21 



family characters, as is the case with the I'altiiea 1 ,^ (Jyclanthacea. 1 . I'aud- 

 anaceae, Musacea3, 2) etc., cases quite analogous to the behavior of intraxylar 

 phloems in some dicotyledons. 3) So we are justified to a certain 

 extent to look upon the peculiar root-structure as one of the must 

 important anatomical characters indicating the special position of the 

 Bambusea3 among the Graminea3. Moreover we have confirmed the presence 

 of a complete uninterrupted pericambium in the root of ever}- species 

 examined." By all means, the root must be regarded as preserving most 

 of the important phyletic characters of the Bambusea?. Then we may a>k 

 whether the root-structure can be applied as a standard of classification 

 among the bamboo-tribe itself? We may answer this question in a positive 

 sense. In fact we are able to distinguish two quite different modes of 

 construction of the roots among the bamboo-plants examined by us : one 

 presented by the species belonging to the genuine Bambusa and to Deivho- 

 calamus, and the other by Arimdinaria and Phyllostachys. It will be 

 convenient to consider the cortex and axial cylinder separately. 



The cortex of the former type is briefly characterised by the persisting 

 epidermis, lignified but never thickened subepidermal cell-la3 T er and a few 

 layers of peripheral sclerenchymatic cells, which are sharply marked off by 

 the thin-walled cortical parenchyma. The latter is divided into two portions, 

 the outer of which consists of large polygonal cells, and the inner of small 

 cells, regularly arranged in radial and concentric rows. The thickness of 

 the inner layer is always a few times greater than that of the outer 

 one. Lastly the endodermic cells are uniformly thickened in their w f alls, 

 bringing about the formation of the so-called O-sheath. The cells of one 

 or two innermost layers of the cortical parenchyma bears on their inner walls 

 irregularly shaped protuberances, raising themselves into the cell-lumina and 

 consisting of almost pure cellulose. This structure may perhaps be regarded 

 as the mechanical support of the cells against the radial pressure. Wester- 

 maier^ has attributed a similar function to the conical processes in the 



1) Reinhardt, Das leitende Gewebe einiger anomal gebauten Monocotylenwurzeln. 

 Jahrb. f. wiss. Bot. Bd. XVI, p. 336. 



Gil lain, Anatomie d. Pal men- und Paudinaceenwurzeln. Bot. Central)). Bd. LXXXVUI. 

 1000, No. 37. 



2) Ross, loc. cit. 



3) Solereder, System atische Anatomie d. Dicotyledonen, p. 97J. 



4) According to Van Tieghem (Ann. d. Sc. nat., 5 e Serie t. XIII), this is not the case 

 with most of other graminaceous plants. 



5) Wester maier, Ueber Bau und Function des pfl, llautsystems. Jahrb. £ \\\±*. Bot. 

 Bd. XIV, p. 04. 



