of Meteorological Obfervations. 249 
lefs force on the mercury of the barometer than 
before, by which means a fall enfues. 
24th. ‘That the caufe of the currents of air, or 
winds, may alfo be this way accounted for: and jn 
very fevere ftorms, where great decompofitions of 
the atmofphere take place, this is particularly evi- 
dent, fuch as occur generally in one or more of the 
Weft India Iflands at one time, a great lofs of 
real gravity, together with a confiderable diminution 
of the {pring of the air immediately: enfues, hence 
a current commences, firft in that direction whence 
the air has moft gravity, or is: moft difpofed. to 
undergo fuch a change; but it being foon relieved 
of its fuperior weight or {pring on that fide, by the 
decompofition going on as faft as the-wind arrives on 
the ifland, it immediately veers to another point, ° 
which then rufhes in moftly with an increafe of 
force; thus it goes on till it has blown more than 
half way round the points of the compas during © 
the continuation of the hurricane. For in this 
manner thefe Weft India phenomena, as well as 
the alteration of the wind during heavy rains in this 
country (fee remark No. 4) can only be properly 
accounted for. 
25th. That the rifing of the barometer may be 
accounted for by the watery vapour and other con- 
ftituent parts of the atmofphere being thoroughly 
concentrated or combined together in the forin of a 
real mixt, by means of the phlogifton, latent heat, 
or electricity communicated by the fun, fire on the 
T7 furface 
