And if we look at the epic literature of other nations, we find much the same. The poem of the Nibelunge, such as we possess it, was composed and written down about A.D. 1200; but we see from the short ballads recording the fates of Sigurd and Brynhild, as preserved to us in the Edda, what the poems may have been like which were used as his materials by the unknown German Homer of the twelfth century.

Firdusi, the author of the Persian national epic, the Shâh Nâmeh, Book of Kings, about A.D. 1000, tells us himself what sources he consulted, and how he travelled about from village to village to collect the materials for his great poem.

We see, therefore, that the constituent elements of the epic poems which we possess the Iliad, the Odyssey, the Nibelunge, and the Shah Nâmeh-existed in oral tradition long before the invention of writing; that they were elaborated during what is called the Mnemonic Period of literature; and that they were reduced to their present form at a time when the art of alphabetic writing had been discovered, and had been applied, not only to inscriptions, but to poetry and other literary compositions; when, in fact, there were not only a few sculptors to engrave official documents, but a large public that could both read and write.

We must distinguish, therefore, between two kinds of epic poetry-one, the Mnemonic, as preserved intact in the Kalewala, and slightly recast in the Kalewipoeg; another, the Literary, which has passed through the hands of later poets, such as Iliad and Odyssey, the Nibelungenlied, and the Shâh Nâmeh. There is a third class of purely Artificial epic poems, written in imitation of these models, such as Virgil's Eneid, Wolfram von Eschenbach's Parzival, the Lusiads of Camoens, and, lastly, poems like Voltaire's Henriade. But these belong to quite a different sphere of poetry.

It is not likely that the number of true epic poems will ever

be increased, and the discovery of two epic poems in India was therefore a great event in the comparative study of ancient literature. These two Sanskrit poems were the Mahâbhârata and the Râmâyana. Leaving here out of consideration the Râmâyana, the author of which, Valmîki, seems to have been more than a mere Diaskeuastes, we have the Mahâbhârata, an ancient Sanskrit poem, which, in its present form, is said to consist of no less than 90,000 couplets, and was published at Calcutta in 1834-39 in four large quarto volumes. In its present form it can hardly be called a poem. It is an enormous poetical treasury, in which everything was thrown that could possibly be brought into connection with its original story. At first the Mahâbhârata was the story of the Great War that had been waged in ancient times in India, but after the popular songs about that "Eighteen Days' War" had been collected, not only ancient stories, but ancient laws, customs, and philosophies were all incorporated into it. The date of its first composition is, of course, unknown, as much as the age of the single Aristeias of Iliad and Odyssey, or of the various Kvidhas of the Edda. All we can say is, that a collected poem, called not only Bhârata, but Mahâbhârata, i.e. the great story of the war of the Bhâratas, is mentioned clearly in the Asvalâyana Sûtras, that is, before the rise of Buddhism in the sixth century B.C. But what the Mahâbhârata was then, we shall of course never know. All we can say is that, when first collected, it must have been a mnemonic poem, composed before the art of writing was known in India, or, at all events, before it had been applied to literary purposes in that country. Extraordinary as it may seem to us that so enormous a poem should have been composed and handed down by memory only, the fact itself can hardly be doubted. It is in India, of all countries in the world, where we must study the facts about the extraordinary, and to us almost incredible, powers

of memory, before the art of writing was known and practised. There is, no doubt, an introductory chapter of our poem which tells us how, with the help of Ganesa, the Mahâbhârata was written. But the absence of that chapter in a large number of MSS., more particularly, as Dr. Winternitz has shown, in the MSS. of the south of India, speaks for itself, and proves that the idea of the poem having been composed on paper was an afterthought, and may, in fact, serve to determine the time when writing for literary purposes became the custom in India. The poet is called Vyâsa, or Krishna Dvaipâyana, a name which means "arranger," from vi and as, to distribute, to extend; and whatever his date may have been, we may be certain that his work, the Mahâbhârata, was a genuine product of the Mnemonic Period of Indian literature.

