21ie structure of Diatoms, By Dr. J, H. L. Fidget 671 



marginal shadows which give the above strongly thickened T lines. 

 A close examination confirms the existence of a very fine line 

 between them. Hence we have here two distinct membranes con- 

 nected by a system of mesh-like walls. The walls commence at the 

 inner membrane very thin, and gradually become thicker as they 

 approach the outer membrane, so that they appear wedge-hke in 

 the section. Similarly the adhesion point at the outer membrane 

 is thickened, and this thickening gradually diminishes over the 

 centre of the chamber. The outer surface of the outer membrane 

 is uncommonly even, so much so, that in a dry condition the valve 

 reflects hght as strongly as a mirror. This fact alone cannot be 

 reconciled with the idea of openings {C. radiatus does not reflect 

 as a mirror), anyhow the openings could not be estimated at more 

 than 1/8 or 1/10 of the chamber diameter in order not to interfere 

 with the smoothness of the membrane. Collodion casts would be 

 very desirable in the investigation of this species. The dots of the 

 areolae in the surface view, formerly described and figured by me 

 (9, p. 86, fig. 6), cannot be seen in the sections. 



§ 4. G. concinnus W. Smith. — This species is much larger and 

 more delicately enveloped than the former, about 0*5 mm. in 

 diameter and equally convex. The section of a valve of a not very 

 large but very finely marked specimen, fig. 31, shows the familiar 

 image of chambers closed on all sides, such as I demonstrated by 

 my Pleurosigma investigations, except that with this Coscinodiseus 

 they are considerably larger (about 1 yu,). The valve reflects light 

 strongly. For the surface view I refer to my former notices and 

 figs. (9, p. 86, fig. 5). After the investigation of these four 

 varieties I believe that in the species of Coscinodiseus we have before 

 ns the gradual transition from the Triceratium type to the Pleuro- 

 sigma type, inasmuch as the small outer opening of the chamber 

 which is still seen in a few varieties, totally disappears in othera 



7. Isthmia. 



It is not at all difiicult to cut this giant amongst the diatoms, 

 but it is very difficult to obtain sections of the requisite degree of 

 fineness. The cell- wall is everywhere of unusual delicacy. After 

 examining some forty-five sections I am enabled to give the follow- 

 ing description of the cell- wall : — 



There is a difference between that portion of the frustule which 

 corresponds to the valves, viz. the sloping ends of the rhomboidal 

 cell, and the middle portion, which is to be regarded as the girdle- 

 band and under which the division takes place. This difference is 

 marked outwards by a strong expression of the areolaD in the end 

 portions in contrast to the delicate markings of small cells in the 

 girdle. With the sections one must always endeavour to determine 



