EXPLORATIONS, WESTERN ATLANTIC, STEAMER BACHE, 1914. 49 



Such a distribution of salinity obviously suggests that water was flow- 

 ing down off the shelf into the ocean deeps, and densities are entirely 

 in harmony with this explanation. Thus the water was decidedly 

 denser — i. e., heavier — over the slope (stations 10258, 10260) than at 

 corresponding depths either on the shelf (stations 10157, 10159, 

 p. 59) or in the ocean basin to the east (station 10161, p. 59) ; hence, 

 would naturally tend to sink. This is further illustrated by the 

 profile (fig. 49), on which all the density curves from the surface 

 down to 500 meters dip sharply toward the ocean basin over the 200- 

 meter contour, and their gradient of about 100 meters in a distance of 

 only 40 miles is steep enough to indicate a very potent dynamic 

 cause for vertical circulation of this type. True, while such a dis- 

 tribution of density suggests a downpour, it does not prove it, be- 

 cause more or less similar densities might result from the opposite 

 process — i. e., an up welling of heavy water from the abyss over the 

 slope. But when we add the facts that this dense water exactly 

 coincides with the fresh tongue just described, and that the tongue 

 is absolutely separated from the abyss by considerably salter water 

 in the middepths, there is no escape from the conclusion that a down- 

 pour or waterfall was actually taking place. If any further con- 

 firmation be needed, it is supplied by the fact that the tempera- 

 ture of the axis of this tongue of 34.5-35''/oo water (station 10158) 

 was almost uniform (11 °-12°) from the surface down to 300 meters — 

 i. e., to almost exactly the depth to which the curve for 35.2°/oo 

 saHnity dips — below which there was a rapid coohng to the consid- 

 erably lower temperatures (4°-5°) of the abyss. Had up welling 

 been active, just the reverse — i. e., a sudden vertical coohng in the 

 upper layers — ^would have obtained. 



The sudden coohng (fig. 47) and the reversal of the vertical change 

 in salinity (fig. 48) at 300-700 meters over the slope (station 10158) 

 marks this zone as the lower limit to the downward flow. The uni- 

 form abyssal temperature (about 4°) and sahnity (about 34.9-35°/oo) 

 was encoimtered here at about 1,200 meters; but in the ocean basin 

 to the east, and, indeed, along the whole line to Bermuda, the upper 

 limit to this abyssal water was at about 1,800 meters (p. 16, fig. 11, 12). 

 So uniform is this water over the north Atlantic as a whole (Kriimmel, 

 1907), and so closely do the curves for 35^/oo and 4° coincide, 

 that this difference in level is only explicable as the result of upwell- 

 ing over the lower part of the continental slope, the first time we have 

 actuaUy been able to demonstrate this type of circulation on any 

 large scale off our coast (1915). So far as true abyssal water is con* 

 cerned, this updraught did not rise above about 1,000 meters; but 

 the close agreement between the salinity and temperature of the 

 bottom water on the slope (station 10160) and of the water of the mid- 

 zone at 1,300-1,400 meters to the east (stations 10161, 10163, 10166) 

 63271°— 17 4 



