496 SHEPAKD [chap. 2(1 



their sides. The delta built by the Rhone into Lake Geneva has a valley which 

 extends the entire length from the river mouth to the plain at the bottom of 

 the lake. Tliis ravine, unlike the marine delta-front gullies, has very large 

 natin-al levees and has a sand floor, contrasting with mud on the levees. R. F. 

 Dill has found currents here. 



7. Shallow Straight-Walled Valleys of Submarine Escarpments 



Along many steep submarine escarpments — many of them definitely fault 

 scarps — there are incised valleys starting at various depths and some of them 

 extending to the base of the slope. These have been described by Buffington 

 (1951) and Emery and Terry (1956). These valleys differ from submarine 

 canyons in lacking the dendritic tributaries and in having relatively straight 

 sides. The depth of these valleys below the surrounding sea floor is mostly less 

 than 100 fm. They do not penetrate into the shelf to any extent. They also lack 

 the marginal levees of the fan valleys and are rather closely similar to the gullies 

 on the front of deltas, although they have much steeper gradients and are some- 

 what more deeply incised. There seems to be little doubt that these straight- 

 walled valleys in escarpments, as well as the gullies of the delta fronts (except 

 in fresh- water lakes), are the product of submarine slumping. They have been 

 explained in this way by almost everyone who has studied them. 



8. Origin of Submarine Canyons 



Many hypotheses have been offered to explain submarine canyons, although 

 most of them seem now to have disappeared leaving only two principal con- 

 tenders : turbidity-current erosion and subaerial erosion followed by down- 

 warping during which the canyons have been preserved by submarine processes. 

 A numljer of wTiters now consider that turbidity currents are the only possible 

 solution, but the principal authors of the turbidity-current hy|3othesis (Daly, 

 1936; Kuenen, 1953) have exj)ressed the opinion that at least some of the 

 canyons were cut by rivers. 



In considering the origin of the submarine valleys it seems useless to consider 

 them as a unit having only one explanation. As has been indicated both by 

 Kuenen (1953) and the present writer there are many different types of valleys 

 and there is no reason for believing that they are all of the same origin. It is 

 felt that little can be gained by just giving the arguments favoring one hypo- 

 thesis along with the objections to the other. For many years Kuenen and the 

 present writer have been discussing the hypotheses both in articles and in 

 personal correspondence. The attempt here will be to point out the various 

 arguments showing how new information bears on them but leaving the reader 

 to judge the merits of the case. 



A. Turbidity Currents as a Cause of the Formation of Canyons 



When Daly (1936) first suggested turbidity currents as the agents which cut 

 submarine canyons there was very little evidence to support his idea. It was 



