308 
EOCENE PERIOD. 
[Ch. XXII. 
chalk in the counties of Wilts and Dorset, that, at no great 
distance from these small elliptical valleys of elevation,, there 
occur several longer and larger valleys, forming deep notches, 
as it were, in the lofty edge of the chalk. These are of similar 
structure to the smaller valleys we have been considering, and 
consist of green-sand, inclosed by chalk at one extremity, and 
flanked by two escarpments of the same, facing each other with 
an opposite dip; but they differ in the circumstance of their 
other and broader extremity being without any such inclosure, 
and gradually widening till it is lost in the expanse of the 
adjacent country. 
The cases I now allude to are the Vale of Pewsey, to the 
east of Devizes, that of the Wily, to the east of Warminster, 
and the valley of the Nadder, extending from Shaftsbury to 
Barford, near Salisbury; in which last not only the strata of 
green-sand are brought to the surface, but also the still lower 
formations of Purbeck and Portland beds, and of Kimmerido-e 
clay. 
It might at first sight appear that these valleys are nothing 
more than simple valleys of denudation; but the fact of the 
strata composing their escarpments having an opposite and out- 
ward dip from the axis of the valley, and this often at a high 
angle, as near Fonthill and Barford, in the Vale of the Nadder, 
and at Oare, near the base of Martinsell Hill, in the Vale of 
Pewsey, obliges us to refer their inclination to some antecedent 
violence, analogous to that to which I have attributed the posi- 
tion of the strata in the inclosed valleys near Kingsclere, Ham, 
and Burbage. Nor is it probable that, without some pre- 
existing fracture or opening in the lofty line of the great chalk 
escarpment, which is here presented to the north-west, the 
power of water alone would have forced open three such deep 
valleys as those in question, without causing them to maintain 
a more equable breadth, instead of narrowing till they end in a 
point in the body of the chalk*. 
Rise and denudation of the secondary roclcs gradual. — To 
* Dr. Buckland, Geol. Trans., 2nd Series, vol. ii. p. 123. 
