80 EWING AND NAFE [(HAP. 5 



Fig. 5 shows four records of the reflections at ranges of 2.6, 3.5, 4.4 and 5.4 sec 

 from Profile 99. Note the sharp change in character and strength of Rn in this 

 relatively short range. The same behavior was noticed in Profile 98 and is 

 a common feature of most sub-bottom reflection j^rofiles in deejj-water 

 areas. 



The 4-5 interface is determined by the refracted arrivals corresponding to 

 lines G4 and by reflected arrivals Rm. Samples of the Rm signals are shown in 

 Fig. 5 in addition to the Rn arrivals. In the case of the 4-5 interface, note that 

 there is perfect agreement between reflected and refracted arrivals (Fig. 1) in 

 that the G4 lines are tangent to the Rm curves. 



As mentioned earlier, it is not unusual to record strong reflections from 

 within the sedimentary column without receiving the corres])onding refracted 

 arrivals. This can be an indication that the reflector is a thin layer, or it could 

 in some cases be explained if the velocity discontinuity is to a lower velocity 

 below. Still another explanation can be that the discontinuity is one of density 

 rather than of velocity, i.e. Ci^C-z ; piCi # P2C2. 



The foregoing has been an attempt to describe the methods of obtaining 

 velocity structure in the sediments in deeji-ocean areas, and some specific 

 results have been shown. Although the cases discussed were confined to com- 

 parison with linear gradients, comparisons have been satisfactorily made with 

 parabolic and exponential curves. It is improbable that any one type of gradient 

 is applicable everywhere, and, indeed, in most cases the best fit might be 

 obtained with a combination, i.e. linear to a certain depth, and parabolic 

 below that. Previously reported results on gradients measured by seismic 

 techniques have been given by Hill (1952), Officer (1955), Katz and Ewing 

 (1956) and Nafe and Drake (1957), in addition to laboratory measurements on 

 artificially compacted sediments by Laughton (1954). These results and others 

 unreported have shown average gradients, principally from Atlantic Ocean 

 profiles, to vary between 0.4 and 2.5 sec^^ for the upper 0.3 to 1.0 km of sedi- 

 ments. The variations are undoubtedly due in part to observational or inter- 

 pretational diff"erences. Some profiles are shot with more appropriate charge 

 sizes or shot spacing than others, and some have much simpler interpretations 

 than others. However, a considerable part of the variations is probably real, 

 indicating diff"erences in sediment t3^es, rates of sedimentation, porosity, 

 grain size, lithification, and other factors affecting seismic velocity (Sutton, 

 Berckhemer and Nafe, 1957). At present there has not been a sufficient number 

 of determinations to permit the classification of areas by sediment velocity 

 structure except possibly, as suggested by Nafe and Drake (1957), to predict 

 that in areas where calcium carbonate deposition is high, sediment velocities 

 will also tend to be high. Most of the better determinations have been made in 

 abyssal plains or other flat-lying areas in order to avoid topographic complica- 

 tions, and results from areas which receive no turbidity'current deposits, if 

 available, might show significant differences. 



Table I gives values of K from several profiles determined by fitting the 

 theoretical curves for linear gradients. The highest value is that measured by 



