TORNADOES. 215 
the direction of the winds in such storms this is purely accidental, though in 
America these winds are mostly toward the east. I say toward the east rather 
than from the west, for the reason that winds are pulled and not pushed—the 
force that creates them is always in advance and not behind. ‘These storms, 
though generally toward the east are not always so; it depends on the location 
of the area of ow. When the area of low barometer is on the land, the storm 
almost always follows this course, for it is natural that the greater force of the 
wind must be in the track of /ow as it advances toward the east from the west. 
For in this case we have not only the force of the wind in proportion to the in- 
tensity of this area of low barometer, but we undoubtedly have added to this the 
progressive force of Jow as it advances towardthe east. So our tornadoes are 
mostly in the track of a storm toward the east—most, but not always, for some 
times, more especially when the area of low barometer is on a high line of lati- 
tude, the storm takes place or begins with a southeast wind. This at first may 
seem contradictory to previous statements, and so also may the statement that, 
relatively to the storm the area of /ow is stationary, but with a little study of the 
weather maps in connection with storms it will be seen that the wind is ever 
changing as the area of /ow is passing over the country. In the east first the 
wind will be toward the west and to the south or north of due west as the area of 
the advancing /ow is on a low or high line of latitude—then, after the passing of 
Jow more or less reversed, or toward the east. ‘Ilo understand this better, let 
any one take a sheet of paper and mark its four sides—North, East, South and 
West. Over the paper sprinkle some iron filings. ‘Then, near enough to attract 
the iron, slowly draw a magnet from the west toward the east. It will be seen 
that the magnet will attract the particles of iron as it advances, and from all 
quarters, and that as it advances it will take up particles of iron with it, and that 
relatively to the iron the magnet is for the time being stationary, though the par- 
ticles of iron will follow the movements of the magnet. This, so far as iron will 
permit, is a fair illustration of the attraction or pulling power of /ow over air, 
though iron being a far more inert substance than air, is not so readily or ex- 
tensively affected by the magnet as the air is by the attractive force of Jow. 
Then, as to the direction of the wind in a tornado, instead of being toward 
the northeast, or toward the northwest, it may in some localities be from the 
northwest, asin Washington, July 4, 1874, and yet in the track of ow. A south 
west wind had been blowing all day—or in other words foward the northeast. 
Suddenly the wind changed to the northwest and blew a terrific storm that up- 
rooted trees and unroofed houses in this locality. 
In the early part of November, 1877, we had a similar storm on Long 
Asland Sound, when the steamer Massachusetts, being caught by it on a lee 
‘shore, came near being a total wreck. In these instances the cause was the same 
_as created the late tornado in Kansas and Missouri, only the area of low barom- 
“eter was in another locality, and,therefore,the wind that caused it must be froma 
different quarter. It was in the track of the area of low barometer. 
