The destructive effects of the tornado are very curious, from the sharp and narrow belt to which the injury is confined. It appears that in the passage of some tornadoes wind-pressures of various amounts, from eighteen to a hundred and twelve pounds per square foot, have been demonstrated by destruction of bridges, brick buildings, etc. The upward pressures are sometimes as great as the horizontal, and even greater. Downward pressures or movements of wind have not been clearly proved. Upward velocities of 135 miles per hour seem not to be unusual, and horizontal velocities of eighty miles have been recorded with the anemometer. The destructive wind-velocities are confined to very small areas. A destruction of fences, trees, etc., is often visible over a path many miles long and a few hundred yards wide, but the path of greatest violence is very much narrower. The excessive cases above referred to are observed only in small isolated spots, less than a hundred feet square, unequally distributed along the middle of the track. Thus, in very large buildings, only a small part is subject to destructive winds. In different parts of this area of _maximum_ severity, the winds are simultaneously blowing in different, perhaps opposite, directions, the resultant tending not to overturn or carry off or crush in, but rather to twist round a vertical axis. Buildings are generally lifted and turned round before being torn to pieces. As the chances are very small that a building will be exposed to the violent twisting action, it is evidently the average velocity of rectilinear winds within the path of moderate destruction that it is most necessary to provide against in ordinary structures. These winds may attain a velocity of eighty miles an hour over an area of a thousand feet broad, and generally blow from the southwest; the next in frequency blow from the northwest. The time during which an object is exposed to the more destructive winds varies from six to sixty seconds. An exposed building experiences but one stroke, like the blow of a hammer, and the destruction is done. Hence, in a suspension-bridge, chimney, or other structure liable to be set into destructive rhythmic vibrations, the _maximum_ winds do not produce such vibrations. The duration of the heavy southwest or northwest winds over the area of moderate destruction is rarely over two minutes. The motion of translation of the central spout of a tornado, in which there is a strong vertical current, is, on ah average, at the rate of thirty miles an hour. Tornadoes mostly occur on sultry days and either in the southeast or right front of cyclones, or in front of the trough of V-depressions. The general character of all tornadoes is so similar that the description of one will do for all. We shall therefore give some of the description furnished by an eye-witness to the United States Signal Office, which is described in the reports as the “Delphos tornado”: