28 RYLANDS, ON MARKINGS OY DIATOMACEA 



with which the cell-like intervals between the septa are 

 filled. Burnt specimens — which I have not — would deter- 

 mine if it be fluid. It is occasionally wanting, and then 

 the cavities are filled with air. In the specimen of Trice- 

 ratium favus referred to, a large proportion of the areohc of 

 the connecting membrane contain air-bubbles. My present 

 impression is that the material is solid throughout. 



There is a peculiarity very common, if not constant, in 

 markings of the Coscinodiscus type, which, in most cases, is 

 sufficient to determine it even from a front view alone. 

 The internal layer, at the base of each areola, is modified 

 internally, so as to produce a dark or light spot, as the focal 

 adjustment of the instrument is changed. 



If the foregoing particulars be applied to the cases ob- 

 served by Dr. Greville and myself, I think the solution is 

 easy. It is not without some hesitation that I venture to 

 throw the slightest doubt upon an observation made by such 

 a man as Dr. Greville. Under ordinary circumstances I 

 should certainly not do so ; but since the publication of his 

 article on Diatomacere in Californian guano, I am inclined 

 to shelter myself under the supposition that his working 

 instrument is not equal to the exhibition of these minutiae. 

 If this were not the case, Ave could hardly have been told 

 that the rays of his Spatangidum Ralfsianum terminate "in a 

 narrow lunate fold of the valve," nor would he have pro- 

 posed to distinguish Spatangidum and Aster 'omphalus by the 

 areolated and the granulated structure of their valves* In 

 the first case the lunate fold has certainly no existence in 

 nature; the appearance results from the elevation of the 

 ray (as in Aulacodiscus) and the sudden depression at its 

 termination. While, in the second case, there is really no 

 structural difference whatever in the markings of the species 

 described, at least as far as regards 8. Ralfsianum and A. 

 flabellatus, both of which follow precisely the Coscinodiscus 

 type; the latter, however, being considerably finer, requires 

 a proportionate increase of power. 



