OF THE GENERA NAVICULA AND CiTMBELLA. 411 



antarctica is found in enormous numbers in slides from the bottom 

 of the Antarctic Ocean at a deptli of 2,000 fathoms. How did 

 it get there ? The explanation was given by my late much- 

 lamented friend, Sir John Murray. It is this : Fragillaria antarc- 

 tica, a freshwater diatom,lives in huge quantities on the Antarctic 

 lands. It is carried away on floes of ice to the open sea, and 

 when the ice melts or breaks up it floats in the ocean as a pelagic 

 plankton, and when it dies, its skeleton falls down to the sea bed, 

 where it goes to form a Diatom Ooze, and where it was found in 

 abundance by the soundings of H.M.S. Challenger. As a general 

 rule, it is in structure, size and outline that the systematist has 

 chiefly to guide himself for the determination of species and 

 varieties. It is therefore also in structure, size and outline that 

 we have to seek for some guiding principle as to synthetic 

 integration. 



The main line upon which it would be perhaps possible to look 

 out for synthetic grouping would seem to lie in a close study of 

 undoubtedly intermediate forms, and in an effort to detect affini- 

 ties between them, seeking a disclosure in these relationships of 

 what may be called " descendancy " or "lineage" evolved 

 through time and ages. Fossil and living forms are to be closely 

 compared and connected with one another. We must try to 

 guide our systematic studies by " genealogical descent." 



In Volume I of the standard Synopsis of Naviculoicl Diatoms, 

 by Cleve (p. 157), is to be found a hint how to approach this 

 problem. There the author seeks to establish two lines of genea- 

 logical descent as regards several species of the genus Cymbella. 

 His hint seemed to me most promising and fruitful, and I have 

 tried to pursue it farther, taking up not the genus Cymbella as 

 a whole, but only some of its tribes. I have tried to establish 

 connecting links within certain limits, showing how intermediate 

 forms seem to pass one into another as regards structure, outline 

 and size. 



In his book Professor Cleve traces relationships between species 

 of Cymbella along two genealogic lines, starting in one case from 

 Navicula dicepJiala, and in the other from a marine Navicula, 

 Navicula Bulnheimii. I have taken up his second line and tried 

 to trace a passage from and connection betvv^een about a dozen 

 species, sketches and microphotographs of which accompany this 

 paper. I do not, however, include in my series the Navicula 



