words which designate those gods cannot be eternal, but must have originated co-evally with the created objects which they denote, since eternal words could not have an eternal connection with non-eternal objects. This difficulty he tries to overcome (ignoring the ground taken by Jaimini, that the Veda contains no references to non-eternal objects) by asserting that the eternal connection of words is not with individual objects, but with the species to which these objects belong, and that Indra and the other gods are proved by the Veda to belong to species. Sankara then goes on to assert, on the authority of Brahma Sutra, i. 3, 28, fortified by various texts from the Vedas and the Smritis, that the gods and the world generally are produced (though not in the sense of evolution out of a material cause) from the word of the Vedas (see pp. 6 and 16) in the form of sphota. This last term will be explained below. This subject above referred to, of the eternal connection of the words of the Veda with the objects they represent, is further pursued in a passage which I have quoted in the Appendix, p. 300, where an answer is given to the objection that the objects denoted by the words of the Veda cannot be eternal, as a total destruction of everything takes place (not, indeed, at the intermediate, but) at the great mundane dissolutions. The solution given is that, by the favour of the supreme Lord, the inferior lords Brahma, etc., retain a recollection of the previous mundane conditions; and that in each successive creation everything is produced exactly the same as it had previously been. I then proceed in p. 105 to adduce a passage from Sayana, the

commentator on the Rig-veda, who refers to another of the Brahma Sutras, i. 1, 3 (quoted in p. 106), declaring that Brahma was the source of the Veda, which Sankara interprets as containing a proof of the omniscience of Brahma. Sayana understands this text as establishing the superhuman origin of the Veda, though not its eternity in the proper sense, it being only meant, according to him (as well as to Madhava; see p. xi.), that the Veda is eternal in the same sense as the æther is eternal, i.e. during the period between each creation and dissolution of the universe.

In opposition to the tenets of the Mīmānsakas, who hold the eternity (or the eternal self-existence) of the Veda, and to the dogmas of the Vedanta, as just expounded, Gotama, the author of the Nyaya aphorisms, denies (Section ix. pp. 108-118) the eternity of sound; and after vindicating the Veda from the charges of falsehood, self-contradiction, and tautology, deduces its authority from the authority of the wise, or competent, person or persons who were its authors, as proved by the efficacy of such of the Vedic prescriptions as relate to mundane matters, and can be tested by experience. It does not distinctly result from Gotama's aphorism that God is the competent person whom he regards as the maker of the Veda. If he did not refer to God, he must have regarded the rishis as its authors. The authors of the Vaiseshika Sūtras, and of the Tarka Sangraha, as well as the writer of the Kusumanjali, however, clearly refer the Veda to Isvara (God) as its framer (pp. 118-133). Udayana, the author of the latter

work (pp. 128-133), controverts the opinion that the existence of the Veda from eternity can be proved by a continuous tradition, as such a tradition must, he says, have been interrupted at the dissolution of the world, which preceded the existing creation. He, therefore (as explained by his commentator), infers an eternal (and omniscient author of the Veda; asserting that the Veda is paurusheya, or derived from a personal author; that many of its own texts establish this; and that the appellations given to its particular śākhās or recensions, are derived from the names of those sages whose persons were assumed by Iśvara, when he uttered them at the creation. In pp. 125 ff. I have quoted one of the Vaiśeshika Sutras, with some passages from the commentator, to show the conceptions the writers entertained of the nature of the supernatural knowledge, or intuition, of the rishis.

Kapila, the author of the Sankhya Aphorisms (pp. 133 -138), agrees with the Nyaya and Vaiśeshika aphorists in denying the eternity of the Veda, but, in conformity with his own principles, differs from Gotama and Kanāda in denying its derivation from a personal (i.e. here, a divine) author, because there was no person (i.e. as his commentator explains, no God) to make it. Vishnu, the chief of the liberated beings, though omniscient, could not, he argues, have made the Veda, owing to his impassiveness, and no other person could have done so from want of omniscience. And even if the Veda have been uttered by the primeval Purusha, it cannot be called his work, as it was breathed forth by him unconsciously. Kapila agrees

with Jaimini in ascribing a self-demonstrating power to the Veda, and differs from the Vaiseshikas in not deriving its authority from correct knowledge possessed by a conscious utterer. He proceeds to controvert the existence of such a thing as sphota (a modification of sound which is assumed by the Mīmānsakas, and described as single, indivisible, distinct from individual letters, existing in the form of words, and constituting a whole), and to deny the eternity of sound.

In the tenth Section (pp. 138-179) I shew (a) by quotations from the aphorisms of the Vedanta and their commentator (pp. 140-145), that the author and expounder of the Uttara Mīmāñisā (the Vedanta) frequently differ from Jaimini the author of the Purva Mīmāmsā in the interpretation of the same texts of the Upanishads. A similar diversity is next (b) proved at greater length (pp. 145-173), by quotations from the aphorisms and commentaries of the Vedanta and the Sankhya, to characterize the expositions proposed by the adherents of those two systems respectively. One quotation is given in pp. 175 ff. to shew (c) that the same is true in regard to the followers of the Vaiseshika philosophy, who distinctly reject the Vedāntic explanations; and last of all (d) I have made some extracts (pp. 177 ff.) from the Bhakti Sūtras of Sandilya to exhibit the wide divergence of that writer from the orthodox views of the Vedanta

regarding the sense of the Vedas. In pp. 173-175 I quote some remarks of Dr. E. Roer, and Prof. Max Müller, regarding the doctrines of the Upanishads, and their relations to the different philosophical schools.

In the facts brought forward in this section we find another illustration (1) of the tendency common to all dogmatic theologians to interpret in strict conformity with their own opinions the unsystematic and not always consistent texts of an earlier age which have been handed down by tradition as sacred and infallible, and to represent them as containing, or as necessarily implying, fixed and consistent systems of doctrine; as well as (2) of the diversity of view which so generally prevails in regard to the sense of such texts among writers of different schools, who adduce them with equal positiveness of assertion as establishing tenets and principles which are mutually contradictory or inconsistent.

In the eleventh Section (pp. 179-207) some passages are adduced from the Nyāya-mālā-vistara, and from Kullūka's commentary on Manu, to show that a distinct line of demarcation is drawn by the scholastic writers between the Vedas on the one hand, and all other classes of Indian scriptures, embraced under the designation of Smriti (including the Darśanas, the Institutes of Manu, the Purānas, and Itihāsas, etc.), on the other, the first being regarded as independent and infallible guides, while the others are (in theory) held to be authoritative only in so far as they are founded on, and coincide with, the Veda. The practical effect of this distinction is, however, much lessened by the fact that the ancient sages, the authors of the Smritis, such of them, I mean, as, like Manu, are recognized as orthodox, are looked upon by Madhava and Sankara as having had access to Vedic texts now no longer extant, as having held communion with the gods,