may we not venture to say that this is evidence, so far as it goes, that the building up of the poem did not commence till at least a century later than the time of the Mahābhāṣya? Such an orthodox work as the Mahābhārata came eventually to be, could not have departed from the much respected orthodox rules, if time had not then made the rules almost obsolete. What is true of the Mahābhārata, is true also in respect of the Rāmāyaṇa, as we now have it. To my humble thinking, the latter shows signs of lateness to a great extent.
Of words formed by imitating natural sounds and underivable from the fixed stock of Sanskrit roots, kolāhala, kilikilā and the like are only found in the eighteen lengthy Parvas of the Mahābhārata. Halahalā, Gadgada and Humbhā (lowing of the cow) are found used in the Rāmāyaṇa; in the 23rd Chapter of the Araṇyakāṇḍa, we find exact sounds of birds used as Sanskrit words. "Chīchīkūchītī vāśyanto babhūbustatra sārikā," would have defiled the purity of language in the second century B.C. This very "chīchīkū" we find also in the Harivamśa. These words, as well as the words Khaṭ-khaṭ, Ṭhan-Ṭhan, Jhan-Jhan, and Raṇaraṇaka of still later literature, have been called Deśī words (words of Provincial origin) by Hemchandra. It is known to all that Hemchandra's Deśī Nāmamāla contains such words as were considered not to have been derived from Sanskrit roots. It is true that Hemchandra has declared such a few words to be Deśī, as are really apabhranśa words, but I must also note that some ingenious attempts have been made at a forced affiliation of many real Deśī words to some recognised roots; I do not however consider it worth while to offer any criticism on this point.
When literature grew, the writers felt the want of words, and were forced to borrow many words from the
