F-5 

 (Stetson, 1949; Gibson et al . , 1968; Ryan et al^. , 1978). 



The origin of submarine canyons has received consider- 

 able attention in the geological literature of the past 50 

 years. Two categories of theories have emerged from a large 

 number which have been proposed (Shepard, 1973, pp. 327- 

 334) : 1) the canyons were originally seaward extensions of 

 now-drowned river valleys which have been maintained by sub- 

 aerial erosion such as turbidity currents; and 2) the river 

 valley walls have been built up by preferential sedimenta- 

 tion. 



Most , if not all, of the east coast submarine canyons 

 can be shown to be extensions of major U.S. rivers, thought 

 to have been more active during late Pleistocene low sea- 

 level stands. In glaciated areas, diversions to meltwater 

 flow were provided by cuestas and moraines (McMaster and 

 Ashraf , 1973) . In some instances, the direct link between 

 river and canyon has no modern bathymetric expression due 

 to post-glacial shelf sediment movements, although shallow 

 seismic studies have revealed buried channels (Pratt, 1967; 

 Knott and Hoskins, 1968; Knebel et al . , 1979). 



All the canyons appear to have high leveed right banks 

 (Pratt, 1967) . Rona (1970) has proposed a sedimentation 

 model for canyons, supported by reflection data in his and 

 other studies, in which intercanyon areas are sedimented 

 while the canyons themselves are swept free of sediment. 

 The resulting strata become convex upward and geostrophic 

 contour currents (Heezen et al., 1966) preferentially depo- 



