The Structure of Diatoms. By Dr. J. H. L. Flogel. 667 



6, CoseinodiscKS. 



Of this species I have examined four varieties. I collected the 

 material myself on the Norwegian coast, and another portion was 

 from the ' Pomerania ' expedition (9). 



§ 1. Coscinodiscus radiatus Ehrenb. — This species gave the 

 best results, because the chambers are of considerable size, and I 

 had a sufficient quantity of them. Among the specimens obtained 

 with the knife one is especially interesting, because it was ob- 

 tained during fission. This is represented in fig. 31. I photo- 

 graphed a particularly successful section as a type-image. This 

 section is very good, but for the study of the sculpture a little too 

 thick ; for this reason I have delineated in fig. 23 the marginal 

 portion of an adjoining fine section under very high magnification. 

 From both images the sculpture of this species may be deduced as 

 follows : — The general form of C. radiatus is a very thin disk like 

 a coin ; both faces are circular and display the pretty dehcate 

 areolae. Vertical sections through the disk give everywhere the 

 same image, the central one included. The cell- wall of these disks 

 consists of a very thin basal membrane, having on the outside 

 very delicate but prominent network. This network is enlarged 

 and thickened at the end, and so much enlarged that the lines 

 almost coalesce. In inferior sections one sees a continuous outline, 

 which is divided from the inner membrane by bars, but in very 

 good sections the lines have the shape of a T. In consequence 

 the network forms prismatic chambers, mostly hexagonal, but also 

 five- sided and four-sided, which without doubt must have an open- 

 ing from the outside, even if it be very small. In the surface view 

 I did not observe this opening until after my attention had been 

 drawn to it by the sections. It is seen with balsamed prepara- 

 tions like a small nodule, having 1/4 to 1/3 of the diameter of an 

 areola, and in superficial examination it may easily be mistaken 

 for the focal image of a chamber. To decide this point search 

 must be made above the acute portion of the chamber-walls. 

 The large number I examined of good sections which all agree 

 when thin enough, and show no continuous outline, prove defi- 

 nitely the existence of a small opening. The opening, in proportion 

 to the size of the chamber, is smaller than the circular opening 

 of the Triceratium chamber, as may be seen from the diameter 

 of the chamber-walls, which have a somewhat different shape from 

 Triceratium. Otherwise they are both very similar, but Coscino- 

 discus has no spines ; the outer surface is therefore smooth. The 

 height of the chamber is equal to its breadth (3-3 • 2 /x). To- 

 wards the edge of the disk both dimensions decrease. The girdle- 

 l^and is seen in most of the sections ; it is without sculpture. 

 Regarding its connection with the two inner newly formed cell- 



"2 Y 2 



