350 



variations which the mean pressure of the dry air undergoes in the 

 different seasons in the part of the globe, where, according to this 

 explanation, considerable variations having particular characters 

 ought to be found. 



2. We have named one of the explanations which have been 

 recently offered of the primary cause of the northern cyclones. 

 Another mode of explanation has been proposed, by assuming the 

 condensation of large quantities of vapour, and the consequent 

 influx of air to supply the place. In such case the phenomena are 

 to be tested in considerable measure by the variations which the 

 other constituent of the barometric pressure, namely, the aqueous 

 vapour, undergoes. 



3. The surface of sea in the southern hemisphere much exceeds 

 that in the northern hemisphere. It is therefore probable that at 

 the season when the sun is over the southern hemisphere, evapora- 

 tion over the whole surface of the globe is more considerable than 

 in the opposite season when the sun is over the northern hemi- 

 sphere. Supposing the pressure of the dry air to be a constant, 

 the difference of evaporation in the two seasons may thus produce 

 for the whole globe an annual barometric variation, the aggre- 

 gate barometric pressure over the whole surface being highest 

 during the northern winter. The separation of the barometric 

 pressure into its two constituent pressures would give direct and 

 conclusive evidence of the cause to which such a barometric 

 variation should be ascribed. It would also follow, that evapo- 

 ration being greatest in the south, and condensation greatest in 

 the north, the water which proceeds from south to north in a state 

 of vapour, would have to return to the south in a liquid state, and 

 might possibly exert some discernible influence on the currents of 

 the ocean. The tests by which the truth of the suppositions thus 

 advanced may be determined, are the variations of the meteor- 

 ological elements in different seasons and months, determined by 

 methods and instruments strictly comparable with each other, and 

 arranged in such tables as have been suggested. A still more 

 direct test would indeed be furnished by the fact (if it could be 

 ascertained), that the quantity of rain which falls in the northern is 

 greater than that which falls in the southern hemisphere ; and by 

 examining its distribution into the different months and seasons 



