OCEANIC AND FLUVIATILE ACTION. 91 



site extremities ; but I do not recollect a case in which their drainage is effected by 

 several parallel outlets. 



A case to illustrate this principle occurs in the valley of Connecticut river. 

 The trap range, called Ilolyoke and Tom (see Plate III), crosses that valley, or 

 nearly so, obliquely; its northern extremity (Holyoke and Norwottuck), turning 

 across the valley almost at right angles. Its crest is crossed by numerous valleys 

 of erosion, of very unequal depth, in one of which Connecticut river now runs. I 

 have no doubt that the drift agency has had a good deal to do with their erosion, 

 yet such is the situation of these valleys, that when the ocean gradually receded 

 from the surface, the waves and tides must have acted with great force in the 

 manner above described. During the last depression of this region below the 

 ocean, the drift agency probably swept over the ridge and modified the small 

 valleys ; but probably the ocean did most of the work long before. I should go 

 more into detail in respect to this case, had I not already done so in my Report on 

 the Geology of Massachusetts, and in my Elementary Geology. I have not, how- 

 ever, in those works advanced the above hypotheses to account for the denudation, 

 but have merely inferred that water and ice must have been the agents. 



5. If the face of a mountain be steep and show marks of denudation ; if it be 

 an outlier; that is, have no corresponding eminence opposite, so as to form a 

 valley, and if there be no evidence of a dislocation of the strata, we must impute 

 the erosion to oceanic agency ; since fluviatile agency is out of the question. But 

 if, while the continent was sinking or rising, its waves and currents beat against 

 the mountain, it might so wear away the strata as to leave a mural face. In this 

 case, however, we must suppose a previous inequality of surface, so as to enable 

 the waves to act upon the shore. 



6. The main force of the ocean is directed towards the axis of mountain chains, 

 although tides and currents will be parallel thereto. Hence the eroded valleys 

 will run towards the crest of the mountain chain. If, therefore, we find valleys 

 running very much oblique to the axis, we may presume them to be formed by 

 rivers. It must be remembered, however, that the direction of the strata will 

 greatly modify the direction of erosions, as in the case of fiords. The unequal 

 hardness of the strata, also, will operate in the same way : so that the application 

 of this distinctive mark will require caution. 



7. In a few cases, where a river has worn a passage through a mountain ridge, 

 and at the same time has, from time to time, made lateral changes in its bed, it 

 might leave a succession of precipices, which were its former banks, on one side, 

 while on the other, they might be worn away. In such a case, however, we must 

 suppose that the stream, after excavating a bed, should suddenly desert it; else, if 

 the lateral change were slow and equable, it would leave on the deserted side, only 

 a uniform slope. 



In suggesting this distinction, I have had a particular case in view,. which I will 

 shortly describe ; but about which I am in doubt, whether to refer it to oceanic 

 or fluviatile action, or to both united. 



Modes and Extent of Erosion by Rivers. — 1. The manner in which rivers are 

 formed, as a continent rises from the ocean, has been described in my paper on 



