192 MAETON AND GENERAL GREENE EXPEDITIONS 



The only other record of an observed temperature which was 

 lower than the maximum temperature to initiate vertical convection 

 to its respective depth in accordance with our assumptions, occurred 

 at GodtliaaVs station 10 at 1,000 meters on May 3, 1928. The sur- 

 vival of this temperature value about 2 14 months subsequent to the 

 coldest part of winter indicates that the water in this region is sub- 

 jected to greater cooling and then less warming than is the water in 

 the vicinity of Meteor^s stations 122 and 121. 



SUMMARY 



The warm, salty west Greenland water progressively sinks as 

 it proceeds northward and westward and bends southward, spread- 

 ing as it goes to furnish the intermediate water of the Labrador 

 Sea between about 500 and 2,000 meters. 



The most nearly motionless water, except perhaps that immediately 

 adjacent to the bottom, occurs at about the 2,000-meter level below 

 which lies the deep water and the bottom water. 



The deep water, according to the foregoing view, is formed dur- 

 ing the colder part of the year largely by mixing of bottom water 

 with Avater from the West Greenland Current which has sunk to deep 

 levels as it travels northward along the Greenland coast and west- 

 ward near the head of the Labrador Basin. The major flow of the 

 deep water is southward along the American side where, off southern 

 Labrador, there is probably some movement toward deeper levels 

 along the bottom, the water flowing down the slope at levels of about 

 3,000 to 3,500 meters. This deep water is probably absorbed into the 

 more central, North American Basin, and thus it compensates for the 

 loss of water at higher levels to the northwestern North Atlantic 

 from the northern branch of the Atlantic Current. 



The bottom water, in our opinion, is formed by wintertime chilling 

 of the surface, intermediate, and deep waters in the northern part 

 of the Labrador Basin in the area off-shore from the rapid currents 

 and roughly bounded on the south by a line from mid-Labrador to 

 Cape Farewell. (See inset fig. 149.) It seems likely in this area 

 the severe winter chilling produces vertical convection to bottom 

 and results, with some mixture of Irminger-Atlantic water, in the 

 coldest bottom water found in the deepest part of the basin and 

 which in summertime is isolated from the cold surface currents by 

 warmer water. The vertical convection which takes place in winter 

 probably sets up steep horizontal density gradients to considerable 

 depths, with a correspondingly increased cyclonic system of circula- 

 tion. With the termination of vertical convection and its resulting 

 heat losses the energizing force for maintaining this vigorous circula- 

 tion is removed and the summertime equilibrium conditions are 

 quickly restored as the marginal water of equal salinity and higher 

 temperature is mixed in to destroy the temporary density gradients 

 and raise the temperature of the intermediate water to the remark- 

 ably uniform value of about 3,2° C. 



If our hypothesis be correct, because of the intermittent nature of 

 the formation of the coldest bottom water and the higher salinity 

 deep water there are horizontal variations in these waters repre- 



