84 MR. T. HOPKINS ON IMPROVEMENT 
not lose its property of gravitation, but as ponderable 
matter must float in and be sustained by the gases. 
Indeed, it may generally be seen there as mist or cloud, 
and as both are sustained by the gases, the weight and 
pressure of the gases must be augmented. 
But from nine to five we have the fact that the barome- 
ter fell considerably, which proves that the gases must 
then have pressed on the surface with diminished force: 
and as they bear the weight of the clouds that float in 
them, they must themselves, from some cause, have been 
rendered so much lighter, as to be less heavy with the 
cloud in them than they were before the cloud was formed. 
That they are reduced in weight, whatever may be the 
cause of the reduction, is proved both by the falling of the 
barometer at the time, and by the kind of disturbance 
which takes place in the atmosphere. Solar heat makes 
the air light at the time, although it is evidently not the 
identical heat which at the same time affects the thermo- 
meter near the surface, It is, however, at the time that 
the barometer falls that the surface is heated, because 
solar heat is also then displayed in heating the surface ; 
but it does not then materially disturb atmospheric press- 
ure unless it finds aqueous matter to unite with. When, 
however, the heat forms an union with water, it passes 
into the upper regions of the atmosphere, where it acts 
in expanding the gases, diminishing atmospheric pressure, 
and producing winds. 
In meteorological registers the tension of vapour is 
sometimes deducted from total atmospheric pressure, and 
the remainder being treated as dry gaseous matter, is 
tabulated and occasionally exhibited in curved lines. 
These lines present an appearance to the eye the 
reverse of the curve of temperature as shown by the 
thermometer; and it seems not improbable that this form 
of exhibiting changes in the atmosphere has contributed 
