11YLANDS, ON MARKINGS OF DIATOMACE^E. 27 



outer layer may be found upon the frustule, but I have 

 never seen them separated ; the force which removes them 

 being apparently sufficient to break them up into single 

 te areolae :" the term in this case is very unfortunate, for 

 they are in fact hemispherical elevations. In other cases 

 the two layers are connected by septa arranged hexagonally, 

 or otherwise, after the fashion of a closed honeycomb. This 

 structure is figured by Dr. Wallich, in the article before 

 referred to, as occurring in his so-called Triceratium fimbria* 

 turn ; but he has overlooked the outer layer forming the lids 

 of the cells. In this case, the surfaces, both external and 

 internal, may be plane, or formed into slightly convex facets ; 

 they may be marked or otherwise. This is the Coscinodiscus 

 type, and will be found with certain modifications in Cosci- 

 nodiscus, Eupodiscus, Aulacodiscus, &c. 



I find it difficult to avoid the conclusion that there are 

 two distinct materials present in the composition of these 

 layers, one of which acts much more powerfully upon light 

 than the other. 



Again, the two layers may be so developed and combined 

 that the septa become, as it were, a skeleton in the more 

 transparent substance. Take the case of Eupodiscus Argus, 

 of which I have one or two most fortunate sectional views. 

 The outer convex surface is formed of a layer of transparent 

 silica produced into radiating lines of hemispherical projec- 

 tions; below this, and imbedded in the same material, are 

 the " areolae," which are composed of the denser substance, 

 and are fragmentary or continuous, according to the size 

 and perfection of the specimen. The internal surface of the 

 valve is marked with minute dots. The following figures 

 represent the forms described : 



Eup. argus. Eupodiscus ? PI. angulation. 



Side of a single areola. 



The small secondary markings in Triceratium are probably 

 simply elevations ; but those of the connecting membrane, 

 like the coarser markings of the sides of the frustules, follow 

 the type of Coscinodiscus. 



I have no positive evidence as to the nature of the material 



