translating several curious passages from one [10] of them.1 I have been still more fortunate in collecting at Benares the text and commentary of a large portion of these celebrated books; and, without waiting to examine them more completely than has been yet practicable, I shall here attempt to give a brief explanation of what they chiefly contain.

It is well known that the original Veda is believed by the Hindus to have been revealed by Brahmá, and to have been preserved by tradition, until it was arranged in its present order by a sage, who thence obtained the surname of Vyása, or Vedavyása: that is, compiler of the Vedas. He distributed the Indian scripture into four parts, which are severally entitled Rich, Yajus, Sáman, and Atharvana; and each of which bears the common denomination of Veda.

Mr. Wilkins and Sir William Jones were led, by the consideration of several remarkable passages, to suspect that the fourth is more modern than the other three. It is certain that Manu, like others among the Indian lawgivers, always speaks of three only, and has barely alluded to the Atharvana,3 without however terming it a Veda. Passages of the Indian scripture itself seem to support the inference: for the fourth Veda is not mentioned in the passage cited by me in a former essay 3 from the white Yajus; nor in the following text, quoted from the Indian scripture by the commentator of the Rich. "The Rigveda originated from fire; the Yajurveda from air; and the Sámaveda from the sun."5

[11]

1 See Preface to Manu, page vi. and the works of Sir William Jones, vol. vi. Manu, chap. 11, v. 33.

3 Essay Second, on Religious Ceremonies. See Asiatic Researches, vol. vii., p. 251.

From the 31st chapter; which, together with the preceding chapter (30th), relates to the Purushamedha, a type of the allegorical immolation of Nárayaṇa, or of Brahma in that character.

Manu alludes to this fabulous origin of the Vedas (chap. 1. v. 23). His commentator, Medhátithi, explains it by remarking that the Rigveda opens with a hymn to fire; and the Yajurveda with one in which air is mentioned. But Kullúkabhaṭṭa has recourse to the renovations of the universe. "In one Kalpa, the Vedas proceeded from fire, air, and the sun; in another, from Brahma, at his allegorical immolation."

Arguments in support of this opinion might be drawn even from popular dictionaries; for Amarasinha notices only three Vedas, and mentions the Atharvana without giving it the same denomination. It is, however, probable, that some portion at least of the Atharvana is as ancient as the compilation of the three others; and its name, like theirs, is anterior to Vyása's arrangement of them: but the same must be admitted in regard to the Itihása and Puránas, which constitute a fifth Veda, as the Atharvana does a fourth.

It would, indeed, be vain to quote in proof of this point the Puráṇas themselves, which always enumerate four Vedas, and state the Itihása and Puránas as a fifth; since the antiquity of some among the Puráņas now extant is more than questionable, and the authenticity of any one in particular does not appear to be as yet sufficiently established. It would be as useless to cite the Manduka and Tápaniya Upanishads, in which the Atharva-veda is enumerated among the scriptures, and in one of which the number of four Vedas is expressly affirmed for both these Upanishads appertain to the A'tharvana itself. The mention of the sage Atharvan in various places throughout the Vedas1 proves nothing; and even a text of the Yajurveda, where he is named in contrast with the Rich, Yajus, and Saman, and their supplement or Brahmana, [12] is not decisive. But a very unexceptionable passage may be adduced, which the commentator of the Rich has quoted for a different purpose from the Chhandogya Upanishad, a portion of the Sáman. In it, Nárada, having solicited instruction from Sanatkumára, and being interrogated by him as to the extent of his previous knowledge, says, "I have learnt the Rigveda, the Yajurveda, the Sámaveda, the A'tharvana, [which is] the fourth, the Itihása and Purána, [which are] a fifth, and [grammar, or] the Veda of Vedas, the obsequies of the manes, the art of computation, the knowledge of omens, the revolutions of periods, the intention of speech 1 Vide Vedas passim. 2 In the Taittiriya Upanishad.

[or art of reasoning], the maxims of ethics, the divine science [or construction of scripture], the sciences appendant on holy writ [or accentuation, prosody, and religious rites], the adjuration of spirits, the art of the soldier, the science of astronomy, the charming of serpents, the science of demigods. [or music and mechanical arts]: all this have I studied; yet do I only know the text, and have no knowledge of the soul.”1

From this, compared with other passages of less authority, and with the received notions of the Hindus themselves, it appears that the Rich, Yajus, and Sáman, are the three principal portions of the Veda; that the Atharvana is commonly admitted as a fourth; and that divers mythological poems, entitled Itihása and Puráņas, are reckoned a sup-[13] plement to the scripture, and as such constitute a fifth Veda.2

The true reason why the three first Vedas are often mentioned without any notice of the fourth must be sought, not in their different origin and antiquity, but in the difference of their use and purport. Prayers employed at solemn rites, called yajnas, have been placed in the three principal Vedas: those which are in prose are named Yajus; such as are in

1 Chhandogya Upanishad, ch. 7, § 1. I insert the whole passage, because it contains an ample enumeration of the sciences. The names by which grammar and the rest are indicated in the original text are obscure; but the annotations of S'ankara explain them. This, like any other portion of a Veda where it is itself named (for a few other instances occur), must of course be more modern than another part to which the name had been previously assigned. It will hereafter be shown that the Vedas are a compilation of prayers, called mantras; with a collection of precepts and maxims, entitled Brahmana, from which last portion the Upanishad is extracted. The prayers are properly the Vedas, and apparently preceded the Brahmana.

