SECT. 2] SUBMAKINE CANYONS 497 



known that the relatively heavy water from rivers which entered certain 

 reservoirs sank below the surface developing an under-flow which sometimes 

 followed the bottom as far as the dam. Later Kuenen and Migliorini (1950) 

 showed in a large experimental tank that turbidity currents could transport 

 sand and other coarse sediments along the sloping bottom without much loss 

 of material. Next came the finding of sand layers on the sea floor at the foot of 

 submarine canyons, first reported by Scripps Institution operations (Shepard, 

 1951) but better documented by Lamont Geological Observatory (Ericson et al., 

 1951, 1961). These sands, along with occasional gravel deposits, indicated that 

 rather appreciable currents must sweep along the submarine canyons in order to 

 carry out such coarse material along the canyon lengths. It has gradually been 

 found that there are huge fans at the base of many of the submarine canyons 

 probably containing much greater quantities of material than would come from 

 the mere excavation of the canyons on the slopes above (Menard, 1960). 



To many investigators the most convincing argument favoring turbidity- 

 current erosion has been the suggestion by Heezen and Ewing (1952) that the 

 successive outM'ard breaks of the Grand Banks cables in 1929 w^ere the result of 

 turbidity currents moving down the slope after the great earthquake of that 

 year. Taking the times and distances between cable breaks they estimated that 

 the currents at first were moving at a speed of 55 knots and then slowed 

 gradually to 12 knots at the outer portion of the breaks. A re-examination of 

 I the data, however, seems to indicate that there is little basis for their estimate 

 of such excessive speeds (Fig. 15). Heezen and Ewing appear to have overlooked 

 the fact that at the time of the earthquake the cables broke instantaneously for 

 a distance of about 100 nautical miles from the epicenter so that turbidity 

 currents could also have started immediately at these considerable distances. 

 With this in mind a more reasonable interpretation appears to be that the 

 times between cable breaks indicate velocities of 15 knots without much, if any, 

 change of speed from top to bottom of the slope. 



If turbidity currents move at even a 15-knot speed, they could well be 

 considered as an erosion agent since this is faster than most rivers flow. There 

 is, however, an alternative explanation for the time intervals between breaks. 

 Terzaghi (1956) has suggested that the advancing front of a wave of spon- 

 taneous liquefaction spread down this slope after the earthquake. Terzaghi 

 considered that the character of the broken ends of the cables later recovered 

 was suggestive of the irregular disturbances accompanying spontaneous lique- 

 faction and that an advance of 15 knots is not out of line with other documented 

 instances of spontaneous liquefaction. Such an explanation was well received 

 by C. S. Lawton, General Plant Engineer of Western Union. It seems possible 

 that the spontaneous liquefaction could have caused turbidity currents which 

 might have acted in conjunction with the spontaneous liquefaction. 



Although Heezen, Ericson and Ewing (1954) have argued that the sediments 

 in the cable-break area support the idea of a tremendous turbidity current, it 

 must be pointed out that their evidence is largely negative, consisting of not 

 having recovered cores and finding bent core tubes in an area near the outer 



