338 



HEEZEN AND LAUGHTON 



[chat, 14 



expect if they were controlled by Coriolis forces. Menard lias put forward an 

 explanation of the left hook by considering the top surface of a turbidity 

 current flowing in a canyon to be tilted and an asymmetry to be produced in 

 the levees built u]) on either side so that the chamiel is continually forced to 

 migrate to the left. He joroposes a general rule for the orientation of leveed 

 turbidity-current channels: "A turbidity current channel flanked by levees 

 has a tendency to deviate from a straight line which varies directly with the 

 amount of the deflecting force of the earth's rotation but is in the opposite 

 direction" (Menard, 1955). However, not all turbidity currents from canyons 

 can be confined to the chainiels on the fans, otherwise the fan may not be built 

 up. Some flows would be too large to be contained by existing channels and 



WNW 



ESE 



ft 

 3000- 

 3300 

 3600 -• 



h 3000 ft 



WNW 



ESE 



fm 

 500 



,-550 

 600 



(b) 



ft 



1500- 

 1800- 

 2100- 



h 3000 ft 



SEE 



(d) 



NNW 



fm 

 -250 

 -300 

 - 350 



ft 

 1200 - 

 1500 -| 

 1800- , 



K3000 ft 



fm 



•200 



250 



300 



(c) 



ft 

 1800- 

 2100- 

 2400- 



W 3000 ft 



NNW 



SSE 



fm 

 •300 

 ■350 

 ■400 



(e) 



Fig. 19. Natural leveetl submarine canyons off Southern California. (After Buffington, 

 1952.) 



ANould spread laterally in the form of a sheet flow. The more rapid decay of 

 the sheet-flow turbidity currents due to the two-dimensional spreading would 

 result in the deposition of their sediment load near the base of the canyon and 

 consequently in the formation of a deep-sea fan. For smaller flows, there might 

 be sufficient topographic control to prevent two-dimensional spreading so that 

 decay would be less rapid and ^\'ell defined channels would be cut. 



B. Mid-Ocean Canyons 



in li>49, four canyon-like de])ressions were found in the abyssal floor east of 

 Newfoundland (Figs. 20 and 21). A subsequent survey showed that these were 

 part of a continuous mid-ocean canyon running for some 1500 mi from the seas 

 around Greenland and feeding into the Sohm Abyssal Plain through a narrow 

 gap in the South East Newfoundland Ridge (Ewing, Heezen, Ericson, Northrop 

 and Dorman, 1953; Heezen et al, 1959). Sections of the canyon ty|:)ically 

 have steep walls, sometimes even vertical, and flat floors. They vary in 



