GrevillEj on New Diatoms. ' 81 



rays, are somewhat irregular in lengtli, and consequently do 

 not form an exact circle. They terminate in one or oc- 

 casionally in two spinous processes, which are evidently 

 analogous to those with which some of the rays in C. armatus 

 and other diatoms are furnished. 



Triceratium. 



The first seven of the following species constitute a very 

 interesting and exceedingly natural little group, and present 

 an excellent illustration of the difficulty of distinguishing 

 between closely allied forms. Without attempting to dog- 

 matise upon the questio vexata of "^ What is a species?^' we 

 may safely venture to figure and describe, wdth benefit to 

 science, such organisms as we have reason to believe exhibit 

 characters by which they may at any time be identified. 

 Such characters are necessarily sometimes minute, but are not 

 thereby of less value. In a systematic work the species 

 about to be described would arrange themselves at once into 

 two sections — the first containing those which have simple 

 (not striated) margins and the central triangular space 

 filled up with radiating lines ; the second those which have 

 striated margins and the central triangular space blank. 

 There is another peculiarity, also, which separates the two 

 sections. In the first the angles of the central triangle are 

 lengthened out until they reach the pseudo-nodule ; in the 

 second the angles are not lengthened out, but each is kept 

 with a short strong line which never reaches the pseudo- 

 nodule, but terminates in a fork more or less connected with 

 other vein-like lines. I have not satisfied myself about the 

 nature of the short line referred to. In T. pulcherrimum 

 (fig. 6) it takes the form of a small spine, distinctly seen 

 within the pseudo-nodule. In T. marginatum it may also be 

 seen, but with some difficulty, through the intervening lower 

 pseudo-nodule. These little spines must be regarded as 

 analogous to the short lines holding a similar relative position 

 to the angles of the inner triangle in the species of the 

 second section. In some instances, especially in T. varie- 

 gatum, I have observed the short line to be slightly raised, 

 suggesting the idea, which is confirmed by the position of the 

 spine in the species of the first section, that this organ 

 belongs properly to the framework of the inner triangle, and 

 that the lines which appear to emanate from it belong to the 

 system of costse or vein-like lines which divide the border of 

 the valve into compartments. 



Triceratium marginatum, Br. — Valve with slightly convex 



