312 Myxophycece 



through the polar extremities of the cell or through the transverse 

 cell-walls. This pore is best seen in the young branches of 

 Stigonema ocellatum, and I have previously pointed out 1 that it is 

 particularly conspicuous if the plants have first been dried and 

 subsequently soaked in water. Fritsch 2 has also described this 

 protoplasmic continuity between the cells of Anabcena. 



In the families Nostocaceae, Scytonemaceae, Stigonemaceae and 

 Rivulariaceae certain special cells, known as heterocysts, occur at 

 intervals along the filaments. They are sparsely scattered between 

 the ordinary vegetative cells, and owing to the absence of pig- 

 mented material they present a very pellucid appearance. Their 

 walls, which are composed of cellulose, are generally thickened and 

 of a pale yellow-brown or yellow-green colour. The heterocysts are 

 frequently larger than the vegetative cells and their walls 

 commonly possess slight polar thickenings. Each polar thickening 

 surrounds the apical pore through the cell-wall, this pore being 

 closed by a minute plate in old heterocysts. 



Heterocysts are developed from ordinary vegetative cells, 

 generally singly from any cell of the filament, or they may be 

 formed from the two cells contiguous to an existing intercalary 

 heterocyst. Under natural conditions heterocysts are almost 

 invariably solitary in all genera except Tolypoihrix and Calothrix, 

 but under unfavourable conditions and in cultures they may 

 become seriate. Both Brandt and Fritsch have described the 

 occurrence of an intercellular substance excreted during the 

 formation of heterocysts. This substance, however, can be 

 frequently observed remote from the heterocysts in Scytonema, 

 and it is also excreted by the cells of certain species of the Oscilla- 

 toriaceae, a family of blue-green Algae in which heterocysts do not 

 exist. 



The function of heterocysts is not thoroughly understood. They 

 have been thought to serve as limitations to the length of the 

 filaments, and they are at times undoubtedly connected with the 

 breaking of the filaments. In normal plants of the genus Anabcena 

 the filaments break readily at all points, and this fracture cannot 

 therefore be controlled by the heterocysts. Neither does the 

 structure of a filament of Stigonema support this view, although 

 in this genus heterocysts limit the hormogones. Hieronymus, 



1 West & G. S. West, 'Welw. Afric. Freshw. Alg.,' Journ. Bot. June 1897, p. 242. 



2 Fritsch in New Phytologist, iii, April 1904, p. 93. 



