or in the side of a glaciated mountain. The valley cirque is
characteristic of calcareous districts. In the Chiltern Hills
especially, and generally along the chalk escarpments, a flat-bottomed
valley with an intermittent stream winds into the hill and ends suddenly
in a cirque. There is an excellent example at Ivinghoe, Buckinghamshire,
where it appears as though an enormous flat-bottomed scoop had been
driven into the hillside and dragged outwards to the plain. In all cases
it is found that the valley floor consists of hard or impervious rock
above which lies a permeable or soluble stratum of considerable
thickness. In the case of the chalk hills the upper strata are very
porous, and the descending water with atmospheric and humous acids in
solution has great solvent power. During the winter this upper layer
becomes saturated and some of the water drains away along joints in the
escarpment. An underground stream is thus developed carrying away a
great deal of material in solution, and in consequence the ground above
slowly collapses over the stream, while the cirque at the head, where
the stream issues, gradually works backward and may pass completely
through the hills, leaving a gap of which another drainage system may
take possession. In the limestone country of the Cotteswold Hills, many
small intermittent tributary streams are headed by cirques, and some of
the longer dry valleys have springs issuing from beneath their lower
ends, the dry valleys being collapsed areas above underground streams
not yet revealed. In this case the pervious limestone is underlain by
beds of impervious clay. There are many of these in the Jura Mountains.
The Cirque de St Sulpice is a fine example where the impervious bed is a
marly clay.

The origin of the glacial cirque is entirely different and is said by
W.D. Johnson (_Journal of Geology_, xii. No. 7, 1904) to be due to basal
sapping and erosion under the _bergschrund_ of the glacier. In this he
is supported by G.K. Gilbert in the same journal, who produces some
remarkable examples from the Sierra Nevada in California, where the
mountain fragments have been left behind "like a sheet of dough upon a
board after the biscuit tin has done its work"; so that above the head
of the glaciers "the rock detail is rugged and splintered but its
general effect is that of a great symmetrical arc." Descending one of
the bergschrunds of Mt. Lyell to a depth of 150 ft., Johnson found a
rock floor c