medium,, so that none of its energy is scattered or dissipated 

 into heat, the wave front spreads regularly over a larger and 

 larger area Thus the total energy in the sound wave is spread 

 even thinner and the concentration of energy in any region de- 

 creases regularly with the distance from the source. 



If the source is assumed to be a point , the intensity of 

 the sound which it radiates should decrease inversely as the 

 square of the range „ In actual practice, however, observed 

 transmission losses can depart widely from the values pre- 

 dicted by the inverse square law This discrepancy, which is 

 sometimes called the transmission anomaly 5 is caused by the 

 fact that (1) the ocean is a bounded medium and sound is re- 

 flected by its surface and bottom; (2) changes in the pressure, 

 temperature and salinity of the ocean may result in changes in 

 the velocity of sound and so change both the sound path and 

 the way sound spreads; and (3) the ocean is something less than 

 a perfect medium, so that energy is lost by absorption and 

 scatteringo 



Research work so far has been unable to produce more than 

 average figures for the value of the transmission anomaly un- 

 der various conditions Even when conditions are seemingly 

 constant^ observed transmission losses are highly variable,, 



Transmission is best when the water is isothermal; that 

 is, when commonly-used measuring devices show no temperature 

 variations with deptho When measurements are made with highly 

 directional supersonic transducers in deep water, bottom re- 

 flection can also be ignoredo 



Under such conditions measurements show that a certain 

 portion of the sound is absorbed In each thousand yards of 

 sound travel. The amount of this absorption is measured in. 

 decibels per thousand yards, a quantity which is sometimes 

 called the attenuation coefficient In water which is iso- 

 thermal from the surface to a depth of 200 feet or more the 

 attenuation coefficient is relatively constant at any one fre- 

 quency and is found to increase with increasing frequency,, 

 Thus it is about 3 decibels per thousand yards at 17 kilo- 

 cycles, 4 decibels per thousand yards at 24 kilocycles and 15 

 decibels per thousand yards at 60 kilocycles,, At very low 

 frequencies, on the other hand, absorption is very weak and 

 the attenuation coefficient seems to be almost negligible , so 

 that sonic sounds trapped in a "sound channel" may travel for 

 thousands of miles and still be heardo 



26 



