332 THEORY OF HURRICANES, ETC. 



gulf of Mexico, and the southern portions of the 

 United States, are also produced in the same 

 way, and are attended by extraordinary floods 

 of rain, with fearful displays of thunder and 

 lightning. An eye witness of the tremendous 

 hurricane which desolated the island of Barba- 

 does on the 10th of August, 1831, informed the 

 author, that it began about ten o'clock at night, 

 with torrents of rain and broad sheets of fire in 

 rapid succession, which threatened to overwhelm 

 the inhabitants in a flood, or sweep them away 

 in a tempest of flame. 



On the other hand, in the tropical portions of 

 the wide Pacific, where the temperature is uni- 

 form, and the winds blow steadily in the same 

 direction, there is but little thunder, lightning, 

 and rain, and no hurricanes. The same is true 

 of the equatorial parts of the Atlantic, which are 

 far from land. It is stated by Capper, in his 

 account of winds and monsoons, that a hurri- 

 cane was never known to occur at St. Helena, 

 situated as it is, nearly midway between Africa 

 and South America. It is near the continents 

 of Asia, Africa, and America, or in the vicinity 

 of large islands, where immense masses of va- 

 pour from the sea, meet with mountains and cold 

 masses of air from the land, that hurricanes of 

 wind, lightning, and rain, are most powerful and 

 frequent, as in the Bay of Bengal, on the coast 

 of Madagascar, Mauritius, &c. In short, there 



