i6 



H. H. GRAN AND K. YENDO. 



M.-N. Kl. 



but have terminal horns rather resembHng those of the first. Sundry- 

 specific names seem to have been given to the local or seasonal vari- 

 ations, and many vexatious identifications have resulted. 



However the form of the frustules or the terminal horns may vary, 

 there are two well-defined, constant peculiarities to be relied upon as 

 good specific characters, viz, a small verrucosa protuberance of the centre 



Fig. 8. Chaetoceras of/we Lauder. 



a. Part of a chain from a specimen from the Kristiania Fiord. >', 360. 



b. Terminal portion of a broader chain from Yeddo Bay. X 360. 



c. Terminal portion of a narrow chain from Yeddo Bay. X 360. 



b and c are drawn from Cleve and Möller's Diat. No. 307. 



of the terminal frustule, and the form of the resting-spores. So far as 

 we could ascertain in the material accessible to us, this minute protuber- 

 ance is confined to the present species. 



The last four species enumerated above are not satisfactorily defined. 

 Cleve himself had recognized the close similarity of C. Ralfsii Cleve 

 to C. javanicum Cleve, and the latter was combined by him with 



