INDIAN ISLANDERS. 213 tioned, are found in the very same ruins where Dewanagari inscriptions are found. At Brambanan I discovered two myself in 1812. They are found engraved both on stone and copper. The character of these inscriptions is an ancient form of the present Javanese, and does not even differ very essentially from it in shape, except that it is rounder. A good deal of it can be read by persons giving their attention to the subject, but there are the best grounds for suspecting the accuracy of the attempts made to render these ancient inscriptions into mo- dern Javanese or the European languages, for no two translations agree. The knowledge of the language is lost in Java, and for faithful translations we have only to look to a better acquaintance with the priests of Bali, among whom it is still the language of re- ligion. The only portion of this character which it can, in our present state of acquaintance with it, be safe to rely upon, is dates, when in written figures^ and perhaps proper names, when these are corrobo- rated by tradition. Trusting to imperfect interpre- tations of the ambiguous and mystical system of no- tation in the matter of dates, which the Javanese have borrowed from the Hindus, several of the Kawi in- scriptions, it is pretended, afford examples of dates which go as far back as the middle of the ninth, nay, in one or two instances, as early as the begin- ning of the sixth centuiy of Salivana. Not one of these is corroborated by a date in legible figures, I
