286 Proceedings of the Royal Physical Society. 



also may be, common to different parts of the stream, such 

 as excessive floods, cycles of climate, changes in the bed of 

 the stream, movements of the surface of the country, etc. ; 

 but in the experience of the writer the great majority of 

 river terraces are not of equal height on opposite banks. 

 The following list of measurements, made chiefly in terraces 

 of this kind throughout six miles' length of the river North 

 Tyne, will indicate that coincidence in height is at least 

 exceptional in our northern valleys. The small figures 

 attached to the larger ones denote several recurrences of the 

 same height at different points. 



Measurements of Terraces between Chollerford and Houxty 

 Burnmouth, on the river North Tyne. 



The flood-level (12 or 13 feet), to which the river builds 

 its banks where it remains long stationary, is naturally repre- 

 sented better than any other. It will be noticed that beyond 

 this there is no disposition in these figures to fall into 

 groups. 



Nothing can better illustrate the unequal and variable 

 height attained by certain lateral terraces — including the 

 variety now specially referred to — than the fact that the 

 same terrace will sometimes be found to vary as much in 

 height as if it were several different ones. The surface, in 

 such cases, has been levelled at different times without the 

 preservation of a terrace high or distinct enough to attract 

 regard. Many crescent-shaped or lobe-shaped haitghs in 

 upland streams are scored with incipient terraces, and yet 

 cannot be mapped as other than a single terrace. The num- 

 ber of amphitheatre terraces that are not more or less 

 composite cannot be great. 



A variety of the amphitheatre terrace is often found to 

 assume the characters of Hitchcock's glacis terrace, its plane 

 having more or less of a tilt away from the stream (Fig. 12). 

 The secondary channels above described often temporarily 



