Maw — On Watersheds. 345 



either below the siu'face by the action of currents, or on the 

 coast line. 



In comparing the form of the sea bottom, and the land surface, it 

 ought not to be overlooked that just as the land has been subject 

 to marine denudation so the general form of the sea bottom may at 

 one time have been influenced by subaerial action, and assimilated 

 thereby to the form of the present dry land. 



There is, however, one essential difference with regard to the 

 much larger proportion of isolated depressions that occur in the 

 ocean bed than on the land surface ; as a rule the land consists of a 

 graduated system of levels leading into each other as a connected 

 series of watersheds, and the exceptions to it in the form of com- 

 plete hollows are exceedingly rare ; the sea bottom on the other 

 hand is full of isolated depressions which would be left, on emer- 

 gence, as unconnected pools and lakes, or isolated seas ; this may be 

 well observed on a miniature scale at low water on almost any 

 shallow coast, and still better on a map of the sea bottom giving 

 with sufficient detail the lines of equal depth. It seems impossible 

 that such hollows should have been produced by any directly 

 denuding force for which a line of approach seems essential. 



Are not, therefore, these close sea valleys, surrounded on all sides 

 by higher ground, invariably the result of accumulation ? Similar 

 depressions are not at all uncommon on the drift surfaces of Shrop- 

 shire and Cheshire, and I believe that the close basins containing 

 the meres and pools of these counties occur exclusively on old sea- 

 bottom-surfaces of drift, or are the result of drift barriers closing up 

 an ordinary watershed valley, and that a hollow of denudation of 

 any extent, withoiit an outlet at its lowest level, if it exists at all, 

 is a phenomenon of the greatest rarity. 



Lake basins, which seem to admit of special explanation, must, of 

 course, be excepted ; also, close valleys and depressions connected 

 with swallow-holes, which are virtually complete watersheds in 

 miniature. 



Apart from the fact that the sea bottom as a rule is subject to 

 accumulation rather than to any denuding process, on what possible 

 theory can any system of marine currents excavate such a delicately 

 graduating and ramifying system of levels and valleys as those 

 forming the land's surface ? In the first place, marine currents, 

 though occasionally diverted by shallow barriers, are not as a rule 

 coiacident with the form of the bottom, and to refer the present 

 shape of the groiind to the action of former mariae currents, you 

 must assume that the greatest force was expended in producing the 

 greatest depths, and this at once presents the difficulty of a system 

 of currents diverging from the greatest depths and ramifying and 

 exhausting themselves in every direction up what are now the river 

 valleys to the lines of watershed ; indeed, to be consistent, you must 

 localise a branch of the supposed submarine current to fit every little 

 depression of the ground that leads up from the watershed valleys. 



Let us now look at the sea as a denuding agent on the coast line. 



K marine action has been the exclusive cause of the moulding of 



