G6 Mr. V. Mcieschkowsky on the 



proved by tlic variability of tliis character in the Naviou- 

 ioidca." as well as in the Nitzscliioiclca? {yit:scJn'a awc/NA/r/v, 

 for example, has a central slit). All the diatoms which 

 possess a slit, and are consequently endowed with movemetit, 

 thus form one natural group, and those which are destitute 

 of slit and of movement form another, 



I therefore divide the diatoms into two great groups — 

 Mobiles and Immobiles. 



I. MODILES. — This group contains two very natural 

 branches or sections. The first is represented by the 

 EaphideaB in its generally accepted sense ; it is characterized 

 by the presence of nodules and the usually central position of 

 the raphe, and is well subdivided in its turn by Perng.\llo 

 into three subgroups : Ileteroidea*, Navieuloidea^, and Tropid- 

 oidcai (comprising in this latter oidy the genera Auricula^ 

 Atnph iprera, iu\d Tr-opi'dotiets*). The second section com- 

 prises the genera Sitcschia, Sitzschiella, IJantcschiu, Cy- 

 lindrotheca^ E2nt}ievx{a, Bhopalodia, Denticida, Surirella, 

 Ct/nuitop/euray Stuiopterotia, Podoci/stis, llydrosilicoit, and 

 Campylodiscus. It is characterized by the absence of nodules 

 (with the exception, however, of Bhopalodia) and the lateral 

 position of the slit (with the exception oi Xit::;>chia parad)X\i, 

 y. anguhsii, and several others), which is always raised on 

 a carina. In accordance with this last character it may be 

 named Carinatae (or Carinophoreie) ; it is divided into the 

 KilZ;chioidta3 and the SurirelloideJS. 



II. Immobiles. — The immobile diatoms can also be sub- 

 divided into two subgroups : the first, comprising the ohl 

 group of Pseudoraphidea.' with the Carinata' exeUuled, is 

 characterized by the elongated or bacillar form of the 

 valves, by the presence of a pseudoraphe (with many excep- 

 tions, however), and the absence of appendages, spines, 

 &c.t To it may be given the name Bacilloideae, divided 

 into Fragilarioideje and Tabellarioidea^, according to the 

 presence or absence of internal partitions. The second sub- 

 group comprises the old group Auaraphideae, which name 

 may be retained for it. It is chaiaeterized by the more or 

 less rounded or angular form of the valve and the presence 

 of processes, awns, or spines. There is another character 



• M. Peragallo (Dint, de Fr. p. 161) has suggested uuitiug tln-se genera 

 with the Nitzschioideai and Surirelloidetv in a single group. This course 

 does not seem to me to be good ; the gnnip thus formed would be com- 

 posed of too heterogeneous forms. A Tropidotuis has certainly more 

 relationship with a Aai-utiiu than with a Xitz^ochiti. The division which 

 I propuee here apiK.'ars to me much more natural. 



+ A \*iv ,-ujalI number of TMlitllarioidea' }x>^so^^ them, however. 



