Abstract: The 33 kinds of bamboos (listed in Table 1) used in this experiment are those generally grown in the south-eastern and south-western parts of this country. For searching new fibre source, the fibre dimension, fibre volume and basic density were studied and arranged in groups according to their size. Such features as fibre length, cell wall thickness and the basic density may play the part in the assessment of bamboo-species suitable for pulping. The results are summarized as follow: 1.The average fibre lengths of the 33 species ranged from 1.70 to 3.19 mm (average 2.52 mm) and the average width ranged from 9.65 to 16.87μ (average 13.2μ) Though the average fibre length is situated between the coniferous (average 3—4mm) and broadleaved woods (average 1.4 mm), it appears mostly to approximate the coniferous woods. To compare with the fibre width of the coniferous woods (average 35μ), the average fibre width of bamboos is always low, even lower than the broadleaved woods (average 24μ). Therefore the bamboo-fibres are apparently slender than that of both coniferous and broadleaved woods. It can be revealed by the L/D ratio (the length to the width). Of these the bamboo-fibres usually stand between 115—290, mostly over 150; while that of the coniferous woods could not yet reach 100, especially that of the broadleaved woods (Table 2). To raise the density of paper the mixture ratio of long and short fibres is more important for the manufacture of chemical paper pulp. Data of the frequency and distribution of fibre-lengths has not yet been available on the pulp industry. From this statistical data of 33 bamboo-fibres the curves of the maximal frequency and distribution of fibre-lengths are shown in Table 3 and Fig. I, 1—33. 2.The 2W/1 ratio should be of particular interest to the paper making industry not only because of its desired strength of paper but also because of its different application. This ratio among 33 bamboos was found to be greater than unity ranging from 1.22 to 6.18 (Table 2). This might be a guide in practice for selecting the bamboosort and determinating the available blending with other than bamboo-fibres. 3.The proportion of fibres, vessels and parenchymas of 33 species was given in Table 4. For an economic returns, the volume of the bamboo should be at least over 30% and probably 50% of fibrous tissue even if the cell dimension and cell wall thickness occur in highly desirable characteristics. For the most part the bamboos appear to have a high ratio of parenchymas to fibres and vessels. 4.The basic density of bamboos was found to vary in different species showing in Table 2. This will be correlated with cell wall thickness as shown in Fig. 4. It also reflects more or less changes in fibre contents. The determination of basic density should be considered as one way for assessing wood quality for pulping. Although the increase in yield has been associated with increase in density, however, some species with lower basic density usually give pulp with higher over-all strength. The yield is only one factor but the behaviour of the fibres is much more important on the suitability of new species for pulping purposes. According to the results of the above mentioned examination 33 species of bamboos, as compared with one another as to thdr mean fibre length, cell wall thickness and basic density, can be classified in 4 groups: namely, 1st group No. 1. (Bambusa rigida) No. 2. (Bambusa sinospinosa) No. 3. (Schizostachyum pseudolima) No. 4. (Sinocalamus affinis) No. 5. (Schizostachyum hainanense) No. 6. (Bambusa lapidea) No. 7. (Phyllostachys bambusoides) No. 8. (Dendrocalamus strictus) No. 9. (Sinocalamus oldhami) No. 10. (Schizostachyum funghomii) No. 11. (Chimonobambusa utilis) No. 12. (Phyllostachys pubescens) No. 13. (Lingnania chungii). 2nd group No. 14. (Sinocalamus minor) No. 15. (Bambusa pervariabilis) No. 16. (Phyllostachys congesta) No. 17. (Semiarundinaria henryi) No. 18. (Lingnania cerosissima) No. 19. (Chimonobambusa quadrangularis) No. 20. (Phyllostachys angusta) No. 21. (Phyllostachys nigra var. henonis). 3rd group No. 22. (Sinocalamus latiflorus) No. 23. (Chimonobambusa mormurea) No. 24. (Bambusa textilis) No. 25. (Sinarundinaria nitida). 4th group No. 26. (Bambusa multiplex) No. 27. (Shibatea chinensis) No. 28. (Phyllostachys virdi-glaucescens) No. 29. (Phyllostachys anguta c. V. solidstem) No. 30. (Pleioblastus amarus) No. 31. (Pseudosasa amabilis) No. 32. (Sinocalamus distegius) No. 33. (Pseudosasa japonica). A total of 33 species representing 12 genera were reported. Among them representatives of 1st and 2nd groups appear to be the most promising in view of yielding high quality pulp, that is also said to be an important source of fibres. After all both the morphological studies and the determination of basic density should be considered to be most significant in indicating those species for the preparation of pulps in high yield and of good strength properties, however, they should not be regarded as the pulping and pulp evaluation studies.