284 Proceedings of the Royal Physical Society. 



siderable time. Meanwhile accretion goes on — sometimes 

 slow and even, sometimes by the addition of cap to cap. 

 Discarded secondary channels remain as grassy furrows on 

 the enlarging hangh ; the selected one is kept open. There is 

 less alluvial deposit in it so long as the deposit is liable to be 

 swept out in flood-time. 



The water that occasionally flows through this selected 

 secondary channel is not entirely inactive. In most cases 

 which the writer has examined, it is thrown, in splitting off 

 from the main channel, against the opposite, i.e., the inner or 

 alluvial bank, of the secondary one. In this and other ways 

 small reversed shifting bends are set up, and work their way 

 along that side. The fresh alluvium, with its cover of turf, 

 is ripe for undermining, and readily forms a scar ; and thus, 

 flood after flood, the furrow is etched out into a step or 

 little terrace. The further it is worked in towards the 

 body or instep of the foot, the higher and more terrace-like 

 it becomes ; and when the main channel, with its enlarging 

 circuit, has progressed to a greater distance and lower level, 

 with new gravel-accretions and other secondary channels, it 

 is left high and dry. Many abortive terraces — represented 

 by crescentic furrows — may be formed for every one that 

 ultimately remains and is truly terrace-like ; but as time goes 

 on the larger of them are selected for preservation, and the 

 lesser are obliterated; and the sloped side of the amphi- 

 theatre becomes benched. Contrary to the fashion of amphi- 

 theatre benches, however, the convexity of these imperfect 

 crescents tends to face the hollow of the opposite bank. 



For this well-marked variety of river terrace, which was 

 not separately indicated by Hitchcock, we venture to use the 

 name Amphitheatre Terrace. 



The larger number of river terraces in this country are of 

 this kind. Of those that have been mapped by the writer in 

 the course of his work in the Geological Survey, 70 per cent, 

 occupy positions in which they may be amphitheatre terraces, 

 being placed, that is, on the sheltered side of more or less 

 pronounced bends of streams. The absolute number of them 

 actually thus produced is probably less. 



Of this kind are the lower of the terraces lying within 



