high up in the limestone cliffs above the Aa- valley, on the left bank, a short distance below Engelberg. Here six or seven small streams issue from the rock, and have worked out a hollow only a few dozen yards wide, yet of the true cirque type.
Fig. 3. — Small Cirque near Engelberg.
A. Limestone Cliffs. B. Shaly bank with some trees &c, out of which the streams break. C. Limestone cliff. D. Cirque. »—> Cascade.
The conditions, therefore, most favourable to the formation of a cirque are : —
(1) Upland glens, combes or terraces, so shaped as to give rise to and to maintain many small streams.
(2) Strata, moderately horizontal, over which these streams fall, which, by their constitution, yield considerably to the other forms of meteoric denudation.
(3) These strata must nevertheless allow of the formation of cliffs ; and thus perhaps the most favourable structure is thick beds of limestone, with occasional alternating bands of softer rock.
Probably some favourable configuration of the ground must be also assumed at the beginning : if a glen ended in a cliff, it would doubtless be more readily cut into a cirque; but whether this is always needed, or what cause has made the cliff, I do not now attempt to investigate. I venture to submit that I have proved that glaciers cannot have produced the cirques, and that (since these cirques cannot be postglacial) they have not, to any great extent, excavated the Alpine valleys. To assign to each agent its due share in the task of erosion in the Alps (and in all other mountain-regions that I have seen) appears too complicated a problem
