Calibration of the - 6 - D. M. Poole and 



Emery Settling Tube W. S. Butcher 



for Sand Analysis 



ro\md sand from a drifting dune near Yuna, Arizona, it is likely that this dif- 

 ference is a difference in the actual diameters as compared to the eq.uiTalcnt, 

 or settling diameters, 



ludv/ick (19^8) collected data follovdng Krumbein' s (l93^) procedure on 

 several southern California beaches, and analyzed them using a composite sample 

 of 8 to compute the coefficient of variation. From these data the percentage 

 probable error in median diameter due to laboratory error (the total error of 

 this paper) can be found. The values range from 0.5J^ to l»5'p '-irith an average 

 of 5 beaches giving 1.0-J^. The median diameters used in the tests in this paper 

 are smaller than those of Ludvrick' s v;ork which ranged from O.I89 to 0.^00 mm. 



The total probable error from Ludwick's data is generally smaller than 

 that found in our work. Eis data show no consistent relation bctv;ccn total 

 probable error and median diameter of the composite sample. The settling tube 

 error must be less than or equal to the total error since the total error is a 

 combination of splitting and settling tube errors. It is probable that there 

 is always some splitting error but its amount depends on the homogeneity of the 

 sand. Ludv;ick's composite sample contained sands from the area covered by his 

 grid on the beach. The different sands varied little in median diameter and 

 sorting and thus a composite sample vmuld be reasonably homogeneous. Conse- 

 quently his total error is small, probably because of a small splitting error. 

 The sands used in the tests of splitting error in this paper were considerably 

 different in median diameter and sorting. The composite sample will thus be 

 less homogeneous a,nd more likely to have a greater splitting error than the 

 more homogeneous mixture. Further, it is difficult to see v;hy the settling 

 tube error should vary significantly vihen the settling tube is used with care 

 by esperienced personnel. The error due to the settling tube from Ludwick's 

 data is thus assumed to be of the same order of magnitude as v;as found here; 



