Coccospheres and the Origin of Coccoliths. 311 



coccospheres there is a residue of stained material left in the central 

 part of the holes, corresponding roughly in shape and position to the 

 D-shaped perforations of the body of the coccoliths. This residue may 

 possibly be the remains of protoplasmic filaments or protrusions once 

 extending from the central region of the coccosphere. Similar granular 

 slimy masses may frequently be seen in stained"* and undissolved cocco- 

 spheres occupying more or less of the central depressions of the cocco- 

 liths ; but their irregularity of occurrence and shape make it impossible 

 to say whether they are proteid material in connection with and de- 

 rived from the central body of the coccosphere, or slime adventitiously 

 deposited from the surrounding liquid. 



Internal Structure. 



By continuing the solution with acid, several points of the internal 

 structure of coccospheres become apparent. The external pellicle 

 gradually disappears, and only a number of disconnected, very minute 

 granules persist to mark its position. The jagged edges of the holes 

 persist the longest, but finally they also dissolve. If the coccosphere 

 is stained in this condition it will appear that immediately within the 

 external pellicle which has now disappeared there is a slimy proteid 

 material, in which the inner valves and bodies of the covering cocco- 

 liths were embedded. Sometimes after treatment with nitric acid and 

 staining with aniline blue this material has a homogeneous appearance, 

 and is imiformly stained, while the positions once occupied by the 

 coccoliths are colourless and transparent. More frequently, however, 

 it is a finely and sparsely granular slime (fig. 9). It gives a faint 

 orange coloration with nitric acid, followed by ammonia. 



This slimy layer, in which the coccoliths are embedded, is bounded 

 internally by a gelatinous, transparent, and sometimes slightly strati- 

 fied membrane (figs. 11 — 14). There is usually no definite demarcation 

 between the two, but the membrane passes imperceptibly into the 

 slimy material outside. The internal valves of the coccoliths rest upon 

 this membrame, and it is often seen to be drawn out into prominences 

 corresponding with the position of the coccoliths on its outside. In 

 optical section discontinuities are sometimes apparent in this mem- 

 brane (figs. 13, 14), and it may be that these represent perforations 

 corresponding with the perforations of the coccoliths, and through 

 which the internal protoplasm communicates with the external sur- 

 roundings. The membrane itself is difiicult to stain, and it, like the 



* I have found Bismarck brown, aniline blue, acid fuchsine, and naethyl green 

 useful stains for bringing out the structure of undissolved coccospheres. Delafield's 

 hsematoxylin sometimes gives very good results when precipitations do not occur. 

 Material stained with the last mentioned is most satisfactory for microtome 

 sectioning. 



