3 .4 Conclusions . From the foregoing discussion of the results of this experi- 

 ment the following conclusions can be drawn: 



1 . The validity of the normal mode theory of acoustic propagation 

 in a surface duct is generally confirmed. Details of the pre- 

 dicted theoretical transmission loss curves may be masked by 

 temporal fluctuations in the received signal strength . 



2. Neither the analysis of transmission loss nor the analysis of 

 signal fluctuations has revealed any significant acoustical dif- 

 ferences between the two experimental areas . (But see Para . 

 3.5.) 



3 . The design of an experiment to determine the influence of in- 

 ternal waves upon acoustic transmission should include con- 

 sideration of the following: 



(i) The use of acoustic source and receiver at a con- 

 stant range and the use of an acoustic frequency 

 for which the transmission loss curve does not 

 vary strongly with range . 



(ii) In deep water, long wave lengths and large ampli- 

 tudes are associated with internal waves, see, e.g., 

 Reference (3). Accordingly, they will affect sur- 

 face duct transmission through the spatial and tem- 

 poral variations they cause of the depth of the duct 

 and the velocity gradient. It is desirable therefore 

 to select an acoustic frequency whose transmission 

 loss depends strongly upon the width of the duct and 

 the acoustic velocity gradient. 



(iii) Measurement of the internal wave structure at 



several points along the acoustic path, simultane- 

 ously with the acoustic transmissions . 



It should be noted that conclusions (i) and (ii) above would restrict the experi- 

 ment to very long range in the channel or to a single frequency for which trap- 

 ping of only the first mode occurred. In consequence the applicability of the 

 results of the experiment would be limited . 



3 .5 Further Analysis of Signal Fluctuations . We have deferred to this final 

 paragraph an additional set of comments on the observed fluctuations in trans- 

 mission loss. While they appear to suggest a possible difference between Area 

 I and Area II, there is inadequate experimental evidence to justify drawing con- 

 clusions at variance with those in the preceding paragraph . 



13 



Arthur 2D3UttleJnr. 



