TORNADOES. 177 



directly along those tracks wliich they appear to fol- 

 low, a comparatively simple problem would be pre- 

 sented to the meteorologist. But this is not found to 

 be the case. At one part of a hurricane's course the 

 storm appears to be travelling with fearful fury along 

 the true storm- c{ at another less furiously directly 

 across the storm - track ; at another, but with yet 

 diminished force, though still fiercely, in a direction 

 exactly opposite to that of the storm-track. 



All these motions appear to be fairly accounted for 

 by the theory that the true path of the storm is a 

 spiral or rather, that while the centre of disturbance 

 continually travels onward in a widely-extended curve, 

 the storm-wind sweeps continually around the centre 

 of disturbance, as a whirlpool around its vortex. 



And here a remarkable circumstance attracts our 

 notice, the consideration of which points to the mode 

 in which cyclones may be conceived to be generated. 

 It is found, by a careful study of different observations 

 made upon the same storm, that cyclones in the 

 northern hemisphere invariably sweep round the on- 

 ward travelling vortex of disturbance in one direction, 

 and southern cyclones in the contrary direction. If 

 we place a watch, face upward, upon one of the 

 northern cyclone regions in a Mercator's chart, then 

 the motion of the hands is contrary to the direction in 

 which the cyclone whirls ; when the watch is shifted 

 to a southern cyclone region, the motion of the hands 



