XX1V INTRODUCTION. 
having now disappeared from the sections, I think it much the safest plan to place the 
Cornstone group below all the groups of Devonshire; especially as it seems in some places 
to pass downwards into the Tilestone of the Ludlow group, and therefore appears to give us 
(what we do not find in Devon and Cornwall) a base to the Devonian series. On this 
hypothesis we might arrange the Devonian series of Herefordshire and the sub-groups of 
the Silurian series in a regularly ascending and unbroken numerical order; but the numerical 
series would be defective in its last term. The phznomena in Scotland seem however to 
be adverse to this hypothesis. 
The Petherwin (or Barnstaple) group is provisionally arranged in the Devonian series. 
Physically, it is best connected with the older groups of Devonshire and Cornwall; but 
palewontologically, it is perhaps better connected with the rocks of the Upper Paleozoic 
division®. Here, therefore, there seems to be a continuous ascending order of deposits, and 
a passage from the Devonian series to the Carboniferous. 
(8) Devonian Series of Scotland. 
Grand as is the development of the Old Red sandstone of Herefordshire, it dwindles into 
insignificance when compared with the rocks which pass under the same name in Scotland. 
They are divided by Miller as follows :— 
1. Great conglomerate and red sandstone. 
2. Bituminous schists—Dipterus, Pterichthys, Coccosteus, &e. 
Red and variegated sandstone. 
These three groups make up the “lower formation” of Miller, and are well seen in Caithness. 
J 4. Gray sandstone, earthy slates, &c. This is the “middle formation” of Miller, and contains a 
peculiar group of fishes, Cephalaspis, &e. 
5. Red sandstone and conglomerate. 
| 6. Impure concretionary limestone. 
7. Yellow siliceous sandstone. 
The last three, called by Miller the ‘‘ upper formation,” are characterized by Holoptychius, &c. 
oo 
Devonian Series. 
This vast North-British series has no true paleontological base; but its “upper for- 
mation” seems to graduate into the Carboniferous series. 
Its lower groups have no known representatives in the Old Red sandstone of Herefordshire, 
and cannot (with our present information) be drawn into a close and unequivocal comparison 
with the “7%lestone” and Fish-beds of the Ludlow rock. But its “middle and upper formation’ 
seem to be represented, though imperfectly, by the “cornstone” and overlying conglomerates, 
&c. of Herefordshire. From this also it seems to follow, that the upper part of the Old Red 
sandstone of Herefordshire is defective in its development; and the conglomerate form of 
> 
* This question is discussed in a paper on the Slate Rocks of Deyon and Cornwall, Quarterly Journal of Geol. Soc. 
Vol. yin. 1852. 
