394 T. C. CHAM BERLIN 



that borders the lane. Here the ice of the sludge is Hkely to be 

 packed and piled up at the surface while the briny water within it 

 has no alternative but to turn downward along the face of the wall 

 of ice at the edge of the lane. The sludge thus driven forward and 

 separated into packed ice and brine exposes new water behind which 

 undergoes the same process, so that the movement edgewise is con- 

 stant until the surface is sheeted over with ice thick enough to sup- 

 press the waves raised by the wind and bring on a new form of action, 

 to be discussed below. When the edge of the sheet of brine is turned 

 downward in front of the bordering ice-wall, the push from behind 

 continues as long as the wind pushes the water or its motion con- 

 tinues by momentum. This push is edgewise, that is, in the line of 

 least resistance. It is also in the line in which it is itself impelled 

 to go by its density, which is greater than that of the adjacent water. 

 Here then is a combination of push and pull in the Kne of least re- 

 sistance, which is felt so long as the force of the wind is felt by the 

 freezing surface. 



It seems reasonable to assume that the briny sheet, once it is 

 turned downward at its edge and projected in that direction, will 

 continue downward, for that is the direction demanded by its den- 

 sity as well as its temporary momentum. Now it is not necessary 

 that this action should continue long, for if the brine sheet moves 

 forward and downward no more than 800 to 1,000 meters — ^which 

 is often less than the width of the lane — ^its forward edge will have 

 crossed both the upper and middle layers and have entered the basal 

 layer. 



This briny layer, while not great in volume, will be exceptionally 

 cold as well as saline, and will carry both low temperature and high 

 salinity into the basal layer, thereby adding Kttle by Uttle the quali- 

 ties it requires to correspond to its observed characteristics. These 

 Httle increments will be added here and there, for the lanes open 

 at haphazard, and the briny sheets are driven ii.to the basal layer 

 at haphazard intervals of both space and time. They will thus be 

 favorably distributed for mixture and diffusion. The abysmal 

 mass is itself in motion, for, as Nansen has shown, it is a current, 

 though probably a very slow one. The action, however, continues 

 year after year for a long period, so that the combination is favor- 

 able to an ultimate homogeneous mass. 



