I can here state a few facts only. There are no inscriptions to be found anywhere in India before the middle of the third century b.c. These inscriptions are Buddhist, put up during the reign of Asoka, the grandson of Kandragupta, who was the contemporary of Seleucus, and at whose court in Patalibothra Megasthenes lived as ambassador of Seleucus. Here, as you see, we are on historical ground. In fact, there is little doubt that Asoka, the king who put up these inscriptions in several parts of his vast kingdom, reigned from 259-222 b.c.

These inscriptions are written in two alphabets—one written from right to left, and clearly derived from an Aramaæan, that is, a Semitic alphabet; the other written from left to right, and clearly an adaptation, and an artificial or systematic adaptation, of a Semitic alphabet to the requirements of an Indian language. That second alphabet became the source of all Indian alphabets, and of many alphabets carried chiefly by Buddhist teachers far beyond the limits of India, though it is possible that the earliest Tamil alphabet may have been directly derived from the same Semitic source which supplied both the dextrorsum and the sinistrorsum alphabets of India.

Here then we have the first fact—viz. that writing, even for monumental purposes, was unknown in India before the third century b.c.

But writing for commercial purposes was known in India before that time. Megasthenes was no doubt quite right when he said that the Indians did not know letters,[268] that their laws were not written, and that they administered justice from memory. But Nearchus, the admiral of Alexander the Great, who sailed down the Indus (325 b.c.), and was therefore brought in contact with the merchants frequenting the maritime stations of India, was probably equally right in declaring that "the Indians wrote letters on cotton that had been well beaten together." These were no doubt commercial documents, contracts, it may be, with Phenician or Egyptian captains, and they would prove nothing as to the existence in India at that time of what we mean by a written literature. In fact, Nearchus himself affirms what Megasthenes said after him, namely that "the laws of the sophists in India were not written." If, at the same time, the Greek travellers in India speak of mile-stones, and of cattle marked by the Indians with various signs and also with numbers, all this would perfectly agree with what we know from other sources, that though the art of writing may have reached India before the time of Alexander's conquest, its employment for literary purposes cannot date from a much earlier time.

Here then we are brought face to face with a most startling fact. Writing was unknown in India before the fourth century before Christ, and yet we are asked to believe that the Vedic literature in its three well-defined periods, the Mantra, Brâhmana, and Sûtra periods, goes back to at least a thousand years before our era.

Now the Rig-Veda alone, which contains a collection of ten books of hymns addressed to various deities, consists of 1017 (1028) poems, 10,580 verses, and about 153,826 words.[269] How were these poems composed—for they are composed in very perfect metre—and how, after having been composed, were they handed down from 1500 before Christ to 1500 after Christ, the time to which most of our best Sanskrit MSS. belong?

Entirely by memory.[270] This may sound startling, but—what will sound still more startling, and yet is a fact that can easily be ascertained by anybody who doubts it—at the present moment, if every MS. of the Rig-Veda were lost, we should be able to recover the whole of it—from the memory of the Srotriyas in India. These native students learn the Veda by heart, and they learn it from the mouth of their Guru, never from a MS., still less from my printed edition—and after a time they teach it again to their pupils.

I have had such students in my room at Oxford, who not only could repeat these hymns, but who repeated them with the proper accents (for the Vedic Sanskrit has accents like Greek), nay, who, when looking through my printed edition of the Rig-Veda, could point out a misprint without the slightest hesitation.

I can tell you more. There are hardly any various readings in our MSS. of the Rig-Veda, but various schools in India have their own readings of certain passages, and they hand down those readings with great care. So, instead of collating MSS., as we do in Greek and Latin, I have asked some friends of mine to collate those Vedic students, who carry their own Rig-Veda in their memory, and to let me have the various readings from these living authorities.

Here then we are not dealing with theories, but with facts, which anybody may verify. The whole of the Rig-Veda, and a great deal more, still exists at the present moment in the oral tradition of a number of scholars who, if they liked, could write down every letter, and every accent, exactly as we find them in our old MSS. Of course, this learning by heart is carried on under a strict discipline; it is, in fact, considered as a sacred duty. A native friend of mine, himself a very distinguished Vedic scholar, tells me that a boy, who is to be brought up as a student of the Rig-Veda, has to spend about eight years in the house of his teacher. He has to learn ten books: first, the hymns of the Rig-Veda; then a prose treatise on sacrifices, called the Brâhmana; then the so-called Forest-book or Âranyaka; then the rules on domestic ceremonies; and lastly, six treatises on pronunciation, grammar, etymology, metre, astronomy, and ceremonial.