-I 



W. M. BALE ON SOME OF THE DISCOID DIATOMS. 21 



C. crassus, so abundant in the Sendai deposit, simply consists of 

 the smaller valves of the equally abundant G. borealis, to which 

 it bears the same relationship that C. biangulatus does to 

 C. asteromphalus. 



In C. marginatus the small valves, with uniform and non- 

 radial areolation, are considered typical, but, as in the above- 

 mentioned species, we find that valves of maximum size have the 

 areolation distinctly radial, with the areolae increasing in size 

 from the central rosette towards the margin. 



In other species similar conditions occur, indicating that the 

 reduction of the differences in size of the areolae is the regular 

 concomitant of the reduction in size of the valves, and showing 

 how little such variations are to be relied on as specific 

 distinctions. 



The presence of a central area may be of specific value in 

 some instances, but in many species it is quite worthless even as 

 a varietal character. Sometimes its disappearance is due to the 

 cellules surrounding it becoming enlarged at its expense. Thus 

 in C. perforatus and C. apiculatus normal valves (if indeed wo 

 are right in considering as normal those valves with separate 

 round markings, which I greatly doubt) have a blank central 

 space, and the cellules surrounding it are in no way different 

 from the rest, but when, by the enlarging of the cellules- 

 generally at the expense of the intervening substance, the 

 structure becomes areolate, the most central cellules often 

 enlarge inwards till they obliterate the area, and thus form a 

 rosette, as in C. Oculus Iridis, etc. 



Far too much importance has been attached to the area in 

 Rattray's monograph, especially in the key. 



The central rosette is one of the most variable of characters. 

 In some cases, as already mentioned, it is conspicuous in the 

 largest valves, dwindling and finally vanishing in the smaller 

 ones ; in others, just alluded to, it results from the obliteration, 

 entire or partial, of the central area. In some no doubt it 

 may be regarded as a fairly constant specific character. 



The tendency in some species for the polygonal areolation to 

 be replaced on a portion of the valve by isolated circular cellules 

 may be briefly referred to. C. perforatus and C. apiculatus are 

 familiar cases in which this modification occurs, either over the 

 whole surface of the valve, or on more or less of one side, while 



