________________
THE UPANIŞADS it has been concluded, with much probability, that the term was in the beginning applied only to these formulas which contain in a nutshell some important truth of Upanişadic philosophy. As an example of them we may instance Tat tvam asi, 'That thou art,' which teaches the ultimate identity of the individual and the cosmic souls. It was these philosophic formulas alone that were once communicated by teacher to pupil, the communication being preceded or followed by expository discourses. The discourses, it is surmised, assumed in course of time a definite shape though not committed to writing yet, giving rise to the Upanişads as we now have them. To judge from the way in which these texts have grown, they contain not the thoughts of a single teacher, but of a series of teachers, and thus represent a growth in which new ideas have mingled with the old. Such a view explains the heterogeneity sometimes seen in the teaching of even one and the same Upanişad. At a later time, when all the ancient lore of the Hindus was brought together and arranged, the Upanişads in this form were appended to the Brāhmaṇas. The significance of such close association of the Upanişads with the Brāhmaṇas is that when this grouping was effected the two were regarded as equally oldso old that neither of them could be referred to any specific authors. Standing thus at the end of the Veda, the Upanişads came to be known as 'Vedānta' or 'end of the Veda'-much as the Metaphysics of Aristotle owed its designation to its being placed after Physics in his writings. A word which at first only indicated the position of the Upanişads in the collection developed later the significance of the aim or fulfilment of Vedic teaching, it being permissible to use anta in Sanskrit, like its equivalent 'end' in English, in both these senses.
The number of Upanişads that have come down to us is very large-over two hundred being reckoned, but all are not equally old. The great majority of them in fact belong to comparatively recent times and hardly more than a dozen are of the period we are now considering. Even among these classical Upanişads, chronological differences are tracePU. P. 20.
* Ch. Up. VI. viii. 7.