28 ANNUAL OSCILLATION OF THE BAROMETER. 



in the direction of the south pole; reasoning from 

 analogy we must expect to find that, all other cir- 

 cumstances being equal, the barometer will be lower 

 on that part of the earth's surface which may be 

 turned in the direction in which the earth moves in 

 its orbit than at the opposite point of the surface. 

 And as at the September equinox the point of sunrise 

 on the tropic of Cancer is in advance; and at the 

 March equinox the point of sunrise on the tropic of 

 Capricorn : these forces should, in the northern hemi- 

 sphere tend to cause a lower barometer at sunrise in 

 September than at sunrise in March: and also a 

 higher barometer at sunset in March than at sunset 

 in September. The reverse of these effects occurring 

 in the southern hemisphere. 



That is to say, that, on the average, the barometer 

 should, in the northern hemisphere show less atmo- 

 spheric pressure in September than in March ; whereas 

 in the southern hemisphere it should show less atmo- 

 spheric pressure in March than in September. And 

 that recorded observations are such as we have thus 

 deduced from theory, appears from the opposite 

 table, which we extract from Dove's l Law of Storms/ 

 page 42, which gives, 'in decimals of an inch, the 

 differences from the mean height of the barometer 

 observed in the Atlantic ocean, from latitude 35 N. 

 to 35 S.' The seasons marked in the table are those 

 of the northern hemisphere, and consequently the 

 table shows a clearly marked oscillation of the 



