74 



MARINE BOTTOM SAMPLES OF LAST CRUISE OF CARNEGIE 



Table 23. Mechanical analyses by the pipette 



19 

 21 

 23 

 31 

 34 

 40 

 43 

 49 

 69 

 70 

 72 

 73 

 74 

 77 

 81 



0.1 



2.8 

 1.0 

 0.1 

 0.1 



tr 

 3.2 

 7.3 



tr 



tr 



12.3 

 10.8 



4.6 



1.6 

 tr 



6.0 

 12.5 



0.3 



2.4 



Average of samples 69, 70, 73, 74, and 77 



8.0 

 8.4 

 8.3 

 1.6 

 0.2 

 5.9 

 5.0 

 1.0 



4.3 



tr 



12.8 

 13.2 



8.8 

 2.3 

 0.9 

 8.0 

 4.4 

 1.0 



tr 



8.0 



(0.8) 

 6.2 



0.35 



0.6 

 3.5 



5.7 



1.06 



Table 24. Statistical constants calculated from pipette analyses 



Sam- 

 ple 



no. 



Ql 



M 



Q3 



In microns 



So 



logSk 



Sk 



Sand 



Silt 



In per cent 



Clay 



lyzed byTrask, log Sk varied between -0.18 and +0.12, 

 the extremes being -1.12 and +1.03. 



The statistical constants advocated by Trask have 

 been adopted by Stetson (1933) and Gripenberg (1934) in 

 their studies of fine-grained sediments from near Spitz- 

 bergen and the Baltic, respectively. There is thus a 

 considerable body of comparable published data ex- 

 pressed in terms of these quantities. Furthermore, the 

 constants are based entirely on the middle half of the 

 size distribution and thus are particularly applicable to 

 analyses of extremely fine-grained samples such as 

 red clays, in which it is not practicable to obtain the 

 complete size distribution. 



On the other hand, each of the three methods of pre- 

 sentation of the data from mechanical analyses, namely, 

 by graphic representation, by tabulation of the percent- 

 ages in various grade sizes, and by tabulation of the 

 statistical constants of the size distribution, possesses 

 certain advantages. For statistical treatment the con- 

 stants, particularly the median diameter, may be used 



as single, easily comparable characteristics of a sedi- 

 ment. In interpreting the statistical constants, however, 

 a more or less regular size distribution must be as- 

 sumed. Irregularities in the size distribution, such as 

 the presence of secondary maxima, can only be indicated 

 by a listing of the percentages in the various size grades, 

 or by a graphical presentation of histograms and of cu- 

 mulative curves, the latter being essentially the integrals 

 of the true frequency curves. The graphical method has 

 the advantage of being easily comparable, and the cumu- 

 lative curves may be constructed from determinations of 

 particle sizes at irregularly spaced intervals. 



Accuracy . The accuray of the pipette method, as in- 

 deed of all sedimentation methods, is dependent on the 

 fulfillment of the following experimental conditions, 

 which are listed by Fisher and Oden (1924): 



"1. Complete dispersion of sediment particles and 

 prevention of any tendency to coagulation. 



"2. Uniform distribution of every class of particles 

 at starting time t=0. 



