14 



the barometer falls as low as 28'25 inches. This may 

 appear extraordinary if we merely take into consideration 

 the actual height of the column, and neglect the quantity 

 of the fall which amounted to T65 inch. The mean 

 height of the greatest observed elevations for the last 

 thirty-eight years is 3O61 inches, and the mean height 

 of the observed depressions for the same period is 28'69 

 inches ; therefore a fall in the mercury of 1'65 inch from 

 the mean of the elevations would give a minimum of 

 28'96 inches ; a depression which is contemporary with 

 violent storms, as it is within three-tenths of the mean 

 of the lowest depressions of the barometer. 



In iking the barometer great care must be taken to 

 fix it perpendicular : a situation should be selected sub- 

 ject to the least change of temperature, for which reason 

 a northern aspect is preferable to a southern ; the height 

 of the cistern of the barometer above the level of the sea, 

 and, if possible, the difference of the height of the mer- 

 cury with some standard, should be ascertained, in order 

 that the observations made with it should be comparative 

 with others made in different parts of the country. Before 

 taking an observation, the instrument should be gently 

 tapped to prevent any adhesion of the mercury to the 

 tube, the gauge should be adjusted to the surface-line 

 of the cistern, and the index of the Vernier brought 

 level with the top of the mercury. If the barometer 

 have a Vernier which admits the light from behind, the 

 lower part of the pointer must make a tangent with 

 the convex part of the mercury in the tube. In reading 

 off the observation the eye should be on a line with the 

 mercury; as by placing it above, the reading would be 



