For the ex-imple cited, in trhich the nieosurenients were taken at a 

 depth of 1^0 feet, it was found that Piats ',■:■; relationships were not valid. 

 In fact, So^ given bj^ the eproirical relationship was almost exactly the 

 square of the standard deviation computed from the wave record, and cK ^ 

 was found to be three times 1 irger than tlie value obtained from the 

 original wave record. Since °^g was small in absolute value, it was 

 considered worthwhile to fit the normal distribution (the symmetric member 

 of the gamma«-type distributions) to the w^ve*period distribution, A s-ur- 

 vey of the results indicates that the approach was valid, and the agree~ 

 ment is in order as shown in figure 8. This observation is a qualitative 

 one and it should be emphasized that quantitative conclusions are not 

 given much weight at this time. However, it appears from analyses of 

 additional pressure records that both the standard deviation and the 

 skewness coefficient are relatively constant for pressure variations with 

 average periods of nine seconds or moreo 



Next, to show the behavior of the ratio of the nujiiber of wave maxima 

 to the nimiber of waves, it is necessary only to obtain the 



Expecriea. YAUrrvber ot WaVe maXt'^o. per ^ec = 



-1 «/ 



4^m^ 



.a*4. 



-f Al-f)a-f 



A 



and form the ratio of this to equation (13), with f\ (t) replaced by 

 Apt'fO • The ratio becomes 



22 



