The subject that I venture to discuss in this paper is one that may seem surprising and even fantastic, and yet, if any orderliness can be introduced into the earliest Indian ages, it can only be attained by examining and co-ordinating all the genealogical and quasi-historical data which have been handed down in Sanskrit books. The subject has been before my mind for many years, and it has been only after long consideration of all the relevant information, which I have been able to collect out of all those books, especially the Epics and Purāṇas, that it has seemed to me some measure of order may be educed out of the chaos of material. That information is condensed in the following pages, and no statement is made without citing the authorities that support it. I may say that the conclusions set out here were not reached from any preconceived ideas, except this one (if it merits that description), that the ancient kṣatriya literature deserves to be examined from a common-sense point of view on the supposition that it may contain genuine tradition, however much distorted in the course of time. It was only after investigating the subject piecemeal, following each detail into other details to which it led and continually rearranging them as their number and mutual relations developed, that something definite seemed at first to emerge out of the chaos, and then gradually the subject seemed to shape itself into some degree of order. Even if my views should not commend themselves to others, yet the material collected here and the method of treatment may, I hope, be of some service to others in elucidating the subject.

The Ahuna-Vairya was so distinguished by the later use which was made of it that it became a formula of unusual moral and ritualistic importance—indeed, more so than, upon our first glance upon it, we should, the most of us, think that it deserved. But, though bearing unmistakable traces of being somewhat artificially constructed, both in its metre and contents, upon closer study the little group of words seems well worthy of its parentage, for it is a succinct cipher of that remarkable manifestation of the moral idea which, as the one point of Zoroastrianism, must have had enormous influence during successive generations among the inhabitantsof Mid-Asia. The Pahlavi form of the name Ahunaver is but a contraction of Ahuna-Vairya, —the nasal n having intruded from thenasal m of an accusative ahūm, or else from mere euphony.

When at Lhasa in 1904 I found, in addition to the two earliest historical Tibetan documents yet known (as notified inthis Journal), a very long inscription of the Chinese emperor Chia-ch'ing, of 1808 a.d., which is of considerable historical interest and importance. It gives an official account of the origin of the Grand Lamaship and of the theory of succession to the same by divine reincarnations; it also prescribes the “Ordeal of the Urn” for the selection of the candidate, one of the steps taken by China to secure political control over the succession to the pontifical throne; and it records the building of the Potala palace at Lhasa as one of “the three Potalas”, and of a fourth “Potala”-academy erected by a Chinese emperor at Jehol, to the north-west of Peking.

Before proceeding to the subject-matter of this paper, I would ask leave to revert for a moment to the preceding article on the Bhāgavata system of incarnations. Several kind Bhāgavata friends have sent me criticisms on points of detail in what I then wrote, which will be utilized when opportunity occurs. I would mention one now, as it affects the question of terminology. On p. 624 I used the name Vibhu or Vibhava Avatāra as the name of one of the forms under which the Supreme manifests Himself. I can give authority for both these names from North Indian literature; but, writing from Mysore in the south, Paṇḍit Gōvindâcārya, the translator of Rāmaňuja's commentary on the Bhagavad Gītā, informs me that the use of Vibhu in this connexion is incorrect. As a technical term of Bhāgavata theology, vibhu means “infinite”, in contradistinction to aṇu, “finite.” For the incarnation, vibhava is the only correct term. In this sense vibhava is explained as vividhēna bhavati, and means literally “many-becomingness”.

[The following article contains a revised translation of Bābar's account of Farghāna, a passage discussed and quoted by many writers on Turkistān. Some mistaken inferences have been drawn from it as it stands in the Memoirs and Mémoires, because these both lacked a pure textual basis and modern local knowledge. I regret that, obeying a Turk in his Turkī, an autobiographer in his style, my wording departs from Mr. Erskine's. The speech of some Englishmen can go straight into Turkī; out of Turkī, Bābar's should go straight into theirs. They are not schooled, nor was he. Neither blurs meaning by complex statement; neither throws “and” into the pause between two thoughts. Mr. Symonds' rule gathers force from the clearness of the mould of Turkā speech: “A good translation should resemble a plaster cast, the English being plaqué upon the original, so as to reproduce its exact form, although it cannot convey the effects of bronze or marble which belong to the material of the work of art.”]

There is much to support the opinion of Rāmānuja, Dr. Thibaut, and many others, that Śaṃkara's doctrine of “illusion” is a biassed rendering of the old Vedānta, Bādarāyaṇik as well as Aupanishadic. If that be granted, it is by no means self-evident that Buddhism has been without influence on Śamkara's speculation; and the last writer on the subject, Vasudev Anant Sukhtankar, a very able pupil of Professor Jacobi, does not conceal his opinion, or his surmise, that Śaṃkara is indebted to Nāgārjuna. That may be true, but I would object that we really know little or nothing about the history of Vedānta, and that conclusions based on philosophical parallels are by no means definitive. Autonomous developments— autonomous if not absolutely independent—are admissible. Nāgārjuna (or his predecessors, the anonymous authors of the oldest Mahāyānasūtras), by the very fact that he proclaims “voidness” to be the real nature of things, was prepared to distinguish the relative truth (saṃvṛtisatya) and the absolute one (pāramārthika); and his nihilism coupled with “idealism” might lead to the Vijñānavāda: “existence of pure non-intelligent (?) intellect.” On the other hand the Aupanishadas, from their main thesis (tat tvam asi, etc.), could derive the distinction of the two brahmans, of the two vidyās. Both developments are natural enough; the conception of the universal void (o) and the intuition of the infinite (∞) are convergent, in the end; but parallel and convergent as they are, these developments do not lose their primitive tinge.

When taking part in the joint discussion, which ensued in connexion with the papers which Messrs. Grierson and Barnett read at Oxford (September, 1908), at the Congress of “History of Religions”, I mentioned what Dr. Grierson has now published in the JRAS. (1909, p. 1122). I quoted then from memory the two Sutras, one relating to Buñ and the other relating to Bun, and cited the following line from Kāśikā Vritti: Na cātra Bun-Buñorviśeṣo vidyate kimartham Vāsudeva grahaṇam, etc.