2 When the study of the Indian scriptures was more general than at present, especially among the Brahmanas of Kanyakubja, learned priests derived titles from the number of Vedas with which they were conversant. Since every priest was bound to study one Veda, no title was derived from the fulfilment of that duty; but a person who had studied two Vedas was surnamed Dwivedi; one who was conversant with three, Trivedi; and one versed in four, Chaturvedi : as the mythological poems were only figuratively called a Veda, no distinction appears to have been derived from a knowledge of them in addition to the four scriptures. The titles above mentioned have become the surnames of families among the Brahmans of Kanoj, and are corrupted by vulgar pronunciation into Dobe, Tiwáre, and Chaube.

metre are denominated Rich; and some, which are intended to be chanted, are called Saman: and these names, as distinguishing different portions of the Vedas, are anterior to their separation in Vyása's compilation. But the Atharvana not being used at the religious ceremonies above mentioned, and containing prayers employed at lustrations, at rites conciliating the deities, and as imprecations on enemies, is essentially different from the other Vedas; as is remarked by the author of an elementary treatise on the classification of the Indian sciences.1

But different schools of priests have admitted some variations in works which appear under the same title. This circumstance is accounted for by the commentators on the Vedas, who relate the following story taken from Puránas [14] and other authorities. Vyása having compiled and arranged the scriptures, theogonies, and mythological poems, taught the several Vedas to as many disciples: viz. the Rich to Paila, the Yajus to Vaiśampáyana, and the Sáman to Jaimini; as also the Atharvana to Sumantu, and the Itihása and Puránas to Súta. These disciples instructed their respective pupils, who, becoming teachers in their turn, communicated the knowledge to their own disciples; until at length, in the progress of successive instruction, so great variations crept into the text, or into the manner of reading and reciting it, and into the no less sacred precepts for its use and application, that eleven hundred different schools of scriptural knowledge arose.

The several Sanhitás, or collections of prayers in each Veda, as received in these numerous schools or variations, more or less considerable, admitted by them either in the arrangement of the whole text (including prayers and precepts), or in regard to particular portions of it, constituted the Sákhás, or branches of each Veda. Tradition, preserved in the Puráņas, reckons sixteen Sanhitás of the Rigveda; eighty-six of the Yajus, or including those which branched from a second re1 Madhusudana-saraswati, in the Prasthanabheda.

velation of this Veda, a hundred and one; and not less than a thousand of the Sámaveda, besides nine of the Atharvana. But treatises on the study of the Veda reduce the S'ȧkhás of the Rich to five; and those of the Yajus, including both revelations of it, to eighty-six.1

The progress by which (to use the language of the Puráņas) the tree of science put forth its numerous branches is thus related. Paila taught the Rigveda, or Bahrṛich, to two disciples, Báshkala and Indrapramati. The first, also called Báshkali, was the editor of a Sanhitá, or [15] collection of prayers, and a S'ákhá bearing his name still subsists: it is said to have first branched into four schools; afterwards into three others. Indrapramati communicated his knowledge to his own son Mánḍukeya, by whom a Sanhitá was compiled, and from whom one of the S'ákhás has derived its name. Veda

mitra, surnamed Śákalya, studied under the same teacher, and gave a complete collection of prayers: it is still extant; but is said to have given origin to five varied editions of the same text. The two other and principal S'ákhás of the Rich are those of Aswalayana and Śánkháyana, or perhaps Kaushitaki: but the Vishnu-purána omits them, and intimates that Sákapúrni, a pupil of Indrapramati, gave the third varied edition from this teacher, and was also the author of the Nirukta: if so, he is the same with Yáska. His school seems to have been subdivided by the formation of three others derived from his disciples.

The Yajus, or Adhwaryu, consists of two different Vedas, which have separately branched out into various S'ákhás. To explain the names by which both are distinguished, it is necessary to notice a legend, which is gravely related in the Puránas and the commentaries on the Veda.

The Yajus, in its original form, was at first taught by Vaisampayana to twenty-seven pupils. At this time, having

1 The authorities on which this is stated are chiefly the Vishnu-purána, part 3, chap. 4, and the Vijayavilása on the study of scripture; also the Charanavyúha, on the S'dkhás of the Vedas.