POLYNESIAN LANGUAGES. 107 to form some rational opinions respecting the nature, character, and extent of the connection between the distant Indian isles and the country of the Hindus. First, The Sanskrit exists in a state of as great pu- rity as the articulation and alphabets of the Archi- pelago would admit, nearly unmixed with any mo- dern dialect of which it is a part, and apparently in a state of original purity. Secondly, It is most pure in the more cultivated dialects. Thirdly, It is abundant in the direct proportion of the im- provement of each language. Fourthly, It is pure and abundant as each dialect of the same tongue is improved, and rare and corrupt as the language is common and popular. Fifthly, Where corruptions of Sanskrit words exist, the same corruption per- vades all the different dialects. It is only from a sober examination of the internal evidence which these prominent facts afford, assisted by the evi- dence which the relics of ancient art and religion lend us, that we can expect to determine the man- ner in which the Polynesian dialects received their infusion of Sanskrit ; for we cannot trust to tradi- tion, and the barbarians, on both sides of the wa- ter, have no historical record of this or any other remote transaction. The singular facts now stated respecting the condition in which Sanskrit exists in the languages of the Indian islands, lead me to imagine that the language was not introduced by conquest, but pr«-
