164 Theory of Marine Currents. 
principal and permanent flow of waters. It appears that a 
small secondary circuit exists in the North Sea, around Ice- 
land. But we may mention more particularly the current 
derived from the Gulf Stream, which M. Duperrey directs to- 
wards the Icy Sea, along the coasts of Northern Europe. 
This current, flowing rapidly towards the north, must enter 
the sea bordering the north of Sibera with considerable ra- 
pidity towards the east, that is to say, towards Behring 
Straits. If the currents which descend from the west of the 
North Sea, indicate a similar movement in the waters on 
the north of America, we may consider the whole of the Icy 
Sea, comprised within the polar ice and the northern limits 
of the Old and New Continent, as having a circulatory mo- 
tion from west to east, and supported by the impulsion of 
masses of water reaching this sea by means of currents ori- 
ginating in lower latitudes. 
It only remains for us to examine the seventh division of 
the terrestrial waters, namely, the circular portion of the 
antarctic seas comprised between the icy regions of the 
south pole, and the southern limits of the three southern 
circuits of the Indian, Pacific, and Atlantic Oceans. It is to 
be supposed that the influence of the movements of these 
three circuits, which all convey their waters from the west 
to the east in the neighbourhood of the antarctic seas, pro- 
duces, by communicating movement to the mass of waters 
composing this sea, a movement likewise directed to the east ; 
this communication of rotatory motion being moreover sub- 
jected to all the circumstances of depth, breadth, friction, and 
obstruction, which in general leave nothing more, in all regu- 
lar and permanent movements, whether primitive or com- 
municated, than one sole law, the law of equal depense, which 
determines and regulates alike the local speed of all perma- 
nent fluviatile currents. 
M. Duperrey’s chart does not appear to us to present any 
thing opposed to this mode of considering the subject. 
By joining, therefore, this current or southern circumpolar 
circuit to the northern circumpolar circuit, and both of them 
to the five great circuits which extend to the equator in one 
of their directions, we shall have descriptively and theoreti- 
