Strictures on Dove's Essay " On the Law of Storms." 145 



that range. Obviously more than one half of the air in such a 

 cylindrical mass would be below the average level of the sum- 

 mits of those mountains. Under such circumstances could it be 

 conceived to rotate about a vertical axis ? 



125. I am aware that various writers have referred to the little 

 transient whirls which are occasionally seen to take place in a 

 windy time, carrying up dust, leaves, and other light bodies, as a 

 support for the idea of whirlwind storms ; and Mr. Redfleld has 

 alleged, " that no valid reason can be given why larger masses of 

 air may not acquire and develope similar rotative movemerits." 



126. It appears to me that there are several valid reasons for 

 not adopting the view of the subject which he has taken. The 

 momentum by which any body is kept in motion, is as its weight 

 multiplied by its velocity, while the expenditure of momentum 

 is ccetcris paribus as its surface. On this account, a globe of 

 which the content in proportion to its superficies is preeminently 

 great, will, in a resisting medium like the air, retain a rotary mo- 

 tion longer than an equal weight, of the matter forming it, in any 

 other shape. The flat cylinder, in diameter about two hundred 

 times its thickness, of which the existence would be necessary 

 to an extensive whirlwind, is a form of which the surface would 

 be very great in proportion to the quantity of matter which it 

 contains. No observer ever noticed any whirl produced as above 

 described, to have a diameter many times greater than its height, 

 or to endure many minutes. Such pigmy whirls appear to be the 

 consequence of eddies resulting from the conflict with each other, 

 or with various impediments, of puffs or flaws of wind. No doubt 

 in this way a deficit of local density is easily caused in a fluid so 

 elastic as the air, and consequently by gravity as well as its elas- 

 tic reaction, a centripetal motion is induced in the surrounding 

 aerial particles. From the confluence and conflict of the air thus 

 put into motion, a whirl may arise. The manner in which light 

 bodies are gathered towards the axis of these whirls, shows that 

 they are accompanied by a centripetal tendency. It is only when 

 the wind blows briskly that such whirls are ever seen to take 

 place, but tornadoes agreeably to universal observation occur 

 when there is little or no wind externally. (See objections, 

 par. 76.) 



127. According to the evidence adduced by the advocates of 

 the whirlwind theory, there is in this respect perfect similarity 



Vol. xliv, No. 1.— Oct.-Dec. 1842. 19 



