100 RAITT [CHAP. 6 



averaging Layer 3 velocities of Table I in two separate groups: (1) the group 

 of 17 stations in which Layer 2 was observed; (2) the group of 48 stations in 

 which Layer '2 was not identified. Group (I) had a mean of 6.8.3 km/sec with a 

 standard deviation of 0.31 km/sec. Group (2) had a mean of 6.65 km/sec with 

 a standard deviation of 0.25 km/sec. Hence Group (2) is significantly lower, as 

 expected. This is additional evidence for the widespread occurrence of Layer 2. 

 It indicates that velocity measurement of Layer 3 is more rehable at stations 

 where Layer 2 has been detected. 



The thickness of Layer 3 depends on the intercept of the travel-time segment 

 for Layer 4. It has been pointed out by Hill (1957) that this intercept is sensitive 

 to the value assumed for the Layer 4 velocity. If the estimate of the velocity is 

 too high, Layer 3 will be estimated too thick. Hence the degree of positive 

 correlation between Layer 3 thickness and Layer 4 velocity will indicate the 

 magnitude of the error in the Layer 4 velocity. Hill showed that there was a 

 positive correlation, which indicated that some of the variability of Layer 4 

 velocity was random error of measurement. 



As a conclusion of this argument, one might obtain more estimates of 

 Layer 3 thickness by using a standard average value of Layer 4 velocity 

 instead of the locally determined velocity for each station. This pro- 

 cedure might be recommended for those stations in which the quality of 

 the Layer 4 arrival is obviously poor. However, it cannot be applied universally 

 for there is good evidence that there are real variations in the velocity of 

 Layer 4. For example, certain areas, such as the Mid-Atlantic Ridge (Ewing 

 and Ewing, 1959), parts of the eastern Caribbean (Officer et al., 1959), and the 

 Easter Island Rise (Raitt, 1956), have Layer 4 velocities significantly lower 

 than normal oceanic values. 



The widespread occurrence of Layer 3 and the remarkable stability of its 

 velocity in widely different parts of the world ocean demonstrate that it is a 

 characteristic feature of the oceanic crusts. The 126 tabulated observations of 

 Layer 3 velocity are all well above the velocity of about 6.0 km/sec charac- 

 teristic of the principal refracting layer on the continents (Gutenberg, 1955). 



There were only 8 stations of the 134 in Tables I to VI at which Layer 3 was 

 not identified. Four of these stations, CR38, CR39, CR20 and CR21, were 

 short-range sonobuoy stations in the eastern North Atlantic and western 

 North Pacific and did not reach velocities greater than 5.75 km/sec. This is 

 probably because of an unusual thickness of Layer 2 at these stations. A 

 similar result was found at station HL18 (Hill and Laughton, 1954), where the 

 maximum velocity attained was 5.80 km/sec. The other three stations have 

 anomalous Layer 2 thicknesses or abnormally low Layer 4 velocities. Hence, 

 there is no evidence at all for the absence of Layer 3, and its existence through- 

 out all oceanic regions is well established. 



This characteristic oceanic property of Layer 3 has caused Ewing and 

 Ewing (1959) to call it the "oceanic" layer. It has also been called the "basaltic" 

 layer, which implies that there is some knowledge of its chemical composition. 

 Until we have identified it by direct sampling with a drill, this description of 



