98 
THE PHYSICAL GEOGRAPHY OF THE SEA. 
But the air is invisible ; and it is not easily perceived hov^ either 
marks or tallies may be put upon it, that it may be traced in its 
paths through the clouds. 
The skeptic, therefore, who finds it hard to believe that the gen- 
eral circulation is such as Plate I. represents it to be, might con- 
sider himself safe in his unbelief were he to declare his willingness 
to give it up the moment any one should put tallies on the wings 
of the wind, which would enable him to recognize that air again, 
and those tallies, when found at other parts of the earth's surface. 
As difficult as this seems to be, it has actually been done. 
Ehrenberg, with his microscope, has established, almost beyond a 
doubt, that the air which the southeast trade-winds bring to the 
equator does rise up there and pass over into the northern hemi- 
sphere, 
158. The Sirocco, or African dust, which he has been observ- 
ing so closely, has turned out to be tallies put upon the wind in 
the other hemisphere ; and this beautiful instrument of his enables 
us to detect the marks on these little tallies as plainly as though 
those marks had been written upon labels of wood and tied to the 
wings of the wind. 
This dust, when subjected to microscopic examination, is found 
to consist of infusoria and organisms whose habitat is not Africa, 
but South America, and in the southeast trade-wind region of 
South Anierica. Professor Ehrenberg has examined specimens 
of sea dust from the Cape de Verds and the regions thereabout, 
from Malta, Genoa, Lyons, and the Tyrol ; and he has found a 
similarity among them as striking as it would have been had these 
specimens been all taken from the same pile. South American 
forms he recognizes in all of them ; indeed, they are the prevail- 
ing forms in every specimen he has examined. 
It may, I think, be now regarded as an estabhshed fact, that 
there is a perpetual upper current of air from South America to 
North Africa ; and that the volume of air which flows to the north- 
ward in these upper currents is nearly equal to the volume which 
flows to the southward with the northeast trade-winds, there can 
be no doubt. 
The " rain dust" has been observed most frequently to fall in 
spring and autumn; that is, the fall has occurred after the equi- 
noxes, but at intervals from them varying from thirty to sixty 
