ZOOLOGY AND BOTANY, MICROSCOPY, ETC. 947 



There is a tendency in most of tlie species to accumulate silex 

 upon the spaces between the areolfe, giving a roughened and irregu- 

 larly granulated appearance to the outer surface of the disk. This 

 condition also interferes with a satisfactory examination of the ' dots ' 

 by causing irregular refraction, &c. For this reason we need to select 

 smooth and evenly marked specimens for one part of the investigation, 

 though for another the roughened examples are most instructive. We 

 find this thickened coating broken away in different degrees ; some- 

 times leaving the shell smooth but perfect ; sometimes taking with it 

 the outer lamina, and leaving only the inner with its delicate punc- 

 tation. The fact that the thickening is upon the interspaces between 

 dots is additional evidence that the latter are areolfe, since they 

 allow the light to pass when the thickened walls around them make a 

 semi-opaque outline approximating the character of the shell in 

 Eupodiscus argus. But among these roughened sjiecimens I have 

 more frequently found the separate inner lamina, and this when once 

 caught by the glass, is always the most convincing proof of the 

 scheme of marking of the valve ; for the film is so homogeneous and 

 even, and the dots upon it are diminished to so fine and regular 

 punctae that the eye is never dazzled, confused, or misled in following 

 its delicate i^attern. 



The examination of the several species last referred to, under the 

 vertical illuminator with higher powers, and as opaque objects with 

 the 1/4 in. objective, is strongly confirmatory of the interpretation I 

 have given. Keflected light may be made to flash from the surface of 

 all the finer examples of C. suhtilis as well as from Adinocyclus and 

 Podosira, so as to show a glassy smoothness, with a play of iridescence 

 in the thinnest specimens. This is true of both the convex and con- 

 cave surfaces of the valves. No trace of projecting spherules can be seen 

 in such an examination, though the dots of the shells are of such appre- 

 ciable magnitude that tliey would be easily visible as protuberances if 

 they were solid spherules. Indeed, with the vertical illuminator and 

 a high power, siliceous fragments of broken sand-grains may often be 

 seen lying upon the surface of these shells, very miich smaller than 

 the areolae, and demonstrating by the ease with which they are seen, 

 that if the dots approached hemispheres in form they also would be 

 perfectly apparent. This, then, is another strong proof that these 

 areolae are contained between smooth and parallel laminaa. 



If, finally, still using the vertical illuminator and a high power, 

 we review the series of valves beginning with the boldest forms of 

 Triceratium and Coscinodiscus and ending with the finest Actinocyclus 

 and Podosira, we find certain appearances consistent throughout the 

 whole range of examination. The areolae, when the surface is care- 

 fully brought into focus and the cover correction accurately adjusted, 

 are always an opaque white or grey, whilst the surrounding wall or 

 solid part is darker, becoming even black when close to a dark back- 

 ground. The comparison which I have already made to ice upon a 

 pond, wlinn part of it is solid and clear and part of it porous, very 

 aptly describes this appearance. There is no break in the series. 

 From coarsest to finest the only change is that the areolao grow smaller 



