944 MISCELLANEOUS GEOPHYSICAL METHODS [Ciiai'. 12 



vertical velocity variation in water is often much the same as that in 

 air but it occurs on a smaller scale. In some experiments described by 

 Swainson,"* the velocity decreased rapidly for the first 50-100 fathoms, 

 then leveled off, and increased again from about 500 fathoms on down 

 (1 fathom = 6 feet). The temperature decreased from 14° to 9° C. to 100 

 fathoms and stayed constant at about 4° from 500 fathoms on down. 

 The salinity increased from 33 to 34.2 thousandths as far as about 200 

 fathoms and stayed constant from that depth on down. These levels 

 change to a certain extent with the seasons. The result of such velocity 

 variations is that the sound rays are first bent away from the horizontal 

 and curve toward it again at greater depth. 



Contrary to air, the energy does not return to the surface by refraction 

 but only by reflection at the ocean floor. Long-range transmission does 

 not occur by the direct path but by multiple reflections on both, the ocean 

 bottom and water surface. According to Swainson's observations, ^^^ the 

 direct ray was recorded up to 20 km, once-reflected waves up to 70 km, 

 twice- to five-times-reflected waves up to 85 km (and possibly more). At 

 that distance the travel time was about 1 minute. These ranges hold for 

 bomb explosions and not for continuous waves. When the depth is small 

 compared with distance (as in most marine-acoustic communication prob- 

 lems, except echo-sounding) the time "delay" t due to the reflection is 

 small compared with the total travel time. According to Swainson,"^ the 

 time difference between the first and fifth reflections at 85 km distance 

 was about 1 second; thus, the delay for each reflection was j second. 

 Although the delay is, strictly speaking, dependent on distance (since the 

 travel-time curves are hyperbolas, see page 557) and varies in value from 

 one reflection to another, it is satisfactory to write 



s = y(t - nx) (12-22) 



for distance determinations from the travel time, for n reflections. 



The velocity of sound waves in water increases somewhat near the 

 source and decreases slightly with an increase in frequency (in the super- 

 sonic range). 



The attenuation of sound waves in water and air is governed by the same 

 relation (formula [12-17]) as far as viscosity damping is concerned. The 

 coeflacient a/f is 1.45 10"" cmsec."^ for air, and 8.5-10"" cmsec."^ for 

 water. For the latter, the attenuation is therefore about 1700 times less. 



i"0. W. Swainson, "Velocity of Sound Waves in Sea Water," U. S. Coast & 

 Geod. Surv., Spec. Rep., Feb. 28, 1936. 

 "s Ibid. 



"^ L. Bergmann, Ultrasonics, Wiley (1939). 



