W. M. BALE ON SOME OF THE DISCOID DIATOMS. 23 



specific character. In some instances these " bright points " are 

 merely the optical expression of a local thickening of the silex ; 

 more generally, however, they are true cellules, differing from the 

 rest in their minute size. They are conspicuous in C. perforatus, 

 and they form the principal ground of distinction between that 

 species and C. apiculatus. But in examining a large series of 

 . perforatus var. ceilulosa I find them by no means so constant 

 as to justify the importance attached to them. While in some 

 valves they appear at the origin of all, or nearly all, the shorter 

 rows of areolae, in others they are much sparser, and in a few 

 cases I failed to detect more than four or five on the whole valve. 

 In such cases, and when, as often happens, the central area is 

 obsolete, it is a critical matter indeed to distinguish the valve 

 from C. radiatus, and in passing I may note that the " C. radiatus" 

 of my Holler's Typen-Platte is just one of these valves of 

 C. perforatus var. ceilulosa, with all its bright points complete. 

 C. obscurus may be mentioned as another species in which the 

 bright points, usually present, may be either totally absent or 

 reduced to a very small number. On the other hand the points 

 often occur in species which are normally without them. I have 

 met with instances of this kind in C. aster omphalus, on a narrow 

 unilateral area where the cellules are separate and rounded. In 

 a slide from Cambridge, Barbados, there are numerous valves of 

 C. excavatus, most of which display these minute cellules, and in 

 some valves not only at the origin of the radial series, but 

 profusely interspersed among the large areolae all over the 

 surface, even in places other than the angles of the areolae. And 

 I have a curious valve of Endyctia oceanica, in which these 

 minute cellules form the principal part of the areolation, the 

 ordinary large cells only existing in scattered groups of four or 

 five, surrounded on all sides by the network of small ones. 



I have referred already to the small importance to be attached 

 to mere differences in the size of the areolation, but I would 

 further remark that it must by no means be assumed that only 

 small differences are to be disregarded. Valves of C. concinnus 

 may have only four cellules in 0*01 mm., while others may have 

 as many as twelve, though the valve may be much larger. And 

 I have seen a frustule of C. excentricus in which one valve was 

 twice as finely marked as the other. Such instances show forcibly 

 the futility of distinctions founded on the size of the areolation. 



