The Cruise of the " Princesse Alice" 263 



unt which falls upon it in the form of rain along with that 

 which is delivered into it by rivers. If shut off by a dam at 

 the Straits of Gibraltar it would shrink in size and increase in 



ness until it attained a condition analogous to that of the 

 Dead Sea. The deficiency due to over-evaporation is made up 

 by the surface current of Atlantic water which enters through 

 the Straits and is so well .known to navigators. But every 

 gallon of Atlantic water brings with it about 6 oz. of salt, 

 which remains in the sea when the water evaporates. As the 

 influx of water at the Straits goes on day after day, year after 



. and century after century, some means must be found 

 fur removing the salt. To effect this nature does exactly what 

 the marine engineer does when he feeds his boiler with sea- 



r she brines her Mediterranean down. The outflow of 

 brine is naturally at the bottom on account of its high density, 

 and after passing the lip of the basin between Capes Span 1 

 and Trafalgar, it must follow the deepest channels outwards 

 until it is lost in the ocean by mixture and diffusion. These 

 rivers "f relatively warm and salt water are probably very 

 narrow ; and it is <jiiite reasonable to suppose that in 1901 the 

 trawl was dropped by the side of one of them, while this year 

 it was dropped into the middle of it. 



Proceeding westwards, two remarkable banks or oceanic 



Is lit nn the route to the Azores, the Gorringe or Gettysburg 

 bank and the Josephine bank. The Gorringe bank was examined 

 by the former "Princesse Alice" on July 25, 1894, and on 

 July 25, 1902, her successor continued the examination, but 

 found nothing very new. It was, however, remarkable that 



presence of the hank was shown juite unmistakably by 



! lippl. - on the surface "I the sea to an extent that 



d nn ;mv nther nee aiiie shoal having the 



depth of \\at< T "\ , I It. Julv J 7 W.I- -p. lit nil t llC JOSCphlUe 



hank h - \\i\\\ HUM be very < Me but. in the time 



is not possible to delimit it. An area of some 



mile- s|u,ue \\a^ Bounded OVCT, -h. \\in- an umiMi.d 



pth. always uithin a fathom or SO of 120 



:.' hank and the Azores the 



\vaS2340fathon :npeiature Of 



