description and illustrations (Kuetz. Sp. Alg. p. 432 and Tav 

 Phyc. VII, t. 88, f. II.) and Thaxter's recent investigations (R. 

 Thaxter Notes on the Str. and Repr. of Compsopogon in Dot. Gaz. 

 Vol. XXIX, P. 259-267, pi. XXI, 1900). Kuetzing illustrates 3 

 cell-layered cross-section of the frond without central cell. He does 

 not clearly mention that the frond is hollow, only stating that 

 ,'articulis in ramis primariis obsoletis." Whether this statement 

 means that the frond is hollow or not is not clear. But Thaxter 

 describes and illustrates that " the corticating cells are separated 



into two well-defined layers which may be increased to 3 



or even 4 in older axes like that shown in Fig. 13, a portion of 

 the axial cell being indicated at x" (Thaxter I.e. p. 262). Consi- 

 dering that Thaxter's material is C. coentleus, it is evident that it 

 has thickly corticated frond with a central cell persisting. Such a 

 structure is not the case in the present plant ; in ours the cortical 

 cells are not more than two layers in the branches thicker than 

 400 n and in the branches of such a thickness the central cells are 

 already disappeared. De Toni's remark saying "cellulae corticales 

 e superficie visa? sunt elongatae, irregulares, acute contiguae " and 

 Kuetzing's illustration representing regularly arranged elongated 

 cortical cells are disproved by Thaxter's description and illustra- 

 tions (i.e. " the corticating cells are irregularly polygonal "). 



Thus, the present plant differs from C. coentleus, as far as I 

 can judge from the literature concerned by the brittle substance, 

 by disappearing of the central cells in older portions and by fewer 

 cortical layers as well as by slenderer habit of the frond. 



PL. CXXXII. Fig. 1 : fronds of Compsopogon Oishii Okam. 

 sp. n. on the leaf of Valisneria spiralis in nat. state and size. 

 Fig. 2: small frond, . fig. 3: portion of fig. 2, marked a' 



