1889.] MICEOSCOPICAL JOURNAL. 75 



way. It is the side view of the diatom that is almost always uppermost, 

 but the front view of the desmid. It is true the side view of the diatom 

 also often exhibits symmetry, owing to the midrib, but this is on account 

 of an internal, not external, fissure, and it also runs lengthwise, and not 

 across the shorter diameter. In the discoidal forms even this median 

 line is wanting. The fissure in the diatoms, I repeat, is internal, not 

 external. This is a better distinction than the color it seems to me, for 

 the diatoms in the young state are also green. Finally, the diatoms 

 move quickly, the desmids slowly, the former reminding one of little 

 tug-boats by their rapid, jerky motions. Let me add right here that in 

 studying the filamentous desmids one should remember that the real 

 desmid is not the filament but the single cell, by repetition of which 

 the filament is produced, and confine his attention entirely to that, en- 

 larging it if possible until it fills the whole field. If he will do this with 

 two or three species he will be rewarded by a far clearer idea of those 

 forms. The filaments are not real filaments ; the connection of the 

 cells is only by chance as it were. The desmid ^Vz each case is the sin- 

 gle cell, not the chain of cells, as in many other algae. Furthermore it 

 should be noted that the two original semi-cells are at the ends of the 

 filament, the growth taking place in the middle and not at the apex of 

 the filament, as is the case in other families of algae. 



Assuming now that we can tell the desmids at sight, and confining 

 ourselves to them, we take up their special classification. The micro- 

 graphic makes five large groups, namely : i, Closterice ; 2, Cosmarice ; 

 3, JDesmidiece ; 4, Ankistrodesmice ; 5, Pediastrece. But the two lat- 

 ter are now referred by the latest authority (Wolle) to other families of 

 algae and may be struck out of the list. We have then nineteen genera 

 arranged in three divisions, which, for convenience, let us place thus : 

 \,Desmidice; 2, Closterice; 3, Cosmarice. Now there is a striking 

 difference between the first of these groups and the others. In t) e first 

 the cells are in filaments. In the second and third the cells are not in 

 filaments. Six genera then may be recognized at once and they are : 1 , 

 Desmidium ; 2, Bambusina ; 3, Sphcerozosma ; 4, Phymatodocis ; 5, 

 Hyalotheca ; 6, Gonatozygon. In all these the cells are united end to 

 end, and the following points will serve to distinguish them : if the cells 

 ai'e toothed and angular it is Desmidium ; if barrel-shaped, Bam- 

 busina ; if deeply notched it is either Sphcerozosma or Phymatodocis , 

 and the squai-e shape will make it the latter, while the scalloped edge 

 will assign it to the former. If there is hardly any notch it is Hyalo- 

 theca ; if none at all, and the cells are much longer than wide, it is 

 Gonatozygon . 



And so the labor narrows down to only thirteen genera, to wit : 7, 

 Closterium ; 8, Penium ; 9, Mesotcenium ; 10, Spirotcenia ; 1 1 , Doci- 

 dium ; 12, Tetmemorus ; 13, Micrasterias ; 14, Euastrum ; 15, Calo- 

 cylindrus ; i6,Cosmarium ; ij , Arthrodesmus ; 18, Xanthidium ; 19, 

 Stanrastrum. 



But here another marked feature comes to our aid, for we can divide 

 these again by the relative leizgth a?id breadth of the cell. In the 

 Closterice the length is much greater than the breadth. In the Cos- 

 maricz the length is not much greater than the breadth. The Clos- 

 terice are all long forms, resembling in this respect the type from which 

 they are named. 



