BY G. I. PLAYFAIR. 



129 



cells. In the projecting sheath of a broken filament in the above, 

 they were plainly visible. Tn large forms, they develop into true 

 septa across the sheath distinct from the cells themselves. Cf. 

 Sydney Water-Supply, PI. Ivi., f.l7. 



Genus Plectonema Thuret. 



Plectonema Nostocorum Bornet. 



Fila. diam. U; cell. alt. 2-4 /z. 



Lismore (30). 



Colour very pale blue, not yellow-green as in Gomont, but in 

 this country yellow-green forms of generally pale blue species are 

 not at all uncommon. I consider that the two colours are prac- 

 tically interchangeable. Filaments very little branched, sheath 

 distinctly observed. In the case of trichomes which had been 

 forced from the sheath, I noted, on more than one occasion, a very 

 interesting phenomenon. The terminal cells, one by one, broke 

 away, and after a short pause became motile. In some instances 

 a portion of the trichome, containing several cells, became dis- 

 connected first, and then broke up. The cells thus set free do 

 not move like Vibrio or Bacillus, but fly through the water, 

 spinning rapidly round their shorter axis. By examination of 

 the base of the aeruginous stratum of Phorniidiuin corium, with 

 which the Plectonema was intermingled, it could easily be seen 

 that the filaments of the latter were originally sessile. Hence 

 these motile cells evidently settle down and form new filaments. 



Genus OsciLf.ATORiA Vaucher. 

 Sheath always present, but generally very delicate. In all the 

 species of Oscillatoria mentioned below, I clearly demonstrated 

 the presence of a sheath — repeatedly in the case of 0. splendida 

 and its various forms. If a fragment of the stratum (or even a 

 drop of water in which filaments are present in any quantity) be 

 thoroughly mashed up on a glass-slip with the thumbnail, the 

 sheath can generally be detected, either as a minute tag at the 

 angles of a broken filament, or as a faint projecting tube, or as a 

 connecting: link between two trichomes on the two halves of a 

 bent filament. It seems doubtful, therefore, whether there is 



