ANALYTICAL KEY. 25 



at some future date have to be split up into smaller 

 genera, but nothing could be more foolish than to do 

 this without first obtaining- a reasonable knowledge of 

 the forms included in these genera. At present our 

 knowledge is insufficient for such a purpose. And 

 when the separation does take place, we do not think 

 that either the relative depth of constriction or the 

 position of the chloroplasts will be the basis upon 

 which they will be subdivided. The arrangement of 

 the chloroplasts is quite unknoAvn in more than half 

 of the species of these genera, and it is a character 

 of such little importance that it has been customary 

 to exclude it from specific descriptions of these 

 Desmids. 



The following is an analytical kev to all the known 



. * . 



genera of Desmids. Five of these, viz. Ancylonema, 



Triploceras, Tchthyocercus, Phymatodocis, and N//r/>/o- 

 t; have not been observed from the British Islands. 



Sub-family I. SACCODERM^E. 



Cell- wall unsegmented, without pores. Point of division of cells 

 not fixed, and unknown previous to the actual division. The young 

 half of the cell is developed obliquely, and its walls are absolutely 

 continuous with the walls of the older half. 



Tribe 1. Gonatozygae. Cells elongate, cylindrical and unci Di- 

 stricted, forming loose filaments. Cell-wall with a differentiated 

 outer layer of which small roughnesses and spines form a part. 



* Chloroplasts axile. 



G-or<atozij(jnn. 



* Chloroplasts parietal and spirally twisted. 



Genicutarid. 



Tribe 2. Spirotaenieae. Cells solitary (except in Anci 

 relatively short and mostly unconstricted. Cell-wall without a 

 differentiated outer layer. There is a periodical growth of the cell 

 until maturity is reached. 



* One chloroplast in each cell. 



t Chloroplast spirally twisted, axile or parietal. 



Spirt;i'niti 



