Notices of Memoirs—sandstones of the Bristol Coal-field. 521 
mixed with the other material to increase the proportion of silica. 
It was pointed out that there were other beds higher up in the Coal- 
measures which would do equally well, and in some cases better. for 
this purpose. The paper next referred to the ‘ Pennant Grit.’ There 
is considerable difficulty in defining the limits of this deposit, but it 
was certainly not 2000 feet thick, as some authors had stated. The 
paper places the thickness at about 970 feet; but the middle or 
Pennant series of Coal-measures, so called on account of the Pennant 
being so extensively developed in this division, was about 2000 feet 
thick, and this, probably, was the origin of the mistake regarding 
the thickness of the Pennant. 
The 970 feet of rock above referred to as the Pennant grit, was 
only entitled to that name as a local distinction. It was nothing 
more than an extraordinary development of a local measure grit : 
the ‘ Dexall grit,’ of the lower series, for instance, was quite as much 
a ‘Pennant,’ if that term is to distinguish a certain class of rock. 
After a careful examination of the rocks of the Coal-field, the author 
had come to the conclusion that, owing to the great similarity of 
Carboniferous arenaceous rocks occurring at different horizons, there 
was risk of serious error, in relying upon them for correlation or 
as stratigraphical landmarks. The proportion of silica could be 
sometimes used as a guide in determining one from another, but 
little reliance could be placed on it over a large area, as so many beds 
contained nearly the same amount. 
The author’s analysis showed the first 50 feet of the Pennant to 
contain 90 per cent. of silica; but after this, for a considerable thick- 
ness, ‘good Pennant’ varied from 84 to 89 per cent. in the propor- 
tion of silica contained in it. 
The paper then referred to changes in the Carboniferous rocks, 
when in contact with carbonaceous inatter. The author found that 
the per-centage of alumina increased, and this mostly as a silicate. 
By comparing the analysis of duns and shale from the district with 
that of the rocks, the same constituents were found to be present, 
the great difference being in the greater proportion of alumina in 
duns and shale. As a rule, the latter beds were to be found near 
coal; but in cases where rock followed, the author found that as it 
neared the coal it became more fissile and argillaceous. 
This change was ascribed to the action of carbonic acid gas, gene- 
rated by decomposing vegetation on silicates. The analysis of the 
rocks given showed them to have been formed from the denudation 
of older silicate rocks, and the action of carbonic acid on such sedi- 
ment would be to readily decompose all silicates with the exception 
of silicate of alumina, which would thus increase in proportion to 
the whole, and give rise to beds of the composition of duns and 
shale.' To this cause the author attributed the formation of the 
latter deposits, and contended that although they may oecur apart 
from carbonaceous matter, there is no proof that it was never present, 
and that it may not have been destroyed by decomposition. 
1 The writer is not now dealing with the hydrocarbons which shales sometimes 
contain. 
