clents, the data for the relative amplitudes for the 9th of De- 

 cember given in table X, and equation (27a), it is now possible 

 to predict one of the components of the internal wave for the 

 9th of December. The values of this component for each of the 

 chosen depths are given in table XII, Finally, as a verifica- 

 tion, it was necessary to compute by harmonic analysis the ac- 

 tual observed components of the amplitudes. These observed am- 

 plitudes are also shown in table XII. In like manner it is pos- 

 sible to compute the other component of the amplitude, 



A comparison of the predicted and observed amplitudes as 

 given in table XII indicates that the results, while promising, 

 are not entirely satisfactory. In general, it is considered 

 that there are two main reasons for the discrepancies in the 

 forecast. First, there are deficiencies in the data themselves. 

 The anchor stations on the )4th and 9th of December were each 

 taken by two ships anchored approximately 30 miles apart. Each 

 ship took 6 observations during the 2[|--hour period of the Ij-th 

 and these were combined to form one set of 12 observations for 

 that one day. The procedure was repeated on the 9th. The fact 

 that the ships were anchored so far apart undoubtedly intro- 

 duced reasonably large errors in the harmonic analyses. Further- 

 more, the anchor stations were located in shallow water and in 

 an area in which reasonably strong currents exist. All these 

 factors tended to contaminate the results. Second, there is a 

 possibility that the incorrect parameter vjas minimized. 'iiiis 

 has already been discussr^d. In the .'uture it is planned to 

 conduct other trials 'nlnimizing other parameters. 



66 



