TORNADOES. 185 



siderably warmer than tlie surrounding lands. From 

 its surface, also, aqueous vapor is continually being 

 raised. When the surrounding air is colder, this 

 vapor is presently condensed, generating in the change 

 a vast amount of heat. We have thus a channel of 

 rarefied air over the Mississippi Yalley, and this chan- 

 nel becomes a storm-track, like the corresponding 

 channels over the warm ocean-currents. The extreme 

 violence of land-storms is probably due to the narrow- 

 ness of the track within whicli they are compelled to 

 travel. For it has been noticed that the fury of a 

 sea-cyclone increases as the range of the " whirl " 

 diminishes, and vice versa. 



There seems, however, no special reason why cy- 

 clones should follow the storm -c| in one direction 

 rather than in the other. We must, to understand 

 this, recall the fact that under the torrid zones the 

 conditions necessary for the generation of storms prevail 

 far more intensely than in temperate regions. Thus 

 the probability is far greater that cyclones should be 

 generated at the tropical than at the temperate end of 

 the storm - ct Still it is worthy of notice, that in the 

 land-locked E"orth Pacific Ocean, true typhoons have 

 been known to follow the storm-track in a direction 

 contrary to that commonly noticed. 



The direction in which a true tornado whirls is 

 invariably that we have mentioned. The explanation 

 of this peculiarity would occupy more space than we 



