Fritsch, Studies on Cyanopliyceae. / 211 



It sliuuld be noted tliat (as exempliiied by Aplianofhcct' and 

 Glo('othece) a cliange in tlie character of the external mucilage 

 was a necessity for tlie formation of the filament. In Oscillaria 

 the external Investment has been discarded altogether and the 

 necessarv rig-iditv is obtained bv the coherence of the cell-sheath: 

 in othei' forms however (Lynyhija) this was not suflicient and 

 a more or less consistent outer sheath was further developed, 

 The heterocystous forms studied have all retained an oiiter 

 sheath and perhaps it is owing to this that coherence of the 

 cell-sheath is not so marked in any of these forms i). Tlu* 

 scheine on p. -Iß is meant to show the way, in which tlie higher 

 lilamentous Cycuwpinjceac arose from the nnicellular forms, as 

 far as can be gathered from present day forms: but ahhongh 

 it indicates relationships it mnst not be regarded as a phylo- 

 genetic series. As will be seen two main series of forms are to 

 be distingiiished : the series Oloeocapsd — Gloeothece — Ana- 

 haena^ in which the cell-sheath can always be recognised as 

 indi^^dllal to each cell, owing to its very marked constriction, 

 even wlien coherent : and the series OsciUar'ia — Lyughya^ which 

 probably developed from a form like Syiiechococeus^ and i^ 

 characterised by the uniform cell-sheatli around the whole lila- 

 ment. Lytiyhya represents a return to the old conditions, in as 

 much as it possesses a well-marked external sheath, which is 

 lacking in Osrillaria. 



Summary. 



It may be well to briefiy summarise the conclusions of the 

 present paper: 



(I) Each cell of the sporogenous filament land probably also 

 of mature vegetative filaments) of an Aiiabaeiia has two enve- 

 lopes, — an inner investment, which completely encloses the 

 jirotoplast, and outside this a special cylindrical sheath, which 

 has been designated the cell-sheath: Avhen division of the ceUs 

 takes place this cell-sheath is simply split into two fresh sheaths 

 by the develo})nient of an intercellular septuni. 



(II) The inner investment, which is possil)ly the only one 

 in young stages, is regarded as a modified plasmic membrane 

 of a viscous, gelatinous nature; the cell-sheath is probably a 

 modified innermost laver of the external mucilao-inous sheath 

 and unlike the inner investment is dissolved l)y chromic acid, 

 except in the almost mature spore. 



ij As already nientioned the hi^liei- heterocystous fonns have not beeil 

 fully exaiiiiued. but it iiiay be well to point out tliat tlie heterocystous 

 forms in Kirchner 's Scijtmiemataceae (98, p. 78, fig. 57 C and D for in- 

 stauce) ajjpear to have the sanie moniliforni stnictiu-e of the cell-sheath of 

 the filaments inside the external sheath as Tohjpotitrix. Plafonenia (loc. cit. 

 11^-. 57 A), which is devoid of heterocvsts on the other hand appears to 

 exhibit a structiire like that of Lyiujhya. All these forms will be treated 

 of snbsequently in greater detail. 



