SECT. 2] SUBMARINE CANYONS 503 



rather surprising that similar features are lacking along the outer portions of 

 such canyons as that of Monterey Bay (Fig. 7). This canyon can be traced sea- 

 ward for 100 miles (including the outer trough) across a zone that must have 

 undergone extensive faulting provided the canyon was submerged in the 

 Miocene. This could be considered likely in view of the Miocene filling of the 

 deeply submerged canyon under the Salinas Valley (Starke, 1956). As a possible 

 reply to this objection one can suggest that there have been displacements from 

 time to time on the sea floor but that the effects of landslides and turbidity 

 currents have re-established the contacts and reconnected the severed canyon 

 sections. Such action on the part of turbidity currents requires far less erosion 

 than does the complete excavation of the canyons. It is also possible that the 

 soundings are not sufficiently detailed as yet to reveal such breaks. Further- 

 more, if the continuity is an objection to the drowned valley hypothesis, it 

 should also be an objection to canyons cut by turbidity cm-rents unless such 

 cutting has occurred in very recent times (see below). 



The idea that the submarine canyons off the east coast of the United States 

 were cut originally by rivers runs into some serious difficulties which have 

 been discussed by Kuenen (1953). As he has emphasized, the upper walls of 

 these canyons are cut through Tertiary sediment, some of it as young as 

 Pliocene. Therefore, he argues, the canyons must have been cut since that time 

 and there is no evidence of any uplift of the coastal margin, at least in such 

 recent periods. The writer made an alternative suggestion that the canyons 

 were cut before the Tertiary deposits were laid down and were kept open as 

 Tertiary material accumulated on the shelf slope (Shepard, 1952), This is 

 rejected by Kuenen on the ground that such a history would be recorded by 

 finding nick points on the wall slopes of the canyons, with gentle slopes above 

 indicative of the post-canyon cutting deposition. Kuenen's profiles do not 

 show such nick points, but the results of using an acoustic probe on the sides 

 of Hudson Canyon (Ewing et al., 1960) show that, near the rim, the upper beds 

 dip toward the canyon but have been covered by other beds that lie nearly 

 horizontal and no nick point is shown along the slope. The writer is inclined to 

 think that the absence of nick points is due to the accumulation of mud on 

 the canyon slopes largely masking the outcrops and developing a slope that is 

 temporarily stable but too steep for permanency. Such steep slopes are well 

 kno%vn on the canyons of southern California where they have been seen by the 

 aqualung divers. Dredging along these mud slopes has encountered occasional 

 rock layers which protrude slightly. In any event the objection offered by 

 Kuenen cannot be dismissed lightly and it may mean that the subaerial erosion 

 hypothesis does not apply to the east coast canyons, as Kuenen believes. As 

 we have seen, these canyons are rather different from the others and may 

 require different explanations. On the other hand, it is hard to agree with 

 Kuenen that the canyons off the French Riviera are drowned river valleys 

 whereas the neighboring canyons off the Gulf of Lyons are due to turbidity 

 currents. Such an explanation requires a large sea-floor displacement at the 

 juncture, for which there is no evidence (Fig. 9). 



