Mr HiKjli Miller on Pdve/r- Terracing. 297 



behind it, there exists a double reason for this division into 

 gorges and compartments. In the North of England, wliere 

 the observations made by the writer chiefiy lie, nearly all the 

 valleys date from before tlie glacial period (a circumstance 

 common to Scotland, Ireland, and other glaciated countries), 

 and still contain quantities of glacial deposits. Towards the 

 coasts the bottoms of these old valleys are found to lie much 

 below the present sea-level ; and even at elevations extend- 

 ing beyond 1000 feet into the interior the country is tracked 

 by drift-filled hollows branching as they rise, that lie deeper 

 than the present water courses, and have all the characters 

 of preglacial stream-channels. The face of the country was 

 not so entirely changed, however, by the erosion and boulder- 

 clay deposit with which the glacial period was accompanied 

 that the rivers were forced to take up entirely new lines of 

 flow after their return. The grouping of the ground remained 

 the same. The same high grounds directed the drainage down 

 npon the same general lines of hollow. 



The streams, therefore, took their way over the surface 

 of deposits under which their former channels lay more or 

 less buried. Sometimes they chanced to regain a position 

 almost mid-valley ; more often they ran more or less to one 

 side of the middle; not imfrequently they lost their w^ay 

 entirely. Their post-glacial excavations accordingly have 

 been conducted ; in the first case, entirely in superficial 

 deposits, with variable reaches of the same on either side ; in 

 the second case, first in boulder clay, etc., and then in the 

 rock of the old valley-bank ; and, in the third, almost 

 entirely in rock. It occasionally happens that in a single 

 sweep of its course a stream may pass over all three. 



It thus comes about that while the narrowing and broaden- 

 ing of the preglacial valleys was determined by the hardness, 

 otr if the term may be allowed, denudeallcncss of the rocks 

 among which they were hollowed, the width of the post- 

 glacial re-excavations has been determined mainly by their 

 relation to their half-buried predecessors. Their wider basins 

 answer partly to the compartments of the older valleys, and 

 partly to spaces in which the river has encountered soft 

 reaches of boulder clay. The glacial period, upon the whole, 



