CORRECTIONS OF SONIC DEPTHS ON ACCOUNT OF ERRORS IN TIMING 



57 



The velocity of sound, determined from experiments 

 in which the source of sound is an explosion, is greater 

 than when the source used is a diaphragm vibrating with 

 constant amplitude. Further, the difference in velocities 

 is dependent on the distance involved and the violence of 

 the explosion. This has been explained as being the re- 

 sult of a sound wave train of normal velocity superim- 

 posed on an explosive wave which suffers great attenua- 

 tion. On this assumption the greater initial velocity is 

 a transient phenomenon, after the disappearance of which 

 the velocity becomes normal. 



Following this line of reasoning, one should expect 

 to find that the soundings taken on the Carnegie with the 

 improvised shotgun would have to be corrected for this 

 effect. On thirty-four occasions the time of echo was 

 measured for more than the first echo. These times of 

 echo were accordingly investigated and are given in 

 table 3. In one case the time was measured from the 

 explosion to the return of the fourth and fifth echoes, and 

 in thirty-three cases times were measured for first and 

 second echoes. Of the thirty-three cases there were two 

 in which the time was recorded as being doubtful, and in 

 another case an error of one second was apparently 

 made in reading the stop watch for the time of the sec- 

 ond echo. In the remaining thirty cases the time for the 

 first echo was subtracted from the time for the first two 

 echoes to obtain the time for the second echo. The dif- 

 ferences between the times for the first and second 

 echoes were then taken. In the case where the times for 

 four and five echoes were measured, it was assumed 

 that the times for the second and succeeding echoes 

 were the same. From this, the time for the first echo 

 was computed and compared with the time for the fifth 

 echo. In these thirty -one cases the average difference 

 between the times for the first and succeeding echoes 

 was 0.113 seconds. 



The echo times have, therefore, been corrected on 

 the assumption that the times of all first echoes, as 

 measured, were too small by this amount. This was 

 done by adding 85 meters to the gross depths based on 

 first echoes alone, 64 meters to the gross depths based 

 on first and second echoes, and 42 meters to the gross 

 depths based on second echoes alone. The shotgun sound- 

 ings adjusted for this correction were then compared 



with the wire depths and depths determined from pres- 

 sure thermometers at the twenty oceanographic stations 

 where such measurements were made. Of the twenty 

 comparisons, the one made on Merriam Ridge (station 

 67) has been omitted because of the steep bottom slope 

 in this vicinity. Because of the great distances between 

 shotgun soundings, there is no way of telling when the 

 bottom was regular and when irregular except in the 

 previously mentioned case of Merriam Ridge. There- 

 fore, as the shotgun soundings and accepted depths are 

 only approximately simultaneous, a greater scatter must 

 be expected than was found in the comparison of sonic 

 depth finder soundings with accepted depths. A greater 

 scatter must also be expected because a stop watch 

 measurement of echo time is not as precise as an echo 

 time measured by the sonic depth tinder. Table 4 gives 

 the approximately simultaneous depths as determined by 

 wire and pressure thermometers and as given by the 

 shotgun soundings corrected for greater intial velocity. 



Comparison of the shotgun depths with correspond- 

 ing wire and thermometer depths shows the shotgun sound- 

 ings consistently greater (see fig. 1). The mean shotgun 

 depth for the nineteen cases, excluding station 67, is 

 3621 meters, and the mean wire depth is 3471 meters, 

 giving a ratio of means of 0.C58+. The mean of the sin- 

 gle ratios is 0.960- with a probable error of +0.006. 



In view of this a timing factor of 0.958 has been 

 adopted and used for the correction of shotgun soundings 

 after they were adjusted for the greater initial velocity 

 resulting from the explosive character of the source of 

 the sound. This makes the assumption that the stop 

 watch had a large gaining rate. In May 1928 the stop 

 watch was compared with a chronometer at the beginning 

 and end of a two-hour run, and it checked so closely that 

 it was considered permissible to use it as a time stand- 

 ard in the calibration of the depth finder timing. The 

 depth finder was then adjusted until it was in agreement 

 with the stop watch over a period of about fifteen minutes. 

 The depth finder was similarly adjusted in February 1929 

 after it had been repaired and overhauled. The same 

 stop watch was used in measuring the echo times for the 

 shotgun soundings. In all three instances, that is, during 

 the period from May 1928 to February 1929, the period 

 from February 1929 to November 1929, and in -the shotgun 



Table 2. Summary of comparison between sonic depths and thermometer or wire depths 



at stations 78 to 162 



