RHODOPHYCE^, OR FLORIDE^ 201 



individual cells, not by their transverse division, or 

 by their median longitudinal division. Portions of 

 the margins of the individual joint-cells may how- 

 ever be cut off by dividing walls, and these then 

 develop into lateral branches. As this process may 

 happen but once, or many times, simultaneously or 

 in succession, varied modes of branching of the cell- 

 filaments are produced. 



Each cell of the filament is provided with a central 

 pit in the cross- wall, and since such pits do not occur 

 in the marginally cut-off cells, their presence or 

 absence enables us to trace in a tissue the genetic 

 connection of the individual cells with their neigh- 

 bours. However, this frequently becomes complicated 

 by the occurrence of secondary pits, formed subse- 

 quently between individual cells of the thallus. The 

 pits are not open pores, permitting free passage of 

 cell-contents such as nuclei and chrornatophores, but 

 they are closed by an exceedingly thin membrane. 

 On both sides of this closing membrane there is a 

 comparatively thick plate of a dense substance, 

 recalling in its characteristics the mucilage-masses 

 of the sieve-tubes of the higher plants. These plates 

 adhere firmly to the closing membrane on both its 

 sides, one in each cell, and they are connected with 

 each other by numerous cords which perforate the 

 closing membrane, mostly round the margin of the 

 pit. The plates stand in direct connection with the 

 protoplasm lining the cell- wall and are, in fact, so 

 coherent with it, that they may be regarded as 

 transformed or rather differentiated protoplasm locally 

 covering the pit. However, it is probable that a thin 



