Professor Joseph Beal Steer (1842-1940) was among the first scholars to acquire some Sinkang manuscripts. These deeds and contracts written in romanized Siraya come from southern Taiwan mostly of the eighteenth century, but they were only recently rediscovered in the Museum of Anthropology of the University of Michigan, whether they had lain since steere’s retirement or death. In February 1999, Professor Henry Wright, curator of archaeology and professor of anthropology at the university, forwarded to me some 20 photocopied pages of the Sinkang manuscripts to determine whether the originals were of any value and therefore worth conserving. My assistant, Miss Chin-wen Chien, and I found that two of them were already contained in Murakami (1933;52, 67), that two were simply lists of personal or place names, that another was written in Chinese, but that the rest were of some interest and had never appeared in print before. These latter are in the Siraya language, they are hand-written documents on heavy hand-made paper, dating somewhere between 1730 and 1818. we have been study these to see if there is anything new to be learned from them. Fortunately we do have a few resources to aid our interpretation of these newly-uncovered Siraya texts: (1) The 21 bilingual Sireaya-Chinese documents contained in Murakmami’s 1933 collection of 101 Sinkang manuscripts: (2) a wordlist and glosses for some of Murakami’s texts prepared by the late Professor Naoyoshi Oagawa; (3) recent linguistic studies of Siraya by Adelaar (1997, 1999, 2000) and Tsuchida (1996, 1999) that provide a good basis for understanding Siraya grammar. Unfortunately, many problems still remain with the texts under consideration. Interpreting the texts was made particularly difficult by the following points: (1) Siraya is an extinct language, and there are no ntaive speakers to consult; (2) many words have never been encountered before and their meanings are unknown; (3) word boundaries are not always evident: (4) spelling is inconsistent: (5) it is difficult to distinguish the letter <d> from <w>, <k> from <h>, <s> from <h>, <i> from <j>, or <a> from <o>; or (6) some of the handwriting is simply illegible. The system of personal names comes out clearly in these documents. The naming system of Siraya is similar to that of Kavalan, a closely related language. As is true for other Formosan languages, Siraya distinguishes between male and female names. Each person has both a given name and surname. The latest Sinakng manuscript was written in 1818; the 80 year-old Siraya women whom Steere consulted in 1873 could only remember a few dozen words, but no sentences. We surmise that the Siraya language became extinct around 1830.