given areas as demonstrated by Klenova and 

 Pakhomova (1940) and Kulikov (1961) and 

 probably other redox potential determining 

 factors, such as the content and type of organic 

 matter. 



The observation of an increase in the nonde- 

 trital Mn/Fe ratio in surface sediments avi^ay 

 from the mouths of the Siberian rivers is con- 

 sistent with deposition of most of the nondetri- 

 tal iron nearer to the points of influx of iron 

 into the Kara Sea. The positive residuals for 

 nondetrital Mn/Fe ratio in the central areas of 

 the Svyataya Anna and Voronin Troughs also 

 support deposition of most of the nondetrital 

 iron in shoaler water. These findings are in ac- 

 cord with the suggestions of Murray and Irv- 

 ing (1895) and more recently Skornyakova 

 (1964) and Strakhov and Nesterova (1969) 

 who attributed a real separation of iron and 

 manganese in marine sediments to the greater 

 chemical mobility of manganese compounds in 

 near-shore sediments because of reduction by 

 organic matter. Alternately Price (1967) and 

 Krauskopf (1957) attributed increased Mn/Fe 

 ratios on going from littoral to pelagic sedi- 

 ments to the ease of precipitation of iron rela- 

 tive to manganese when river water, charged 

 with soluble divalent ions of these metals, en- 

 ters the sea. Both processes lead to a seaward 

 (pelagic) enrichment in manganese and thus 

 increased Mn/Fe ratios in pelagic sediments. 



Although the Siberian rivers are the most 

 logical source of nondetrital iron and manga- 

 nese in Kara Sea sediments, submarine volcan- 

 ism, which has been postulated (literature re- 

 viewed by Bostrom, 1967) as a major source 

 for nondetrital iron and manganese accumula- 

 tions in sediments, may theoretically have con- 

 tributed significant amounts of these elements 

 to Kara Sea sediments. However, this possibil- 

 ity is unlikely because there is no evidence 

 (Saks, 1948; Saks and Strelkov, 1961) of re- 

 cent submarine volcanic activity in the Kara 

 Sea or the adjacent Barents and Laptev Seas 

 and Arctic Ocean (Strakhov, 1966). Further- 

 more it seems unlikely that the volcanic activ- 

 ity associated with North Atlantic ridge sys- 

 tems has contributed large amounts of iron 

 and manganese to Kara Sea sediments and not 

 to the sediments of adjacent seas (e.g., Nor- 

 wegian Sea) which are closer to this obvious 

 volcanic source. 



The observation of an expandable clay min- 

 eral, which is probably montmorillonite, in 

 Kara Sea sediments (table 2) is not prima 

 facie evidence of a volcanic regime. Applying 

 the criteria of Griffin and Goldberg (1963) for 

 distinguishing volcanic from nonvolcanic 

 montmorillonites in the marine environment to 

 the Kara Sea, it can be established that mont- 

 morillonite (1) is not the most abundant min- 

 eral in the less than 2 micron fraction, (2) is 

 not associated with the zeolite phillipsite and 

 (3) is not associated with abundant volcanic 

 shards or any of the other products of submar- 

 ine volcanic effusions such as are reported 

 from the Pacific Ocean by Bonatti and Nayudu 

 (1965). Therefore, montmorillonite in Kara 

 Sea sediments is almost certainly continentally 

 derived. 



Variability in the rates of influx of iron and 

 manganese into the Kara Sea, regardless of the 

 source (s) or point (s) of influx, cannot satis- 

 factorily account for the observed distribution 

 of these elements with depth in cores (fig. 8). 

 If this distribution were explained in terms of 

 variability in the rate of influx of either of 

 these elements, it would imply that they are 

 geochemically much more dissimilar than is 

 known to be the case (Rankama and Sahama, 

 1950). Although there is considerable evidence 

 of preferential, or early, precipitation of iron 

 in soil-forming processes, preferential dissolu- 

 tion and transport of iron unaccompanied by 

 manganese rests on extremely flimsy evidence 

 (Krauskopf, 1957). Thus we would expect the 

 Mn/Fe ratio in fluvial waters to remain rela- 

 tively constant even if the absolute amounts of 

 each element changed. 



Eustatic control of the deposition of ferro- 

 manganese compounds has been suggested 

 (Goodell, personal communication) to account 

 for cyclic brown banding in delta and near- 

 shore sediments. Under this model rates of de- 

 trital sedimentation at low stands of sea level 

 are fast (relatively) in near-shore areas 

 (Huang and Goodell, in press) and chemical 

 elements associated with the finest suspended 

 matter are highly diluted and concentrate only 

 in pelagic areas. At high stands of the sea, 

 rates of sedimentation are comparatively 

 slower in former, i.e., relict, near-shore areas 

 permitting such elements as iron and manga- 

 nese to concentrate in these areas. During sub- 



10 



