The story of the Louisville tornado serves to well illustrate all the peculiar features of the local whirlwinds produced by great storms. They seldom travel more than thirty miles; usually much less. Sometimes as large as two miles in diameter, they seldom exceed five hundred yards; and one of but fifty yards in diameter may be powerful enough to wreck a house. Often it is possible to trace the path of a tornado through the forest a century or more after its passage; for the reason that trees once destroyed are usually replaced by different varieties. But the tornado usually originates in the open country, though after its formation it may sweep through heavy timber. [Illustration: TORNADO FOLLOWED BY RAIN STORM.] So far as loss of life is concerned, the tornado is much more to be feared than lightning. About two thousand people have been killed in this country within ten years by these rotary storms. Yet, all over the land, people put up rods that are expensive, and often worse than useless, as a precaution against lightning, when a small “cyclone cellar” could be dug that would be far more useful, and less expensive. While intense electrical displays accompany the tornado, there is no authentic record of lightning striking during one; and as will be seen in another place, the amount of electricity present seems to be rather an effect than a cause: for rapid motion of gases may be made to produce powerful electric currents. While the tornado is justly feared in this country, yet, as a destructive agent, it is far surpassed by a number of [Illustration: Instantaneous View of a Tornado.] others whose ravages are less dreaded. It would be comparatively easy to show, we think, that more persons have been killed in one way or another by railways in ten years past than by tornadoes. The one that has been so carefully examined must not be considered as the worst our country has known. An examination of records of the past century will show a number that were more destructive to life and property. Doubtless, an account of some of these would interest the reader. Place is given to a few. The tornado has been observed, to some extent, in this country for more than a century: but only when our central states were well peopled did it attract very great attention. It is not common in the eastern states, and but one has ever been recorded west of Dodge City, Kansas. It is not unknown in Europe, though far less common than with us, having been noticed a few times in France. In general, it is so rare that a tornado that passed through Monville in 1845 attracted such attention as to be noticed in French text-books on physics. To the American, there is nothing unusual in the conduct of this storm. Perhaps the earliest detail of a storm of this sort among us is that of a double one in South Carolina, on the afternoon of May 2, 1761: