This paper examines the notations of two pieces from the tenth-century flute-score, Hakuga no fue no fu. Both are from the Togaku repertory of Japanese Court Music: Shoenrakuis a short piece in one movement; Banshiki Sangun, a suite in 26 movements. Neither piece has survived to the present day; indeed, both are known only from Hakuga no fue no fu. Six different systems of notation may be distinguished in Hakuga no fue nof u. The notations of Shoenrakuan d BanshikiS angunr epresent one such system. The inclusion of a number of different notational styles appears to result from Hakuga's (Minamoto no Hiromasa) stated method of compilation, namely, that he copied pieces from a number of earlier scores, preserving the original notational styles.<BR>Through analysis of the notations of Shoenrakua nd BanshikiS angun, a nd examination of textual evidence from the Heian period, values have been deduced for notational symbols. First, the question of the relationship between tablature signs and pitch is considered; secondly, meanings are determined for metrical and rhythmical elements of the notation. By applying, in reading the notations, the values thus deduced, transcriptions of ShOenrakaun d BanshikiS angunh ave been made. The melodiest hus yielded, are not only strikingly attractive, but, in the case of BanshikSi angun, are perhaps the most interesting, from a metrical point of view, of all Togaku tunes. Five different metres are used. In a majority of movements, one metre predominates, but in some, two metres are strongly in evidence. Bar lengths tend to contract as the suite continues, from 12-beat measures in the first three movements, to 4-beat measures in the last. Other unusual metrical features include, the elongation, as a type of metrical cadence, of the final measure in some movements, and the introduction of new metres at the end of movements, before they are taken up in the next movement. Notations of movement-types, previously known only from textual descriptions in Chinese sources of the Tang dynasty, are included in Banshiki Sangun.<BR>Finally, the question of tempo and melodic structure is discussed. Work published by Picken and by Condit suggests that in earlier times the tempo of Togaku was considerably quicker, and the melodic structure quite different, from that of present Togaku. Evidence from Banshiki Sangun, that supports this view, is presented.