PURAIT'A (literally "old," from the purl, before, past), is the name of that class of religious works which, besides the Tantras (q.v.) is the main foundation of the actual popular creed of the Brahminical Hindus (see HINDU RELIGION under INDIA). According to the popular belief, these works were compiled by YytIsa (q.v.), the sup posed arranger of the Vedas (q.v.), and the author of the .1fahilbheirata (q.v.), and possess an antiquity far beyond the reach of historical computation. A critical investigation, however, of the contents of the existiny works bearing that name must necessarily lead to the conclusion, that in their present form they do not only not belong to a remote age, but can barely claim an antiquity of a thousand years. The word pura n'a occurs in some passages of the the law-books of tljuavalkya and Manu (q.v.); it is even met with in some Upanishads and the great Braman' a portion of the White-l'aijur Veda; but it is easy to show that in all these ancient works it cannot refer to the existing com positions called Puran'a, and, therefore, that no inference relative to the age of the latter can be drawn from that of time former, whatever that may be. Nevertheless, it must be admitted that there are several circumstances tending to show that there existed number of works called Puran'a, which preceded the actual works of the same name, and were the source whence these probably derived a portion of their contents. The oldest known author of a Sanskrit vocabulary, Amara-Sadie, gives as a synonym of Pnran'a the word panchalakshan'a, which means "that which has five (panchan) char acteristic marks" (lakshan'a); and the scholiasts of that vocabulary agree in stating that these lakshan'as are: 1. Primary creation, or cosmogony; 2. Secondary creation, or the destruction and renovation of worlds; 3. Genealogy of gods and patriarchs; 4. Jleanwan tetras, or reigns of Manus; and 5. The history of tile princess of the solar and lunar races. Such, then, were the characteristic topics of a Puran'a at the not of Amara-Sinha himself—which is probable—at least of his oldest commentators. Yet the distinguished scholar, most conversant with the existing Puran'as, who, in his preface to the transla tion of the Vizehn'u-Purcin'a, gives a more or less detailed account of their chief contents (prof. H. H. Wilson), observes, in regard to the quoted definition of the corumeutators
on Amara-Sinha, that in no one instance do the actual Puran'as conform to it exactly; that "to some of them it is utterly inapplicable; to others, it only partially applies." To the Viskit'it-PuiyinW, he adds, it belongs more than to any other Purilu'a; but even in the case of this Puran'a he shows that it cannot he supposed to be included in the term explained by the commentators. The age of Amara-Sinha is, according to Wilson, the last half of the century preceding the Christian era; others conjecture that it dates some centuries later. On the supposition, then, that Amara Sinha himself implied by pancha btkshau'a the sense given to this term by his commentators, there would have been Puran'as about 1900 or 1600 years ago; but none of these have descended to our time in the shape it then possessed.
Val ions passages in the actual Puran'as furnish proof of the existence of such elder Puran'as. The strongest evidence in this respect is that afforded by a general descrip tion given by the Matslia-Punin'a of the extent of each of the (which are uni formly stated to be 18 in number), including itself; for, leaving aside the exceptional case in which it may be doubtful whether we possess the complete work now going by the name of a special Puran'a, prof. Wilson, in quoting the description from the -Vaisya Panin'a, and in comparing with it the real extent of the great majority of Punin'as the completeness of which, in their actual state, does not admit of a reasonable doubt, has conclusively shown that the Maisija-Paran'a speaks of works which are not those we now possess. We are then bound to infer that there have been Puran'as older than those preserved, and that their number has been 18, whereas, on the contrary, it will be hereafter seen that it is very doubtful whether ire are entitled to assign this number to the actual Pura/1'a literature.