150 feet, where periods less than 7 seconds are not represented. Therefore, 

 the limits of integration for equation (1?) are 7 and Oo • Thus, the 

 average wave period given by equation (17) is TL-IA.15 seconds. This 

 closely approximates the value "V — jSl.To seconds already obtained from a 

 direct analj'sis of the wave datao 



Assuming a gamma-type distribution, prediction curves for wave period 

 distributions can be determined from the mean period, the standard devia- 

 tion, and a measure of the skewness. This type of analysis has been done 

 previously by Puts (1952). In fact, Putz gives relationships in terms of 

 the average period for both the standard deviation S^ and the skewness 

 ®^2 •■ These expressions are : 



a^ - 0.313 T- 0.15S ^ 



3 



A comparison of these empirical prediction curves with the period distri- 

 butions from several pressure records is illustrated in figure 6 and 

 figure 7« 



The period for Patz's prediction curves has been defined in a different 

 manner than the definition used in this report. Putz has defined a wave 

 period as twice the time it takes for the record to complete a half cj'cle 

 (trough to peak). This difference in measuring periods Td.ll have great 

 effect in some cases. It is believed that the good agreement for tlie 

 records shov/n is due to the relatively deep depths at which Putz's measure- 

 ments were made; for the em.pirj.cal curves. As will be shown later, at these 

 depths the ratio of the number of vrfve maxima to the number of waves ap- 

 proaches one. 



19 



