the importance of similar analyses for the pur- 

 pose of data "smoothing" and interpretation. 



Because of inequable sample spacing the 

 geochemical Jata were divided into three geo- 

 graphic groupings, the East Novaya Zemlya 

 Trough (A^=15), the St. Anna Trough 

 (A^ = 39) and the Voronin Trough (A^=13). 

 Three-dimensional regression surfaces were 

 developed for each of the three data groupings 

 and for selected combinations of the groupings 

 by using the expression Z = f(x,y)^', where x 

 and y, the geographic coordinates of a sample 

 location measured from an arbitrary origin, 



are the independent variables, and each geo- 

 chemical attribute (z) is the dependent varia- 

 ble. The independent variables are raised to 

 higher and higher polynomials, with the coef- 

 ficients of the polynomials being determined 

 through standard macrix techniques on the 

 digital computer using least squares analysis. 

 A statistical analysis of variance was used to se- 

 lect the polynomial which accounted for the 

 largest significant reduction in the variability 

 of the data (Goodell, 1967). For this study no 

 polynomial higher than linear proved to be sig- 

 nificant at the 95 percent confidence level. 



Results 



Color 



The color variations with depth in the 20 

 cores from the Edisto-Eastwind collection are 

 given in figures 4-7. Although no two cores 

 show exactly the same variations, there are 

 marked similarities, especially in the upper 10 

 to 15 cm. With the exception of core E-32 

 which is greenish-gray throughout, every core 

 is some shade of brown in the upper layers and 

 grades sharply or gradually into gray or green 

 lower layers. Several of the Edisto-Eastwind 

 cores (table 7, e.g., E-8, 20 to 38.5 cm.) show 

 gray-green layers which are mottled by black 

 lumpy material similar to the substance de- 

 scribed by Klenova (1948, cited in Strakhov, 

 1966) and identified as hydrotroilite (FeS n 

 H.O). 



The generalized color sequence (table 1) 

 given by Yermolayev (1948o,b), brown to dark 

 brown to bright orange (brittle) to gray to 

 brown (brittle) is best represented in cores 

 E-8, E-12, E-18, E-25, E-26, E-31, E-34, and 

 E-35. The rest have layers which can be iden- 

 tified with one of Yermolayev's layers but the 

 sequence is incomplete or out of order. 



The cores from the Northwind collection 

 also show similar color variations with depth 

 (Andrew, personal communication) and many 

 display the generalized color sequence of Yer- 

 molayev (table 1 and fig. 8, for typical se- 

 quences). 



Texture 



The distribution of gravel, sand, silt and 

 clay with depth in the eight cores from the No- 



vaya Zemlya Trough and two cores from the 

 Svyataya Anna Trough is given in figures 5-7. 

 Almost without exception the cores consist of 

 silty clay (classification after Shepard, 1954) 

 throughout. The exceptions are core E-28 

 which contains appreciable sand and gravel in 

 the lower layer and core E-21 which has sandy 

 layers. 



The 10 cores (fig. 4) not analyzed below the 

 surface layer for grain size distribution like- 

 wise consist of silty clay in the surface layers 

 except for cores E-14 and E-15 which contain 

 considerable sand and some gravel (app., table 

 8). There is no visible evidence table 7) of 

 marked shifts in the proportions of sand, silt 

 and clay below the surface layers of these 

 cores. 



The statistical parameters (table 9) support 

 the observation of little variability in the grain 

 size distribution with depth in the cores. Al- 

 most without exception the mean size of the 

 sediments corresponds with very fine silt to 

 clay on the Wentworth size scale and the sedi- 

 ments are very poorly sorted, near-symmetri- 

 cal to coarse-skewed and mesokurtic. The few 

 departures from these characteristics are also 

 reflected in the distribution of gravel, sand, silt 

 and clay (figs. 5-7). 



The percentages of each size fraction show a 

 marked increase in the amount of material 

 finer than 1.0 micron below the surface layers. 

 The quantity of this material increased from 

 ca. 30 percent in the surface layers to ca. 40 

 percent on entering the subsurface gray-green 

 layers and did not decrease again on entering 

 secondary brown layers. Only where the per- 



6 



