90 GENERAL OBSERVATIONS ON THE If on the philological principles here assumed the Javanese form of words is to be considered as approaching the nearest to the speech of the an- cient race whom I have supposed to have dissemi- nated its language and civilization over the other nations and tribes of the Archipelago, to enable us to consider that language consistent with itself, we must look upon it from very early times as a writ- ten language ; for it is a fact fully understood, that oral utterance and the ear are altogether in- adequate to the preservation of the integrity of sounds ; a fact nowhere more amply and satisfac- torily illustrated than among the languages of the Indian islands, where those that have a written cha- racter preserve a surprising consistency, while the more barbarous wanton in the wildest and most fantastic corruptions. Two examples will suffice. In every cultivated, or, which is the same thing, every written language, the moon is invariably WU' Ian or bulan, but when they cease to be written we have the following variations : in the Lombokj ulan ; in the Gorongtalo, ulano ; in the Ceram^ tmlante ; in Bima, uiirah; and in the Menado, thoroughly mangled, leleho7i. In the greater num- ber of the written languages *wulu or bulu is a hair ; in the unwritten we have the following whim- sical corruptions : in the Butung, ivelua ; in Go- rongtalo, xvoJio ; in Minado, wukuk ; in Ceram, whura ; in Ende, abbreviated as well as corrupted
