W. M. BALE ON SOME OF THE DISCOID DIATOMS. 19 



have expressed approval of this proposal, but none have adopted 

 it, nor are any likely to do so. 



In surveying the various characters by which species may be 

 defined, the outline will naturally be the first to be considered. 

 This in the Coscinodisci, however, is of little assistance, as, except 

 in a few aberrant species, the circular form prevails. Passing to 

 the surface contour, we have a character which has been utilised 

 by Grunow, Rattray, and others, but by no means so fully as 

 might be. Thus neither of these observers, in differentiating 

 C. asteromphalus from C. centralis, refers to the fact that the 

 former has usually the centre depressed, while the latter is 

 convex throughout. In several cases the absence of information 

 on this point in Rattray's descriptions just renders the diagnosis 

 doubtful. And this is the more important from the fact that 

 even a good figure does not always bring out this special point. 

 At the same time it may be observed that it is not rare for 

 individuals of a given species to depart from the normal character 

 in regard to surface contour, and further, that in particular 

 localities this variation may prevail. This refers especially to a 

 tendency for the surface to be more depressed than is normally 

 the case, and does not apply to C oscinodiscus only. Thus in some 

 of the Oamaru deposits we find that Aulacodiscus margaritaceus, 

 A. amoenus and the large forms of the Triceratium favus group 

 are all characterised by the unusually depressed surface of the 

 valves. 



It may be noted, further, that it is not safe to describe the 

 surface contour of a species without examining both valves. 

 Rattray describes C. superbus as convex, but in reality one valve 

 is convex, while the other has the centre depressed. Several 

 species, such as C. tumidus, have the surface concentrically undu- 

 lated, while in a series of forms, described by Grunow as Pseudo- 

 Stephanodiscus, there is an asymmetrical inflation of the surface. 

 The inflations and depressions in C. excavatus are also familiar 

 examples of specialised areas. 



Variations of the radial symmetry, other than those men- 

 tioned, are rare. A notable instance is that of C. cocconeiformis, 

 which has the markings bilaterally arranged. 



In the great majority of cases the form, size and arrangement 

 of the cellules or puncta which cover the surface are the prin- 

 cipal or sole ground relied upon for specific distinction, many 



