142 Prof. W. J. Sollas on Sponge-spicules in Chert. 



Of slices C 22 and C 23 it may be safely said that the 

 chief determinable constituents are sponge-spicules ; they make 

 up the larger part of the chert. 



Rhombohedra of dolomite, precisely similar to those de- 

 scribed by Prof. Penard in his paper on the Carboniferous 

 Phthanites*, are also to be seen in most of the sections. The 

 presence of these crystals can readily be explained in accord- 

 ance with the theory of the formation of flint lately advocated 

 by me. It is well known that most limestones contain a trace 

 of magnesian carbonate, probably in the form of dolomite. 

 This is very much less soluble in acids than calcite, so that it 

 is left as a residue on dissolving limestone in dilute hydro- 

 chloric acid. The silicic acid which dissolved and replaced 

 the calcic carbonate would therefore act with greater difficulty 

 on the dolomite, and, indeed, would in all probability have no 

 action upon it at all ; and so, while the mass of the limestone 

 underwent silicification, the rhombohedra of dolomite would 

 remain unaltered in the midst of the chert, just as we now 

 find them. 



That the formation of the chert continued, if it did not com- 

 mence, some time subsequent to the formation of the limestone 

 is proved by a curious fact observable in the section of coral 

 shown in the centre of the section represented in fig. 1 of 

 Prof. Hull's plate. Some of the vesicular spaces at the edge 

 of this are completely lined by crystals of calcite (dog-tooth 

 spar) which has not undergone silicification, while the interior 

 of the spaces is occupied by crystalline grains of quartz 



(%. 2). 



Fig. 2. 



Part of section of coral from cliert (fig. 1, pi. iii. Hull's paper) : Q e, ex- 

 ternal wall, consisting of radiating crystalline fibres of quartz ; S, 

 septum ; C, calcite lining marginal vesicle ; Q i, quartz within the 

 vesicle. (x60diam.) 



It is clear that the coral had been dead and exposed to the 

 • Bull, de I'Acad. roy. Belg. 2* ser. t. xlvi. 



