THF. BRISTOL COALFIELD. 



68 



silicates is to decompose them, with the exception of silicate of 

 alumina,* the bases forming soluble carbonates; but from the 

 occurrence of hydrated oxide of alumina in nature t along with the 

 silicate, a proof that it also undergoes decomposition under certain 

 conditions. The alumina, however, is incapable of forming a car- 

 bonate, and the hydrated oxide of alumina formed is less soluble than 

 the carbonates of the other bases arising from the decomposition. 



It is then plain that the result of this action would be the removal 

 of all silicates, with the exception of silicate of alumina, and thus 

 the latter would increase in proportion to the whole, and where it is 

 decomposed the proportion of alumina would increase owing to the 

 insolubility of the hydrated oxide. This is just what we have found, 

 and it appears to me to be conclusive evidence that shales, and those 

 semi-argillaceous rocks which I have included under the name of duns, 

 are formed from the action of carbonic acid gas upon arenaceous rocks 

 such as the grits or the sediment of which they are composed.* 



One fact, which appears contrary to what we might expect, is the 

 proportion of iron which remains in those beds which are in contact 

 with carbonaceous matter. I would, however, point out that the iron is 

 not in combination with silica, but exists as an oxide or carbonate, the 

 natural product of the decomposition of the silicate. This fact, coupled 

 with the presence of manganese, sulphur, and in one instance phos- 

 phorus, seems to show that carbonaceous matter has the power of 

 secreting certain inorganic substances. 



That this is the case with iron was demonstrated to me in the 

 analysis of fossil Calamites from a bed of Pennant grit. 

 The following is the analysis : — 



Analtis of the Pennant bed in which the Calamites occurred. 



Silica 87-GU 



Alumina 5*23 



Iron 2-73 



Lime 240 



Carbonaceous matter 1-70 



Carbonic Acid 



Water -90 



Analysis op the Fossil Calamite. 



* Bischof's Chemical and Physical Geology, vol. I., page 1. For the decompo- 

 sition of silicate of iron see page 3. 



t This we have found ; the hydrated oxide of alumina being that portion 

 soluble in acid. 



t These rocks occnr in coal-bearing strata, but apart from carbonaceous 

 matter ; still this is no argument against what I have said, as the carbonaceous 

 matter may have been removed by decomposition, 



