[ 383 ] 



heat 48^,396, which, fubtraded from 87P gives the temperature 

 at that height, 38,664. The mean heat of this column is there- 

 fore 62^,8, which exceeds the temperature of 32^ by 30^,8. 

 Now, according to Sir George Schuckburgh's calculation, 1000 

 feet of air, by one degree of heat above 32^, gain 2,43 feet, con- 

 fequently 21800 feet of air fhould gain 52,97 feet. And if one 

 degree of heat gives an increafe of 152,97 feet, 30,8 degrees fhould 

 give 1631,476 feet j confequently the mercury fhould rife in it to 

 the fame elevation as if the barometer had been placed 1631,476 

 lower, that is at the height of 21800 — 1631,476= 20168,524 

 feet = 3361,420 fathom, which fubtraded from log. 30, as above, 

 gives the logarithmic number 1409792, .correfponding with the 

 natural number 13,835. At this number of inches then the mer- 

 cury would ftand in the intr-atropical air. The denfity ^of the 

 fuperior air incumbent upon it exceeds that of the air incumbent 

 over the extra-tropical air of the fame height over the earth by 

 ,838 of an inch of mercury, or about ^ of the whole weight of 

 each column. It muft therefore flow over, or into the rarer ex- 

 tra-tropical air. 



This overflow takes place at even far lower heights in the at- 

 mofphere ; for by a fimilar calculation it will be found that the 

 denfity of the intra-tropical air exceeds that of the extra-tropical 

 by tV of an inch, even at the height of 8000 feet. 



3 B 2 - This 



