RADIO ACOUSTIC RANGING—ADAMS 233 
accuracy the average temperature of the water through which the 
sound wave passes must be known within approximately 1° C. 
During a hydrographic survey controlled by R. A. R., frequent tem- 
perature observations are made. Observations from the surface to 
the bottom are made at various places and times, supplemented by 
more frequent observations at the surface and the bottom. Tempera- 
tures are measured with one or more reversing thermometers attached 
to a sounding wire and lowered to the desired depth. The thermome- 
ter reverses as soon as it starts upward and breaks the column of mer- 
cury so that the value registered at the greatest depth can be read 
after the thermometer has been brought to the surface. 
A bathythermograph, a comparatively new instrument, is also used 
to measure water temperatures in the upper 75 fathoms, where the 
greatest variation occurs. This instrument records automatically and 
graphically the temperatures with reference to depths. 
The variation in salinity normally encountered affects velocities of 
sound only slightly as compared with temperature, but salinity does 
vary and its value must be determined. In the Coast and Geodetic 
Survey the salinity is determined indirectly by measuring with a 
hydrometer the specific gravity of a water sample. An accuracy of 
about one-tenth part per 1,000 may thus be obtained. 
Velocity of sound as determined from the physical characteristics of 
the water and from tables is obviously the velocity of propagation of 
the sound wave, irrespective of direction. The effective horizontal 
velocity is required in R. A. R. If the sound wave is refracted, or is 
reflected from the bottom one or more times en route to the receiving 
station, as is explained later, it is apparent that the theoretical velocity 
alone will not suffice for use in R. A. R. In such cases, the measured 
time intervals are greater than they would be if the sound traveled 
by a direct horizontal path. 
Under certain conditions the effective horizontal velocity of sound 
at a place can be determined experimentally. If a bomb is exploded 
at a known horizontal distance from a receiving station and the time 
interval from the explosion to its receipt at the receiving station is 
measured, the distance divided by the time interval will give the effec- 
tive horizontal velocity of sound between the source and the receiver 
for the temperature and salinity of the water at that place and time, 
irrespective of the path of the sound wave. Where the depths of water 
permit, it is customary to make such tests throughout an area being 
surveyed and at intervals during the survey. For a survey in uniform 
depths where the temperature and salinity are relatively constant, 
results of such tests can be subsequently used in determining R. A. R. 
positions. Where the temperature and salinity change frequently, 
the velocity of sound determined by tests can be modified to take into 
