202 E. M. NELSON ON DIATOM STRUCTURE. 



assume the appearance of a somewhat oval areolation with two 

 tree-like processes in it. 



In Fig. 4 the tree-like expansions at the end of the processes 

 are much diminished, and in some the boughs are altogether want- 

 ing, the processes then taking the form of a pollard willow. 



In Fig. 5 there is no expansion at all on the top of the process. 

 The process is a mere spike projecting into the oval areolation. 



In Fig 6 the process or spike has disappeared, leaving only a 

 small indentation at the margin of the areolation. 



In Fig. 7 you have a clean oval and regular areolation which 

 resembles that on Isihmia nervosa, but without secondary structure. 



Hitherto I have been either unable to rind, or unable to resolve, 

 secondary structure in this diatom. It should, however, be very 

 carefully watched, as it is undoubtedly an interesting point, whether 

 the areolation is first formed and the secondary structure is a sub- 

 sequent growth or not. 



In the first instance we have an illustration of the increase of 

 silex and of the diatom generally, and in the other six the growth, 

 so to speak, of the areolations. 



It may be said that Fig. 7 shows the elementary structure, and 

 Fig. 2 the most highly developed. I have no evidence to bring 

 that that is not the case, but I can only give it as my own opinion 

 that the reverse is what has really taken place. 



