33 8 The National Geographic Magazine 



ties in that ocean. These charts all 

 unite in showing a steady easterly drift 

 in the temperate latitudes of both the 

 North and the South Atlantic, and an 

 equally steady westerly drift in the 

 tropical latitudes, the generalized cur- 

 . rent system of either ocean thus consist- 

 ing of a vast eddy about some central 

 point, the direction of the circulation be- 

 ing anti-cyclonic in either hemisphere — 

 i. e. , with the diurnal motion of the sun, 

 as observed in that hemisphere, just as 

 in the case of the prevailing winds. 



Taken collectively, the lines of drift 

 of floating bottles in the Pacific again 

 show, precisely as in the case of the 

 Atlantic Ocean, that the general direc- 

 tion of the drift is eastward in the higher 

 latitudes, westward in the lower. At 

 least this is so for the North Pacific. 

 For the South Pacific evidence of the 

 eastward motion of the extratropical 

 waters is lacking. None of the drifts 

 reveal the existence of the equatorial 

 counter-current flowing eastward be- 

 tween the westward-moving equatorial 

 currents of the southern and northern 

 hemispheres. The average velocity of 

 the easterly drifts is 4.4 miles per day, 

 of the westerly drift 10 miles per day, or 

 more than twice as great, which is again 

 in accordance with the results for the 

 Atlantic Ocean. The highest velocity 

 attained was that of a bottle thrown 

 overboard from the steamer IVarriiuoo 

 January 23, 1897, in latitude 4° N., 

 168° W., and found March 6, 1897, on 

 one of the Gilbert Islands, having drifted 

 1,100 miles in 42 days, or at an average 

 rate of 26 miles per day. None of these 

 velocities makes any allowance for the 

 time during which the bottle may have 

 lain undiscovered on the beach. The 

 longest drift was that of a bottle which 

 was thrown overboard near Cape Horn 

 June 18, 1896, and found near Cape 

 York, on the northern coast of Queens- 

 land, Australia, after the lapse of nearly 

 three years. The shortest practicable 

 route which it could have pursued meas- 



ures 10,100 miles in length, or nearl}^ 

 two-thirds of the total distance around 

 the earth in the latitude of its path, 

 giving an average velocity of 10. i miles 

 per day. The actual distance traversed 

 was probably much greater than this. 



My main object in directing attention 

 to these drifts is to suggest the idea that 

 they illustrate an apparent paradox. 

 The bottles themselves float upon the 

 surface, but if I were asked whether the 

 lines drawn upon a chart to show their 

 course represented the surface currents 

 of the sea, the currents with which the 

 navigator has to deal, I should s^.y em- 

 phatically no. The actual surface cur- 

 rents present no such uniformity, either 

 in direction or velocity. As an example 

 of this, take the currents actuallj^ ob- 

 served in the five-degree square of the 

 North Atlantic Ocean bounded by the 

 parallels 35 to 40 degrees north and the 

 meridians 65 to 70 degrees west (off the 

 coast from Hatteras to Sandy Hook), in 

 the heart, therefore, of what is ordina- 

 rily known as the Gulf Stream, concern- 

 ing which the popular impression is that 

 it fiows along steadily like a might}^ 

 river. For any given month, say Sep- 

 tember, the currents actually observed 

 within this square were as follows : 



Setting northeast, 32 per cent of the 

 whole number of observations ranging 

 from 6 to 70 miles in 24 hours. 



Setting southeast, 23 per cent of the 

 whole number of observations ranging 

 from 8 to 65 miles in 24 hours. 



Setting southwest, 27 per cent of the 

 whole number of observations ranging 

 from 6 to 76 miles in 24 hours. 



Setting northwest, 18 per cent of the 

 whole number of observations ranging 

 from 9 to 63 miles in 24 hours. 



Evidently here there is none of the 

 uniformity presented by the drifts and 

 which the mind ordinarily associates 

 with the Gulf Stream. 



To get at the true meaning of these 

 lines of drift beyond the fact that they 

 represent the resultant of the traverse 



