DESMIDIACE^. 



195 



both on account of the remarkable manner in which it 

 takes place, and because it unfolds the nature of the pro- 

 cess in other families, and furnishes a valuable addition to 

 our knowledge of their structure and physiology. 



Fig. 108. 



5, Micrasterias dentlculata. 6, Didymoprium Borreri. 

 8, Staurastrum hirsutum. 



7, Cosmariitm Ralfsii. 



The compressed and deeply-constricted cells offiiiastrum 

 offer most favourable opportunities for ascertaining the 

 manner of their division ; for although the frond is really 

 a single cell, yet this cell in all its stages appears like two, 

 the segments being always distinct, even from the com- 

 mencement. As the connecting portion is so small, and 

 necessarily produces the new segments, which cannot arise 

 from a broader base than its opening, these are at first 

 very minute ; though they rapidly increase in size. The 

 segments are separated by the elongation of the connecting 

 tube, which is converted into two roundish hyaline lobules. 

 These lobules increase in size, acquire colour, and gradually 

 put on the appearance of the old portions. Of course, as 

 they increase, the original segments are pushed further 

 asunder, and at length are disconnected, each taking with 

 it a new segment to supply the place of that from which 

 it has separated. 



It is curious to trace the progressive development of 

 the new portions. At first they are devoid of colour, and 

 o 2 



