Vol. 58.] ALPINE VALLEYS IN RELATION TO GLACIERS. 699 
does not harmonize with the work of heat and cold, of rain and rivers, 
guided, of course, by the results of earth-movements.’ 
Personal examination of every part of the Alps, of the Pyrenees, 
the Apennines, Scandinavia, Auvergne, and many other hill- or 
mountain-regions, has led me to the following conclusions: 
(1) That the action of permanent snow is more conservative 
than destructive. 
(2) That the erosive effect of a glacier is at a minimum beneath 
a comparatively level névé, such as 1s often found at the base of a 
rock-wall, not seldom from 300 to 600 or 700 feet high. 
(3) That cirques, corries, and bowl-lke heads of valleys are mainly 
the work of water, their forms depending on local circumstances : 
they are made on a small scale by rain and its runlets, on a larger 
by streams.” They are not restricted to glaciated regions, though 
permanent snow-beds, other things being equal, favour the formation 
of the first and second. ‘The principal cirques in the Alps exist in 
regions where the peaks do not rise so high as 11,000 feet, and are 
sometimes considerably below this; their floors are not at very high 
levels—under rather than over 4000 feet. In other words, cirques 
occur on the grandest scale where the ice would have the smallest 
extension, the shortest duration, and the least erosive action. 
(4) That while running water is very obviously competent (with 
the requisite environment) to carve out cliffs on a grand scale, no 
proofs have been brought to show that a glacier can do this on any 
but a small scale. Indeed, in all cases that I have seen, it appeared 
more probable that the ice had only modified a cliff already in 
existence, than that it had hewn one out by itself. We need also 
something better than vague assertions about ‘ broken-bedded 
valleys ’ before we can regard it as proved that a glacier can hew 
its rocky floor into great steps. We may fairly appeal to the Unter- 
Grindelwald Glacier, for its ice-curtain has been partly withdrawn 
during the last 40 years; when we examine the steps which it has 
exposed, these appear to have preceded the ice, and to have had 
only their edges rounded off by it. Indeed, my imagination will 
not enable me to picture a glacier in the act of cutting steps in a 
subjacent mass of rock. 
Again, the obstacles which projecting rocks, or ridges like those 
of Sion, present to Prof. Davis’s excavating glaciers, cannot be 
? The features of the Saxon Switzerland are often thoroughly Alpine. There 
are little gorges in the glens leading to the Elbe, and lines of castellated summits, 
which in a photograph might be taken for scenes in the Dolomites, yet these 
are seldom more than about 1500 feet above sea-level, so that glaciers, if ever 
they formed in this region, would be very short-lived. 
* See my paper on ‘The Formation of Cirques’ Quart. Journ. Geol. Soc. 
vol. xxvii (1871) p. 812. During the 30 years which have elapsed since that 
paper was written, I have revisited some of these cirques, have examined many 
others, great and small, and have been continuously observing the forms of 
valleys, and I am more than ever convinced that I have rightly interpreted the 
teaching of Nature. Such contours as are represented by Mr. Harker (Trans. 
Roy. Soc. Edinb. vol. xl, pp. 232-41) are to be found associated with, but are 
not characteristic of, glacial action. 
