This paper considers Kazuo Mabuchi`s view of Chinese character used as Manyo Kana or Japanese phonetic alphabet in Manyoshu. The result leads us why we have to widen our study to studies and data witten in various languages including Far-Eastern Asia`s old languages. The main conclusions are as follows: (1) Mabuchi insists that there is no Chinese charaters in rime 虞 and 尤 for /nu/, and 農 is used only by Yamanoueno Okura. However, he has fallacies in data interpretation, because there are 10 characters in rime 有 with rising tone and 5 ones in rime 宥 with departing tone, and 農 is also used by Yamabeno Akahito. (2) He emphasizes on Okura`s not differenciating between /nu/ and /no/. But we do not agree with him from the two viewpoints. One is that Okura differenciates between /nu/ written with 奴·農 and /no/ recorded with kun reading or explanatory reading, and the other is that he did not need to make use of Manyo Kana only for one word with /no/ used just only for one time. (3) Mabuchi lays stress on a theory that Chinese literary men, not Korean peninsula ones, took part in recording Japanese works. This theory is avaluable only for On Kanas or phonetic alphabets, but is not from the standpoint of literary stylistics in which Old Japan and Baekje, not China, had common.