400 THE DEPTHS OF THE SEA. [chap. viit. 



return currents are very visible on the chart taking 

 this direction, indicated by marked deflections of 

 the isothermal lines. The most marked is the 

 Labrador current, which passes down inside the 

 Gulf-stream along the coasts of Carolina and New 

 Jersey, meeting it in the strange abrupt ' cold 

 wall,' dipping under it as it issues from the Gulf, 

 coming to the surface again on the other side, 

 and a portion of it actually passing, under the Gulf- 

 stream, as a cold counter-current into the Gulf of 

 Mexico, 



Fifty or sixty miles out from the west coast of 

 Scotland, I believe the Gulf-stream forms another, 

 though a very mitigated, 'cold wall.' In 1868, 

 after our first investigation of the very remarkable 

 cold indraught into the channel between Shetland and 

 Fseroe, I stated my belief that the current was entirely 

 banked up in the Faeroe Channel by the Gulf-stream 

 passing its gorge. Since that time I have been led 

 to suspect that a part of the Arctic water oozes down 

 the Scottish coast, much mixed, and sufficiently 

 shallow to be affected throughout by solar radiation. 

 About sixty or seventy miles from shore the isother- 

 mal lines have a slight but uniform deflection. 

 Within that line types characteristic of the Scandi- 

 navian fauna are numerous in shallow water, and 

 in the course of many years' use of the towing net 

 I have never met with any of the Gulf-stream 

 pteropods, or of the lovely Polycystina and Acantho- 

 metrina which absolutely swarm beyond that limit. 

 The difference in mean temperature between the 

 east and west coasts of Scotland, amounting to 

 about 1°C., is also somewhat less than might be 



