428 PROTOPHYTA 



Sub-class 1. Nostochinese. 



It is convenient to group together under this head those families of 

 Cyanophyceae, the Nostocacetz, Rivulariacea, Scytonemacea, and Osdl- 

 lariacece, which are characterised by the individual consisting of a cel- 

 lular or pseudocellular filament, reproduced by motile hormogones, and, 

 in some of the families, forming heterocysts. 



i Borzi, who has closely examined the structure and life-history of the 

 Nostochineae, considers that they display only a rudimentary cell-structure. 

 The cell-wall is extremely thin, is inseparable from the contained proto- 

 plasm, and not sharply differentiated from it in its chemical reactions. 

 The cells do not contain true starch, its place being taken, in spores 

 and in the constituent elements of the hormogones, by a substance to 

 which he gives the name cyanophycin of a granular nature, apparently 

 formed from the substance of the cell-wall, possibly a carbohydrate, and 

 identical with the gelatinous substance of which the sheath which 

 encloses the filaments is composed. The only indication of a nucleus 

 is a local fragmentation and concentration of proteinaceous matter. 

 Borzi states that in all the families of Nostochineae the cells of a fila- 

 ment are in communication with one another by pores in their transverse 

 walls, through which pass strings of protoplasm, or of a substance allied 

 to cellulose. When heterocysts are about to be formed, this intercellular 

 communication is suspended, and the pores are closed by a thickening 

 of the cyanophycin or other cell-contents. In the Nostocaceae, Scyto- 

 nemaceae, and Rivulariaceae, the filaments are enclosed in a gelatinous 

 sheath, which is much thinner in the Oscillariaceae, and altogether want- 

 ing in Borzia (Cohn). This mucilaginous sheath may be continuous or 

 septated, and may be composed of a single layer or of several ; in the 

 latter case it not unfrequently assumes the form of a funnel at the end 

 of the filament. Gomont differs to a certain extent from these conclu- 

 sions, previous observers having, according to him, confounded the 

 envelope proper of the cell with the mucilaginous sheath of the trichome. 

 By the action of chromic acid he was able to separate the one from the 

 other, and finds the very thin perfectly transparent envelope proper of 

 the cell to possess properties intermediate between those of "fungus- 

 cellulose and of vegetable cutin. Bornet and Flahault call the string 

 of cells or pseudocysts the trichome^ the trichome with its envelope the 

 filament. 



When in a purely vegetative condition, all the cells of a filament are 

 alike in form and size, and multiply by repeated bipartition. Scott and 

 Wille have been able to demonstrate the presence of a nucleus in 



