52 



must also be attached to the object-glass, and on examining as an 

 opaque object any one of the disks, the slide can be turned over, and 

 the same disk readily be recognised by means of the lines around it ; 

 and thus, however minute, both sides may be inspected. 



In viewing these disks as transparent objects mounted in balsam, 

 it is by no means easy to determine whether they have or have not 

 a central aperture, and polarized light renders no assistance in doing 

 so ; but on mounting them as above described, or on a black disk, 

 and viewing as opaque objects, it becomes indisputably evident that 

 a central opening is present. When in situ, these openings are par- 

 tially covered internally by a delicate cup-like process, so as to form 

 a species of valve ; hence, when the shells are attached to the Alga, 

 and are viewed even as opaque objects previously to their undergo- 

 ing any sort of manipulation, it is not satisfactorily shown that the 

 apertures exist in every specimen. I consider it highly probable 

 that at these openings the nutriment is taken in, and from the pecu- 

 liar radiated arrangement of the siliceous parts immediately around 

 them, it seems not improbable that the animals may be provided 

 with organs somewhat resembling in arrangement the tentacula of 

 Madrepores. This is supposing the objects to be of an animal 

 nature, of which I confess I think there is very little doubt. 



With regard to the two annular valves previously mentioned, they 

 are situated between the two discoid ones, and thus the whole shell, 

 when in its natural state, resembles, like those of some other genera, 

 a circular snuff-box. These valves consist of a siliceous ring, within 

 which (extending a slight distance towards the centre) is an annular 

 membrane, and when in situ the valves are placed so that the mem- 

 branes of each valve are in contact, and thus the space included be- 

 tween the two discoid valves is partially divided into two (not three) 

 chambers. The membrane is so thin, that when mounted in bal- 

 sam it is not visible ; I have, however, been able to detect it upon 

 one or two occasions, when it has got accidentally discoloured. It 

 can, however, be readily seen by breaking away the upper valve from 

 a specimen when attached to the Alga, and also when mounted be- 

 tween the two pieces of thin glass. Plate XI., fig. 3. 



It is somewhat difficult to make out whether there are two, or only 

 one, of these annular valves, when the shell is entire ; they can, how- 

 ever, under favourable circumstances, be defined, even while attached 

 to the Alga, but if -this latter be soaked in water, a few of the shells 

 can sometimes be removed entire, and by examining them between 



