DEDUCTIONS AND REMARKS. 135 



to favor an idea that has been advanced, that there is a vast eddy extending from 

 the western shore of the Gulf of Mexico to the eastern shore of the Atlantic — that 

 the easterly trade-winds of the Atlantic Ocean, when they strike the American 

 Continent, veer northwardly and then toward the north-east, and thus recross the 

 Atlantic and follow down the coast of Portugal and Africa till they complete the 

 circuit. Though, on the whole, the evidence is against this theory. 



"We wait with interest for the results of the investigations, now going on under 

 the auspices of the Smithsonian Institution, in Oregon, California, and the terri- 

 tories west of the United States. When they shall be received, no doubt this 

 article will require modification. 



10. Near the limits which', divide this zone from the polar -winds on the north, 

 and from the equatorial on the south (particularly the latter), the progressive 

 motion is very small. The reader will notice the shortness of the arrows in South 

 Carolina, Georgia, Alabama, Mississippi, and other places along the line, as com- 

 pared with those farther north, on Plates VII. and VIII. The same thing is very 

 noticeable on the Atlantic Ocean, and, in some degree, in Europe. The only mate- 

 rial exception in the latter is Mafra, in Portugal, and it is exceedingly doubtful 

 whether that place is properly located. And not only is the progressive motion 

 small, but the direction is very uncertain. The different results obtained at 

 Augusta and Savannah, in Georgia, in different years, could hardly be more diver- 

 sified, and those of the upper and lower currents at Eutaw, Alabama (Nos. 169 and 

 170), are almost directly opposite each other. In Pekin, too, in China, which is 

 near the line, the results obtained by the French missionaries in the last century, 

 differ entirely from those of the recent Russian ones, as shown by the two arrows 

 at that place on Plate VII. 



11. The progressive motion is less in Europe than in America, as may readily be 

 seen by comparing the length of the arrows. 



12. There seems to be some approach to parallelism between the mean direction 

 of the wind in any part of the belt, and the direction in which that j)art of the belt 

 runs, so that the mean directions incline to make a constant angle with meridians 

 drawn through the pole of the belt.^ Thus, the winds are more southerly in the 

 eastern part of the Atlantic than in the western part — more so in western Europe 

 than in America or Asia. In eastern Siberia it is even north-westerly, if we may 

 rely upon the results at Yacoutsk and Nertchinsk, and the prevailing testimony of 

 navigators seems to be that the winds of the extreme North Pacific are also north- 

 westerly, though the observations at Iluluk, south of Behring's Strait, do not indi- 

 cate it. 



The following table shows the latitudes at which the limits of this zone cross the 

 different meridians, at intervals generally of 10°; the direction in which they run 

 reckoned eastwardly, and the region of country, &c., where they cross. 



* This remark is thrown out rather as a conjecture, which future observations may or may not verify. 



