540 



NA TURE 



[Ai'RiL 6, 1899 



•diflferences of level caused by off- and on-shore winds, 

 inequalities of density, &c. In this second case the 

 •current may, and often does, move in a direction forming 

 a considerable angle with the wind ; and the float may 

 in most cases follow the current. That it does so in 

 •certain cases Dr. Schott has shown ; but nearly ever)' 

 paper on this subject contains "erratics," and this 

 is no exception. Several records are distinctly un- 

 satisfactory ; the most flagrant case is, perhaps, that of 

 two identical tloats started together in i^ 44' X. lat. and 

 27° 16' W. long., one of which was found on the coast of 

 Nicaragua, and the other on the coast of Siena Leone. 

 ■Compare this with Mr. Russell's results on the east 

 •coast of Australia. " In view of the well-known southerly 

 current on this coast, it is remarkable that so few of the 



With regard to deductions as to the speed of a current 

 based on records from floats, we are almost inclined to 

 go further than Ur. Schott, and to regard such as prac- 

 tically valueless. The whole tendency of recent investi- 

 gation has been to show that steady forward movement of 

 surface water only occurs when there is a distinct " head " 

 of water strongly controlled by the shape of the land ; the 

 best examples being the north and south currents moving 

 polewards on the eastern sides of the great land masses. 

 These currents come to an end as soon as they get clear 

 of the land, and their waters are distributed by " drift " 

 currents controlled primarily by the prevailing winds, 

 but subject to continual variation, according to the rela- 

 tive amounts of denser and lighter water supplied by the 

 true currents. But the movements of the drift currents 



papers found seem to go with it, and that the majority of 

 the papers found go against the current." Evidently 

 great caution is necessary in applying the method. 

 What seems most likely is that the relation of the 

 "movement of the float to that of the wind and of the 

 surface water is really a function of the strength 

 ■of the wind and the sea disturbance, and of the 

 density of the water. A float may make headway 

 with a current against a light wind ; but if it meets 

 with a cyclone, it, and perhaps a skin of water with 

 it, may be removed from the surface of the current, a 

 merely local disturbance transferring it to another 

 tnember of the oceanic circulation : yet the " record " 

 ■of this float would show the two currents as a continuous 

 stream. 



NO. 1536. VOL. 59] 



ririft during the winter i 



are irregular in the extreme ; the water goes backwards 

 and forwards, the horizontal course of any particle of 

 water, perhaps, resembling that of a particle of the 

 air in the winds which drive it. The track of the Fram 

 across the Polar Sea is probably a generalised form of 

 such a course. In effect, there is no '■ river in the ocean," 

 and we are not m general justified in jdiuing the begin- 

 ning and end ol the course of a float by a continuous 

 line. Under tlicse circumstances, one is inclined to 

 suspect that the wind has a considerable share in the 

 high average \ clocities sometimes obtained for surface 

 drifts from float observations. 



Dr. Schott says little or nothing about the precise 

 limitations «hii h are to be observed in the interpreta- 

 tion of the rloat records, and it seems a pity that here 



