ing to his mind the original Tonga in which it was spoken. Sometime afterwards I found the first, and was much pleased to discover so little difference between them, that they appeared almost like copies, which sufficiently evinced the correctness with which he remembered the original Tonga, and at the same time furnished an instance of the characteristic uniformity of his expression in his own language. Two or three months afterwards I reminded him of the propriety of writing down in the Tonga language all that he knew of their popular tales, speeches, songs, &c. while they were fresh in his memory; he did so, and at a subsequent period when the dictionary of the language was in a state of forwardness, I translated them literally with his occasional assistance, and had a new proof of the correctness, as to sense, of what he had before given me of Finow's speech, the English copies of which I had all along kept in my possession. As a retentive memory was a quality particularly essential to Mr. Mariner, under the circumstances in which he was placed, it is worth mentioning, that even when I be-
