270 Proceediiifjs of the Royal Physical Society. 



Professor James Geikie * have recognised rivers as seeking 

 down from level to level without necessary aid from changes 

 at the sea- coast. Terrace-making is connected, by Dr Arch. 

 Geikie, with the winding and looping of the river, prolonged 

 erosion carrying it to levels from which it can no longer 

 reach its former flood-plains, whereupon it works into new 

 ones. Methods of river-terracing receive some admirable 

 illustration from C. Barrington Brown's paper "On the 

 Ancient Eiver Deposits of the Amazon " (Quart. Jour. Geol. 

 Soc, 187^, p. ?S3). 



Hitchcock's view of the terraces of the Connecticut has 

 since been more clearly stated, without his unfortunate 

 diagram, by Upham,-f* who confirms the non-correspondence 

 of the lower terraces on the opposite sides of the valley. 

 He accounts for the highest terraces, which he finds to stand 

 more strictly vis-a-vis, by the great floods of the Champlain 

 or Glacier-melting period. In British Columbia G. A. 

 Dawson finds similar facts to be accounted for. Automatic 

 river-action explains the lower terraces ; but he believes flood- 

 ing by submergence necessary to explain the uppermost 

 level. J An admirable reference to the subject of river- 

 terracing, evincing the widest knowledge, is also made in 

 G. K. Gilbert's " Geology of the Henry Mountains," a 

 section of which contains a truly scientific treatise upon 

 denudation in general. 



The Huttonian view of the planing-out of river terraces 

 at successive levels has found several rivals, which can 

 scarcely yet be considered out of the field. Mr Gilbert, in 

 the work just cited, points out that it is a mistake to suppose 

 that the higher terraces were made by a great volume of 

 water when the valley was as open as now. Mr A. Tylor, 

 however, has maintained that the flint-bearing terraces of tlie 



* Preliistoric Europe, 1881, p. 131. In tlie diagrammatic sections given by 

 the three authors last named, the river terraces are represented in opposite 

 pairs. 



+ W. Upham. The northern part of the Connecticut Valley in the Cham- 

 plain and Terrace periods (Amer. Jour, of Science, 3d series, vol. xiv., p. 459, 

 1877). 



X Superficial Geology of British Columbia (Quart. Jour. Geol. Soc, xxxviii., 

 p. 274). 



