The detrital mineralogical observations of 

 the present study do not reflect any significant 

 changes in the clay mineralogy (table 2) or 

 any readily apparent change in the detrital 

 components of the coarse fractions across the 

 upper layers of Kara Sea sediments. The ob- 

 served downward increase in montmorillonite/ 

 illite ratio across the upper brown layers and 

 in the vicinity of the secondary brown layers 

 in core E-26 would not contribute to the brown 

 color of these layers because neither mont- 

 morillonite nor illite are brown clay minerals 

 (Keller, 1953). The possibility exists that a 

 mixed layer clay incorporating an expandable 

 minei-al and iron hydroxide (Rich, 1968) has 

 formed in the Fe-rich layers but the evidence 

 for this is weak. 



The effects of particle size and size distribu- 

 tion can also be eliminated as responsible for 

 the color variations in the upper layers of 

 Kara Sea sediments because there are no 

 marked textural changes across the upper lay- 

 ers (figs. 5-7). The decrease in fine clay con- 

 tent in the upper layers is probably related to 

 authigenic cementation in this fraction by the 

 high content of nondetrital ferromanganese 

 oxides (fig. 8) and not to any real difference in 

 the grain size distributions between surface 

 layers and lower layers. Although particle 

 packing increases with depth in sediments 

 under natural compaction, in view of other fac- 

 tors to be discussed subsequently, packing does 

 not appear to be significant in determining 

 color in the upper layers of Kara Sea sedi- 

 ments. 



In the Kara Sea the nondetrital sedimentary 

 constituents, the stabilities of which depend 

 upon the sufficiency or insufficiency of oxygen, 

 and low abundance of aquatic life, would ap- 

 pear to exert the greatest influence on sedimen- 

 tary color. Nondetrital iron and manganese 

 compounds in varying proportions are the con- 

 stituents of Kara Sea sediments which impart 

 the brown color to the upper layers (Klenova, 

 1938; Brujevicz, 1938a,6; Trofimov, 1939; 

 Klenova and Pakhomova, 1940 ; Yermolayev, 

 1948o,6; Gorshkova, 1957; figs. 5-7, this 

 study). The sediment assumes a gray-green 

 color typical of illite, chlorite and other clay 

 minerals (Keller, 1953) when nondetrital iron 

 and manganese are least abundant (fig. 8) and 

 after mild acid leaching. That the brown color 

 is nondetrital is also evidenced by the brown 



coatings on detrital mineral grains in the 

 coarse fractions from brown layers. 



Any interpretation of the distribution of 

 nondetrital iron and manganese in Kara Sea 

 sediments must consider (1) the possible 

 sources of iron and manganese, (2) the possi- 

 bility of variability in the rates of influx and 

 primary deposition of iron and manganese, and 

 (3) the postdepositional mechanisms which 

 could concentrate iron and manganese in some 

 layers and not others in the absence of any 

 variability in the rates of influx or primary de- 

 position. 



The trend analyses of the distribution of 

 nondetrital iron and manganese in surface sed- 

 iments (figs. 9, 10 and 11) are an attempt to lo- 

 cate the source(s) or point(s) of influx of 

 these elements into the Kara Sea. Such an ap- 

 plication of trend analyses assumes that nonde- 

 trital elemental concentrations will increase to- 

 wards the source(s) or point(s) of influx, 

 when account is taken of dilution by detrital 

 constituents. In the present study dilution ef- 

 fects are minimized by sampling within a com- 

 paratively narrow water depth range (200 to 

 600 meters), by small sample-to-sample varia- 

 bility in grain size distribution (app., tables 10 

 and 11; Andrew, in preparation), and by ana- 

 lyzing specifically for the nondetrital compo- 

 nent of the total iron and manganese content. 



The observation of a regional increase in the 

 concentration of nondetrital iron in surface 

 sediments toward the Siberian mainland 

 (southward) is consistent with a possible 

 source for this element in the Siberian rivers. 

 In order for the concentration of a nondetrital 

 component to show a relative increase in the 

 direction of a detrital source there must also be 

 an increase in the absolute quantity of the non- 

 detrital component to compensate for detrital 

 dilution in the direction of the source. The 

 spread of continental runoff across the Kara 

 Sea from the river mouths to north of 80° 

 (Milligan, 1969) also supports river discharge 

 as the source of the nondetrital iron. 



On the basis of the trend surface analyses 

 alone, it is not possible to determine the 

 point(s) of influx of nondetrital manganese 

 into the Kara Sea. Each deeper area reflects a 

 different source, suggesting that factors other 

 than distance from source influence the distri- 

 bution of nondetrital manganese. These factors 

 include the nature of water masses present in 



