an examination of such passages as have occurred to me on occasionally referring to his text, pointing out the readings both where he is right, and where he is wrong. The differences between the two texts may appear trifling, but I shall not avail myself of that plea. On the contrary, I quite agree with those scholars who hold that in truly critical scholarship there is nothing trifling. Besides, it is in the nature of the case that what may, by a stretch of the word, be called various readings in the Veda, must be confined to single letters or accents, and can but seldom extend to whole words, and never to whole sentences. I must therefore beg my readers to have patience while I endeavour to show that the text of the Rig-veda, as first published by me, though by no means faultless, was nevertheless not edited in so perfunctory a manner as some learned critics seem to suppose, and that it will not be easy to supplant it either by a collation of new MSS., such as are accessible at present, or by occasional references to the Prâtisâkhya.

I begin with some mistakes of my own, mistakes which I might have avoided, if I had always consulted the Prâtisâkhya, where single words or whole

passages of the Veda are quoted. Some of these mistakes have been removed by Professor Aufrecht, others appear in his transcript as they appear in my own edition.

I need hardly point out passages where palpable misprints in my edition have been repeated in Professor Aufrecht's text. I mean by palpable misprints, cases where a glance at the Pada text or at the Sanhitâ text or a reference to Sâyana's commentary would show at once what was intended. Thus, for instance, in vi. 15, 3, vridhé, as I had printed in the Sanhitâ, was clearly a misprint for vridhó, as may be seen from the Pada, which gives vridháh, and from Sâyana. Here, though Professor Aufrecht repeats vridhé, I think it hardly necessary to show that the authority of the best MSS. (S. 2. alone contains a correction of vridhó to vridhé) is in favour of vridháh, whatever we may think of the relative value of these two readings. One must be careful, however, in a text like that of the Vedic hymns, where the presence or absence of a single letter or accent begins to become the object of the most learned and painstaking discussions, not to claim too large an indulgence for misprints. A misprint in the Sanhitâ, if repeated in the Pada, or if admitted even in the commentary of Sayana, though it need not be put down to the editor's deplorable ignorance, becomes yet a serious matter, and I willingly take all the blame which is justly due for occasional accidents of this