(2.65) and have been computed for d n in the range 0.1 through 2.0 mm. 

 Their tables also employ the Corey shape factor (Equation 6). By using 

 the median value of 24.2 seconds for the 1-meter fall result shown in 

 Figure 5, one determines from Zeigler and Gill's tables that the median 

 nominal diameter of the sample, for shape factor 0.7, at 26° C. , is 

 0.289 mm. 



The following nonparametric statistical measures were used for 

 estimating mean settling velocity M v , and mean nominal diameter M z , 

 from the analyzer curves 



My or Mz = P2 ° + P5 ° + P 80 (15) 



3 



where P20, ?50> and p 80 are the 20th, 50th, and 80th percentiles, re- 

 spectively, and P is expressed in seconds for determining M Vj or nominal 

 diameters,for determining M z . Sorting of the distribution curve for 

 nominal diameters was estimated with the following equation for the sort- 

 ing coefficient: 



S = P 80 ~ P 20 (16) 



P50 



Equation 16 was first used on similar data by Miller and Zeigler (1958). 

 The closer S is to zero, the better is the sorting of the sample, for 

 the spread of the size distribution decreases as S approaches zero. 



Determination of Settling Velocity in Nature . Because it was 

 desirable to obtain values for the settling velocities occurring in the 

 natural environment, it was necessary to prepare curves (Figure 6) of 

 settling velocity versus water temperature for representative samples of 

 Virginia Beach sands from the study area. 



Ten samples representative of the swash, breaking wave, and shoaling 

 wave zones were run in fresh water and in sea water of 15.9, 26.8, and 

 38.9 0/00 salinity. Temperatures ranged from 39° to 120° F. Mean 

 settling velocity values were determined from percentiles of the WHRSA 

 curves using the statistic: 



p 10 + P30 + P50 + P70 + P90 

 My = (1?) 



McCammon (1962) indicated that this statistic is 93 percent effi- 

 cient, where the statistical efficiency of a graphic measure is the ratio 

 of the variance of the distribution of the corresponding efficient 

 estimate and the variance of the limiting distribution of the graphic 

 measure . 