Life seems too short to read such a gigantic poem as a whole, and various attempts have been made from time to time to reduce it to its original dimensions. The most promising was to follow the indications of an old Index or Anukramanikâ, incorporated in the poem itself. But if we add the number of couplets of each book, as stated in that preface, we are still left with a poem of no less than 85,000 couplets. It is impossible to enjoy such a poem, however beautiful it may be in some of its stories; nor can we imagine that the original poet, or the original collector even, contemplated such a monstrous production. The original subject was the war between two branches of the same royal family, the Bhârata princes. These Bharatas are mentioned as an ancient and most powerful race in the north of India as early as the Brâhmana period, for in the Satapathabrâhmaṇa, xiii. 5, 4, we read, "The greatness of the Bhâratas neither the men before nor those after them have ever attained; nor did the seven tribes of men, even as mortal man does not touch the sky with his flanks."

It is easy to see how round the nucleus of this war an immense mass of poetry, both popular and artificial, was accumulated, but it was not so easy a task to sift this enormous mass, and to extract from it what may have been the original story. This task has been boldly undertaken and carried through, as far as I can judge, with great success by Mr. Romesh Dutt in his "Mahabharata condensed into English Verse.' He has himself given an account of the principles by which he was guided in his work. He has, as much as possible, taken a number of verses of the original and rendered them faithfully into English. He has left out on the very largest scale, but he has not added; and the impression which his bold undertaking leaves on the reader is certainly that something like what we here read in English may have been recited in India when the war between the Kurus and the sons of Pându was first sung by the ancient bards of the country. If Vyâsa, the collector, was in any sense the author of some parts of the poem, it is important to observe that in the poem itself he is represented as the kinsman, though under peculiar circumstances, of the royal family of the Bhâratas.

Mr. Romesh Dutt tells us that without exaggeration two hundred millions of Hindus cherish to the present day the story of their ancient epics. "The Hindu," he writes, "scarcely lives, man or woman, high or low, educated or ignorant, whose earliest recollections do not cling round the story and the characters of the Mahâbhârata," listening to it either in the original, or to a translation into one of the many vernaculars. Considering how many people there are now in India who know English better than Sanskrit, there can be little doubt that this English translation and condensation by Mr. Romesh Dutt will be most welcome to his own compatriots. It will be equally welcome to many English students of literature who wish to gain an idea

of that wonderful epic. The original, in its colossal bulk, can be consulted by Sanskrit scholars only. It is true it has been translated from beginning to end into English by my late friend Protap Chandra Roy and his assistants, but even that translation will hardly meet with many patient readers in England. To the student of Sanskrit literature the Mahâbhârata, so far as it reflects the ancient life of India, will always remain an invaluable treasury. Quite apart from the story of the ancient heroic war, the great bulk of the later accretions also is full of interest and instruction. Of course there exists always one great difficulty; we cannot tell which period of Indian history is represented to us in each of its various component parts. Every race in the north, the south, the east, and west of India wished to see some account of its own heroes inserted in the national epic. Philosophers claimed some mention of their doctrines as a proof of their antiquity, and as a kind of brief of nobility. Moral and legal authorities clamoured naturally for the same mention; and even mere customs, rules about caste, marriage, and inheritance had to be placed under the shield of the ancient epic. The geography, the ethnology, the migrations of races in India, all may be studied in the Mahâbhârata. The dominating religious cult, not at the time of the Great War itself, but at the time of the final redaction of the poem, was clearly the worship of Krishna, and this is generally referred to a time after the disappearance of Buddhism from the soil of India, so that the final redaction of the epic can hardly be placed before the Renaissance period of Indian literature, that is, before the fourth century A.D. But much of this is still uncertain, and we must carefully guard against premature and positive assertions when we treat of the chronology of Indian literature, and of its epic literature in particular. Every generation of scholars builds up its own system of chronology, and the next generation