400 Miscellanies. 



face by boiling prevented any very marked action of the whirl. As the 

 risinqr fluid evidently ascended the clear space in the centre of the cone or 

 column, it was certain that the column was hollow, and that within the 

 whirl there was a diminished atmospheric pressure. During the times that 

 the water mounted the highest, (which was between four and five inches,) 

 there was a violent agitation of the surface in the immediate vicinity of, 

 or beneath the base of the rotating column, and a careful examination 

 showed that small pieces of the foam were occasionally wrested from the 

 upper part of the rising water and instantly disappeared. It could not be 

 seen that there was any distinct rotation to the elevated water, which 

 swayed and bent whh the column of steam. 



It appeared to us that from this incident, simple and trifling as it may 

 appear in itself, some valuable inferences may be drawn. The origin of 

 water spouts, in connection with whirlwinds, and the laws that regulate 

 the ascent of water, were well e.xhibited. That water should ascend to 

 the height it evidently does in water spouts at sea, by atmospheric pres- 

 sure alone, is not to be supposed ; but it is atmospheric pressure that 

 forces the water into the hollow at the base of the cone, and places it in a 

 position to be first acted upon and then lifted by the whirling air. When 

 once the upward current is established, there seems to be no difficulty in 

 continuing it; and, as the water thus lifted must return to the earth by 

 being thrown without the upper circumference of the whirl, or when the 

 column is suddenly separated, by pouring downward with the same volume 

 with which it was rising, it accounts for the deluges of water that at times 

 accompany water spouts. 



The action or ascent of the water within the tube also showed th-^t the 

 atmospheric pressure was greatly lessened or removed in the interior of 

 the whirl, and thus explained satisfactorily many of the phenomena that 

 accompany tornadoes or whirlwinds. Thus it has always been found in 

 violent tornadoes, that the windows or gables of buildings that were near 

 the centre of the line of the whirlwind, are almost invariably burst out- 

 wards, and frequently directly in the face of the advancing storm. This 

 was particularly noticed at the destruction of Natchez, and at Shelbyville, 

 and cannot be well accounted for in any other way than by the violent 

 expansion of the air within the buildings, when the outer pressure is sud- 

 denly taken off". 



In many storms or tornadoes, the thunder does not appear in distinct 

 explosions, nor the lightning in separate flashes. On the contrary, there 

 is a continued blaze of fire in the cloud and the roll of thunder is inces- 

 sant. In such cases, effects are observed which indicate in the line of 

 the storm the continued action of electric energy, and give reason to sup- 

 pose that the ascending column produced by the whirl forms a perfect 

 conductor, along which the electric fluid descends continuously and not 

 in successive masses. Thus in most tornadoes the trees within their 



