Vol. 6$.] AND COEALLIAN KOCKS OF BRILL. . 39 



The ' white ' spicules, on the other hand, are more or less dusky 

 throughout, and show by ordinary light a well-marked concentric 

 and a less distinct radial structure : both structures being marked 

 out by alternations of brown and clear material, although no part 

 is so clear as the central mass of a ' blue ' spicule. The concentric 

 layers show no cuspation. Between crossed nicols, they exhibit a 

 regular radial structure in much duller tints than those of the ' blue ' 

 spicules. 



The foregoing descriptions seem to indicate that the ' white ' 

 spicules are in a less altered state than the ' blue,' and that the 

 latter have either undergone recrystallization, or that all but their 

 outermost shells have been dissolved away and the cavity so formed 

 afterwards refilled by secondary chalcedony. This view obtains 

 some confirmation from the structure of the calcareous variety of 

 the stone, described below. 



The matrix in which the spicules and sand-grains are embedded 

 is not easy to determine. In section, it varies from faint yellow 

 to deep brown in colour, is speckled with brown (? limonite), and 

 between crossed nicols shows a cryptocrystalline to amorphous 

 structure. 



In the ' calcareous Arngrove Stone ' of Warren Farm the matrix 

 is minutely-crystalline calcite, with cloudy impurities. The three 

 constituent grains of the typical chert are now represented by (1) 

 calcite-grains, (2) siliceous grains, and (3) sand-grains, as before. 

 The first consist each of a single crystal or a small number of crystals 

 of calcite, and can therefore be more easily understood as formed by 

 the filling of a cavity than by true pseudomorphic replacement of 

 chalcedony. The siliceous spicules (which can be examined whole 

 in the insoluble residue of the rock as well as in section) differ from 

 the ' white ' spicules of the chert by showing little or no concentric, 

 though a fairly-distinct radial, structure, and by their perfect iso- 

 tropism. They are probably, therefore, the globate spicules quite 

 unaltered. The sand- grains are similar to those in the chert, and a 

 few flakes of mica also occur. (See PI. I, fig. 3.) 



All the conditions in which the spicules of ffliaccella occur in these 

 rocks were described long ago by Dr. Sorby x in the similar rocks 

 of Yorkshire. I have not seen a specimen of the Yorkshire chert ; 

 but the descriptions of it agree very closely with that of the 

 Arngrove Stone, and the age of the two must be very nearly the 

 same. We have already seen that, in describing this stone, Green 

 (writing mainly from Mr. Polwhele's note) states that 



' immediately below this stone we find Oxford Clay with Gryphcea dilatata, and 

 it is therefore without doubt the bottom bed of the Calcareous Grit.' 2 



This statement probably refers to the escarpment of Pan's Hill, 



1 Quart. Journ. Greol. Soc. vol. vii (1851) pp. 1-6. 



2 ' G-eology of the Country round Banbury, &c.' Mem. Geol. Surv. 1864 ? 

 p. 44. 



