BOTTOM WATERS 143 



also very great. In other words, they are not unique 

 for the configuration of the bottom near the Romanche 

 Deep. 



In discussing these results in a paper about to be 

 published, Koczy has entered on the more general prob- 

 lem offered by the interchange between bottom water 

 and sediment. The following abstract is given with his 

 permission. 



Near the bottom, apart from hummocks and mounds, 

 there is a laminary current, that is, one without tur- 

 bulence. On mounds and ridges, on the other hand, a 

 turbulent motion followed by chemical solution seems 

 to prevail. 



The most soluble elements like Mn, Ni, Fe, Na, and 

 Ca are, therefore, liable to displacement, their content 

 being reduced in certain localities and increased in 

 others. This does not apply to Ti. Owing to the reduced 

 turbulence, the oxygen content can locally get so low 

 that in hollows in the bottom a reducing milieu may 

 arise, with corresponding changes in solubility, pre- 

 cipitation, and exchange of elements. 



It is necessary, however, to return once more to the 

 question of the bottom currents. As mentioned in an 

 earlier chapter, the dominating currents in very great 

 depths are set up by temperature differences produced 

 through the cooling of the surface waters in high 

 northern and, more especially, high southern latitudes, 

 which cause the water to sink to the bottom. From the 

 Antarctic Continent great currents spread northward 



