must be separated strictly by statistics and a knowledge of the approximate 

 configuration of the bedrock surface gained from drill-hole study. Bedrock 

 velocities vary over the study area, but most are between 15,000 and 19,000 

 feet per second. The 19,000 feet per second velocities are in the Castle 

 Neck-Essex area. The bedrock is largely Salem Gabbro Diorite and Cape Ann 

 Granite CClapp, 1921). 



Seismic velocities can be summarized as follows: 



Material Velocity (feet per second) 



dry sand 800-2500 



water-saturated sediments 4700-5300 



(sand, gravel, and sandy till) 



compact till 6000-9000 



bedrock 15,000-19,000 



There are situations where a continuous spectrum of velocities exists from 

 the driest sand to the most compact till. For this reason, it is difficult 

 if not impossible to ascertain the texture of underlying material by simply 

 determining its velocity. 



3. Graphic Interpretation 



Interpretation techniques and procedures were standarized during the 

 reduction of the study data. Interpretation of dipping bedrock was studied 

 in detail. A back projection (on the time-distance plot) from the first 

 phone showing a bedrock return, using the true velocity of the bedrock, 

 yielded the time intercept used in the computation of depth. On every 

 time-distance plot with arrival times from both sediments and bedrock it 

 is usually obvious which phone received the first bedrock arrival. The 

 bedrock returns lie along a line of flatter slope, and are usually scattered 

 about this line in a more random fashion than the sediment arrival times. 

 In homogeneous sands, it is not unusual for the arrival times to lie pre- 

 cisely along one line. When in doubt about which phone carried the first 

 bedrock arrival, it is acceptable to choose the next phone from the shot- 

 point as the basis for the back projection. Figure 14 shows a back pro- 

 jection of true bedrock velocities on a time-distance plot that might 

 otherwise be difficult to interpret by the time-intercept method. This 

 technique can also be used for a dipping till-layer that is yielding an 

 apparent till velocity rather than a true one. 



Irregular bedrock surfaces are a problem. Relief on bedrock topography 

 can be computed by a method presented by Meidev (1960) and reproduced in 

 Figure 15 . Empirical results for this study area suggest that use of 

 1/2 AT rather than AT may yield better estimates of bedrock relief. Dipping 

 and irregular bedrock surfaces caused a lack of agreement between adjoining 

 seismic lines. Seismic lines which share a common shot hole should yield 

 depths of close agreement. However, where irregular bedrock surfaces pre- 



22 



