THE GULF SlTvEAM. 27 



84. The Niagara is an " immense river descending into a plain." 

 The Niagara. Biit instead of preserving its character in Lake On- 

 tario as a distinct and well-defined stream for several hundi'ed miles, 

 it spreads itself out, and its waters are immediately lost in those of 

 the lake. Why should not the Gulf Stream do the same ? It 

 gradually enlarges itself, it is true ; but, instead of minghng with 

 the ocean by broad spreading, as the " immense rivers " descending 

 into the northern lakes do, its v/aters, like a stream of oil in the 

 ocean, preserve a distinctive character for more than three thousand 

 miles. 



85. Moreover, vv'hile the Gulf Stream is rmming to the north 

 A current counter to h-'om its supposod elcvatcd Icvcl at the south, there 

 the Gulf Stream. jg ^ qqI^ cmTcnt comiug dowu firom the north ; 

 meeting the vrarm waters of the Gulf midway the ocean, it divides 

 itself, and runs by the side of them right back into those very re- 

 servoirs at the south, to which theory gives an elevation siifticient 

 to send out entirely across the Atlantic a jet of warm water said 

 (§ 75) to be more than three thousand times greater in volume than 

 the Mississippi Eiver. This cmTent from Baffin's Bay has not only 

 no trade-winds to give it a head, but the prevailing Avinds are un- 

 favourable to it, and for a great part of the way it is below the sur- 

 face, and far beyond the propelling reach of any wind. And 

 there is every reason to believe that this, with other polar cur- 

 rents, is quite equal in volume to the Gulf Stream. Are they not 

 the efiects of like causes ? If so, what have the trade-winds to do 

 with the one more than the other ? 



86. It is a custom often practised by seafaring people to throw 

 Bottle chart. a bottlc ovcrboard, with a paper, stating the time 



and place at which it is done. In the absence of other informa- 

 tion as to currents, that afforded by these mute little navigators is 

 of great value. They leave no tracks behind them, it is true, and 

 their routes camiot be ascertained. But knowing where they 

 were cast, and seeing where they are found, some idea may be 

 formed as to their course. Straight lines may at least be drawn, 

 showing the shortest distance from the beginning to the end of 

 their voj^age, with the time elapsed. Captain Becher, E.N., has 

 prepared a chart representing in this way the tracks of more than 

 one hundred bottles. From this chart it appears that the waters from 

 every quarter of the Atlantic tend toward the Gulf of Mexico and 

 its stream. Bottles cast into the sea midway between the Old 

 and the New Worlds, near the coasts of Europe, Afi^ica, and America, 



