490 J. G. Ooodchild — On the Origin of Coums. 



Similar remarks apply also in part to many other such recesses in 

 like positions. 



Now, as an ordinary glacier scoops, necessarily, only in a downward 

 and outward direction, we ought to find, if the theory of the origin 

 of coums by means of glaciers were the true theory, that all traces 

 of horizontal prominences have been ground off by the ice ; and, in 

 place of terraces extending at right angles to the path the ice must 

 have taken, we ought to find whatever furrows the ice left, in, or 

 nearly in, the line of motion of each part of the bottom of the glacier. 

 Yet, it is a well-known fact that where the rocks that form the coum 

 consist of a nearly horizontal set of alternations of beds of different 

 degrees of hardness, each separate hard bed forms an amphitheatric 

 shelf or terrace, which is separated from those above and below it by a 

 horizontal interval, often of considerable extent ; so that the general 

 effect resembles, on a gigantic scale, the tiers of seats in a great 

 amphitheatre. This is especially noticeable in the case of many of 

 the cirques in the Jura, some instances of which have been mentioned 

 by the Eev. T. G. Bonney in his paper on the origin of Cirques. It 

 is clearly impossible that any such ledges of rock can be due to the 

 erosive power of ice acting vertically ; and therefore, as very perfect 

 coums exist to which the theory of glacier erosion will clearly not 

 apply? we are f° rce( l to conclude with Mr. Bonney that simple 

 glaciers have had little, if they have had anything, to do with the 

 formation of the greater number of the coums, either here, or on 

 the Continent. 



There are some other and perhaps not less weighty objections 

 against the glacier origin of coums. It has been before re- 

 marked that the greatest regularity of form is often to be found 

 only near the higher parts of the recess ; while, in their lower parts, 

 many of the coums do not differ very much from the higher parts of 

 ordinarv valleys, except that in a few cases they are flatter, or rather 

 wider than one usually finds the head of a similar valley where no 

 coum exists ; and that rock basins on a small scale often occur at the 

 foot of the steeper slopes. It seems quite clear that if the coums are 

 due to the scooping out of a valley head by the long-continued action 

 of a small glacier, the greatest amount of erosive force must have 

 been exerted in those parts of the coum that had the greatest over- 

 burden of ice — in the higher parts, where the neve was hardly 

 sufficiently consolidated to deserve the name of ice at all, the amount 

 of erosive force exerted must be very small indeed. Yet the lower 

 parts are those that, in many instances, do not differ very much from 

 ordinary valley heads ; while at the higher parts, where the glacier 

 ice can have exerted little or no erosive power, the configuration of 

 the surface plainly shows that the erosive agents have acted with the 

 greatest effect. Again, where tiers of rock ledges occur one behind 

 another, as they do in some of the English coums, and more strikingly 

 in those of Switzerland, it is also obvious that the coums cannot be 

 due to the undermining of the higher rocks by a glacier that was 

 grinding away the soft beds at the lower part of the recess. Such 

 •action, where the rock is of a uniform lithological character, might 



