THE EARTH. 261 



surface of the basin below. The height at which 

 it is known to stand in the tube, is usually about 

 twenty-nine inches, when the air is heavy ; but 

 not above twenty-six, when the air is very light. 

 Thus, by this instrument, we can with some ex- 

 actness determine the weight of the air ; and, of 

 consequence, tell before-hand the changes of 

 the weather. Before fine dry weather the air 

 is charged with a variety of vapours, which float 

 ,in it unseen, and render it extremely heavy, so 

 that it presses up the quicksilver ; or, in other 

 words, the barometer rises. In moist, rainy wea- 

 ther, the vapours are washed down, or there is 

 jiot heat sufficient for them to rise, so that the 

 air is then sensibly lighter, and presses up the 

 quicksilver with less force ; or, in other words, 

 the barometer is seen to fall. Our constitutions 

 seem also to correspond with the changes of the 

 weather-glass ; they are braced, strong, and vigo- 

 rous, with a large body of air upon them; they 

 are languid, relaxed, and feeble, when the air is 

 light, and refuses to give our fibres their proper 

 tone. 



But although the barometer thus measures the 

 weight of the air with exactness enough for the 

 general purposes of life, yet it is often affected 

 with a thousand irregularities, that no exactness 

 in the instrument can remedy, nor no theory ac- 

 count for. When high winds blow, the quick- 

 silver generally is low: it rises higher in cold 

 weather than in warm ; and is usually higher at 

 morning and evening than at mid-day : it gene- 

 rally descends lower after rain than it was before 



