30 THE PHYSICAL GEOGRAPHY OF THE SEA. 



formed the tour of the Gulf is all but conclusive. And there is 

 reason to suppose that some of the "bottles of the admiral's chart 

 have also performed the tour of the Gulf Stream ; then, without 

 being cast ashore, have returned with the drift along the coast of 

 Africa into the inter-tropical region ; thence through the Caribbe- 

 an Sea, and so on with the Gulf Stream again. (Plate VI.) 



Another bottle, thrown over off Cape Horn by an American 

 master in 1837, has been recently picked up on the coast of Ire- 

 land. An inspection of the chart, and of the drift of the other 

 bottles, seems to force the conclusion that this bottle too went 

 even from that remote region to the so-called higher level of the 

 Gulf Stream reservoir. 



13. Midway the Atlantic, in the triangular space between the 

 Azores, Canaries, and the Cape de Verd Islands, is the Sargasso 

 Sea. (Plate VI.) Covering an area equal in extent to the Mis- 

 sissippi Valley, it is so thickly matted over with Gulf weeds {fucus 

 natcms)^ that the speed of vessels passing through it is often much 

 retarded. When the companions of Columbus saw it, they thought 

 it marked the limits of navigation, and became alarmed. To the 

 eye, at a little distance, it seems substantial enough to walk upon. 

 Patches of the weed are always to be seen floating along the outer 

 edge of the Gulf Stream. Now, if bits of cork or chaff, or any 

 floating substance, be put into a basin, and a circular motion be 

 given to the water, all the light substances will be found crowding 

 together near the centre of the pool, where there is the least mo- 

 tion. Just such a basin is the Atlantic Ocean to the Gulf Stream ; 

 and the Sargasso Sea is the centre of the whirl. Columbus first 

 found this weedy sea in his voyage of discovery ; there it has re- 

 mained to this day, moving up and down, and changing its position 

 like the califts of Cancer, according to the seasons, the storms, and 

 the winds. Exact observations as to^ its limits and their range, 

 extending back for fifty years, assure us that its mean position 

 has not been altered since that time. This indication of a cir- 

 cular motion by the Gulf Stream is corroborated by the bottle 

 chart, by Plate VI., and other sources of information. If, there- 

 fore, this be so, why give the endless current a higher level in 

 one part of its course than another ? 



