SECT. 1] CONTINUOUS REFLECTION PROFILING 53 



frequencies than at high, hydrophones having a high-pass frequency charac- 

 teristic response are sometimes selected. Sensitive hydrophones are required 

 but there is no need in the present state of the art for especially low noise 

 properties compared with the quietest ambient levels for underway 

 observations. Recently, especially in the work of John Ewing and his co- 

 workers at Lamont, the so-called "slacking technique" has been used success- 

 fully to obtain records of high quality in deep water from a ship proceeding at 

 10 knots. The slacking technique, long familiar in oceanic seismic-refraction 

 research, is to slack the hydrophone towing cable seconds before the desked 

 wave train is to be received, thus placing the hydrophone nearly still in the 

 water. The Lamont procedure while profiling is to shoot small charges of 

 exf)losive on a two-minute schedule, slacking the hydrophone cable anew for 

 each shot and then hauling in cable before the next. 



A long array of hydrophones designed to be uniformly neutrally buoyant 

 has also been employed by several groups as a means of improving the signal- 

 to-noise ratio at high towing speeds. The recordings in Figs. 14 and 15 were 

 made with a 40-ft array of five hydrophones. By such arrays useful observations 

 have been made in the deep ocean at towing speeds of 8 knots. This speed does 

 not appear to be an upper limit. 



C. Pre-amplijiers 



Conventional broad-band amplifier designs have been used ; they are being 

 altered and improved almost continually. Hersey (1957) describes an early 

 design by Willard Dow. 



D. Filters 



Commercially available passive and electronic filters have usually been 

 used. For good transient response it is well to avoid filters relying on sharply 

 resonant circuits for sharp cut-off and, of course, care must be taken to match 

 the filters to surrounding circuitry. In using the seismic profiler we have pre- 

 ferred, whenever possible, to split the pre-amplifier output feeding it to one 

 high-frequency and one low-frequency filter. The high-frequency band is 

 chosen either to discriminate between side-echoes from a rough bottom and 

 sub-bottom reflections, or more rarely for high resolution of shallow echoes. 

 The low-frequency band is chosen for maximum penetration. The two bands 

 are recorded on a dual-channel graphic recorder. 



E. Recorder Drivers 



Little detailed information has been 2:)ublished about these units. The 

 characteristic problem of seismic exploration recording is the very large con- 

 trast in amplitude between high-energ}^ arrivals travelling only a short distance 

 and the faintest reflections, some of which may have short travel times as well. 

 Some kind of compression is almost mandatory (otherwise the recording paper 



