The Structure of Diatoms. By Dr. J. II. L. Flogel. 669 



could be accomplished with not wholly soluble but only suspended 

 colours a further clear proof would be furnished of the existence 

 of the openings of the chambers. Fossil specimens of Coscino- 

 diseus radiatus from marl slate from Oran, for which I have to 

 thank Prof. Dippel, appear in their section-images quite in harmony 

 with the above. As a matter of course here is no longer cellulose, 

 and the valves and the girdle-bands are isolated. In the surface- 

 image the opening to the chamber is a little larger than with the 

 other specimens, which may be attributed to the loss of the cellulose. 

 Putting the outer surface (for instance, line 1, fig. 23) under a 

 high power we see these openings as delineated in fig. 26. By 

 slightly raising the slide so that line 2 is seen, the inner ring dis- 

 appears and distinct chamber-walls are seen, fig. 27. One esti- 

 mates the thickness of these walls in the surface-image too high, a 

 circumstance connected either with the magnification of the outer 

 marginal thickening of these walls, or which must be explained by 

 the reflection of the light from the walls.* 



In a few specimens, not far from the edge, I observe unusually 

 large chamber- walls, fig. 28 ; the cavity is apparently quite 

 spherical, and only at some distance ofi" does it take the prismatic 

 form. It seems to me quite unimaginable that the membrane should 

 have raised itself as an annular wall until quite above or nearly 

 80, and then have closed itself again over such a cavity (the pointed 

 shaped opening seems really sometimes to be wanting altogetherj ; 

 and for this second reason in support of my former conjecture — 

 according to which the cavities develop and enlarge themselves in 

 the substance of the walls, and in this case open by the resorption 

 of the girdle — I give preference to my explanation rather than to 

 Muller's. 



§ 2. C. Ocidus iridis Ehrenb. — The works of Slack, Stephen- 

 son, and Morehouse relating to this species (25, 26, and 13), I know 

 only through Just's ' Jahresbericht.' According to Pfitzer's refer- 

 ence, Stephenson describes an outer layer of deep hexagonal cells 

 which are one and a half times as deep as broad, and which, judging 

 from the positive images which they give of outer objects, are either 

 open on both sides or closed by nearly plane membranes. The 

 latter becomes more probable through the appearance of small 

 depressions in the base of the cells, the edge of which is undulated. 

 The inner layer Stephenson describes as a thin hexagonal areola 



* On thia subject compare my work on Pleurosiijina, p. 507, where the reflec- 

 tion appearances witliin vertical cliamber walls are fully discussed. With a 

 stage having an opening covered })y a glass \>hd(i which is sliglitly tilled 

 these experiments can be controlled. Since outwards of the edge-line of a wall 

 black always api»eais first and white within, it becomes evident that witii small 

 dimensions an apparent thickening of such vertical membranes must always 

 result, because the eye naturally tak" s the middle of the black line as the 

 liiuit. 



