LOCAL VARIABILITY IN MARINE SEDIMENTS 723 



lor analysis. Some 60 locations on a traverse across Paget Sound have 

 been sampled, making available for comparison the data obtained from 

 the analysis of 180 cores. (Fig. ?> gives the location of these samples.) 

 The sampling of approximately 18 to 20 locations in the mid-channel 

 and on the near shore slopes is planned for the near future. 



The vertical sectioning of the cores in the manner just described 

 yielded well over 500 individual samples. Each sample ■was analyzed 

 for particle size by standard methods. The material courser than 1/16 

 mm. in diameter was sieved mechanically, separating it into standard 

 fractions. The material finer than 1/16 mm., previously separated from 

 the sample by wet sieving through a 1/16 mm. sieve, was submitted to a 

 pipette analysis for size determination. This latter method is based 

 upon Stokes Law; that is upon the settling velocities of spheres of a 

 given radius suspended in a fluid of known viscosity. 



Since this study deals with the variation in the results obtained 

 in the size analysis of adjacent core samples, it is essential to know the 

 analysis error. The standard method briefly described above satisfies 

 this condition in allowing an impartial duplicate analysis of the same 

 sample to be run under essentially standard laboratory conditions. When 

 the reproducibility of the results are known, it is possible to differentiate 

 between inconsistencies due to analysis and the actual variation between 

 samples. 



Duplicate analysis of 20 samples indicated that the analysis error 

 was very small, with an average reproducibility for any one size deter- 

 mination of ±0.2 percent. There was very little divergence from this 

 average, with only about 20 percent exceeding the average by a factor 

 of 2. The probabilities are thus in favor of actual sample variation ac- 

 counting for percentage differences between adjacent samples which ex- 

 ceed ±0.4 percent. 



Initial Results 



In order to present a summation of the results obtained, the east- 

 west traverse of the Sound has been divided into three general areas. 

 The eastern and western slope area from sea level down to about 500 

 feet below sea level represents one division. The lower portion of both 

 slopes, from about 500 feet below sea level to the base of the slopes and 

 adjacent channel areas, represents the second division. The last division 

 is the mid-channel region. 



The following table presents some of the differences in grade size 

 percentages between the samples from the three cores obtained simul- 

 taneously at one location. The percent differences shown are thus in- 



