in a Voyage across the Atlantic. 125 
and to the want of that refreshing influence in the other; in 
short, to the trade-wind, which probably lowers the temperature 
of the sea’s surface, by evaporation; for it is dry, while equa- 
torial winds are damp. 
The trade-winds, which blow over the middle Atlantic, are 
ascribed to the rarefaction of the atmosphere by the sun’s heat. 
Their proximate cause seems to be the diminution of the air’s 
specific gravity by absorption of moisture. No doubt they are 
a supply of denser air taking the place of that which is lighter, 
and which rises therefore in the atmosphere. 
Diffusion of moisture, like that of temperature, through air, 
is attended with internal commotion. Particles of air, becoming 
lighter by accession of warmth or of moisture, from a warm or 
a humid surface, rise and are succeeded by others ; which in 
their turn become light, and similarly ascend. This has been 
long since shewn in respect of diffusion of temperature in fluids; 
and may be presumed in regard to humidity, as a more 'con- 
sistent explanation than that which supposes moisture derived 
by affinity of a less saturated portion of air above, from a more 
saturated one beneath; and so enabling this to dissolve more, 
which is in like manner transmitted. 
While evaporation is going on, humid air is continually 
rising. It quits the evaporating surface, previous to full satu- 
ration, because it earlier becomes specifically lighter than drier 
air immediately above it; and rises therefore through superior 
strata, until it reaches an elevation suited to its levity: and there 
gradually parting with excess of heat, parts likewise with 
moisture which it can no longer sustain. At that elevation, 
therefore, vapour is condensed from it; and clouds are 
seen. 
The atmosphere over the sea, examined as close to the sur- 
face as waves permit, manifests no saturation, though sensibly 
moister than at a little height above it. For instance, a differ- 
ence of one degree of dryness has been observed at an interval 
of eleven feet, when the temperature at both places has been 67°, 
and the point of dew was 59° at the upper station, and 60° at 
