_ 6 — 



Towards the periphery these hyphse-like cell-threads are more 

 richly ramified and here pass evenly into the radially placed 

 assimilation-filaments, which are arranged in small groups (B"ig. 3 A). 

 They are dichotomously divided, moniliform , consisting of oval 

 cells which are about 28^ long and of about half that in breadth. 



In the upper end of the cell lies the chromatophore but often 

 it also fills up more or less of the whole cell; it is irregularly star-like, 

 consisting of numerous strands which radiate from the centre of 

 the cell in all directions (Fig. 3i'^); when these strands meet each 

 other at the periphery of the cell they grow together and form a 

 clathrate layer with large and numerous openings within the cell- 

 wall. In the central body of the chromatophore a pyrenoid is 

 present which was densely stained by means of haematoxyline. 

 The structure of the chromatophore seems thus to be in good 

 accordance with that of Nemalion muUifidum according to Wolfe's 

 description (''Annals of Botany", vol. XVIII, 1904, p. GIO) with the 

 exception that he has not found a pyrenoid in his material. And 

 further it also seems to agree well with that of Nemalion luhricum 

 according to the statement of L. Kurssanow in '"Flora", 99. Bd, 

 4tes Heft (1909) p. 311. It is here pointed out in accordance with 

 the opinions of earlier investigators and in contradiction to Wolfe's 

 statement that a pyrenoid is present in the middle of the chro- 

 matophore. 



Very seldom the vegetative cells terminate in a rather long 

 hair (Fig. 3 A). 



Most often the main cell-thread in the middle of the above 

 mentioned groups of assimilation-filaments terminate in a carpo- 

 gonium and further also often 2 — 4 or more side-branches like- 

 wise bear terminal carpogonia (Fig. 3 A). Fig. 3 C shows a branch 

 dichotomously divided, which always has the one branch terminating 

 in a carpogonium. 



The carpogonic branch is composed usually of 4 cells but 

 sometimes only 3 occur, occasionally even 5. This is in agreement 

 with Nemalion muUifidum where the number of cells also is some- 

 what variable. Bo met et Thuret^) give the number to be 3 — 4, 

 Janczewski^) indicates it to be composed of 3 cells and Wolfe 



^) Bornet, E. et G. Thuret, Recherches sur la fecondation des Floridees 

 (Ann. sc. nat. Botanique, V. ser., t. 7, 1867 p. 141). 



-j Janczevvski, E.. Notes sur le developpement du cystocarpe dans les Flori- 

 dees. Mem. de la Soc. Nat. d. Sc. de Cherbourg, vol. XX, p. 109. 



