F r i t s c li , Studies on Cyanopliyceae. 209 



t'roin tJiat of tlie related unicellular fornis in any essential 

 respects. It is of e.special interest because it is one of tlie uni- 

 cellular Cyanoplnjceae^ from whieli the higher ülanientous forms 

 were in all probability derived. In this genus the cells are 

 capable of dividing in three directions at right angles to one 

 another, and if we -suppose, as seems most likely. that the 

 ancestor of such unicellular forms was capable of dividing along 

 all directions (a parallel for which is found in Ajiltanocap.sa) 

 Glorocap.sa represents a fairly primitive type, — more primitive 

 probably tlian a form like Mcrismopedia^ which is only capable 

 of division along two dii'ections. If we imagine the capacity 

 for division along one of these planes to become lost (Äphano- 

 tkece, Gheothcce), we reach the conditions, necessary for the 

 development of the hlamentous stage and with that pass on to 

 the higher forms: in correspondence with this tendency for 

 division solely along one direction we find that the spherical 

 cell of a Gloeocapsa or Merismopedia acquires a more or less 

 pronounced cylindrical form (in Oloeothece, etc.j. In Glomtltece 

 and Aphauothece however the hlamentous tendency is still ojiposed 

 by the consistency of the elli^Dsoidal or spherical mucilage enve- 

 lopes, which effectively prevent the formation of a row of cells 

 by this uniaxial division. The phenomenon, which we have 

 before us here, is well described by Xägeli (49, p. 57) in the 

 following passage: „In der Mutterblase (i. e. the enveloping layer 

 of mucilage ) liegen die beiden Tochterzellen nach der Teilung 

 hintereinander. Sie dehnen sich dann in die Länge; ist die 

 Blase weich, so folgt sie anfänglich dem Drucke, reißt aber, 

 wenn die Tochterzellen ihre eigenen Blasen bilden (fig. 2 c). 

 Besitzt dagegen die Mutterblase zweier Individuen nicht so viel 

 Elastizität, um dem Drucke der Ausdehnung dieser letzteren 

 folgen zu können, so werden dieselben mechanisch von der ur- 

 sprünglichen Richtung abgelenkt (hg. 2b; lig. Bb, c). Mit dem 

 weiteren AVachstum und der Bildung der eigenen Hüllmembranen 

 weichen sie zuletzt so sehr von der anfänglichen Stellung ab, 

 daß sie mehr oder weniger parallel neben einander liegen 

 (fig. 3d, e)." — The same applies to Aplxuiothece. — In the 

 genus Synecliococcuü we have a stage, which takes us onwards 

 a little way towards the filamentous series. As in Äpf/anof/iecc 

 und G/ocofhecc we have cylindrical cells, dividing in one direction 

 only, but the colonies are only surrounded by a loose difHuent 

 mucilage. As a rule division apparently follows the lines of 

 that in Gloeocapsa^ G/ocof/iccr etc., mucilage being formed abun- 

 dantly botween the jn-oducts of division and the cell-sheath no 

 doubt forming a complete envelope round each individual cell; 

 occasionally however the intercellular mucilage is less developed 

 and the products of division form short rows of cells. It remains 

 to be Seen whether the cell-sheath is then individual to each 

 cell or coherent. I have not l)een able to obtain Si/iirchorornis 

 for this investigation, but judging from memoj'v and the publishcd 

 hgures, it seems probable that the former will be the case. Tho 



