130 SEA AND LAND 



have as yet l)ul little decided information concerning this 

 ocean-stream, save that afforded by tlie movements of the 

 berg and floe, we can readily see how it affects the journey 

 of these wanderinix fragments from the vast Greenland 

 olaciers. Though somewliat inconstant, this current is toler- 

 ably steady, setting south through the wide channel which 

 separates the shores of Greenland from those of the many 

 islands which beset the northeast coast of the American con- 

 tinent. By this southward-moving water the ice is propelled 

 out into the open sea. The stream continues to \X\v. south, 

 but widens and diminishes in the energy of its flow. It shortly 

 comes in contact with the Gulf Stream, which it somewhat 

 affects, and b}' which it is much affected. As we have already 

 noticed, a part of the southward-setting current passes down 

 along the shore of Labrador as a superficial stream of no 

 great width or speed. Another, and perhaps the larger part, 

 flows beneath the (julf Stream, and in tinie joins the great, 

 slow-moving procession of Arctic waters which, following the 

 bottom of the deeper sea, in the end attain the equatorial 

 district. For a considerable distance southeast of Greenland 

 there are thus two distinct currents in the ocean waters — a 

 lower, moving southwardly, and an upper, or superficial 

 stratum, creeping toward the north. The thin tloc-ice, l1oat- 

 ing altogether within a hundred feet of the surface, is beaten 

 back against the Labrador shore hy the surface stream ; but 

 the icebergs, because of their greater depth, are driven for- 

 ward by the under-current in a southwardly direction. Owing 

 to this peculiarity we sometimes may observe the bergs 

 l)lougliing their way through vast fields of floe-ice as stead- 

 fastly as a steamship when it breaks its way in the new-formed 

 ice of a harbor. This southward journey of the bergs is 



