349 



know that when comparison is made between places in the same 

 latitude, and having the same, or very nearly the same, differences of 

 temperature in summer and in winter, the differences between the 

 summer and winter pressures of the dry air are found to be subject 

 to many remarkable anomalies. The variations in the pressure of 

 the dry air do not therefore, as might be at first imagined, depend 

 altogether on the differences between the summer and winter tem- 

 peratures at the places where the variations themselves occur. The 

 increased pressure in the hottest months appears rather to point to 

 the existence of an overflow of air in the higher regions of the 

 atmosphere from lateral sources ; the statical pressure at the base 

 of the column being increased by the augmentation of the super- 

 incumbent mass of air arising from an influx in the upper portion. 

 Such lateral sources may well be supposed to be due to excessive 

 ascensional currents caused by excessive summer heats in certain 

 places of the globe (as, for example, in Central Asia). Now the 

 lateral overflow from such sources, traversing in the shape of cur- 

 rents the higher regions of the atmosphere, and encountering the 

 well-known general current flowing from the equator towards the 

 pole, has been recently assigned with considerable probability (de- 

 rived from its correspondence with many otherwise anomalous 

 phenomena already known, and which all receive an explanation 

 from such supposition) to be the original source or primary cause 

 of the rotating storms or cyclones, so well known in the West Indies 

 and in China under the names of hurricanes and typhoons. A 

 single illustration may be desirable. Let it be supposed that such 

 an excessive ascensional current exists over the greatly heated parts 

 of Asia and Africa in the northern tropical zone, giving rise, in 

 the continuation of the same zone over the Atlantic Ocean, to a 

 lateral current in the upper regions ; this would then be a current 

 prevailing in those regions from east to west : and it would encounter 

 over the Atlantic Ocean the well-known upper current proceeding 

 from the equator towards the pole, which is a current from the 

 south-west. An easterly current impinging on a south-west current 

 may give rise, by well-known laws, to a rotatory motion in the atmo- 

 sphere, of which the direction may be the same as that which cha- 

 racterises the cyclones of the northern hemisphere. To test the 

 accuracy of this explanation, we desire to be acquainted with the 



