located at approximately 80 feet and sources and receivers at approximately 

 300 feet. The paths therefore involved transmission through the thermocline. 

 The following sets of experimental data were compared with the results noted: 



Area 1: 700 cps (Curves 1 and 16) 

 1300 cps (Curves 2 and 17) 

 3000 cps (Curves 3 and 18) 



Area 2: 700 cps (Curves 19 and 34) 

 1300 cps (Curves 20 and 35) 

 3000 cps (Curves 21 and 36) 



Good - less than 5 dbs difference 



Fair - less than 10 dbs difference 



Poor - less than 20 dbs difference 



Good - less than 5 dbs difference 



Fair - less than 10 dbs difference 



Good - less than 5 dbs difference 



Since the cases where there is disagreement between theory and experiment 

 represent a relatively small fraction of the experimental results it is clear that 

 the experiment has confirmed the validity of the normal mode theory of surface 

 duct transmission. Its value is particularly clear in predicting the effect of fre- 

 quency on transmission loss (e.g., Curves 13-15). Ray-tracing theories can 

 not, in general, predict the frequency dependence that is demonstrated here, 

 theoretically and experimentally. 



3 .2 Comparison Between Results Obtained in Area I (Boundary Region) and 

 Area II (Transition Water Region) . Comparison between equivalent runs is 

 obtained from experimental curves: 



1 - 6 (Area I) and 19 



7 - 12 (Area I) and 15 



13 - 18 (Area I) and 31 



24 (Area II) 

 30 (Area II) 

 36 (Area II) 



Since the thermal structures (from the bathythermograph data) of the two areas 

 were very similar, the theoretical curves show only minor differences. Hence 

 any comparison between the areas must rely principally upon the experimental 

 results. With the exception of Curves 3 and 21, the experimental data differ by 

 much less than 10 dbs. In several cases, particularly where theory predicts a 

 simple curve, experimental agreement is closer than 5dbs. In general, agree- 

 ment between corresponding sets of experimental data is better than between 

 theory and experiment. 



3 .3 Study of Signal Fluctuations . The preceding comparison of theory with 

 experiment and of experimental results in the two ocean areas has been based 

 principally upon the averaged experimental data. In an attempt to discover 

 more detailed differences, the signal fluctuations were analyzed as described 

 in paragraph 2.3. The results are presented in Table 1 and Figures 2 and 3. 

 The behavior of the A's as a function of frequency is given by the o's calculated 

 from Curves 60 (700 cps), 61 (1300 cps), and 62 (3000 cps). While the absolute 



11 



3rtlmr 2).Hittle,3nr. 



