114 
THE PHYSICAL GEOGRAPHY OF THE SEA. 
or at least to that degree which obstructs or prevents, we all 
know. 
206. For example, open the window of a warm room in winter, 
and immediately there are two currents of air ready at once to set 
through it ; viz., a 'current of warm air flowing out at the top, and 
one of cold coming in below. 
But the brown fields in summer afford evidence on a larger- 
scale, and in a still more striking manner, of the fact that, in na- 
ture, columns, or streamlets, or curdles of air do readily move 
among each other without obstruction. That tremulous motion 
which we so often observe above stubble-fields, barren wastes, or 
above any heated surface, is caused by the ascent and descent, at 
one and the same time, of columns of air at different tempera- 
tures, the cool coming down, the warm going up. They do not 
readily commingle, for the astronomer, long after nightfall, w^hen 
he turns his tele&cope upon the heavens, perceives and laments 
the unsteadiness they produce in the sky. 
207. If the air brought down by the northeast trade-winds differ 
in temperature (and why not?) from that brought by the southeast 
trades, we have the authority of nature for saying that the two 
currents would not readily commingle. Proof is daily afforded 
that they would not, and there is reason to believe that the air of 
each current, in streaks, or patches, or curdles, does thread its way 
through the air of the other without difficulty. Now, if the air of 
these two currents differs as to magnetism, might not that be an 
additional reason for their not mixing, and for their taking the di- 
rection of opposite poles after ascending 1 
208. Therefore we may assume it as a postulate which nature 
concedes, that there is no difficulty as to the two currents of air, 
which come into those calm belts from different directions, cross- 
ing over, each in its proper direction, without mingling. 
209. Thus, having shown that there is nothing to prevent the 
crossing of the air in these calm belts, I return to the process of 
reasoning by induction, and offer additional circumstantial evi- 
dence to prove that such a crossing does take place. Let us 
therefore catechise, on this head, the waters which the Mississippi 
pours into the sea, inquiring of them as to the channels among the 
clouds through which they were brought from the ocean to the 
fountains of that mighty river. 
